Personalized Nutrition

9281_C000.
fm Page i Thursday, July 26, 2007 7:30 PM
PERSONALIZED
NUTRITION
Principles and Applications
Edited by
Frans Kok
Laura Bouwman
Frank Desiere
Boca Raton London New York
CRC Press is an imprint of the

Taylor & Francis Group, an informa business
9281_C000.fm Page ii Thursday, July 26, 2007 7:30 PM
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
2008 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works

Printed in the United States of America on acid-free paper
10 9 8 7 6 5 4 3 2 1
International Standard Book Number-10: 0-8493-9281-0 (Hardcover)

International Standard Book Number-13: 978-0-8493-9281-8 (Hardcover)
This book contains information obtained from authentic and highly regarded sources. Reprinted
material is quoted with permission, and sources are indicated. A wide variety of references are
listed. Reasonable eorts have been made to publish reliable data and information, but the author
and the publisher cannot assume responsibility for the validity of all materials or for the conse-
quences of their use.
No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any
electronic, mechanical, or other means, now known or hereafter invented, including photocopying,
microlming, and recording, or in any information storage or retrieval system, without written
permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.
copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC)
222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-prot organization that
provides licenses and registration for a variety of users. For organizations that have been granted a
photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and
are used only for identication and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
Personalized nutrition : principles and applications / editors, Frans Kok, Laura

Bouwman, and Frank Desiere.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-8493-9281-8 (alk. paper)
1. Nutrition--Genetic aspects. 2. Genomics. I. Kok, Frans. II. Bouwman,
Laura, 1971- III. Desiere, Frank. IV. Title.
QP144.G45P47 2007
612.3--dc22 2007012503
Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
9281_C000.fm Page iii Thursday, July 26, 2007 7:30 PM
Table of Contents
Section I
Scientific Principles of Personalized Nutrition .....................................................1
Chapter 1 Nutrigenomics and Transcriptomics: Applications

of Microarray Analyses in Nutrition Research ...................................3
Lydia A. Afman and Michael Mller
Chapter 2 Exploring the Proteome for Markers of Health.................................13

Manuela Rist and Hannelore Daniel
Chapter 3 Metabolomics and the Personalized Metabolic Signature ................23

Michael J. Gibney, Marianne Walsh, and Lorraine Brennan
Chapter 4 Nutrition, Genomics, and Cardiovascular Disease Risk ...................33

Jose M. Ordovas and Dolores Corella
Chapter 5 Nutrigenomics and Chronic Inflammation ........................................49

Lynnette R. Ferguson and Martin Philpott
Chapter 6 Personalized Prevention of Type 2 Diabetes.....................................61

Hans-Georg Joost
Chapter 7 Toward Personalized Nutrition for the Prevention

and Treatment of Cancer ...................................................................75
John C. Mathers
Chapter 8 Nutrigenomics and Angiogenesis in Obesity ....................................89

Aldona Dembinska-Kiec
Chapter 9 Metabolic Programming during Pregnancy: Implications for

Personalized Nutrition......................................................................101
Simon C. Langley-Evans
9281_C000.fm Page iv Thursday, July 26, 2007 7:30 PM
Chapter 10 Taste as the Gatekeeper of Personalized Nutrition .........................115

Toshiko Tanaka, Danielle R. Reed, and Jose M. Ordovas
Chapter 11 Personalized Nutrition and Public Health .......................................133

Pieter vant Veer, Edith J.M. Feskens, and Ellen Kampman
Section II
Personalized Nutrition and Stakeholders in Society........................................149
Chapter 12 Imminent Applications of Nutrigenomics:

A Stakeholder Analysis....................................................................151
J.T. Winkler
Chapter 13 The Personal Factor in Nutrition Communication ..........................169

Laura I. Bouwman, Maria A. Koelen, and Gerrit J. Hiddink
Chapter 14 A Marketing and Consumer Behavior Perspective

on Personalized Nutrition ................................................................185
Hans C.M. van Trijp and Amber Ronteltap
Chapter 15 U.S. Consumer Attitudes toward Personalized Nutrition................205

David B. Schmidt, Christy White, Wendy Reinhardt Kapsak, Josh Conway,
and Elizabeth Baily
Chapter 16 Ethics of Personalized Nutrition......................................................221

Michiel Korthals
Section III
The Future of Personalized Nutrition ...............................................................235
Chapter 17 International Efforts on Nutrigenomic Health

for Individuals in the Global Community .......................................237
Jim Kaput
Chapter 18 The Future of Foods ........................................................................261

Heribert J. Watzke and J. Bruce German
Index ......................................................................................................................279
9281_C000.fm Page v Thursday, July 26, 2007 7:30 PM
Foreword
PERSONALIZED NUTRITION: HISTORY
AND PROMISES?
Personalized nutrition is an emerging but controversial new discipline among nutri-
tion scientists. Due to its possible genetic component, it has become an attractive
topic for the media. So far, the science fiction stories of a lady doing food shopping,
guided by a PDA containing her genome information are not taken very seriously.
Of course, in the developed world, personalized food choice is taken for granted.
Did you realize that while ordering Italian three-taste gelato, you can choose from
over a million combinations and variations, if you consider tastes, sizes, toppings,
and packaging? The same is true for coffee or take-away pizzas.
In the context of this book, personalized nutrition is related to a differentiated
health effect, with the key questions targeted at the possible scientific basis for this
phenomenon. In order to put this in the right perspective, let us have a look at the
historical development of the diet and health relationship. Until the early twentieth
century, nutritional sciences dealt with hygiene and processing. This was followed
by the incorporation of biochemistry and (molecular) physiology: the vitamins were
described, energy metabolism was understood, etc. During the second half of the
twentieth century, knowledge of cellular and molecular biological processes allowed
for further optimizing of our diet toward a balanced macro- and micronutrient
composition. This contributed to a prolongation of life span and produced a com-
mercial wave of supplements.
Gradually, nutrition scientists indicated the involvement of diet in a large number
of diseases and disorders (e.g., colon cancer, cardiovascular disorders, type 2 diabetes
mellitus, a number of inflammation-related health problems, and many more. During
the 1990s, this development triggered the introduction of functional foods, dietary
components with added health value. So far, these functional foods, apart from
the aforementioned supplements (vitamins, cofactors, minerals, and so forth, which
I do not consider to be real functional foods), have resulted in only a limited number
of successful products (cholesterol-lowering stanols, probiotics, a number of specific
fatty acids). Except for the cholesterol-lowering products, where an established
biomarker (the relationship between CVD and HDL/LDL cholesterol) was estab-
lished by the biomedical research community, it was very difficult to obtain scientific
proof of efficacy for most other functional foods. In many cases, the advertisement
and PR efforts by far outreached the scientific efforts accompanying the market
introduction.
Why does nutrition science have such a hard job in providing evidence for health
claims related to a dietary component? In pharmacological and biomedical research,
bioactive compounds are developed to treat a well-characterized disease. In contrast,
nutrition deals with prevention of disease and optimizing health. Here, biomarkers
9281_C000.fm Page vi Thursday, July 26, 2007 7:30 PM
quantifying the health status are essentially missing, and much of the nutrition
research (the large observational and intervention cohorts) rely on disease endpoints
instead of health endpoints. Also in the gold standard of diet and health research,
the (cross-over) dietary intervention trial, the quantification of the effect is a major
issue. Usually, the observed effects are small, and great efforts have to be made to
unravel the intervention effect from potential confounding variables. In other words,
the confounding parameters have a larger impact than the intervention effects. The
recent omics-related observations in human intervention studies, using very accurate
biomarkers, confirm that intraindividual variation is much smaller than interindividual
variation. Differences between study subjects may be much larger than differences
affected by the dietary treatment.
The keys for personalized nutrition actually are these confounders that make
the life of nutrition scientists so difficult. Age, gender, lifestyle (e.g., exercise),
phenotype (e.g., body mass index), genetic makeup, and epi-genomic imprinting
all possibly determine our nutritional needs, the way we respond to nutrition, and
thus our personal diet-and-health relationship. Infant nutrition clearly differs from
a sports diet. Now, a triad of questions arise:
1. To what extent is this personal diet-and-health relationship practically

valid?
2. How can nutrition science demonstrate this?
3. What is the proposition of stakeholders in society, including the con-
sumer?
This book at least attempts to answer these questions. My personal opinion is that
indeed this relationship exists to a much greater extent than assumed until now, and
that nutrition science will need to do a much better job in accurately identifying and
quantifying the subtle differences in health status related to dietary treatment. A
complete merger of nutrition with a number of fundamental scientific disciplines
(molecular biology, biochemistry, bioinformatics, statistics, etc.) will be essential to
reach the goal.
Now, what about this specific and overemphasized part of personalized nutrition,
the impact of genetic differences? Of course, we know the examples of cholesterol
and folate, where genetic predisposition partly determines plasma concentration and
related health effects. In the case of cholesterol, a phenotypic assessment of LDL
and HDL provides a solid phenotypic biomarker, which makes genetic testing (at
least partly) unnecessary. In fact, this phenotypic readout is the gross result of at
least 20 genetic different makeups. The HDL/LDL relationship partly determines
biomedical intervention and the consumers food choice. In the case of folate, the
occurrence of a genetic polymorphism in one of the enzymes involved in folate
biotransformation results in a relatively low plasma folate concentration. In extreme
cases, this leads to embryonic deformation (spina bifida) due to impaired DNA
synthesis. Folate is essential for DNA synthesis as it provides the methyl group for
nucleotide synthesis. Folate is cheap, and some countries have turned to folate
supplementation to the general population (e.g., by fortification of bread). In other
words, a genetically based differential dietary advice is overruled by mass treatment.
9281_C000.fm Page vii Thursday, July 26, 2007 7:30 PM
Possible negative consequences in persons efficiently dealing with folate (high folate
levels have been shown to overmethylate DNA and are thus potentially mutagenic
and carcinogenic) were considered nonrelevant. Time will tell whether a personalized
approach would be better.
A number of disorders related to nutrition are packed with genetic variation, but
the effect of nutritional modulation of the phenotypic outcome of these variations
is difficult to assess as yet. Nutrition is the worst-case scenario for this approach in
science as multiple minor genetic differences can occur, possibly modulated by
multiple food bioactives, usually with low receptor affinity, and resulting in multiple
minor changes in gene expression and resulting phenotypic expression.
Eventually, nutrition science may very likely determine a large number of per-
sonalized nutrition and health relationships. However, this is only a small part of
the equation. Food consumption nowadays is only partially driven by health con-
cerns. It is still much more driven by convenience and price. Food is pleasure
rightfully is a universally held credo, and science will have a hard job in promoting
healthy diet if this aspect is compromised. Therefore, a personalized diet needs to
be both optimized toward personal health and also for personal convenience, plea-
sure, and affordability. What a challenge!
Ben van Ommen

Senior Research Fellow, Nutritional Systems Biology
Director, European Nutrigenomics Organisation (NuGO)/TNO Quality of Life
Zeist, The Netherlands
9281_C000.fm Page viii Thursday, July 26, 2007 7:30 PM
9281_C000.fm Page ix Thursday, July 26, 2007 7:30 PM
Preface
The primary role of diet is to provide an individual with optimal levels of nutrients
to meet metabolic requirements, and to give the consumer a feeling of satisfaction
and well being through the pleasure of eating. In addition, particular diets, foods
and food components can provide additional physiological, cognitive and psycho-
logical benefits, and result in biological activities beyond their widely accepted
nutritional effects. Moreover, diet not only helps to achieve and maintain optimal
health and development but can also play an important role in reducing the risk of
specific diseases and disorders.
However, a diet that serves the health of one individual does not necessarily
work for all individuals due to differences in (genetic) predisposition, and environ-
mental and lifestyle factors. Hence, diagnostic tools that reflect both the overall
health status and (genetic) predisposition of a particular person at a given time need
to be developed. The knowledge obtained thereby will facilitate appropriate nutri-
tional advice to individuals according to their immediate and long-term health needs
and, ultimately, we can foresee the time when nutritional products will be tailored
for individual consumers.
The success of the introduction of personalized nutrition in society depends on many
factors. The scientific background for this must be established, the regulatory system
must be in place, ethical issues need to be addressed, services and products with clear
health benefits must be available, and the accompanying communication must be clear.
This textbook has the aim of defining the area of personalized nutrition both
from a biomedical and social science perspective. A selected group of leading
scientists in the field will comprehensively address the molecular, physiological,
epidemiologic, and public health aspects of personalized nutrition, highlighted with
examples from major diseases. Another group of well-known social scientists will
discuss the behavioral, ethical, and consumer perspectives that will influence a
legitimate successful introduction of personalized nutrition.
We expect our book will be useful for the education of students in several
disciplines, for example, nutrition, behavioral and communication science. More-
over, stakeholders involved in personalized nutrition in government, health care, and
business may use the book as a reference guide.
Our book is divided in three sections: (1) Scientific Principles of Personalized
Nutrition, (2) Personalized Nutrition: Consumer and Society, and (3) Future Per-
spectives on Personalized Nutrition.
In the first section, Scientific Principles of Personalized Nutrition, the state of
the art of nutrigenomics technologies, including transcriptomics, proteomics, and
metabolomics, are discussed. Subsequently, the use of genomics technology for a
better understanding of the molecular mechanisms involved in major diet-related
chronic disorders i.e., chronic inflammation, cardiovascular disease, type 2
diabetes, cancer, and obesity is addressed.
9281_C000.fm Page x Thursday, July 26, 2007 7:30 PM
Next, the consequences of metabolic programming during pregnancy and the

potential of new molecular biomarkers, as well as the gatekeeper role of taste in
personalized nutrition, are covered. In the last chapter in this section, the epidemi-
ologic aspects of genomic testing and the health benefits for the individual vs. the
population are described.
Besides providing the most recent knowledge, all authors discuss opportunities
and pitfalls of nutrigenomics in the development of personalized nutrition.
The second section, Personalized Nutrition: Consumer and Society, starts with
a stakeholder analysis in which five interest groups (scientists, food companies,
consumers, competitive athletes, and health-care providers) are compared regarding
their practices and opinions of personalized nutrition. Based on behavioral change
strategies, the next chapter is focusing on whether and how personalized nutrition
will contribute to a healthier dietary pattern.
Two chapters specifically target the consumer perspective. The first reviews the
marketing potential of personalized nutrition; in the second, consumer attitudes are
discussed. The section is completed with a contribution on ethical issues related to
personalized advice on diet and health and the customization of food products.
In the final section, Future Perspectives on Personalized Nutrition, the challenges
of nutrigenomics research and its applications are presented in two chapters. One
chapter focuses on the need to address the humanitarian issues related to developing
countries and issues a call for international efforts to develop best practices, fostering
international collaborations and sharing data sets. The final chapter provides an
outlook of personalized nutrition in the context of ongoing innovations in food
technology, nutrigenomics, and food delivery systems.
We sincerely hope that you enjoy reading this guide to a scientific underpinning
of personalized nutrition.
Frans J. Kok
Laura Bouwman
Frank Desiere
9281_C000.fm Page xi Thursday, July 26, 2007 7:30 PM
The Editors
Frans J. Kok PhD, is full professor of nutrition and health and head of the Division
of Human Nutrition at Wageningen University, The Netherlands. Moreover, as dean of
Wageningen Graduate Schools, he is responsible for securing the quality of academic
research and training at the university.
Dr. Kok was trained in human nutrition at Wageningen University and in epi-
demiology at Harvard University, Cambridge, Massachusetts. His scientific expertise
covers topics such as energy balance and body composition, diet in disease preven-
tion, and sensory science. He has authored more than 250 original scientific publi-
cations and is a member of the Dutch Health Council and the Academic Board of
Wageningen University. As a faculty member, he has lectured in many courses on
nutrition and health, including the European Nutrition Leadership Program, as well
as in courses in the U.S., Africa, and Asia.
Laura Bouwman is a PhD researcher in the department of communication science at

Wageningen University. Her research is focused on whether and how innovations in
personalized nutrition can be used for the development of more effective nutrition inter-
ventions. Two topics are central: 1) how stakeholders in science, healthcare, health
education, government, and industry perceive their roles and responsibilities in the inno-
vation process, and 2) the representation of personal factors in food choice, with emphasis
on the representation of genetic background. Laura is involved in the development of
communication activities for diverse stakeholder-groups regarding personalized nutrition
in the European Nutrigenomics Organisation. Before her current PhD research, Laura
worked for six years at a communication bureau for the oils and fats industry.
Frank Desiere, PhD, MBA, started his scientific career in the earlier days of biotech-
nology doing research in fermentation technology, microbiology, and biological conver-
sion of xenobiotics, and later joined the Nestl Research Center in Lausanne, Switzerland,
to work on the ecology and evolution of bacteriophages in milk fermentation. Later, he
initiated a bioinformatics effort at Nestl and worked on microbial genomics, transcrip-
tomics, and proteomics. Dr. Desiere also initiated Peptideatlas (www.peptideatlas.org)
for the integration of proteomics data with human and other genomes. He was then
responsible for a project involving personalized nutrition at Nestl that initiated consumer
and patient applications of personalized nutrition concepts in the marketplace. As a result
of this initiative, his interests evolved into the commercial application of scientific con-
cepts in science and nutrition, management of innovation and new business ideas. After
having completed an MBA at IMD Lausanne, Dr. Desiere now works as business
development manager at Roche Diagnostics in Basel, Switzerland. His scientific activities
have resulted in more than 30 peer-reviewed publications and numerous presentations
at international scientific conferences.
9281_C000.fm Page xii Thursday, July 26, 2007 7:30 PM
9281_C000.fm Page xiii Thursday, July 26, 2007 7:30 PM
Contributors
Lydia A. Afman Hannelore Daniel
Nutrition, Metabolism, and Genomics Molecular Nutrition Unit
Division of Human Nutrition Center of Life and Food Sciences
Wageningen University Technical University of Munich
Wageningen Freising-Weihenstephan
The Netherlands Germany
Elizabeth Baily Aldona Dembinska-Kiec

International Food Information Council Department of Clinical Biochemistry
Washington, D.C. Jagiellonian University
Medical College
Laura I. Bouwman Krakow
Wageningen University Poland
Sub-Department of Communication
Science, Communication
Management Group Lynnette R. Ferguson
Wageningen Discipline of Nutrition
The Netherlands University of Auckland
Grafton, Auckland
New Zealand
Lorraine Brennan
UCD School of Biomolecular
and Biomedical Science Edith J.M. Feskens
UCD Conway Institute Division of Human Nutrition
University College Wageningen University
Dublin Wageningen
Ireland The Netherlands
Josh Conway
Cogent Research, LLC J. Bruce German
Cambridge, Massachusetts University of CaliforniaDavis
Davis, California
Dolores Corella
Department of Preventive Medicine Michael J. Gibney
University of Valencia UCD Agriculture and Food Science
Genetic and Molecular Epidemiology Centre
Unit University College
Valencia Dublin
Spain Ireland
9281_C000.fm Page xiv Thursday, July 26, 2007 7:30 PM
Gerrit J. Hiddink Simon C. Langley-Evans

Wageningen University University of Nottingham
Sub-Department of Communication School of Biosciences
Science, Communication Sutton Bonington, Loughborough
Management Group United Kingdom
Wageningen
The Netherlands John C. Mathers
Human Nutrition Research Centre,
School of Clinical Medical Sciences
Hans-Georg Joost University of Newcastle
German Institute of Human Nutrition Newcastle-on-Tyne
Potsdam-Rehbruecke United Kingdom
Nuthetal
Germany Michael Mller
Nutrition, Metabolism, and Genomics
Division of Human Nutrition
Ellen Kampman
Wageningen University
Division of Human Nutrition
Wageningen
Wageningen University
The Netherlands
Wageningen
The Netherlands Jose M. Ordovas
Nutrition and Genomics Laboratory
Wendy Reinhardt Kapsak Jean Meyer USDA Human Nutrition
International Food Information Council Research Center on Aging
Washington, D.C. Tufts University
Boston, Massachusetts
Jim Kaput Martin Philpott

Laboratory of Nutrigenomic Medicine Discipline of Nutrition
Department of Surgery University of Auckland
University of Illinois Chicago Grafton, Auckland
Chicago, Illinois New Zealand
Danielle R. Reed
Maria A. Koelen Monell Chemical Senses Center
Wageningen University Philadelphia, Pennsylvania
Sub-Department of Communication
Science, Communication Manuela Rist
and Innovation Group Molecular Nutrition Unit
Wageningen Technical University of Munich
The Netherlands Freising-Weihenstephan
Germany
Michiel Korthals Amber Ronteltap
Applied Philosophy Marketing and Consumer Behavior
Social Sciences Group
Wageningen University Wageningen University
Wageningen Wageningen
The Netherlands The Netherlands
9281_C000.fm Page xv Thursday, July 26, 2007 7:30 PM
David B. Schmidt Marianne Walsh

International Food Information Council UCD Agriculture and Food Science
Washington, D.C. Centre
University College
Toshiko Tanaka Dublin
Nutrition and Genomics Laboratory Ireland
Jean Mayer USDA Human Nutrition
Research Center on Aging
Tufts University Heribert J. Watzke
Boston, Massachusetts Nestl Research Centre
Lausanne
Hans C.M. van Trijp Switzerland
Marketing and Consumer Behavior
Group
Wageningen University Christy White
Wageningen Cogent Research, LLC
The Netherlands Cambridge, Massachusetts
Pieter vant Veer

Division of Human Nutrition J. T. Winkler
Wageningen University London Metropolitan University
Wageningen London
The Netherlands United Kingdom
9281_C000.fm Page xvi Thursday, July 26, 2007 7:30 PM
9281_S001.fm Page 1 Wednesday, July 18, 2007 4:31 PM
Section I
Scientific Principles
of Personalized Nutrition
9281_C001.fm Page 3 Thursday, July 26, 2007 7:04 PM
1 Nutrigenomics
and Transcriptomics:
Applications of Microarray
Analyses in Nutrition
Research
Lydia A. Afman and Michael Mller
CONTENTS
1.1 Introduction ......................................................................................................3

1.2 Nutrients and Transcription .............................................................................5
1.2.1 Nutrient-Activated Transcription Factors ............................................5
1.2.2 Role of the Transcription Factor PPAR ............................................6
1.3 Transcriptomics ................................................................................................6
1.3.1 Technology ...........................................................................................6
1.3.2 Transcriptomics and Study Design......................................................6
1.4 Application of Trancriptomics in Human Studies ..........................................7
1.4.1 Requirements for Human Transcriptomics Studies.............................7
1.4.2 Application of Adipose Tissue Transcriptomics..................................8
1.4.3 Application of Muscle Transcriptomics ..............................................8
1.4.4 Application of Blood Cell Transcriptomics ........................................9
1.4.5 Biomarker Profiling............................................................................10
1.5 How Can Transcriptomics Be Used in Personalized Nutrition?...................10
1.6 Future Perspectives for Personalized Nutrition.............................................11
Key Readings...........................................................................................................11
References................................................................................................................11
1.1 INTRODUCTION
This chapter describes how nutrients can influence gene expression and describes
how this knowledge can be applied in nutritional research by utilization of transcrip-
tional profiling. Two research approaches are discussed: a mechanistic approach and
3
4 Personalized Nutrition: Principles and Applications
a biomarker profiling approach. The mechanistic approach is a commonly applied

method in transcriptomics research and has been extremely useful in the identifica-
tion of the molecular pathways and signaling routes involving nutrients. The biom-
arker profiling approach is under development but will have significant impact,
especially in human research. For both methods, nutrition-specific pitfalls are dis-
cussed, in addition to an outline of potential applications of gene expression profiling
in humans and animal models, with examples of nutrition-related transcriptomics
studies published so far. Furthermore, transcriptional profiling is discussed in relation
to its current and future applications in personalized nutrition.
A major drawback in the investigation of molecular mechanism of nutrients in
human studies is the accessibility of organ and tissues material in healthy volunteers,
hampering the investigation of organ-specific functions of nutrients. In case tissue
material is available, factors such as genetic diversity present among even quite isolated
populations, in addition to environmental and lifestyle factors, such as physical activity,
dietary behavior, and stress, provoke such great variability in nutrient-induced
responses that large, expensive study populations are required. Therefore, evidence-
based nutritional research generally exploits model systems such as animal models.
One of the major advantages of the use of animal models in dietary research is that
dietary intake and environmental factors can be precisely controlled. In order to
minimize the variability caused by genetic background, genetically identical inbred
strains can be used. Moreover, transgenic and knockout models are only available in
animal models and are of great benefit in identification of functional effects of nutrients.
For example, mice with an overexpression or knockout of genes encoding nutrient-
activated transcription factors can be used for functional genomics studies to charac-
terize nutrient-specific induced changes in gene expression and the specific involvement
of the particular transcription factor (Kersten, Seydoux et al. 1999). Another advantage
of animal models is the possibility of studying molecular mechanisms of nutrients in
specific organs or tissues. The enormous amount of data derived from transcriptomics
studies can be used to identify nutrient-specific changes in pathways, networks, and
signaling routes. Ultimately, this type of study will facilitate the unraveling of the
underlying molecular mechanism of a wide variety of nutrients, with the goal of
improving dietary advice via evidence-based nutrition. Another advantage of such a
large data set is the identification of nutrient-induced changes in organ-specific poten-
tially secreted proteins. These proteins are excreted by the organ in body fluids such
as plasma and urine, making them promising biomarker candidates that can easily be
verified in human intervention studies and consequently be used as markers for organ-
specific function and health.
Another application of transcriptomics, which is not feasible yet, is the identi-
fication of predisease biomarkers of nutritional-related chronic diseases such as
cardiovascular diseases, type 2 diabetes, and cancer. In order to identify those
markers, monitoring of transcriptional changes during the development of diet-
induced chronic diseases is required, enabling the identification of early biomarker
profiles (Afman and Mller 2006). As the development of chronic diseases require
long-term dietary exposure, this approach requires time- and cost-consuming long-
term nutritional intervention studies that will not result in identification of early
biomarkers in the near future.
Nutrigenomics and Transcriptomics 5
1.2 NUTRIENTS AND TRANSCRIPTION

1.2.1 NUTRIENT-ACTIVATED TRANSCRIPTION FACTORS
It is widely recognized that nutrients (macro- and micronutrients) can have profound
effects on gene regulation and gene transcription. These important modulatory effects
of dietary components can be monitored on a genomewide scale by using state-of-
the-art oligonucleotide microarrays. The exact mechanisms of how nutrients influ-
ence gene transcription is not completely known yet; however, knowledge regarding
this is emerging (Mller and Kersten 2003; Desvergne, Michalik et al. 2006).
Nutrients and their metabolites, such as fatty acids and certain eicosanoids, choles-
terol metabolites (bile salts, oxysterols), or vitamin A metabolites (retinoic acid),
serve as ligands for certain ligand-activated transcription factors such as PPARs,
FXRs, LXRs, or RARs and RXRs, all members of the large superfamily of nuclear
receptors (Zhang and Mangelsdorf 2002; Mller and Kersten 2003). During ligand
binding, nuclear receptors undergo a conformational change that results in the
coordinated dissociation of corepressors and the recruitment of coactivator proteins
to enable transcriptional activation. These nutrient-sensing receptors (PPARs, FXRs,
LXRS) form heterodimers with RXR and bind to specific nucleotide sequences
(response elements) in the promoter regions of a large number of genes. The response
elements are specific for the nuclear receptor, and each nutrient will induce a
characteristic set of genes, also depending on the recruited coactivators (Figure 1.1).
Through these transcription factors, nutrients are able to influence a wide variety of
cellular functions.
Systems-biology
databases and bioinformatics
Nutrient
Metabolomics and
Functions functional
Transporter genomics
Molecular-
biology tools Transcription
Tet-on + factors Molecular
Proteins Proteomics
biomarkers
Transgenics mechanisms
RNAi and targets
tdnAd
mRNA Transcriptomics Disease

prevention
DNA
Nucleus
FIGURE 1.1 The smart combination of molecular nutrition and nutrigenomics.

1.2.2 ROLE OF THE TRANSCRIPTION FACTOR PPAR

Unsaturated fatty acids, for example, can activate gene transcription by activating
the fatty acid sensor PPAR, which is expressed in liver, heart, muscle, and small
intestine and other organs. In the liver PPAR controls the transcription of up to
3000 genes involved in numerous metabolic and regulatory processes, including
fatty acid oxidation, ketogenesis, gluconeogenesis, amino acid metabolism, cellular
proliferation, and the acute-phase response (Kersten, Seydoux et al. 1999). Hepatic
PPAR has an important role during fasting, when increased levels of free fatty
acids are released from adipose tissue and enter the liver. These fatty acids bind to
PPAR and control as a consequence several important processes, including their
degradation, TG synthesis, lipid transport, or, on the other hand, gluconeogenesis.
In particular, the last process that in part uses glycerol also released from the adipose
tissue (Patsouris, Mandard et al. 2004) as a precursor for glucose synthesis is of
crucial importance for glucose homeostasis (Kersten, Seydoux et al. 1999).
The role of hepatic PPAR for dietary fats and in obesity is less clear but most
likely relevant to our understanding of the obesity-linked pathophysiology of type
2 diabetes (Patsouris, Mller et al. 2004). Visceral obesity is linked to increased free
fatty acid levels (Moller and Kaufman 2005), and these molecules may be recognized
by the liver as being in need of glucose signals, resulting in increased gluconeo-
genesis, which may contribute to a more pathophysiological phenotype, particularly
under conditions of hepatic insulin resistance.
1.3 TRANSCRIPTOMICS
1.3.1 TECHNOLOGY
Transcriptomics is both a sensitive and well-validated technique, which employs
high-density oligonucleotide microarray analysis to study the number of mRNA
copies per gene for virtually all actively transcribed genes. This technique enables
determination of expression levels of genes of the whole genome, meaning hundreds
to thousands of genes at the same time and within one study. As nutrients influence
transcription of genes, this technique can be applied to measure diet-induced changes
in every tissue of interest. A challenge in nutritional research is the relatively small
effect of dietary interventions on physiological parameters. Similarly, the effects of
nutrition on gene expression patterns are also rather small and not easy to detect.
Consequently, the development of a highly sensitive and validated microarray anal-
ysis platform is required along with a robust research hypothesis that is adequately
addressed by the experimental design to avoid so-called fishing expeditions.
1.3.2 TRANSCRIPTOMICS AND STUDY DESIGN

In order to draw causal relations between nutrients and transcriptional changes,
dietary adjustments should be limited when two or more diets are compared within
the experiment. In addition, circadian rhythms are prevalent in several organs and
tissues and potentially influence biological pathways; therefore, collection of
tissue material is of crucial importance in increasing the signal-to-noise ratio
(Ptitsyn, Zvonic et al. 2006; Zvonic, Ptitsyn et al. 2006). Another undesirable factor
is pooling of samples, but increasing the number of biological replicates will enhance
accuracy by decreasing the false-positive rate and result in reliable data (Kendziorski,
Irizarry et al. 2005). Once a transcriptomics study is performed, RNA quality and
quantity should be verified and subsequent labeling and hybridization of RNA should
preferably be conducted by the same technician to keep variation between arrays
low. Preprocessing of data requires specific attention, as the small gene expression
changes in dietary studies require a highly reliable algorithm (Irizarry, Warren et al.
2005). The most challenging part of the process occurs after the array data are
obtained, when the data have to be analyzed in terms of their biological meaning.
Several commercial and noncommercial tools have been developed that assist with
the extraction of significantly changed genes and the visualization of changed path-
ways or related networks (Curtis, Oresic et al. 2005).
1.4 APPLICATION OF TRANCRIPTOMICS

IN HUMAN STUDIES
1.4.1 REQUIREMENTS FOR HUMAN TRANSCRIPTOMICS STUDIES
The execution of a transcriptomics study requires the accurate collection of organ
or tissue material for the extraction of RNA, which is a bottleneck in human studies.
Biopsies can be obtained from both unhealthy as well as surrounding healthy tissue
of patients who have to undergo surgical procedures. The disadvantage of this
approach is that the pathological state of diseased tissue is often reflected in the
surrounding nondiseased tissue as well. A major drawback in studies with healthy
volunteers, which is quite common in nutrition research, is the very limited acces-
sibility of tissues and organ material. The only way to obtain material from healthy
volunteers is via biopsies from tissues such as muscle and subcutaneous adipose
tissue. Although other metabolically relevant organs such as small intestine and liver
are not accessible with this approach, transcriptional profiling of muscle and adipose
tissue biopsies are extremely useful as these organs fulfill important physiological
functions, have a central role in energy metabolism, and play crucial roles in certain
diet-related diseases. A disadvantage of the collection of muscle and adipose tissue
biopsies is the invasiveness of this method and the low yield of tissue and, conse-
quently, low RNA quantity. A less invasive alternative is the use of blood cells for
transcriptional profiling. These cells travel through the whole body and react on
internal threats to body homeostasis, and may therefore serve as functional markers
for whole-body health. Moreover, disease-specific gene expression patterns have
been identified in blood cells (Martin, Graner et al. 2001; Valk, Verhaak et al. 2004),
and although interindividual variety in gene expression is high, the much smaller
within-person variation makes them suitable for use in dietary intervention studies
(Whitney, Diehn et al. 2003; Radich, Mao et al. 2004; Cobb, Mindrinos et al. 2005).
An additional difficulty in human studies is the diverse genetic background, requiring
a powerful study design to obtain relevant and consistent data. One of the features
of such a design is sampling within a person before and after intervention, with
preferably a cross-over design. The invasiveness of tissue biopsy collection makes
this method less suitable for application in short or time-course intervention studies.
Another characteristic of an appropriate nutrigenomics design is a high number of
replicates; the accompanying high costs, however, are often the limiting factor in
this respect. A third aspect of a sound study design is the magnification of the signal-
to-noise ratio, leading to an increased change on selecting the correct but rather
small diet-induced effects. Such an effect could be achieved by magnifying the
response through a nutritional challenge of the body.
The number of human studies that describe nutrition-related transcriptome pro-
filing is limited. Most of the papers describe differential gene expression in muscle
and adipose tissue biopsies in nutrition-related chronic diseases.
1.4.2 APPLICATION OF ADIPOSE TISSUE TRANSCRIPTOMICS

In a transcriptomics study on adipose tissue, gene expression profiles were analyzed
of subcutaneous white adipose tissue from obese individuals before and after a very
low caloric diet intervention of 28 d (Clement, Viguerie et al. 2004). Microarray
analyses were performed before and after 2 and 28 d of caloric restriction. Two days
of diet intervention did not result in any considerable changes; 28 d of caloric
restriction, however, induced significant gene expression changes that were mostly
related to inflammatory processes. Genes related to this process could be classified
into upregulation of anti-inflammatory genes and downregulation of proinflamma-
tory genes. Other changed genes were involved in cellcell and cellmatrix adhesion
or in cell proliferation, growth, and differentiation; the latter two were mostly
downregulated on caloric restriction. This study also managed to locate the genes
to specific adipose tissue fractions; i.e., most genes were expressed in the stromavas-
cular fraction of adipose tissue. Microarray analysis of adipose tissue may explain
why caloric restriction delays onset of conditions frequently associated with an
increase in basal inflammation, such as type 2 diabetes and cardiovascular disease,
or with conditions in which cell proliferation is of pivotal importance, such as cancer.
Another often-used approach is to search for gene expression changes in animal
models followed by subsequent confirmation of changes in genes of interest in
human volunteers. As an example, microarray analyses of white adipose tissue of
diet-induced obese mice revealed changes in several markers of inflammation that
were also differently expressed in adipose tissue of obese and lean subjects. (Weisberg,
McCann et al. 2003; Xu, Barnes et al. 2003). This transcriptomics study revealed
that adipose tissue is not only of importance as a storage source of energy but that
it might also play an essential role in the development of inflammatory conditions
associated with obesity, such as metabolic syndrome and type 2 diabetes. Further
studies are required to elucidate how dietary interventions, besides caloric restriction,
can reduce, or preferably prevent, the inflammation in adipose tissue.
1.4.3 APPLICATION OF MUSCLE TRANSCRIPTOMICS

Microarray profiling studies with human muscle tissue are often associated with
insulin resistance and type 2 diabetes. A study on gene expression profiles in muscle
tissue demonstrated a reduced expression of genes involved in oxidative phospho-
rylation and mitochondrial function in diabetic and insulin-resistant subjects
(Mootha, Lindgren et al. 2003; Patti, Butte et al. 2003). Moreover, they showed that
these changes have most likely been mediated by the peroxisome proliferator-
activated receptor coactivator-1 (PGC1), which was found to be downregulated in
these subjects as well. Similarly, microarray analyses performed on muscle biopsies
before and after 3 d of high-fat feeding showed a downregulation in expression of
genes involved in oxidative phosphorylation or mitochondrial function. This high-
fat feeding intervention also resulted in the downregulation of PGC1 and PGC1,
similar to the findings in diabetic muscle tissue (Sparks, Xie et al. 2005). To confirm
this, a comparable intervention study was performed with C57Bl/6 mice fed a high-
fat diet for 3 weeks. The gene expression changes due to high-fat feeding in the
mouse study were of a greater magnitude than those observed in the human study,
but were all related to a decrease in the expression of genes involved in oxidative
phosphorylation and mitochondrial biosynthesis and its master regulator, PGC1. The
authors suggest that high-fat diet may explain the reduction in oxidative phospho-
rylation previously observed in muscle of type 2 diabetes patients and prediabetic
individuals (Kelley, He et al. 2002). In conclusion, this is a good example of a
nutrigenomics study demonstrating that diet-induced changes in inbred mice that
are genetically identical animals are of a greater magnitude than comparable inter-
ventions performed in humans with different genotypes.
With the use of transcriptomics techniques, the aforementioned studies were
able to identify pathways involved in different stages of the development of type 2
diabetes. In addition, these studies were able to demonstrate that these pathways
were changed in the same direction after intervention with a high-fat diet, implying
that this pathway can be influenced by diet, which points toward a mechanism in
diet-induced type 2 diabetes. The question still remains whether transcriptional
analyses of this pathway can be used as a diagnostic tool for the early predisease
state and whether and which dietary adjustments can lead to upregulation in oxidative
phosphorylation, mitochondrial biosynthesis and PGC1, and ultimately result in the
prevention of type 2 diabetes.
1.4.4 APPLICATION OF BLOOD CELL TRANSCRIPTOMICS

Although transcriptomics studies with muscle and adipose tissue have proved to be
of great value in identification of nutrition-related pathways, a major drawback of
these tissues is the difficult accessibility in healthy human volunteers. In the last
few years, several studies have shown disease-characteristic gene expression pat-
terns, or signatures, in human blood cells (Martin, Graner et al. 2001; Valk, Verhaak
et al. 2004). In addition, injection of bacterial endotoxin in healthy human volunteers
displayed time-dependent changes in blood cell gene expression (Calvano, Xiao et al.
2005). Although proved to be useful as potential diagnostic tool under these condi-
tions, the question still remains whether blood cells can be used for the identification
of nutrition-specific gene expression patterns. So far, nutrigenomics-related studies
have not yet employed circulating blood cells. The advantage of blood cell profiling
in comparison to tissue biopsies profiling in nutritional studies is the less invasive
way of collecting blood, enabling application in time-course experiments or chal-
lenge studies.
1.4.5 BIOMARKER PROFILING

The ultimate opportunity for blood cell profiling in nutrigenomics research is the
identification and subsequent application of gene expression patterns as early biom-
arkers for a predisease state. Commonly used biomarkers are often indicative of the
onset or presence of diseases, and once the disease is diagnosed, drug treatment is
unavoidable. The identification of early biomarkers that are indicative for a phenotype
of an out-of-balance system would be very valuable in the development of nutritional
strategies to regain normal healthy homeostasis. The latter is of specific relevance in
the assessment of nutritional effects on disease outcome, as this requires long-term
and expensive human intervention studies. The ultimate goal of the characterization
of early biomarkers is the identification of subjects in the predisease state with the
long-run prevention of disease outcome, achieved by adequate dietary interventions.
In summary, to our knowledge only a few nutrition-related human intervention
studies have been carried out, pointing to the infant state of nutrigenomics research,
in particular in humans. The aforementioned microarray studies performed in muscle
and adipose tissue, especially the ones that applied the smart combinations between
animal and human studies, provide the first evidence for the feasibility of nutrige-
nomics-related research in human intervention studies.
1.5 HOW CAN TRANSCRIPTOMICS BE USED

IN PERSONALIZED NUTRITION?
Although the term personalized nutrition implies tailor-made nutritional advice,
currently this is not a reality and may not even be possible in the near future. More
promising results, however, can be expected from exploring the influences of specific
nutrients on cellular function and their ultimate effects on health and disease. With
such information, better knowledge-based dietary advice can be given to specific
groups in the population, including elderly people, pregnant women, athletes, chil-
dren, etc. In addition, this knowledge can lead to the development of subpopulation-
specific functional foods. A following step is the identification of biomarkers that
are specific for either the healthy or predisease state. One of the characteristics of
the predisease state is that the body homeostasis is out of balance. A way to identify
an out-of-balance system is by studying the flexibility or resilience of the body, i.e.,
the capability of the body to cope with challenges. For example, in subjects in the
predisease state the impact of a challenge may be higher and the recovery time
longer than in healthy individuals. The use of transcriptional profiling during such
a challenge will assess the flexibility of the body by determination of both the
magnitude of the gene expression changes and the time required for regaining the
basal gene expression profile. Similarly, transcriptional profiles from diseased indi-
viduals may differ from predisease subjects through a more pronounced or an even
completely different profile, in addition to a longer recovery time than predisease
people. The response identified with transcriptomics can serve as a phenotype specific
for an out-of-balance system representing a predisease state. For a more compre-
hensive phenotyping, metabolomics and proteomic parameters should be combined
with transcriptomics data, preferable before and after a nutritional challenge.
1.6 FUTURE PERSPECTIVES FOR PERSONALIZED

NUTRITION
The identification of early biomarkers of the predisease state requires time-consum-
ing and expensive long-term nutritional intervention studies that will not result in
identification of early biomarkers in the near future. Therefore, results in the near
future concerning biomarker profiling will mainly be expected via animal studies,
with subsequent translation of these animal biomarkers to the human situation.
Promising results are, however, also expected from the previously described chal-
lenges studies, in which the flexibility of the human body is tested by measuring
the adaptation and recovery of the body after intervening with dietary stressors. The
latter has still to be proved useful and demonstrates that the biomarker profiling
approach is currently under development. Nevertheless, the identification of early
biomarkers will have high future impact, especially in human research.
Although early biomarkers will be essential for the identification of subjects in
a predisease state, additional knowledge is required about the underlying molecular
mechanisms of nutrients for an adequate nutritional intervention with the ultimate
goal of regaining normal homeostasis. Undoubtedly, the combination of evidence-
based nutrition and early biomarker-based diagnostics will be mandatory to enable
accurate, tailor-made nutritional advice in the future.
KEY READINGS
Afman, L. and Mller, M. (2006). Nutrigenomics: from molecular nutrition to prevention of
disease. J Am Diet Assoc 106(4): 56976.
Desvergne, B., Michalik, L. et al. (2006). Transcriptional regulation of metabolism. Physiol
Rev 86(2): 465514.
Mller, M. and Kersten, S. (2003). Nutrigenomics: goals and strategies. Nat Rev Genet 4(4):
31522.
Mutch, D.M., Wahli, W., and Williamson, G. Nutrigenomics and nutrigenetics: the emerging
faces of nutrition. FASEB J 19(12): 160216, October 2005.
REFERENCES
Afman, L. and Mller, M. (2006). Nutrigenomics: from molecular nutrition to prevention of
disease. J Am Diet Assoc 106(4): 56976.
Calvano, S.E., Xiao, W. et al. (2005). A network-based analysis of systemic inflammation in
humans. Nature 437(7061): 10327.
Clement, K., Viguerie, N. et al. (2004). Weight loss regulates inflammation-related genes in
white adipose tissue of obese subjects. FASEB J 18(14): 165769.
Cobb, J.P., Mindrinos, M.N. et al. (2005). Application of genome-wide expression analysis
to human health and disease. Proc Natl Acad Sci USA 102(13): 48016.
Curtis, R.K., Oresic, M. et al. (2005). Pathways to the analysis of microarray data. Trends
Biotechnol 23(8): 42935.
Desvergne, B., Michalik, L. et al. (2006). Transcriptional regulation of metabolism. Physiol
Rev 86(2): 465514.
Irizarry, R.A., Warren, D. et al. (2005). Multiple-laboratory comparison of microarray plat-

forms. Nat Methods 2(5): 34550.
Kelley, D.E., He, J. et al. (2002). Dysfunction of mitochondria in human skeletal muscle in
type 2 diabetes. Diabetes 51(10): 294450.
Kendziorski, C., Irizarry, R.A. et al. (2005). On the utility of pooling biological samples in
microarray experiments. Proc Natl Acad Sci USA 102(12): 42527.
Kersten, S., Seydoux, J. et al. (1999). Peroxisome proliferator-activated receptor alpha medi-
ates the adaptive response to fasting. J Clin Invest 103(11): 148998.
Martin, K.J., Graner, E. et al. (2001). High-sensitivity array analysis of gene expression for
the early detection of disseminated breast tumor cells in peripheral blood. Proc Natl
Acad Sci USA 98(5): 264651.
Moller, D.E. and Kaufman, K.D. (2005). Metabolic syndrome: a clinical and molecular
perspective. Annu Rev Med 56: 4562.
Mootha, V.K., Lindgren, C.M. et al. (2003). PGC-1alpha-responsive genes involved in oxi-
dative phosphorylation are coordinately downregulated in human diabetes. Nat Genet
34(3): 26773.
Mller, M. and Kersten, S. (2003). Nutrigenomics: goals and strategies. Nat Rev Genet 4(4):
31522.
Patsouris, D., Mandard, S. et al. (2004). PPARalpha governs glycerol metabolism. J Clin
Invest 114(1): 94103.
Patsouris, D., Mller, S. et al. (2004). Peroxisome proliferator activated receptor ligands for
the treatment of insulin resistance. Curr Opin Invest Drugs 5(10): 104550.
Patti, M.E., Butte, A.J. et al. (2003). Coordinated reduction of genes of oxidative metabolism
in humans with insulin resistance and diabetes: potential role of PGC1 and NRF1.
Proc Natl Acad Sci USA 100(14): 846671.
Ptitsyn, A.A., Zvonic, S. et al. (2006). Circadian clocks are resounding in peripheral tissues.
PLoS Comput Biol 2(3): e16.
Radich, J.P., Mao, M. et al. (2004). Individual-specific variation of gene expression in periph-
eral blood leukocytes. Genomics 83(6): 9808.
Sparks, L.M., Xie, H. et al. (2005). A high-fat diet coordinately downregulates genes required
for mitochondrial oxidative phosphorylation in skeletal muscle. Diabetes 54(7):
192633.
Valk, P.J., Verhaak, R.G. et al. (2004). Prognostically useful gene-expression profiles in acute
myeloid leukemia. N Engl J Med 350(16): 161728.
Weisberg, S.P., McCann, D. et al. (2003). Obesity is associated with macrophage accumulation
in adipose tissue. J Clin Invest 112(12): 1796808.
Whitney, A.R., Diehn, M. et al. (2003). Individuality and variation in gene expression patterns
in human blood. Proc Natl Acad Sci USA 100(4): 1896901.
Xu, H., Barnes, G.T. et al. (2003). Chronic inflammation in fat plays a crucial role in the
development of obesity-related insulin resistance. J Clin Invest 112(12): 182130.
Zhang, Y. and Mangelsdorf, D.J. (2002). LuXuRies of lipid homeostasis: the unity of nuclear
hormone receptors, transcription regulation, and cholesterol sensing. Mol Interv 2(2):
7887.
Zvonic, S., Ptitsyn, A.A. et al. (2006). Characterization of peripheral circadian clocks in
adipose tissues. Diabetes 55(4): 96270.
9281_C002.fm Page 13 Wednesday, July 18, 2007 1:47 PM
2 Exploring the Proteome

for Markers of Health
Manuela Rist and Hannelore Daniel
CONTENTS
2.1 Introduction ....................................................................................................13

2.2 Techniques Employed in Proteomics.............................................................14
2.3 Biofluids as a Source of Marker Discovery...................................................16
2.4 Examples of Biomarker Discovery Approaches
in Body Fluids by Proteome Analysis............................................................18
2.5 Future Perspectives ........................................................................................20
References................................................................................................................21
2.1 INTRODUCTION
Proteome analysis of a particular cell or organ, or of body fluids such as plasma,
urine, and others, makes it possible to identify biomarkers that respond to alterations
in diet or any other treatment, and that may have a predictive value for the assessment
of a humans health status. To date, proteomics has not been applied on a large scale
in human studies related to nutrition, yet it has obvious advantages over transcrip-
tome profiling techniques in that it directly assesses the molecular entities that carry
out the biological functions. This chapter summarizes the different approaches in
proteomics research and provides some examples that demonstrate the state of the
art of biomarker discovery based on proteome analysis and their implications for
the development of personalized nutrition approaches.
Human metabolism can be described as the orchestrated interplay of a huge
number of proteins that perform different functions as receptors, enzymes, signaling
molecules, and transporters or structural components, and that are expressed at the
level of the cell, the organ, or whole organism. Proteins are the transmitters of
metabolic functions, and therefore changes in the level of an individual protein or
sets of proteins may indicate an altered metabolic condition. Protein synthesis and
protein degradation, and thereby metabolic functions, can be affected by the overall
composition of the diet, the levels of specific nutrients and nonnutrient components
of foods and, indeed, many other aspects of lifestyle, in the most complex ways.
Analysis of the proteome as the whole protein complement of the genome expressed
in a particular cell, an organ, or body fluid also called proteomics allows
13
changes in protein expression patterns, and the identity of the proteins themselves
to be determined simultaneously. Moreover, for individual proteins, posttranslational
modifications that may be crucial for protein function, or even amino acid sub-
stitutions (polymorphisms), can be detected. The potential value of proteomics for
assessing markers of health has been recognized for some years [1], but only very
few studies so far have employed proteome analysis as a tool for research in human
nutrition and health [2].
2.2 TECHNIQUES EMPLOYED IN PROTEOMICS

Proteome analysis is beginning to emerge as a second high-throughput tool for
nutrition research. The revival of the two-dimensional gel electrophoresis (2D-
PAGE) with improved resolution, the advanced instrumentation, elegant software
tools, databases, and the enormous advancements in mass spectrometry have made
proteomics applications a practical alternative screening method that complements
transcript profiling approaches. There are three different principal strategies for
proteome analysis: (1) classical 2D-PAGE that separates proteins in a gel prior to
analysis of individual proteins by peptide mass fingerprinting using mass spectrom-
etry, (2) the so-called shotgun approaches, in which all proteins contained in a sample
are first digested by proteases, followed by chromatographic separation of the
obtained peptide fragments, which then are analyzed by mass spectrometry, and
(3) a solid-phase-based enrichment of proteins depending on their different physico-
chemical characteristics mainly done on chips followed by mass analysis that
usually does not identify the proteins but provides mass patterns (Figure 2.1). The
latter technique is predominantly used in clinical diagnostics.
The 2D-PAGE technique separates proteins according to charge (isoelectric
point: pI) by isoelectric focusing (IEF) in the first dimension and according to size
(molecular mass) in the second dimension. Therefore, it has a unique capacity to
resolve complex mixtures of proteins, permitting the simultaneous analysis of hun-
dreds or even thousands of gene products [3]. However, not all proteins are resolved
and separated equally well by 2D-PAGE. Very alkaline, hydrophobic, and integral
membrane proteins as well as high-molecular-weight proteins are still a problem.
In some cases, a prefractionation according to cellular compartments (membranes/
microsomes, cytosol, mitochondria) or according to protein solubility by classical
means may be necessary [4]. In addition, proteins of low cellular abundance, which
may be particularly important in view of their cellular functions, for example, in
signaling pathways, are still very difficult to be resolved in the presence of large
quantities of housekeeping proteins [5]. However, new techniques are constantly
being developed that employ, for example, tagging techniques [6] or the enrichment
of the minor proteins prior to separation in two-dimensional gels. The most common
procedure for the identification of a protein spot in a gel is currently the peptide
mapping or fingerprint analysis method, but other techniques and approaches can
also be applied. For peptide mapping, protein containing spots are excised from the
gel before the gel is altered chemically to make the protein accessible for hydrolysis
by a protease such as trypsin [7]. Based on this site-specific enzymatic hydrolysis,
a distinct and characteristic pattern of peptide fragments of a given protein is
Exploring the Proteome for Markers of Health 15
Biouid sample
2D-gel Tryptic
separation digestion
Peptide mixture
Mixture of proteins
LC-separation
Column
Excised protein
Loading on chips
Identication of the protein

with
dierent properties Peptide LC-prole
Tryptic digestion
Identication of the protein
Mixture of peptides
Time of ight Time of ight

or electrospray or electrospray
mass spectrometry mass spectrometry
Intensity
Mass spectrum Mass spectrum

Low Mass/Charge High
FIGURE 2.1 The different approaches to proteome analysis.
generated. The mixture of peptides isolated from the gel spot after digestion with
the protease is usually submitted to matrix-assisted laser desorption/ionization time-
of-flight mass spectrometry (MALDI-TOF-MS) analysis to determine the corre-
sponding peptide masses that are characteristic for a given protein and serve as the
peptide mass fingerprint. The obtained mass spectrum is evaluated with computer
programs that apply various algorithms for interpretation of the peptide pattern and
predict the protein identity based on a comparison with masses predicted by virtual
digestion of identified open-reading frames in a given genome [8].
In contrast to gel-based proteomics analysis, shotgun proteomics is a gel-free
approach that uses different liquid chromatography techniques for separation of
complex peptide mixtures coupled with mass spectrometry identification [9]. It starts
with the isolation of the protein-containing sample, which may also be prefraction-
ated by chromatographic or electrophoretic techniques followed by a digestion step
to generate a complex mixture of peptides. Multidimensional chromatographic tech-
niques are subsequently used to separate and further fractionate the obtained peptide
mixtures, which are then submitted for analysis by electrospray ionization (ESI) or
MALDI-based mass analysis. Sophisticated mathematical algorithms are used to
compare the peptide sequence data derived from mass spectra with sequence infor-
mation in protein databases to identify the proteins.
Major technological advancements in protein chemistry have led to chip-based
protein sample arrays with different chromatographic surfaces to selectively bind
proteins with specific chemical properties. One of the evolved technologies is the
surface-enhanced laser desorption ionization time-of-flight (SELDI-TOF) platform
for high-throughput analysis of proteins in complex biological specimens [10]. It
uses a protein chip array system with different chromatographic properties, including
hydrophobic, hydrophilic, anion exchange, cation exchange, or immobilized-metal
affinity surfaces. After addition of an energy-absorbing matrix to the chip, the
proteins are analyzed by MALDI-TOF mass analysis. The output from the proteome
analysis based on chips via SELDI-TOF is a trace of relative intensity vs. m/z (mass-
to-charge ratio) of the detected proteins, providing a pattern that can be compared
from different samples or that can be used for cluster analyses for significant protein
abundance differences between samples. The future of more simple proteome anal-
ysis tools may be the use of antibody libraries that contain specific antibodies raised
against an expressed open-reading frame and making proteome analysis similar to
high-throughput microplate assays that allow essentially every protein to be identi-
fied and quantified easily [11].
2.3 BIOFLUIDS AS A SOURCE OF MARKER DISCOVERY

The analysis of the effects of nutrition on human health requires easily accessible
biological samples that provide information on the metabolic status. Plasma or serum
represents the fluid compartment that carries metabolites and signaling molecules
from and to cells in the various body organs and therefore serves as a distribution
and messenger system in interorgan metabolism (Figure 2.2). Blood is readily
available in sufficient quantities, and its molecular composition appears to reflect
Eux Proteome of
blood cells
Eux
Inux Plasma
proteome
Eux
Eux Urine
proteome
FIGURE 2.2 The importance of human blood as a source of biological information with the
possibilities of analyzing plasma, blood cell proteomes, and urine as easily obtainable samples.
the condition of individual organs and the physiological processes associated with
it. Moreover, proteins in the blood are actively participating in many life-sustaining
body functions such as the transport of nutrients and metabolites, inflammation,
hormone signaling, and many more. The blood proteome therefore promises to be
an ideal biological entity for assessing markers of health or disease.
Analysis of blood proteins, however, depends on many variables, most impor-
tantly, on the methods of sampling, sample processing, and storage. A particular
problem is that plasma/serum contains around 22 abundant proteins that represent
99% of the protein mass. The most prominent protein is albumin, with around
45,000 g/ml, whereas the concentration of other proteins, such as the alpha-
fetoprotein, may be as low as 0.005 g/ml [12]. This huge concentration difference
makes analysis of the plasma proteome very difficult and requires the removal of
the most abundant proteins from the sample to allow a reliable analysis of the minor
components. During the course of the Human Proteome Project (HPP), a group of
38 expert labs has focused on the human plasma proteome. The Human Plasma
Proteome Project (HPPP) has worked out standard operational procedures and has
made recommendations for sample analysis and processing [13]. A wealth of bio-
logical information is already available from the HPPP and is in the public domain
(http://www.plasmaproteomedatabase.org/). A total of 2446 unique gene products
and more than 4900 protein entities were identified when specific isoforms were
taken into account. In silico methods were employed to determine the origin and
nature of the various plasma proteins (see Figure 2.3). This analysis delivered a
surprisingly high number of unexpected protein classes, including a large number
The Human Plasma Proteome: Origin and Nature of The Proteins
Cytoplasma 53%
Cellular proteins 58% Membrane-associated 19%
Other extracellular or secreted 14%
Immunoglobulins 8% Nuclear 9%
Cytokine and related 5%
Unknowns 21%
Common circulatory proteins 28%
Classic plasma proteins 13% Coagulation and complement factors 26%
Binding proteins 14%
Inhibitors 13%
Proteases and other enzymes 19%
Classication According to Biological Functions

Signal transduction and cell communication 24%
Unknown biological processes 25%
Energy metabolism 9.5%
Protein metabolism 8%
Transport processes 7.4%
Immune response 4.3%
Nucleic acid/nucleoside/nucleobase metabolism 13.6%
Cellular organization and biogenesis 8.6%
FIGURE 2.3 The origins and nature of the proteins identified in human plasma by the HPPP.
of intracellular proteins, and even proteins of nuclear origin. Whether they are
indicators of normal tissue renewal processes or a measure of tissue (organ-specific)
damage is currently under investigation. In this context, the isoform-specific proteins
(in case of selected enzymes, these are well-established clinical markers) in plasma
are interesting for further exploration. The nutritional status is known to affect the
levels of distinct plasma proteins, and organ-specific processes appear also to be
detectable in plasma as some of the proteins found in plasma are derived from
individual organs. In the HPPP, 362 of the plasma proteins were of hepatic origin
and 345 proteins were cardiovascular related, and about 110 proteins could be
related to inflammation such as cytokines, adhesion molecules, and chemokines [14].
Plasma peptides are usually too small in molecular mass for standard proteome
analysis. However, they are of particular interest because of their specific functions
in biological processes in which they act as hormones, growth factors, adipokines,
etc. The plasma peptidome defined as the fraction with peptides of < 10,000 or
15,000 Da therefore requires special techniques for separation and analysis.
Usually, a prefractionation of plasma based on ultrafiltration techniques with mem-
branes of different cutoff values is used and then further chromatographic separation
steps combined with mass analysis are employed to characterize the low-molecular-
weight peptides in the sample. The plasma peptidome is considered to represent up
to 5000 different peptides [15].
Proteome analysis may also be applied to circulating blood cells. The fraction
of mononuclear cells can easily be collected from blood samples by commercial
separation kits, and these cells may serve as reporter cells as they are in circulation
for extended time periods and reach different body compartments. Proteome analysis
of human peripheral mononuclear cells has not been performed in the context of
biomarker discovery or clinical diagnostics. In contrast, several studies have targeted
the platelet proteome, and a high-resolution platelet proteome map that comprises
more than 2000 different protein features has been generated [16,17].
Other body fluids such as urine, saliva, tears, seminal fluid, synovial fluid, and
cerebrospinal fluid have also been submitted to proteome analysis in search for
biomarkers of disease and toward improved clinical diagnostics [18].
2.4 EXAMPLES OF BIOMARKER DISCOVERY

APPROACHES IN BODY FLUIDS
BY PROTEOME ANALYSIS
Proteome analysis techniques are used to search for changes in the relative abundance
of individual proteins or protein clusters that may be indicative of a certain disease
state. Despite the fantastic accomplishments in proteome analysis in recent years,
there remain a number of important challenges that will need to be addressed to
pave the way for the acceptance of proteomics as a relevant diagnostic tool. The
most advanced state of analysis in this respect has been reached in cancer diagnostics.
Numerous studies have screened biosamples in cancer patients, including
serum/plasma for identification of specific markers and for validation of the diagnostic
power of the proteome approaches as compared to established clinical diagnosis
strategies [19]. The ultimate goal here is to develop biomarkers that are able to
reliably and accurately predict outcomes during cancer development, management,

and treatment, and it is envisioned that there might in the future be a serum or urine
test available for every phase of cancer that may guide clinical decision making and
which may replace the invasive techniques used today [20].
Beyond cancer proteomics, there are not too many studies that have taken
proteome analysis into the clinical or diagnostic environment. However, a recent
study has assessed candidate protein biomarkers of rheumatoid arthritis in synovial
fluid, which is found in small amounts in joints, bursas, and tendon sheaths. A total
of 33 candidate biomarkers for rheumatoid arthritis were identified from a total of
418 detected proteins. Among the proteins that were elevated in the synovial fluid
of patients were C-reactive protein (CRP) and six members of the S100 protein
family of calcium-binding proteins [21]. Cerebrospinal fluid has also been used as
a sample, and it has been hypothesized that alterations in the protein composition
of CSF may reflect abnormalities in protein expression associated with trauma,
neurodegenerative disorders, or multiple sclerosis. A number of proteomic approaches
have recently been used to find differentially expressed proteins in patients with
Alzheimers disease [22] or multiple sclerosis [18]. Although none of the putative
protein markers have been validated for these diseases, yet the efforts seem very
promising.
Urine proteome analysis has the advantage that obtaining the sample does not
need any invasive approaches such as drawing blood. Because it is anticipated that
altered urine protein patterns or protein concentrations may reflect mainly impaired
kidney functions, urine proteome analysis is currently predominantly used to charac-
terize acute and chronic renal toxicity of compounds or for diagnostics of diseases
that are known to affect kidney function. The power of urine proteomics as a diagnostic
tool was recently demonstrated by identification of a large number of polypeptides
in spot urine samples of type 1 diabetes patients with different stages of nephropathy
[23]. Specific clusters of 54 polypeptides were only found in type 1 diabetes patients.
A major challenge of urinary proteomics is also to reduce the complexity of the
urinary proteome to allow low-abundance proteins of special pathophysiological
significance to be detected. Urinary proteins can originate from glomerular filtration
of low-molecular-weight plasma proteins or by the secretion of renal tubular soluble
proteins. In addition, whole-cell shedding and apical plasma membrane shedding by
nonspecific or apoptotic processes will lead to protein excretion in urine. Recent
studies have demonstrated that small vesicles, so-called exosomes, which represent
internal vesicles of multivesicular bodies are delivered to the tubular fluids and urine
via fusion of the bodies with the apical plasma membrane of renal tubule epithelial
cells. Exosomes were shown to contain numerous proteins of clinical importance that
may also serve as markers of kidney function in health and disease [24].
Regarding the effects of nutritional factors and putative markers of diet-
dependent diseases, there are only very few examples of proteome analysis being
used as a discovery tool in human studies. In a controlled dietary intervention,
MALDI-TOF spectra were generated from serum samples of 38 human volunteers.
In a randomized controlled trial, the participants ate during separate feeding periods
either a basal diet devoid of fruits and vegetables or a basal diet supplemented with
cruciferous vegetables (broccoli). At the end of each 7-d feeding period, serum
samples were obtained and large, abundant proteins removed. Bioinformatic analysis
of MALDI-TOF spectra revealed two significant peaks that could classify partici-
pants based on diet with 76% accuracy. One of the peaks (at 2740 m/z) was identified
as the B-chain of -2-HS glycoprotein, a serum protein previously found to vary
with diet and thought to be involved in insulin resistance and immune function [25].
The adipocyte fatty acidbinding protein (A-FABP) is considered to be a cyto-
solic fatty acid chaperone. However, mice with a targeted disruption of the A-FABP
gene show protection from insulin resistance, hyperglycemia, and atherosclerosis,
suggesting that A-FABP could have additional functions. 2D-gel-based separations
of proteins from cultured adipocytes identified the cellular fatty acid-binding protein
as a prominent secretion product. When obese patients were concomitantly screened
for plasma levels of A-FABP, it was observed that concentrations of A-FABP were
significantly higher in overweight and obese as compared to lean persons and after
adjustment for age and sex, serum A-FABP concentrations correlated significantly
with blood pressure, fasting insulin, waist circumference, and dyslipidemia. A-FABP
therefore appears to be a new biomarker that is associated with obesity and the
metabolic syndrome [26].
2.5 FUTURE PERSPECTIVES

The recent advancements in proteomic technologies, including improved two-
dimensional gel electrophoresis, new mass spectrometric techniques, and convenient
chip-based sample treatment techniques with simplified analysis combined with the
advancements in bioinformatic tools should foster the discovery of new biomarkers
that are both sensitive and specific to distinguish a healthy individual from a diseased
one. The advantage of proteomics is that it assesses the abundance of proteins, which
are the prime entities that carry out and mediate biological functions. This advantage,
however, is currently counterbalanced by the expensive technologies and the rather
difficult techniques. In view of proteome analysis as a tool for assessing health status
in personalized nutrition concepts, simpler, faster and more convenient methods are
needed. Protein microarrays are considered to meet these requirements for a simul-
taneous analysis of the abundance, function, and interaction of proteins on a sys-
temwide scale [27]. The power of chip-based miniaturized protein assays lies in the
potential to investigate in parallel large numbers of analytes in a variety of biological
samples for innovative in vitro diagnostic applications. It can be expected that a
range of customized and targeted equipment will become available in the next years
for easy-to-use proteomics approaches in lifestyle monitoring and for the assessment
of disease progression. As the field is rapidly developing with respect to the analysis
of marker proteins in all kinds of body fluids (plasma, tears, saliva, urine, etc.), one
can also expect that the consumers themselves will collect those fluids easily and
send them via their doctors or specialized health-care providers to a proteomics
laboratory for analysis. Initial applications will probably target the diagnostics of
common diseases such as diabetes, arthrosclerosis, allergies, or autoimmune dis-
eases. In diabetics, nonenzymatic glycosylation of target proteins may be measured
routinely to assess the maintenance of healthy blood glucose profiles (similar to
analysis of hemoglobin A1) or by monitoring the urine proteome, which may indicate
disturbed kidney function at an early stage as the disease progresses. Detecting serum
autoantibodies in autoimmune diseases or quantitative measurements of serum aller-
gen-specific IgE levels in type I allergy are already possible via proteome analysis
in chip formats with minute amounts of serum. A detailed analysis of the protein
components of the lipoproteins of the LDL and HDL classes may provide better
diagnostics for arthrosclerosis and heart disease, in particular, when different iso-
forms related to different organs of origin or because of genetic variations are taken
into account. However, before proteomics moves into the real world and these
new and more convenient assay systems become available, nutritional science needs
to explore with existing technologies the power of proteomics in characterizing
human metabolism and diet responses in well-defined studies.
REFERENCES
1. Chambers, G. et al., Proteomics: a new approach to the study of disease, J Pathol,
192, 280, 2000.
2. Kussmann, M. et al., Proteomics in nutrition and health, Comb Chem High Through-
put Screen, 8, 679, 2005.
3. Gorg, A. et al., The current state of two-dimensional electrophoresis with immobilized
pH gradients, Electrophoresis, 21, 1037, 2000.
4. Cordwell, S.J. et al., Subproteomics based upon protein cellular location and relative
solubilities in conjunction with composite two-dimensional electrophoresis gels,
Electrophoresis, 21, 1094, 2000.
5. Corthals, G.L. et al., The dynamic range of protein expression: a challenge for
proteomic research, Electrophoresis, 21, 1104, 2000.
6. Adam, G.C. et al., Profiling the specific reactivity of the proteome with non-directed
activity-based probes, Chem Biol, 8, 81, 2001.
7. Conrads, T.P. et al., Utility of accurate mass tags for proteome-wide protein identi-
fication, Anal Chem, 72, 3349, 2000.
8. Chamrad, D.C. et al., Interpretation of mass spectrometry data for high-throughput
proteomics, Anal Bioanal Chem, 376, 1014, 2003.
9. Wolters, D.A. et al., An automated multidimensional protein identification technology
for shotgun proteomics, Anal Chem, 73, 5683, 2001.
10. Issaq, H.J. et al., The SELDI-TOF MS approach to proteomics: protein profiling and
biomarker identification, Biochem Biophys Res Commun, 292, 587, 2002.
11. LaBaer, J. and Ramachandran, N., Protein microarrays as tools for functional pro-
teomics, Curr Opin Chem Biol, 9, 14, 2005.
12. Qian, W.J. et al., Comparative proteome analyses of human plasma following in vivo
lipopolysaccharide administration using multidimensional separations coupled with
tandem mass spectrometry, Proteomics, 5, 572, 2005.
13. Rai, A.J. et al., Hupo plasma proteome project specimen collection and handling:
towards the standardization of parameters for plasma proteome samples, Proteomics,
5, 3262, 2005.
14. Ping, P. et al., A functional annotation of subproteomes in human plasma, Proteomics,
5, 3506, 2005.
15. Richter, R. et al., Composition of the peptide fraction in human blood plasma: database
of circulating human peptides, J Chromatogr B Biomed Sci Appl, 726, 25, 1999.
16. Garcia, A. et al., Applying proteomics technology to platelet research, Mass Spectrom
Rev, 24, 918, 2005.
17. ONeill, E.E. et al., Towards complete analysis of the platelet proteome, Proteomics,
2, 288, 2002.
18. Veenstra, T.D. et al., Biomarkers: mining the biofluid proteome, Mol Cell Proteomics,
4, 409, 2005.
19. Alaiya, A. et al., Clinical cancer proteomics: promises and pitfalls, J Proteome Res,
4, 1213, 2005.
20. Chatterjee, S.K. and Zetter, B.R., Cancer biomarkers: knowing the present and pre-
dicting the future, Future Oncol, 1, 37, 2005.
21. Liao, H. et al., Use of mass spectrometry to identify protein biomarkers of disease
severity in the synovial fluid and serum of patients with rheumatoid arthritis, Arthritis
Rheum, 50, 3792, 2004.
22. Davidsson, P. and Sjogren, M., Proteome studies of CSF in AD patients, Mech Ageing
Dev, 127, 133, 2006.
23. Meier, M. et al., Identification of urinary protein pattern in type 1 diabetic adolescents
with early diabetic nephropathy by a novel combined proteome analysis, J Diabetes
Complications, 19, 223, 2005.
24. Pisitkun, T. et al., Identification and proteomic profiling of exosomes in human urine,
Proc Natl Acad Sci USA, 101, 13368, 2004.
25. Mitchell, B.L. et al., Evaluation of matrix-assisted laser desorption/ionization-time
of flight mass spectrometry proteomic profiling: identification of alpha 2-HS glyco-
protein B-chain as a biomarker of diet, Proteomics, 5, 2238, 2005.
26. Xu, A. et al., Adipocyte fatty acid-binding protein is a plasma biomarker closely
associated with obesity and metabolic syndrome, Clin Chem, 52, 405, 2006.
27. Kung, L.A. and Snyder, M., Proteome chips for whole-organism assays, Nat Rev Mol
Cell Biol, 2006.
3 Metabolomics
and the Personalized
Metabolic Signature
Michael J. Gibney, Marianne Walsh,
and Lorraine Brennan
CONTENTS
3.1 Introduction ....................................................................................................23

3.1.1 Phenotype, Endogenous Metabolism, and Nutrients
as Determinants of Metabolomic Signatures ....................................27
3.1.2 Food Nonnutrients as Determinants
of Metabolic Signatures .....................................................................28
3.1.3 The Colonic Microflora as Determinants
of Metabolic Signatures .....................................................................29
3.2 Characterizing Dietary Patterns Using Metabolomics...................................29
3.3 Bottlenecks in the Human Nutrition Metabolomics
Research Road Map .......................................................................................30
3.4 What Is the Future for Metabolomics in Personalized Nutrition?................31
Key Readings...........................................................................................................31
3.1 INTRODUCTION
Whereas traditional nutrition and metabolism research sets out a priori the metab-
olites of interest in the hypothesis to be tested, nutritional metabolomics attempts
to capture the presence of the maximum number of small chemicals regardless of
their identity or relevance to any advance hypothesis. It is thus a hypothesis-
generating exercise. The totality of all signals in a sample is reduced in a two-stage
process to one point in two-dimensional space, and metabolic signatures that are
comparable will cluster closely together and, as the metabolomic profiles differ, the
points separate in proportion to the differences. Metabolomic profiles of many
diseases have been studied, and distinct clusters have been identified. Similarly, met-
abolic profiles arising from pharmacological intervention have also been characterized
and are found to cluster together for similar drugs. To date little or no serious work
has been published on nutritional metabolomics, but a number of clear challenges
exist. In nutrition, metabolic noise will be relatively high because food contains
23
many nonnutrients, which are absorbed and which will be detectable in biofluids.
This, coupled with the fact that the effect of diet on metabolomic profiles is likely
to be milder than that of disease or drugs, makes the challenge of nutritional
metabolomics very special. However, whether it is through targeted metabolomic
profiles or universal metabolomic profiles, nutritional metabolomics offers great poten-
tial for nutrition research.
Traditionally, nutritional science has proceeded in a hypothesis-driven manner
in which a priori, end points were known, were part of the hypothesis, and were
the basis of experimental measure. Although that will remain so, new advances in
biology have led to the development of a parallel strategy that is hypothesis gener-
ating rather than hypothesis testing. Metabolomics is part of that new biology. As
ever in science, terminology becomes important, and two terms coexist in this field,
metabolomics and metabonomics. For completeness we will present the original
definitions of both and leave it to the reader to distinguish between the two. Meta-
bonomics is defined as the quantitative measurement of the time-related multi-
parametric metabolic response of living systems to pathophysiological stimuli or
genetic modification. On the other hand, metabolomics has been defined in the
literature as the comprehensive and quantitative analysis of all metabolites .
Throughout this chapter we will use the term metabolomics exclusively but remind
the reader that either term can be used.
In relation to human studies, metabolomics is the measurement of the maximum
number of metabolites present in a biofluid. A range of analytical methods can be
used to generate data. However, mass spectrometry (MS) in its various hyphenated
derivations (LC-MS, GC-MS) and high-resolution 1H nuclear magnetic resonance
spectroscopy (1H NMR) are the dominant platforms. Both platforms have their
advantages and disadvantages, and best practice is to apply both, if possible. For
the purpose of this chapter we will focus more on NMR applications but in no way
wish to claim it is a superior technique. In the case of 1H NMR, each biofluid has
a characteristic spectroscopic fingerprint that reflects the metabolic composition (see
Figure 3.1 for 1H NMR spectra of different biofluids). The intensities of peaks, in
an NMR spectrum, belonging to a certain metabolite are determined by its concen-
tration, and assignment can be made based on chemical shifts (position in spectra),
spin multiplicities, and addition of authentic material. Typical metabolomic studies
generate large amounts of data that cannot easily be analyzed in an unbiased way.
To overcome this problem, a number of pattern recognition techniques have been
applied to robustly analyze NMR spectra.
The most commonly used method to analyze metabolomics data is through
principal component analysis (PCA). This statistical method reduces a large number
of variables to a smaller number of variables, called principal components. The
principles underlying the reduction of a complex multisignal NMR output to a single
point in two-dimensional space through PCA is best illustrated with a 3-peak output
rather than with the several hundred peaks normally found in NMR spectra.
Figure 3.2A and Figure 3.2B show how these three peaks can be processed: first,
the multipeak output is summarized by one point in multidimensional space and
then reduced to a lower two-dimensional space. In the resulting scores plot, each
point represents all the data contained in the spectrum belonging to one individual.
Metabolomics and the Personalized Metabolic Signature 25
// //
Creatinine
TMAO
Citrate
DMA
Glycine
Alanine
C
4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0
Chemical shift (ppm)
//
Propionate
Propionate
Acetate
Lactate
Pyruvate Alanine
Choline
B
4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0
Lactate
Valine
Alanine
Acetoacetate
Lactate Glutamine
Acetate
A
4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0
FIGURE 3.1 Typical 1H NMR spectra of biofluids obtained from a healthy volunteer: (a) plasma,
(b) saliva, and (c) urine. TMAO, trimethylamine-N-oxide; DMA, dimethylamine. (From Am
J Clin Nutr, 2006, 84:531539. With permission.)
C B
b
FIGURE 3.2A Three NMR or MS peaks a, b, and c, must be reduced to one point in multidi-
mensional space. Three coordinates A, B, and C are used to achieve this. The signal strength
of peak a is laid on coordinate A proportionate to the size of the signal. Peak b is now placed
parallel to coordinate B, and similarly peak c is set parallel to coordinate C, each proportionate
to the size of the relevant signal. The final point is the only position these three peaks can give
in three-dimensional space, and the same holds true for k peaks in K-dimensional space.
PC 1
PC 2
FIGURE 3.2B The figure on the left should be imagined as a bunch of grapes in multidi-
mensional space. The center (nonfilled) grape is found, and a pencil is inserted into the bunch
of grapes through the center multiple times from all possible angles until maximum variance
to the bunch of grapes is found. A second pencil is similarly inserted at a right angle to the
first. Half of each pencil is lost, and the distance each grape is from each pencil [principal
components (PC) 1 and 2] is measured as shown for one point with an arrow on the right
figure. Now, a two-dimensional plot retaining maximum information on multidimensional
space is developed.
Hence, observation points that cluster close together have similar spectra, whereas
those that are far apart will have very different spectra. The spectral variables
responsible for the positioning of the observations in the scores plot can be obtained
from the loadings plot. Although PCA is the initial unsupervised data analysis
carried out in most metabolomic studies, it is usually followed by more sophisticated
supervised techniques such as Partial Least Squares Discriminant Analysis (PLS-DA).
The power of these statistical techniques to reduce complex NMR profiles to
single points in two-dimensional space is shown in Figure 3.3A and Figure 3.3B.
Figure 3.3A gives typical NMR spectra for urine samples collected in the morning
or in the evening by the same individual. Clearly, the samples taken at different time
points exhibit differences in the NMR profiles which are clearly discernible to the
Night
Morning
11 10 9 8 7 6 5 4 3 2 1
FIGURE 3.3A 1H NMR spectra of urine from the same subject sampled first thing in the
morning and last thing in the evening.
Morning 12 Morning
Night 10 Night
10 8
6
0 4
2
t(2)
t(2)
0
10 2
4
6
20 8
10
12
20 10 0 10 10 0 10
t(1) t(1)
PCA PLSDA
FIGURE 3.3B Principal component analysis (PCA) and partial least squares discriminant
analysis (PLS-DA) of NMR data from human urine samples collected both in the early
morning and late evening.
human eye, as shown by asterisks. When PCA and PLS-DA, in particular, are applied,
quite clear separation of the nightmorning samples is seen because the samples
taken at each time point have comparable signatures but nighttime and morning-
time signatures are quite different. To date, metabolomic profiling of biofluids have
shown unique signatures for cases and controls in many chronic diseases such as
cardiovascular disease, cancers, osteoarthritis, and several neurodegenerative diseases.
The metabolomic profiles of human biofluids are determined by a number of
identifiable factors, although the quantitative effect of each individual factor remains,
as yet, unknown. It is evident from later sections that the presence or absence of a
disease or, if present, the severity of the disease is one determinant of a metabolic
signature. However, for the purposes of this section of the chapter, the focus will
be on physiological factors determining the nature of the metabolomic profiles in
human biofluids. For clarity in what can be a rather confused landscape, environ-
mental factors, including stress, will not be considered. This is not to say that these
are unimportant but, rather, it is advantageous to focus on core determinants of
metabolomic profiles that are amenable to dietary influence. Thus, this chapter will
focus on the effects of phenotype and endogenous metabolism on food with an
evaluation of the effects of both nutrients and nonnutrients, and also on the role of
the colonic microflora.
3.1.1 PHENOTYPE, ENDOGENOUS METABOLISM, AND NUTRIENTS

AS DETERMINANTS OF METABOLOMIC SIGNATURES
It is the essence of life that biological systems turn over constantly with a simulta-
neous set of anabolic and catabolic pathways. The balance of these is determined
by a variety of endocrine-based factors. The anabolic phase is dominant in growth,
reproduction, and weight gain, whereas the catabolic factors are dominant in aging,
illness, and weight loss. This constant entry and exit of metabolites from plasma
plays a major role in determining the metabolic profile of that biofluid in the normal
sampling state of fasting. Phenotype therefore must play a very significant role in
shaping fasting plasma metabolomic profiles. For example, protein turnover is
strongly related to lean body mass and, thus, per unit of plasma, the rate of exit and
entry of amino acids into plasma will be far higher in individuals with a higher as
compared to a lower lean body mass per unit weight. Height and girth will reflect
skeletal differences, and greater skeletal mass will mean greater traffic of related
metabolites in and out of plasma. Clearly, as percentage body fat increases, the
constant anabolism and catabolism in adipose tissue determines the flow of nones-
terified fatty acids, glycerol, glucose, and triacylglycerol-rich lipoproteins and their
remnants to and from the liver via the plasma metabolome. Diet can clearly alter
the nature of the plasma metabolome in the postprandial phase, and the nature of
that effect will be dependent on the nature of the ingested meal. However, this is an
acute event that in the short term will not influence the nature of the fasting plasma
metabolome. If a modification to diet is sustained such that it can be described as
chronic, then it will lead eventually to a readjustment of gene expression and to a
redefining of the parameters of anabolic and catabolic pathways, and that will
influence the fasting plasma metabolome. Thus, an acute change of the ratio of
polyunsaturated to saturated fatty acids in the diet will have a transient effect on the
postprandial metabolome, whereas a chronic change will lead to a sustained read-
justment of the expression of many genes, resulting in altered fasting levels of plasma
cholesterol fractions, haemostatic functions, insulin responsiveness, and the like.
Phenotype will also influence urinary metabolomic profiles in that age, sex, height,
weight, and the endocrine environment all shape the constituents of urine.
3.1.2 FOOD NONNUTRIENTS AS DETERMINANTS

OF METABOLIC SIGNATURES
There is no doubt from an extensive literature that chronic changes in the human
diet significantly alter metabolic homeostasis and that the altered metabolic homeo-
stasis will in turn alter the acute response to a nutrient load. There is, however, a
major difference in the relative effects of food and nutrients on urinary and plasma
metabolomic profiles. In the case of plasma, nutrients are preferentially retained
unless fairly high renal thresholds are exceeded. In contrast, xenobiotics, which
abound in the nonnutrient element of food, are preferentially secreted into urine. An
extensive review of the metabolism of plant chemicals or phytochemicals shows an
average half-life for excretion into urine of about 8 h, with a maximum of about 20
h. These phytochemicals do not enter normal metabolic pathways, but they do
operate as receptorligands, triggering a number of metabolic responses such as is
encountered with onion peeling, or the consumption of coffee, or the habit of
cigarette smoking.
In effect, the plasma metabolomic profile is dominated by the body composition
dimension of phenotype and by metabolites of a multitude of pathways. It can be
altered by diet, but only after chronic exposure has adjusted gene expression and,
thus, the management of metabolic pathways. In the case of urine, phenotypic
characteristics will play a significant role, but they will be subject to considerable
interference by recently ingested nonnutrients. In one study of urinary metabolomic
profiles in subjects consuming chamomile tea, a significant number of outliers were
characterized by high levels of trimethylamine oxide (TMAO) indicative of the recent
consumption of seafood. A recent study by the authors has shown that strict and
total control over food choice for 24 h prior to biofluid collection has no effect on
reducing variability in plasma or saliva, but does have a very significant effect in
reducing between-person variability in urinary metabolomic profiles. The authors
speculated that this is best explained by the reduction in between-person variability
in nonnutrient intake rather than any change in metabolic homeostasis. Thus, in
human nutrition, urinary collections might have to be collected with strict control
over the prior 24-h food choice, an issue for discussion by international groups setting
standards for this area of research (www.nugo.org).
3.1.3 THE COLONIC MICROFLORA AS DETERMINANTS

OF METABOLIC SIGNATURES
A final diet-related factor to be considered is the large-bowel microflora. A case has

been made that the continual production of metabolites by the colonic microflora
will significantly influence the shape of the plasma and urinary metabolome. It is
certainly the case from experimental studies that the transition from a gnotobiotic
rodent status to a conventional rodent status is characterized by dramatic shifts in
urinary metabolomic profiles. Equally, within established colonies of conventional
animals, metabolomic subsets can be identified, which have been attributed to
variation in the colonic microflora. However, to date, no human studies are available
to ascertain whether the use of different prebiotics and probiotics would lead to
detectable changes in the metabolomic profiles of different human biofluids.
3.2 CHARACTERIZING DIETARY PATTERNS

USING METABOLOMICS
Although the early days of nutrigenomics heralded the dawn of personalized nutri-
tion, literally at the level of the individual, it is becoming increasingly evident that
such an approach is overambitious. Unquestionably, nutrigenomics will move us
from the present paradigm where data at population epidemiology level is used to
infer events at the individual level. However, it is more likely to move from whole
population to subgroup rather than to the level of the individual. We have seen that
metabolomics can be used to allocate an unknown metabolic signature to a given
cluster of a clinical condition that has similar characteristics as the unknown signa-
ture. This is now moving from the individual to a similar subgroup of the population
in terms of metabolomic profiles. Could the same be achieved for dietary patterns?
One possibility to envisage would be to take population dietary patterns and to use
cluster analysis to identify 8 to 12 dietary patterns that together explain 90% of
dietary variability. Then, healthy volunteers varying in phenotype (age, sex, and body
composition) could be employed to create reference metabolomic profile databases
based on plasma and urinary samples. Then, unknowns could be run with the
original databases appropriate for the given phenotype to ascertain where in the spec-
trum of dietary variability the unknown lies. Thus, the dietary profile of the subject
would be characterized. This does not quantify food and nutrient intake patterns and
will not replace such areas of investigation. However, it does overcome the enormous
bottleneck that exists in dietary phenotyping, namely, energy underreporting, in
which subjects misreport the consumption of a food or, if consumed, its frequency
of consumption and portion size. In this scenario, the true overall pattern of food
and nutrient intake will have been ascertained, and quantification at that point may
lead to reduced manipulation of the truth, because the truth will have been ascertained
by objective modern technologies.
A parallel question that will arise is whether, once a metabolomic profile has
been established, any inference might be made as to the biological future of someone
with that metabolic profile, such as their likely metabolic response, favorable or
otherwise, to some change in diet. Until recently, this has been seen as belonging to
the realm of nutrigenetics, which studies how common genetic variants (single-
nucleotide polymorphisms) influence the responsiveness to dietary change. However,
a recent study in rats has shown that knowledge of the metabolic profile of an animal
prior to pharmacological intervention can accurately predict responsiveness to the
test drug. At first this seems somewhat unexpected, but on reflection, the metabolomic
profile is the ultimate signature of the product of phenotype as well as gene expression
and single-nucleotide polymorphism and, as such, is the universal biomarker.
3.3 BOTTLENECKS IN THE HUMAN NUTRITION

METABOLOMICS RESEARCH ROAD MAP
Using metabolomics to characterize the human diet or to predict any biological
consequence of a given metabolomic profile is at present within the realm of vision-
ary thinking. In reality, many bottlenecks exist in this field that need to be addressed
before that path can be realistically followed. Foundation studies that determine the
relative strengths of the various determinants of the human nutrition metabolome
need to be put in place in order to increase our confidence in interpreting the results
of such studies. Such determinants include the acute, chronic, and acute-on-chronic
effects of diet, including nutrients and nonnutrients, the role of phenotype and
endogenous metabolism, and the role of the colonic microflora. There is also another
major research challenge to human nutritional metabolomics. At present, most
metabolomics research uses nontargeted approaches, that is, a total data capture
approach. When two treatments or conditions are seen to differ using PCA, the
relative importance of different parts of the NMR or MS spectra can be examined
to ascertain which parts make the greatest contribution to the difference between
treatments. Once the main differential compounds have been identified, a hypothesis
can be generated to explain why these compounds might play such a role. In effect,
this moves from the hypothesis-generating mode to the hypothesis-testing mode.
Although this approach will continue to prosper, in human nutrition there is a
commitment to a parallel approach in which the human nutrition metabolome is
identified and in which there is an aspiration for the development of a new technology
that would not only identify all the metabolites of this human nutrition metabolome
but also quantify them in the various biofluids. The European Nutrigenomics Organ-
isation (www.nugo.org) is committed to defining the human nutrition metabolome
and to creating an open-source database on the biological details of all compounds
in the human nutritional metabolome (www.nugowiki.org). This database will link
with that of the Human Metabolome Database (www.metabolomics.ca), which is a
major project of Genome Canada. The basic philosophy is that one day we will have
access to a list of all nutrients present in plasma, for example, that are important in
human nutrition and the pathways they regulate. Additionally, the factors controlling
the balance of these nutrients, such as genetic factors, will be described in detail,
and the normal and abnormal concentration ranges will be known. Thus, normal
values for plasma cholesterol are well established, and so we can recognize low and
high levels. In an ideal world we could recognize normality for the concentrations
of the totality of plasma nutrients and related compounds in a metabolomic signature
and, if that can be achieved, we should be able to recognize metabolic perturbations
at an early stage.
3.4 WHAT IS THE FUTURE FOR METABOLOMICS

IN PERSONALIZED NUTRITION?
Although metabolomics offers much potential for our understanding of diet and
health, the subject is too immature at present to stake any special claim as a unique
tool in personalized nutrition. Many basic foundation studies are needed to under-
stand the relative strengths of the many determinants of the nutritional metabolome
and its change with diet. Only when we have a firm understanding of these forces
will we be able to plan and interpret the studies that will allow the true potential of
metabolomics in personalized nutrition to be established. However, conceptually,
metabolism is so close to nutrition that the likelihood that metabolomics will play
a significant role in personalizing nutrition is significant. Recently, metabolomic
profiling has been used to predict responsiveness to drug therapy in rodents. How-
ever, humans are far more heterogeneous than their rodent counterparts, and nutrient-
based signals are much softer than those for pharma. However, the road map to test
the role of metabolomics in personalized nutrition is clear and, thus far, uncluttered.
The next 5 years will be critical.
KEY READINGS
Bollard, M.E., Stanley, E.G., Lindon, J.C., Nicholson, J.K., and Holmes, E. NMR-based meta-
bonomic approaches for evaluating physiological influences on biofluid composition.
NMR Biomed 18: 14362, 2005.
German, J.B., Hammock, B.D., and Watkins, S.M. Metabolomics: building on a century of
biochemistry to guide human health. Metabolomics 1: 39, 2005.
Gibney, M.J., Walsh, M.C., Brennan, L., Roche, H.M., German, B., and van Ommen, B.
Metabolomics in human nutrition: opportunities and challenges. Am J Clin Nutr 82:
497503, 2005.
Keun, H.C. Metabonomic modeling of drug toxicity. Pharmacol Ther 109: 92106, 2006.
Lamers, R.J., van Nesselrooij, J.H., Kraus, V.B., Jordan, J.M., Renner, J.B., Dragomir, A.D.,
Luta, G., van der Greef, J., and DeGroot J. Identification of a urinary metabolite
profile associated with osteoarthritis. Osteoarthritis Cartilage 13:7628, 2005.
Nicholson, J.K., Holmes, E., and Wilson, I.D. Gut microorganisms, mammalian metabolism
and personalized health care. Nat Review Microbiol 3: 4318, 2005.
4 Nutrition, Genomics,
and Cardiovascular
Disease Risk
Jose M. Ordovas and Dolores Corella
CONTENTS
4.1 Introduction ....................................................................................................33

4.2 How Nutrients Communicate with Genes.....................................................35
4.3 Observational Studies: The Current Workhorse of Nutrigenetics.................36
4.4 Interventional Studies in Nutrigenetics .........................................................42
4.5 GeneDiet Interactions in the Postprandial State .........................................42
4.6 Conclusions and Steps Forward ....................................................................43
Acknowledgments....................................................................................................44
Suggested Lectures/URLs .......................................................................................44
References................................................................................................................45
4.1 INTRODUCTION
The existence in the population of hyper- and hyporesponders to dietary modification
and to any other therapeutic approach is a well-known phenomenon. Whereas some
individuals are able to improve their plasma lipid profile or body weight using the
current dietary guidelines, others are unable to gain any significant benefit. The basis
for these interindividual differences are complex, and current knowledge suggests
that it is partially genetic. During the last two decades, evidence of interaction
between traditional lipid metabolism candidate genes, dietary components, and
cardiovascular risk factors (i.e., plasma lipids and obesity-related traits) has been
slowly gaining momentum. Several experimental approaches have been used for this
research, including observational and interventional studies. However, the replication
among studies is still limited, mostly driven by less-than-optimal experimental
designs. Current studies are starting to incorporate experimental approaches (i.e.,
increased coverage of the genetic variation at each locus) that could provide more
solid results. However, other limitations, such as small populations that limit the
statistical power especially to examine higher-level interactions (genegenediet
behavioral factors) have yet to be incorporated in nutrigenetic studies. Moreover,
the current quality of the dietary and behavioral information adds another major
33
limitation, specifically for observational studies. These needs can be met through
the collaboration of experts in the different fields involved, ranging from basic
scientists to computational biologists and behavioral scientists. Once more solid
evidence is achieved, nutrigenomics could be applied toward primary prevention
and treatment of cardiovascular diseases in the general population and subjects at
high risk.
The effect of dietary changes on cardiovascular disease (CVD)-relevant traits
(i.e., plasma lipid measures, glucose levels, body weight, and blood pressure) differs
significantly among individuals [1]. Some individuals appear to be relatively insen-
sitive (hyporesponders) to dietary intervention, whereas others display enhanced
sensitivity (hyperresponders) [1]. This phenomenon has been more extensively
researched in relation to changes in dietary fat and plasma lipid concentrations for
the prevention of CVD compared to other pathological conditions. Although common
knowledge associates low-fat diets with reductions in total and plasma low-density
lipoprotein cholesterol (LDL-C), the clinical evidence shows dramatic interindividual
differences in response and may be one of the underlying causes of the limited
success of dietary recommendations in the prevention of diseases observed by recent
randomized clinical trials [2].
A growing body of data supports the hypothesis that the interindividual vari-
ability in response to dietary modification is determined by genetic factors [3].
Indirect evidence comes from the general observation that the phenotypic response
to diet is determined partly by the baseline value of the phenotype, which is itself
affected by genetic factors [1]. Some of the main challenges are the following:
(1) how to uncover and elucidate the many potential genediet interactions; (2) how
to include in the experimental designs the ability to test for genegenediet inter-
actions; (3) how to investigate more complex and realistic scenarios, including the
contributions of other behavioral factors (i.e., physical activity, smoking, socioeco-
nomic status, traditions, and social environment); and (4) how to collect reliable and
standardized information of relevant covariates [4]
Several studies have found specific candidate genes to be associated with the
variability in response of LDL-C levels in response to changes in dietary fat, but so
far, the humbling reality is that these findings have been highly inconsistent. These
conflicting outcomes reflect both the complexity of the mechanisms involved in
dietary responses and the limitations of the current experimental designs used to
address these problems, some of which are due to the challenges listed in the previous
paragraph. In addition, we need to keep in mind that the practical translation to the
public of any nutrigenetic finding needs to be supported by a more holistic approach
to health rather than applying the tunnel-vision approach of focusing exclusively on
the changes in one particular risk factor. To this effect, it is true that low-fat diets
can result in reduced plasma LDL-C levels, but also in lowering of HDL-C and
increased triacylglycerol (TAG) concentrations, which may be particularly harmful
for some persons. For example, it has been shown that individuals with a predom-
inance of small, dense LDL particles (subclass pattern B), a phenotype that is
associated with an increased risk of coronary heart disease, benefit more from a
low-fat diet [5] than do those with the subclass pattern A (larger LDL). A significant
proportion of the latter group unexpectedly exhibited a more atherogenic pattern B
Nutrition, Genomics, and Cardiovascular Disease Risk 35
subclass after consuming a low-fat diet. Intervention studies are increasingly focus-
ing on the interindividual differences in response to diet rather than on the mean
effect analyzed for a population. Moreover, new evidence indicates that the variabil-
ity in response is an intrinsic characteristic of the individual, rather than being the
result of differing dietary compliance with the experimental protocols. Jacobs et al.
[6] found that individual TAG responses to a high-fat or to a low-fat diet were vastly
different, suggesting that many patients with hypertriglyceridemia are not treated
optimally if general advice for either a low-fat or a high-fat diet is given. Under-
standing the mechanisms driving this variation will not only allow us to better grasp
the biology of dietary response but also to identify individuals who can benefit from
a particular dietary intervention.
4.2 HOW NUTRIENTS COMMUNICATE WITH GENES

There has been an ongoing debate about the correct term that would encompass the
overarching goal of understanding the interaction between nutrients and genes and
how different individuals may modulate the interaction according to their genome.
Thus, the terms nutrigenomics, nutrigenetics, nutritional genomics, among others,
have been used in the literature, and the ontology of each term has not been
conclusively determined. In this work we will adhere to the convention used in the
pharmacological field that we have previously described [7]. Therefore, before
presenting some of the current nutrigenetic evidence in the area of lipid metabolism
and CVD, it is helpful to gain an understanding of how nutrients and other chemicals
in the diet may influence gene expression and drive genediet interactions, which
covers, according to the previously indicated convention, the subject proper of
nutrigenomics, which seeks to understand genediet interactions in the context of
the total genetic makeup of each individual. Past technological limitations restricted
investigators to a piecemeal approach: one gene, one gene product, and one nutrient
at a time. Conceptual and technological advances are changing the playing field.
For the first time, researchers can cast a wide net in the form of microarrays that
can potentially capture the information about each one of the genes expressed in a
specific cell or tissue of interest. Despite these advances, the challenges are not
trivial, given the chemical complexity of food and our incomplete knowledge about
the various bioactive components present in food grown in different climates at
different times of the year. Moreover, our ability to carry out mechanistic studies in
humans gets impaired by (1) our inability to assay gene expression in the most
appropriate target tissues in humans and (2) the still-prohibitive cost of performing
experiments using adequate number of samples, assay replicates, and time points or
intervention levels.
Regulation of expression of genes involved in fatty acid metabolism may occur
when a dietary fat or metabolite binds to and activates specific fatty acid transcription
factors. These dietary chemical-regulated transcription factors are members of the
nuclear receptor superfamily. This gene family consists of 48 mammalian transcrip-
tion factors that regulate nearly all aspects of development, inflammation, and metab-
olism. Two subclasses, the peroxisome proliferator-activated receptors (PPARs) and
liver X receptors (LXRs), are lipid-sensing receptors that have critical roles in lipid
and glucose metabolism [8]. PPARs are among the best-studied fatty-acid-regulated
nuclear receptors [9]. After uptake into target cells, a subset of them are transported
to the nucleus in association with fatty acid binding proteins (FABPs), which facil-
itates their interaction with PPARs. Several PPAR subtypes have been described.
PPAR-alpha (PPARA) plays a key role in lipid oxidation and inflammation, whereas
PPAR-gamma (PPARG) is involved in cell (adipocyte) differentiation, glucose lipid
storage, and inflammation. PPAR-delta (PPARD, also known as PPAR-beta) may
play an important role in development, lipid metabolism, and inflammation [10].
Many of the previously published nutrigenetic [i.e., single-gene/single-nucleotide
polymorphisms (SNPs)] studies focused on genes that are the subject of regulation
by PPARs and other nuclear receptors [11]. Polymorphisms in promoter regions of
these genes may disrupt, or at least alter, the communication with these transcription
factors, which would have significant consequences in a persons response to dietary
factors that are ligands (i.e., PUFAs) of the transcription factors. It is also obvious
that polymorphisms within the transcription factors themselves will have a significant
impact on the way each one of us responds to dietary components [1214].
The quantitative evidence for genediet interactions between common SNPs at
candidate genes and dietary factors related to lipid metabolism is increasing. How-
ever, we reiterate that caution is needed before applying these results to clinical
practice for three primary reasons: (1) the meaning of statistically significant
results is subject to differing interpretations and often depends on the study design,
(2) many initial genenutrientphenotypes associations were not replicated in sub-
sequent studies, and (3) gene variations may influence phenotypes differently in
individuals from different ancestral backgrounds owing to genegene (epistatic)
interactions.
4.3 OBSERVATIONAL STUDIES: THE CURRENT

WORKHORSE OF NUTRIGENETICS
Observational studies have the advantage of allowing the collection of data on large
numbers of subjects and the ability to estimate long-term dietary habits. However,
the level of evidence of the results obtained from these studies has traditionally been
considered to be lower than that of experimental studies, and this topic has been
extensively reviewed [7,15].
The median population size for recent observational studies is approximately
below 1000. This sample size may be informative for traditional single geno-
typephenotype association studies but, considering the higher measurement error
of dietary intake in comparison with experimental studies, it may not have enough
statistical power to address properly the complexity of geneenvironment interac-
tions, and the replication of results is still very low. The main reasons for the
disparities are not much different from those already indicated previously; however,
another factor to consider is that the interpretation of the studies may be much
affected by differences in the analytic strategies and study populations. Investigators
have reported different results for association of the Pro12 allele and T2DM by
PUFA intake when they used different modeling methods (or statistical packages).
If such differences are found within the same population, it is expected that different
modeling methods may have an even more pronounced effect in diverse population
groups. So, it is not surprising that when different study populations are combined with
different analytical techniques, there are so many inconsistencies in the literature [14].
So far, most of the initial descriptions of genediet interactions have been based
on observational studies; regardless of the apparent solidity of any initial reported
finding, every genediet interaction needs to go through a series of verification steps,
which should at least include (1) its replication on several other properly sized
populations, (2) its testing on an interventional study in people preselected on the
basis of their genotypes for the SNP under consideration, and (3) additional in vitro
and ex vivo tests of functionality to provide the mechanistic knowledge that will
support the reported statistical interactions. Moreover, genotype-nutrient-phenotype
analyses may be improved by determining ancestral backgrounds of each study
participant. These additional data are necessary because SNPs may be expressed
differently among individuals of differing ethnicities because of varying gene-gene
and gene-nutrient interactions.
The enumeration of those studies focusing on the interaction between genetic
variants at candidate genes, nutrients, and CVD risk factors have been the subject
of recent reviews [7,15,16], and we will not reproduce their contents in this chapter.
Rather, we will focus on capturing the current trends in nutrigenetic research.
We will begin by focusing on one of the rising genes from the last few years:
apolipoprotein A5 (APOA5). Totally unknown until 2001, APOA5 has become a
major player in lipoprotein metabolism and specifically in triglyceride-rich lipopro-
tein (TRL) metabolism [17]. Five common APOA5 SNPs have been reported in
several populations: 1131T > C, 3A > G, 56C > G IVS3 + 476G > A, and 1259T >
C, and these APOA5 gene variants have been associated with increased TAG con-
centrations [1821]. With the exception of the 56C > G SNP, the SNPs are reported
to be in significant linkage disequilibrium [22]. Despite the relative good consistency
with plasma TAG concentrations, the association between APOA5 SNPs (or haplo-
types) and CVD risk is still controversial [19,23,24]. Learning from previous expe-
rience, it is reasonable to implicate gene environment interactions modulating the
association with CVD risk factors as well as with the disease risk itself. Consistent
with this hypothesis, several intriguing interactions are beginning to emerge. In the
Framingham Heart Study, we have reported a genediet interaction between the
APOA5 gene variation and the polyunsaturated fatty acids (PUFA) in relation to
plasma lipid concentrations and lipoprotein particle size [25]. Furthermore, we have
demonstrated that obesity modulates the effect of the APOA5 gene variation in
carotid intimal medial thickness (IMT), a surrogate measure of atherosclerosis bur-
den. This association remained significant even after the adjustment for TG [26].
More recently, we have found a genediet interaction between the APOA5 gene and
total fat intake in relation to body mass index (BMI), overweight, and obesity risk
[27]. The homogeneity of this interaction was detected in both men and women and
followed a dose-response relationship. In comparison with wild-type subjects, the
carriers of the variant C allele at the 1131T > C SNP appeared to be protected
from an increase in body weight when consuming more dietary fat. Interactions
between BMI and APOA5 had been previously reported by Aberle et al. [28]. More-
over, some interesting peculiarities are emerging that deserve further examination.
One relates to the specificity of the interactions in terms of the location of the APOA5
tagging SNPs, one in the promoter region (1131T > C) and the other affecting the
primary structure of the protein (56C > G). We have reported that the complex
interactions between APOA5 SNPs, dietary intake, BMI, and IMT appear to be
related to the promoter polymorphism, whereas the APOA5 56C > G SNP appears
to associate with the different traits independently of environmental triggers [2527].
The second aspect to consider is the specificity of the interactions in terms of dietary
fat, being different for saturated, monounsaturated, and polyunsaturated n-6 and n-3.
We investigated the interaction between these APOA5 SNPs and dietary fat in
determining plasma fasting TGs, remnant-like particle (RLP) concentrations, and
lipoprotein particle size in Framingham Heart Study participants. Significant
genediet interactions between the 1131T > C SNP and PUFA intake were found
in determining fasting TGs, RLP concentrations, and lipoprotein particle size, but
these interactions were not found for the 56C > G polymorphism [25]. The 1131C
allele was associated with higher fasting TGs and RLP concentrations in only the
subjects consuming a high-PUFA diet (>6% of total energy), regardless of gender.
Interestingly, when we further analyzed the effects of n-6 and n-3 fatty acids, we found
that the PUFAAPOA5 interactions were specific for dietary n-6 fatty acids. Thus,
higher n-6 (but not n-3) PUFA intake increased fasting TGs, RLP concentrations, and
VLDL size and decreased LDL size in APOA5 1131C carriers, suggesting that n-6
PUFA-rich diets are related to a more atherogenic lipid profile in these subjects. This
could shed some light into the data coming from postprandial studies [2933] that will
be discussed in the following text. Similar specificities in terms of SNP, exclusive for
the 1131T > C, and dietary fat, primarily monounsaturated fat, were found in relation
to the interaction between fat intake and BMI within the Framingham Study [27].
These observations suggest some new hypotheses regarding some of the
observed geographical differences in CVD risk as well as the changes in risk
observed following Westernization of lifestyles. In this regard, it is important to pay
attention to the differences in allele frequencies of the APOA5 SNPs identified so
far. These frequencies appear to be higher in Asian compared to White populations
[18]. If the interactions observed in Framingham apply to other geographical areas,
we should expect that an increased fraction of the Asian population will be placed
at higher CVD risk because of the current trend toward higher dietary fat intakes,
especially if they are driven by increases in PUFA n-6. Moreover, we found in the
Framingham Study that some individuals (carriers of the C allele at the 1131T >
C SNP) may consume a high percentage of fat, especially MUFA, in their habitual
diet without adversely affecting their BMI and their risk of overweight and obesity [27].
The studies described earlier provide evidence of the current transition from the
traditional approach of studying one SNP per gene to a more comprehensive exam-
ination using multiple SNPs and haplotypes. Moreover, they represent a sample of
the exploding interest for obesity-related phenotypes and their complications. Given
the current public attention to this problem, it is not surprising that increasing number
of researchers aim their focus on the investigation of genediet and obesity-related
parameters.
Some of the recent reports examine obvious but previously ignored candidate
genes. This is the case with the loci encoding the carnitine palmitoyltransferase I
(CPT1), a key enzyme in beta-oxidation of fatty acids. Robitaille et al. [34] screened
the genes for CPT1A (11q13) and CPT1B (22qter) for common mutations in a
French-Canadian population and selected nine variants within CPT1B and one
variant within CPT1A for further analyses in relation with obesity based on the
minor allele frequency (>10%) or on potential functional impact. A significant
association between obesity phenotypes (BMI, weight, and waist girth) and CPT1B
c.282-18C > T and p.E531K variants was observed. When subjects were divided
into six groups according to p.E531K genotypes and further on the basis of fat using
the median value as a cutoff point (34.4% of energy), BMI, weight, and waist girth
were higher in E531/K531 on a high-fat diet compared to E531/K531 subjects under
a low-fat diet. There was no difference among E531/E531 and K531/K531. This
interaction is complex to interpret as the statistically significant interaction was
driven by heterozygotes, and no differences were observed for the homozygous at
each of the extremes. Similar results were obtained with the CPT1A p.A275T variant
as BMI and waist girth were higher in A275/A275 on a high-fat diet compared to
A275/A275 subjects on a low-fat diet. Among carriers of the T275 allele, obesity
indices were not affected by fat intake. These findings suggest that obesity-related
traits might be modulated by an interaction between CPT1 variants and fat intake,
but their interpretation is not straightforward.
Other studies were initially designed to examine the genetics of obesity and to
identify significant and valuable nutritional interactions. This was the case with
NUGENOB [35]. This study examined 42 SNPs at 26 candidate genes for obesity
in 549 adult obese women recruited from eight European centers in a case-only
study. The nutritional variables assessed in this study were the dietary fiber intake
(grams per day), the ratio of dietary polyunsaturated fat to saturated fat (P:S ratio),
and the percentage of energy derived from fat in the diet as calculated from a weighed
3-d food record (%E). Some weak, confirmatory evidence for genotype-by-nutrient
interactions in obesity was detected in this case-only study [i.e., hepatic lipase gene
(LIPC); adiponectin (ADIPOQ) and PPARG3]. Although this design was initially
considered useful for detecting interactions, the authors state that few genotype-by-
nutrient interactions have been suggested in obese European women after completing
the initial goals of the study.
Our own studies focused on the gene for human perilipin (PLIN), which is on
chromosome 15q26, in the region of a linkage locus for diabetes, hypertriglyceridemia,
and obesity. In a study to examine the regulation of lipolysis in obese and nonobese
women [36], it was found that obese women exhibited two- to fourfold higher basal
rates of lipolysis in fat cells obtained through a biopsy in the abdominal subcutaneous
area. Similar differences in lipolysis rates were observed following noradrenaline
induction. It was also noted that adipocyte perilipin content was reduced in obese
women. This study also examined the effect of a polymorphism (11482G > A) in
intron 6 of the perilipin locus in relation to perilipin content as well as the rates of
lipolysis. Adipose tissue from women who were homozygous for the A allele at
position 11482 exhibited 50 to 100% higher rates of lipolysis and 80% lower perilipin
content. Following this study, several polymorphisms at the perilipin locus have been
examined for potential associations with BMI or obesity risk. We have examined
four SNPs (PLIN1: 6209T > C, PLIN4:11482G > A, PLIN5: 13041A > G, and
PLIN6:14995A > T) in several geographically and ethnically diverse populations to

determine the patterns of linkage disequilibrium in the various populations and the
associations between these polymorphisms with obesity-related traits. Our first report
included 1589 white patient samples from a general Spanish population [37]. The
aforementioned SNPs were common, with minor allele frequencies ranging from
0.26 to 0.38. Two of these (6209T > C and 11482G > A) were in strong linkage
disequilibrium (D 0.96). The presence of one or both of these SNPs was associated
with lower BMI and reduced risk of obesity in women. A significant genegender
interaction was observed, however, and the effect was seen only in women. In
addition, the 11482G > A polymorphism also showed significant associations with
fasting glucose and triglyceride concentrations. Conversely, a trend was observed
for the other two SNPs (13041A > G and 14995A > T) to be associated with increased
BMI and risk of obesity. These same polymorphisms were then examined in 734
white patients (373 men and 361 women) who were attending a residential lifestyle
modification program in California [38]. As with the study in Spain, the effects were
seen primarily in women. As in Spanish women, the 6209T > C and the 11482G >
A polymorphisms were associated with lower BMI. These associations, however,
did not reach statistical significance. Instead, the presence of the rare alleles at
positions 13041 and 14995 were associated with increased BMI. At this stage, it
appeared that the rare alleles at positions 6209 and 11482 at the 5 end of the perilipin
locus had opposite effects on BMI when compared with the minor alleles at positions
13041 and 14995, at the 3 end of the perilipin locus. Whereas the presence of the
former was associated with reduced BMI, the latter were associated with increased
BMI. In line with this notion, the haplotype that included the common alleles at the
6209T > C, 11482G > A SNPs and the rare alleles at the 13041A > G and 14995A
> T SNPs had the greatest effect on obesity risk in the population in California. We
subsequently proceeded to examine these SNPs in a multiethnic population living
in Singapore comprising 2763 Chinese, 726 Malays, and 598 Asian Indians [39].
One interesting finding was that the intragenic linkage disequilibrium structure of
the perilipin locus differed between populations that were white compared with
populations of Asian extraction. Specifically, in whites, the 11482G > A polymor-
phism was in strong positive linkage disequilibrium with the 6209T > C polymor-
phism and negative linkage disequilibrium with the 13041A > G and 14995 A > T
polymorphisms. In Chinese, Malays, and Asian Indians living in Singapore, the
situation was reversed, with the 11482G > A polymorphism in negative linkage
disequilibrium with the 6209T > C polymorphisms and in positive linkage disequi-
librium with the 13041A > G and the 14995A > T polymorphisms. Given the
differences in patterns of linkage disequilibrium observed, one would expect that
the rare allele for the 11482G > A polymorphism, which was associated with lower
levels of obesity in the white population, would be associated with increased levels
of obesity in the Asian populations. In fact, this is exactly what was observed.
Moreover, as in the two previous studies, the significant associations were observed
only in females. They were only present in Malays and Asian Indians, however, not
in Chinese. It is also worth noting that another study including 1120 individuals
from France failed to find a statistically significant association between the 11482G
> A polymorphism [40].
It is unclear at this time why no association was observed in French and Chinese
ethnic groups despite being reasonably powered to detect an association, if present.
If indeed complex traits such as obesity are a consequence of interactions between
genetic and environmental factors [41], such geneenvironment interactions could
explain the discrepant findings among populations. It could simply be that the French
and Chinese are exposed to different environments when compared with whites in
Spain or California, and Malays and Asian Indians in Singapore. Do such gene
environment interactions operate in relation to the association between polymor-
phisms at the perilipin locus and obesity? In fact, the interaction between polymorphisms
at this locus and gender has already been described in whites and some of the Asian
groups. Recently, one study [42] has reported that individuals carrying the A allele
for the 11482G > A polymorphism are resistant to weight loss with caloric restriction.
Another [43] reported resistance to weight gain following treatment with the PPARG
agonist rosiglitozone in individuals carrying this polymorphism. Therefore, over-
weight and obese individuals carrying this allele may not benefit from traditional
dietary approaches to losing weight, and alternative approaches may be needed. In
this regard, an interesting observation was made regarding the potential regulatory
effects of a plant extract on perilipin, hormone-sensitive lipase, and other parameters
of lipid metabolism in obese women [44].
Up to this point, we have dealt primarily with associations between polymor-
phisms at the perilipin locus and their associations with the degree of obesity per
se, as opposed to the downstream effects of obesity, such as insulin resistance. Jang
et al. [45] studied 177 obese Korean men and women who were treated with caloric
restriction (300 kcal per day) and found that the presence of the A11482 allele was
associated with greater weight loss in response to caloric restriction and that this
weight loss was primarily in the visceral fat compartment. Because visceral obesity
is generally associated with increased plasma-free fatty acid concentrations, one
may expect a greater reduction in plasma-free fatty acids in these individuals. Instead,
these individuals experienced an increase in plasma-free fatty acids despite a greater
reduction in visceral fat mass.
It is important to remember that fatty acids serve not only as a source of energy
but also as signaling molecules. Free fatty acids result in impaired glucose homeostasis
and may increase the risk of type 2 diabetes mellitus. To test the hypothesis that
dietary fat may interact with polymorphisms at the perilipin locus to modulate dia-
betes-related traits, we reexamined data from the Singapore population, taking into
consideration dietary macronutrient intake [46]. We found evidence of an interaction
between dietary fat (specifically saturated fat) intake, polymorphisms at the perilipin
locus (11482G > A and 114995A > T), and insulin resistance. Greater intake of energy
from saturated fats was associated with increased insulin resistance among individuals
who were homozygous for the rare alleles but not individuals who carried the common
allele. This effect was independent of BMI and consistent with the associations with
anthropometric measures; these interactions were observed only in women.
Overall, observational studies have provided substantial support to the concept of
genediet interactions modulating cardiovascular risk factors; however, very few find-
ings have been consistently replicated; among several concerns related to this approach,
we want to underscore the need for more reliable assessment of dietary intake.
4.4 INTERVENTIONAL STUDIES IN NUTRIGENETICS

Interventional studies in which subjects receive a controlled dietary intake provide
the best approach to conducting genenutrientphenotype interaction studies.
However, well-controlled feeding studies have several important logistical limita-
tions, most importantly, the small number of participants and the brief duration
of the interventions. Scores of interventional studies examining genediet inter-
actions on different parameters of lipid metabolism have been published. However,
similar to the situation described for observational studies, the level of replication
among studies analyzing the same genetic variation tends to be low. The lack of
replication is most likely due to the different characteristics (ethnicity, physical
condition, age, and lifestyle differences) of study subjects, length of intervention,
sample size, and heterogeneity in the design. In an initial systematic review,
Masson et al. [47] identified 74 relevant articles, including dietary intervention
studies, in which the lipid and lipoprotein response to diet in different genotype
groups were measured and 17 reviews on genediet interactions. After a compar-
ative analysis of the individual findings, they concluded that there is evidence to
suggest that (1) variations in the APOA1, APOA4, APOB, and APOE genes con-
tribute to the heterogeneity in the lipid response to dietary intervention and
(2) all of these genes are regulated directly or indirectly by PPARA or other nuclear
receptors. However, the evidence suggested by Masson et al. in relation to the
foregoing genes comes from meta-analyses of the published data and described
the average effect. The absence of total consistency of results among individual
studies should be noted.
More recently, we [7,15,16] reviewed this topic extensively and included addi-
tional studies reported after 2002. The median for the sample sizes in these more
recent studies was in the range of 60 subjects. These small sample sizes highlight
one of the bases for the lack of reproducibility: the statistical power is very low,
specially when the allele groups have not been balanced by a priori selection of the
study subjects. In addition, the composition of the dietary intervention in these
studies varied considerably. We propose that the design of future intervention studies
be standardized for key dietary intake variables and phenotype measurements. A
minimum set of variables would include patients physical and genetic characteris-
tics, medications, composition and length of the dietary treatment, and sample size.
Such standardization would allow better comparison among studies and the possi-
bility of conducting meta-analyses, which is not possible under current experimental
conditions.
4.5 GENEDIET INTERACTIONS

IN THE POSTPRANDIAL STATE
Human beings living in industrialized societies spend most of the waking hours in
a nonfasting state because of meal consumption patterns and the amounts of food
ingested. Postprandial lipemia, characterized by a rise in TAG after eating, is a
dynamic, non-steady-state condition [48]. Over 25 years ago, Zilversmit [49] proposed
that atherogenesis was a postprandial phenomenon because high concentrations of

lipoproteins and their remnants following food ingestion could deposit into the
arterial wall and accumulate in atheromatous plaques. Several studies have investi-
gated the potential interaction between some polymorphisms in candidate genes and
diet on postprandial lipids (for a review see Reference 7). In postprandial studies,
subjects usually receive a fat-loading test meal that has differing compositions
depending on the nutrients to be tested. After the test meal, blood samples are taken
to measure postprandial lipids to compare with preprandial levels [48]. Consistency
among studies is still very low, and replication of findings is a major necessity.
Postprandial studies often have low numbers of subjects (usually <50), with the
added complexity that their designs may add even more bias than those seen in other
experimental approaches [50].
Postprandial studies have also shown the specificity of the different dietary fat
as drivers of genediet interactions. Thus, the slightly different outcomes regarding
the interaction between postprandial metabolism and the APOA5 SNPs [2933] made
us hypothesize that an impaired postprandial response in carriers of the C allele at
the 1131T > C SNP could be triggered by PUFA n-6 rather than by saturated fatty
acids.
This type of experimental approach faces some logistical problems, including
standardization of the protocols (type of meals used for the fat load and time of
sample collection). Moreover, it requires keeping the participating subjects for sev-
eral hours at the study site. Conversely, a major advantage is that significant infor-
mation might be gained in the space of 1 d of intervention vs. the weeks or months
required to examine changes in lipids in the fasting state.
4.6 CONCLUSIONS AND STEPS FORWARD

Despite the excitement due to an increasing number of findings related to nutri-
tional genomics, progress in the field is hampered by the inadequacy of the current
experimental approaches to efficiently deal with the biological complexity of the
phenotypes, the complexity of dietary intakes, differing genetic background
among participants, and the limitations of low statistical power of the studies.
We and others have proposed that only a comprehensive, international nutritional
genomics approach [51,52] will yield short- and long-term benefits to human
health by (1) revealing novel nutrientgene interactions, (2) developing new
diagnostic tests for adverse responses to diets, (3) identifying specific populations
with special nutrient needs, (4) improving the consistency of current definitions
and methodology related to dietary assessment, and (5) providing the information
for developing more nutritious plant and animal foods and food formulations that
promote health and prevent, mitigate, or cure disease. Achieving these goals will
require extensive dialogue between scientists and the public about the nutritional
needs of the individual vs. groups, local food availability and customs, analysis
and understanding of genetic differences among individuals and populations, and
serious commitment of funds from the public and private sectors. Nutritional
genomics researchers are seeking collaborations of scientists, scholars, and
policymakers to maximize the collective impact on global poverty and health by

advancing our knowledge of how genetics and nutrition can promote health or
cause disease.
Although the current evidence from both experimental and observational nutri-
genetics studies is not enough to start making specific personalized nutritional
recommendations based on genetic information, there are a large number of examples
of common SNPs modulating the individual response to diet as proof of concept of
how genediet interactions can influence lipid metabolism. It is critical that these
preliminary studies go through further replication and that subsequent studies be
properly designed with sufficient statistical power and careful attention to phenotype
and genotype. The many challenges that lie ahead are evident. This review has
examined the vast world of nutrigenetics and nutrigenomics only through the small
keyhole of lipid metabolism-related genes and dietary fat. These initial steps in
understanding nutrigenomics will likely lead to fundamental breakthroughs that will
both clarify todays mysteries and pave the way for clinical applications. However,
to arrive at the point where it is possible to assess the modulation by specific SNPs
of the effects of dietary interventions on lipid metabolism, well-designed, adequately
powered, and adequately interpreted randomized controlled studies (or their equiv-
alent) of greater duration than current studies are needed, with careful consideration
given to which patients to include in such studies. Moreover, research must also
investigate the potential mechanisms involved in the genediet interactions reported
by nutrigenetic studies [51]. These imperative needs can be achieved only through
the collaboration of experts in the different fields involved, which must include
nutrition professionals [52]. In addition, a number of important changes in the
provision of health care are needed to achieve the potential benefits associated with
this concept, including a teamwork approach, with greater integration among phy-
sicians and nutrition professionals. Once more experience is gained from patients
and individuals at high risk, these approaches could be applied toward primary
prevention of CVD.
ACKNOWLEDGMENTS
This work was supported by NIH grants HL72524, HL54776, and DK075030 and
by contracts 53-K06-5-10 and 58-1950-9-001 from the U.S. Department of Agri-
culture Research Service. (JMO), and contracts 53-K06-5-10 and 58-1950-9-001
from the U.S. Department of Agriculture Research Service (JMO), and the Spanish
Ministerio de Educacin y Ciencia (PR2006-0258) and the Spanish Ministerio de
Sanidad (CB06/03/0035).
SUGGESTED LECTURES/URLS
www.nugo.org.
http://www.athero.org/.
http://www.isnn.info/.
http://www.nutragenomics.com.
REFERENCES
1. Katan, M. B.; Beynen, A. C.; de Vries, J. H.; et al. Existence of consistent hypo- and
hyperresponders to dietary cholesterol in man. American journal of epidemiology
123: 221234; 1986.
2. Prentice, R. L.; Caan, B.; Chlebowski, R. T. et al. Low-fat dietary pattern and risk
of invasive breast cancer: the Womens Health Initiative Randomized Controlled
Dietary Modification Trial. Jama 295: 629642; 2006.
3. Loktionov, A. Common gene polymorphisms and nutrition: emerging links with
pathogenesis of multifactorial chronic diseases (review). Journal of nutritional bio-
chemistry 14: 426451; 2003.
4. Greenwood, D. C.; Gilthorpe, M. S.; Cade, J. E. The impact of imprecisely measured
covariates on estimating gene-environment interactions. BMC medical research meth-
odology 6: 21; 2006.
5. Krauss, R. M. Dietary and genetic effects on low-density lipoprotein heterogeneity.
Annual review of nutrition 21: 283295; 2001.
6. Jacobs, B.; De Angelis-Schierbaum, G.; Egert, S.; et al. Individual serum triglyceride
responses to high-fat and low-fat diets differ in men with modest and severe hyper-
triglyceridemia. The Journal of nutrition 134: 14001405; 2004.
7. Ordovas, J. M.; Corella, D. Nutritional genomics. Annual review of genomics and
human genetics 5: 71118; 2004.
8. Jump, D. B. Fatty acid regulation of gene transcription. Critical reviews in clinical
laboratory sciences 41: 4178; 2004.
9. Clarke, S. D. The multi-dimensional regulation of gene expression by fatty acids:
polyunsaturated fats as nutrient sensors. Current opinion in lipidology 15: 1318; 2004.
10. Tobin, J. F.; Freedman, L. P. Nuclear receptors as drug targets in metabolic diseases:
new approaches to therapy. Trends in endocrinology and metabolism: TEM 17:
284290; 2006.
11. Mandard, S.; Muller, M.; Kersten, S. Peroxisome proliferator-activated receptor alpha
target genes. Cellular and molecular life sciences 61: 393416; 2004.
12. Soriguer, F.; Morcillo, S.; Cardona, F. et al. Pro12Ala polymorphism of the PPARG2
gene is associated with type 2 diabetes mellitus and peripheral insulin sensitivity in a
population with a high intake of oleic acid. The journal of nutrition 136: 23252330; 2006.
13. Tai, E. S.; Corella, D.; Demissie, S.; et al. Polyunsaturated fatty acids interact with the
PPARA-L162V polymorphism to affect plasma triglyceride and apolipoprotein C-III con-
centrations in the Framingham Heart Study. The journal of nutrition 135: 397403; 2005.
14. Nelson, T. L.; Fingerlin, T. E.; Moss, L. K.; et al. Association of the peroxisome
proliferator-activated receptor gamma gene with type 2 diabetes mellitus varies by
physical activity among non-Hispanic whites from Colorado. Metabolism: clinical
and experimental 56: 388393; 2007.
15. Corella, D.; Ordovas, J. M. Single nucleotide polymorphisms that influence lipid metab-
olism: interaction with dietary factors. Annual review of nutrition 25: 341390; 2005.
16. Ordovas, J. M.; Corella, D. Genes, diet and plasma lipids: the evidence from obser-
vational studies. World review of nutrition and dietetics 93: 4176; 2004.
17. Pennacchio, L. A.; Olivier, M.; Hubacek, J. A.; et al. An apolipoprotein influencing
triglycerides in humans and mice revealed by comparative sequencing. Science 294:
169173; 2001.
18. Lai, C. Q.; Tai, E. S.; Tan, C. E.; et al. The APOA5 locus is a strong determinant of
plasma triglyceride concentrations across ethnic groups in Singapore. Journal of lipid
research 44: 23652373; 2003.
19. Lai, C. Q.; Demissie, S.; Cupples, L. A.; et al. Influence of the APOA5 locus on
plasma triglyceride, lipoprotein subclasses, and CVD risk in the Framingham Heart
Study. Journal of lipid research 45: 20962105; 2004.
20. Talmud, P. J.; Hawe, E.; Martin, S.; et al. Relative contribution of variation within
the APOC3/A4/A5 gene cluster in determining plasma triglycerides. Human molecular
genetics 11: 30393046; 2002.
21. Talmud, P. J. Rare APOA5 mutations-Clinical consequences, metabolic and functional
effects: an ENID review. Atherosclerosis; 2007.
22. Pennacchio, L. A.; Olivier, M.; Hubacek, J. A.; et al. Two independent apolipoprotein
A5 haplotypes influence human plasma triglyceride levels. Human molecular genetics
11: 30313038; 2002.
23. Talmud, P. J.; Martin, S.; Taskinen, M. R.; et al. APOA5 gene variants, lipoprotein
particle distribution, and progression of coronary heart disease: results from the
LOCAT study. Journal of lipid research 45: 750756; 2004.
24. Dallongeville, J.; Cottel, D.; Montaye, M.; et al. Impact of APOA5/A4/C3 genetic
polymorphisms on lipid variables and cardiovascular disease risk in French men.
International journal of cardiology 106: 152156; 2006.
25. Lai, C. Q.; Corella, D.; Demissie, S.; et al. Dietary intake of n-6 fatty acids modulates
effect of apolipoprotein A5 gene on plasma fasting triglycerides, remnant lipoprotein
concentrations, and lipoprotein particle size: the Framingham Heart Study. Circula-
tion 113: 20622070; 2006.
26. Elosua, R.; Ordovas, J. M.; Cupples, L. A. et al. Variants at the APOA5 locus,
association with carotid atherosclerosis, and modification by obesity: the Framingham
Study. Journal of lipid research 47: 990996; 2006.
27. Corella, D.; Lai, C. Q.; Demissie, S.; et al. APOA5 gene variation modulates the
effects of dietary fat intake on body mass index and obesity risk in the Framingham
Heart Study. Journal of molecular medicine 85: 119128; 2007.
28. Aberle, J.; Evans, D.; Beil, F. U.; et al. A polymorphism in the apolipoprotein A5
gene is associated with weight loss after short-term diet. Clinical genetics 68:
152154; 2005.
29. Martin, S.; Nicaud, V.; Humphries, S. E.; Talmud, P. J. Contribution of APOA5
gene variants to plasma triglyceride determination and to the response to both fat
and glucose tolerance challenges. Biochimica et biophysica acta 1637: 217225;
2003.
30. Kim, J. Y.; Kim, O. Y.; Koh, S. J.; et al. Comparison of low-fat meal and high-fat
meal on postprandial lipemic response in non-obese men according to the 1131T >
C polymorphism of the apolipoprotein A5 (APOA5) gene (randomized cross-over
design). Journal of the American College of Nutrition 25: 340347; 2006.
31. Jang, Y.; Kim, J. Y.; Kim, O. Y.; et al. The 1131T>C polymorphism in the
apolipoprotein A5 gene is associated with postprandial hypertriacylglycerolemia;
elevated small, dense LDL concentrations; and oxidative stress in nonobese Korean
men. The American journal of clinical nutrition 80: 832840; 2004.
32. Masana, L.; Ribalta, J.; Salazar, J.; et al. The apolipoprotein AV gene and diurnal
triglyceridaemia in normolipidaemic subjects. Clinical Chemistry and Laboratory
Medicine 41: 517521; 2003.
33. Moreno, R.; Perez-Jimenez, F.; Marin, C.; et al. A single nucleotide polymorphism
of the apolipoprotein A-V gene 1131T>C modulates postprandial lipoprotein metab-
olism. Atherosclerosis 189: 163168; 2006.
34. Robitaille, J.; Houde, A.; Lemieux, S.; et al. Variants within the muscle and liver
isoforms of the carnitine palmitoyltransferase I (CPT1) gene interact with fat intake
to modulate indices of obesity in French-Canadians. Journal of molecular medicine

85: 129137; 2007.
35. Santos, J. L.; Boutin, P.; Verdich, C.; et al. Genotype-by-nutrient interactions assessed
in European obese women: a case-only study. European journal of nutrition 45:
454462; 2006.
36. Mottagui-Tabar, S.; Ryden, M.; Lofgren, P.; Faulds, G.; Hoffstedt, J.; Brookes, A. J.;
Andersson, I.; Arner, P. Evidence for an important role of perilipin in the regulation
of human adipocyte lipolysis. Diabetologia 46: 789797; 2003.
37. Qi, L.; Corella, D.; Sorli, J. V.; et al. Genetic variation at the perilipin (PLIN) locus
is associated with obesity-related phenotypes in White women. Clinical genetics 66:
299310; 2004.
38. Qi, L.; Shen, H.; Larson, I.; et al. Gender-specific association of a perilipin gene
haplotype with obesity risk in a white population. Obesity research 12: 17581765;
2004.
39. Qi, L.; Tai, E. S.; Tan, C. E.; et al. Intragenic linkage disequilibrium structure of the
human perilipin gene (PLIN) and haplotype association with increased obesity risk
in a multiethnic Asian population. Journal of molecular medicine 83: 448456; 2005.
40. Meirhaeghe, A.; Thomas, S.; Ancot, F.; et al. Study of the impact of perilipin poly-
morphisms in a French population. Journal of negative results in biomedicine 5: 10;
2006.
41. Manolio, T. A.; Bailey-Wilson, J. E.; Collins, F. S. Genes, environment and the value
of prospective cohort studies. Nature reviews 7: 812820; 2006.
42. Corella, D.; Qi, L.; Sorli, J. V.; et al. Obese subjects carrying the 11482G > A poly-
morphism at the perilipin locus are resistant to weight loss after dietary energy restric-
tion. The journal of clinical endocrinology and metabolism 90: 51215126; 2005.
43. Kang, E. S.; Cha, B. S.; Kim, H. J.; et al. The 11482G > A polymorphism in the
perilipin gene is associated with weight gain with rosiglitazone treatment in type 2
diabetes. Diabetes care 29: 13201324; 2006.
44. Abidov, M. T.; del Rio, M. J.; Ramazanov, T. Z.; et al. Effects of Aralia mandshurica
and Engelhardtia chrysolepis extracts on some parameters of lipid metabolism in
women with nondiabetic obesity. Bulletin of experimental biology and medicine 141:
343346; 2006.
45. Jang, Y.; Kim, O. Y.; Lee, J. H.; et al. Genetic variation at the perilipin locus is
associated with changes in serum free fatty acids and abdominal fat following mild
weight loss. International journal of obesity (2005) 30: 16011608; 2006.
46. Corella, D.; Qi, L.; Tai, E. S.; et al. Perilipin gene variation determines higher
susceptibility to insulin resistance in Asian women when consuming a high-saturated
fat, low-carbohydrate diet. Diabetes care 29: 13131319; 2006.
47. Masson, L. F.; McNeill, G.; Avenell, A. Genetic variation and the lipid response to
dietary intervention: a systematic review. The American journal of clinical nutrition
77: 10981111; 2003.
48. Ordovas, J. M. Genetics, postprandial lipemia and obesity. Nutrition, metabolism,
and cardiovascular diseases 11: 118133; 2001.
49. Zilversmit, D. B. Atherogenesis: a postprandial phenomenon. Circulation 60:
473485; 1979.
50. Perez-Martinez, P.; Perez-Jimenez, F.; Bellido, C.; et al. A polymorphism exon 1
variant at the locus of the scavenger receptor class B type I (SCARB1) gene is
associated with differences in insulin sensitivity in healthy people during the consumption
of an olive oil-rich diet. The journal of clinical endocrinology and metabolism
90: 22972300; 2005.
51. van Ommen, B.; Stierum, R. Nutrigenomics: exploiting systems biology in the nutri-
tion and health arena. Current opinion in biotechnology 13: 517521; 2002.
52. Kaput, J.; Ordovas, J. M.; Ferguson, L.; et al. The case for strategic international
alliances to harness nutritional genomics for public and personal health. The British
journal of nutrition 94: 623632; 2005.
5 Nutrigenomics
and Chronic Inflammation
Lynnette R. Ferguson and Martin Philpott
CONTENTS
5.1 Introduction: the Nature of Chronic Inflammation .......................................49

5.2 Chronic Inflammation and Disease................................................................51
5.3 Genes and Inflammation-Prone Diseases ......................................................51
5.4 Other Genes That Impact Nutrient Requirements
in Chronic Inflammation................................................................................53
5.5 Dietary Modulation of Chronic Inflammation and Downstream Effects .....53
5.5.1 Recruitment and Growth Stimulation of Immune Cells ...................54
5.5.2 Dietary Antioxidants ..........................................................................55
5.5.3 Cell Signaling Molecules Mediating Inflammatory Response .........55
5.6 GeneDiet Interactions in Chronic Inflammation .........................................56
5.7 Future Perspectives ........................................................................................57
Key Readings...........................................................................................................57
References................................................................................................................57
5.1 INTRODUCTION: THE NATURE

OF CHRONIC INFLAMMATION
Increased understanding of the basic pathologies of several noncommunicable diseases
suggests that chronic inflammation may be a common underlying mechanism.
Chronic inflammation has been linked with asthma, Alzheimers disease, cardio-
vascular disease, and cancer and diabetes, among others. Inflammatory processes,
including the release of proinflammatory cytokines and formation of reactive oxygen
and nitrogen species, are an essential part of the immune system. Polymorphisms
in the genes affecting immune response will have predictable effects on susceptibility
to disease, while other genes affecting antioxidant effects, apoptosis, and cell signaling
will interact. Each of the stages involved in inflammation is likely to respond to a
different dietary intervention, and a complementary approach may be necessary to
optimally reduce symptoms. Understanding the genetic polymorphisms involved
may aid in developing more precise dietary strategies, in order to reduce chronic
inflammation and its long-term implications.
49
Inflammatory processes lead to the release of proinflammatory cytokines and

chemokines, growth and angiogenic factors, as well as formation of reactive oxygen
species (ROS) and reactive nitrogen species (RNS). They provide an essential part
of the immune systems response to a pathogen. Chronic inflammation is defined as
an inflammatory response of prolonged duration weeks, months, or even indef-
initely whose extended time course is provoked by persistence of the causative
stimulus to inflammation in the tissue(medweb.bham.ac.uk/http/depts/path/Teaching/
FOUNDAT/CHRONINF/chronic.html). This may occur through progression from
acute inflammation because of persistence of the original stimulus, or following a
number of episodes of acute inflammation.
Most cases of chronic inflammation are generally thought to occur through
infectious agents that are somehow protected from host defenses, either through
endogenous resistance mechanisms, or through being maintained in a protected
environment. However, nonliving foreign material that cannot be broken down will
have a similar effect. There are a number of examples of chronic inflammation being
stimulated through a normal tissue component, because of a defect in the regulation
of the bodys immune response to its own tissues. Examples here are autoimmune
diseases such as rheumatoid arthritis or lupus. There are also classes of diseases that
have an underlying component of chronic inflammation, such as the inflammatory
bowel diseases, Crohns disease (CD) or ulcerative colitis (UC), where there is
increasingly strong evidence for an interplay between genes and environment in the
etiology.
A schematic depicting some of the known factors in inflammation is provided
in Figure 5.1. ROS and NOS play a central role in enhancing the process, through
the activation of stress kinases (including JNK, MAPK, and p 38), or redox-sensitive
Pro-inflammatory genes
Infiltration/ Inflamed and mediators
recruitment tissue (COX-2, INOS, IL)
Inflammatory
immune cells
Pro-inflammatory mediators
(e.g., TNF-, IL-1)
Activation of upstream kinases

ROS, (MAPK, PKC, etc)
RNS
Damage to DNA, protein or Activation of transcription factors

lipids (NF-B etc)
FIGURE 5.1 Overview of the process of chronic inflammation, resulting in damage to host
tissues and activation of proinflammatory genes.
Nutrigenomics and Chronic Inflammation 51
transcription factors such as nuclear factor-B (NF-B) and AP-1. Although ROS
and RNS may act directly to produce damage to DNA, lipid, or proteins, they also
commonly act through secondary products such as the peroxidation product, 4-
hydroxy-2-nonenal. NF-B, a family of ubiquitously expressed transcription factors,
regulates the expression of genes encoding proteins essential for stress response,
maintenance of intercellular communication, cell cycle control, and apoptosis.
5.2 CHRONIC INFLAMMATION AND DISEASE

Chronic inflammation may be a common underlying mechanism of several noncom-
municable diseases, including Alzheimers disease (AD), cardiovascular disease,
various types of cancer, diabetes, and rheumatoid arthritis.
Cancer may be the best studied of these examples, because chronic inflammation
has long been considered to be a significant risk factor for a variety of epithelial
cancers, including cancers of the prostate, cervix, esophagus, stomach, liver, colon,
pancreas, and bladder.1,2 A primary cause may be ROS and RNS, leading to DNA
damage and mutation, hence contributing to cancer initiation. However, the release
of growth factors and of proinflammatory cytokines from activated macrophages
will also enhance cell growth and promote later stages of the process. NF-B also
acts in the tumor microenvironment, enhancing the survival of premalignant epithe-
lial cells, thereby promoting tumor growth and vascularization.35
The release of cytokines not only drives inflammation, but may also lead to
tissue destruction in diseases such as rheumatoid arthritis, psoriasis, chronic obstruc-
tive pulmonary disease, and atherosclerosis.6 Two key cytokines, which have been
targeted clinically for treatment of such diseases, are tumor necrosis factor- (TNF-
) and interleukin-1 (IL-1). Toll-like receptors (TLRs), which recognize self vs.
nonself molecular patterns, may also be pivotal in the regulation of cytokine-driven
inflammation. Although there is still speculation as to whether targeting these or
their downstream signaling pathways will be beneficial therapeutically, the action
of each of these is likely to be impacted by nutrition.
The most common form of dementia is Alzheimers disease (AD), whose pathol-
ogy is characterized by activated microglia clustered around aggregated amyloid
beta (Abeta1-42) peptide-containing plaques. This 42-mer peptide is associated with
high levels of ROS, and is derived from the beta-amyloid precursor protein (APP).
5.3 GENES AND INFLAMMATION-PRONE DISEASES

Polymorphisms in the genes affecting immune response will have predictable effects
on susceptibility to disease, while other genes affecting cell signaling, apoptosis,
and antioxidant effects will interact. A review by Yamada and Ymamoto 7 summarizes
some of the recent work, focusing on the postulated function of the genes and level
of evidence of involvement in disease.
In keeping with the recognized importance of cytokines, a considerable number
of genetic polymorphisms in genes associated with immune processes will be of
obvious significance to the development and progression of inflammation-related
disease. The bulk of functional polymorphisms thus far identified occur in upstream
promoter sequences, leading to changes in the expression of the respective pro- or
anti-inflammatory gene products. Interleukin Genetics is an example of a company
that focuses attention on polymorphisms in a single gene. As pointed out by Kornman,8
variants in the genes for the family of interleukin 1 (IL-1) proteins often lead to
modified inflammatory response, and may result in increased susceptibility to a range
of diseases, including coronary artery disease, AD, gastric cancer, and periodontitis.
Similarly, they may modify the onset and progression of the deterioration associated
with aging.
A number of other such loci have also been associated with multiple disorders.
For example, polymorphisms in the cytotoxic T-lymphocyte-associated 4 (CTLA-4)
gene have been associated with autoimmune thyroidosis9 and type I diabetes melli-
tus,10 and have been suggested as a general susceptibility factor in autoimmune
disease.11 Caspase recruitment domain-containing protein 15 (CARD15) functions
as an intracellular receptor for bacteria, signaling the presence of pathogens by the
production of NF-B. Ogura et al.12 provided evidence for certain polymorphisms
in CARD15 significantly increasing susceptibility to Crohns disease in humans,
and others have more recently shown a key involvement in colorectal cancer.13 The
products of the HLA (human leukocyte antigen; HLA) family of genes play a critical
role in regulation of the immune response.14 There are recognized associations
between several different HLA polymorphisms and a range of inflammatory diseases.
Thus, individuals carrying the HLA-DQA1*0501, DQB1*0201 genotype show a
more than 200-fold increased risk of developing celiac disease, whereas other related
factors may be involved in the development of malignant lymphoma in individuals
who already have celiac disease. Similar considerations pertain to the development
of rheumatoid arthritis.14
The recognition of ROS involvement in chronic inflammation and tissue destruc-
tion suggests that either genes or nutrients involved in regulating antioxidant
response will play a significant role in modulating the development and progression
of inflammation. At a cellular level, redox homeostasis is generally achieved through
the balance of ROS generation with mechanisms for their removal, including dietary
antioxidants and intracellular antioxidant defense mechanisms. Chen and Kunsch15
describe the antioxidant response element (ARE) as one of a class of cytoprotective
regulatory genes. This element is located in the regulatory regions of a number of
genes, including phase II detoxification enzymes proteins such as glutathione-S-
transferases and gamma-glutamylcysteine synthase. It is bound by the transcription
factor Nrf2, leading to activation of the expression of other leucine-zipper-containing
transcription factors that have profound effects on immune and inflammatory
responses. Factors that decrease expression of these genes enhance inflammatory
responses to ROS, and also enhance the progression and development of autoimmune
disease.
There are several polymorphisms in the PS-1 and APP genes that increase
production of Abeta (1-42), and are likely to be the major causes of early onset
familial AD.16 High levels of ROS and RNS, and increased vulnerability to oxidative
stress appear to be a consequence of elevated levels of Abeta (1-42), and may also
contribute significantly to neuronal apoptosis and death in familial early-onset AD.17
The end point of a number of cell signaling pathways is apoptosis, or pro-

grammed cell death, which functions to remove cells from the body. It provides an
essential function in growth and development, but also in disease processes such as
cancer. PDCD1 (programmed cell death 1) is involved in the regulation of apoptosis,
and is a strong candidate gene for systemic lupus erythematosus (SLE).18 There is
also evidence that this gene may be associated with diabetes mellitus I.19
5.4 OTHER GENES THAT IMPACT NUTRIENT

REQUIREMENTS IN CHRONIC INFLAMMATION
As well as genes that are directly involved in chronic inflammation, polymorphisms
in genes affecting uptake or efficacy of nutrient absorption will themselves play
important roles in controlling the detrimental effects of inflammation-related disease
processes.
An increasing range of genes affecting selenium uptake and selenoprotein func-
tion have been identified. A SNP at codon 593 of human Glutathione peroxidase-1
(hGPX1) changes a cytosine (C) to a thymine (T), resulting in a substitution of
leucine for proline at position 198. Individuals carrying this variant show decreased
activity of glutathione peroxidase, an antioxidant enzyme, in cell culture studies20
and increased susceptibility to breast cancer, lung cancer, and prostate cancer.21
Individuals carrying GPx proteins differing by a single amino acid at codon 198
respond differently to increasing selenium in the diet. They show different baseline
levels of glutathione peroxidase activity per gram of hemoglobin, and these levels
are not significantly affected by 6 months of selenium supplementation.21
Curran and coworkers22 described different genetic polymorphisms in the gene
for selenoprotein S that directly affected inflammatory response in two different
population groups. Selenoprotein S (SEPS1, also called SELS or SELENOS) is a
gene that appears to be involved in stress response in the endoplasmic reticulum,
and in control of inflammation. They genotyped 13 SNPs in 522 individuals from
92 families, and considered whether groups of variants had increased or decreased
evidence for inflammation, as measured by plasma levels of interleukin 6 (IL-6),
interleukin 1L-1 (IL-1), and TNF-. They demonstrated a direct mechanistic link
between the incidence of a variant allele in one particular SEPS1 and the production
of inflammatory cytokines. From these data, they speculated that SEPS1 plays an
important role in mediating inflammation. It is still unclear as to whether dietary
supplementation with increased Se can overcome the defect in the group of indi-
viduals carrying the variant allele.
5.5 DIETARY MODULATION OF CHRONIC

INFLAMMATION AND DOWNSTREAM EFFECTS
Different nutrients or nonnutrients, so called for their lack of a defined role in
nutrition, particularly with respect to insufficiency disease, may modulate the various
processes leading to chronic inflammation and to increased disease susceptibility
(see Figure 5.2). These have been classified into three main groups, of which a few
representative examples follow.
NSAIDS
Amino Acids Omega-3 Fatty Acids
Micronutrients Probiotics/Prebiotics
Pro-inflammatory genes
Infiltration/ and mediators
Inflamed
recruitment (COX-2, INOS, IL)
tissue
Inflammatory
immune cells
Pro-inflammatory mediators
(e.g., TNF-, IL-1)
Activation of upstream kinases

ROS, (MAPK, PKC, etc)
RNS
EGCG
Plant Polyphenols Resveratrol
Vitamin C, Vitamin E Curcumin
Damage to DNA, protein or Activation of transcription factors

lipids (NF-B etc)
FIGURE 5.2 Key points within the process of chronic inflammation where nutrients and
nonnutrients interact, reducing the likelihood of chronic inflammation leading to a disease
end point. Various micronutrients and amino acids are required for an essential early inflam-
matory response, whereas diverse nutrients and nonnutrients can inhibit later events and reduce
concomitant damage to host tissues during chronic inflammation.
5.5.1 RECRUITMENT AND GROWTH STIMULATION OF IMMUNE CELLS

We have previously identified the role of various nutrients in immune response and
cancer.23 Various vitamins and minerals are particularly important to the immune
system, as they are either preferentially used by the system or are rapidly consumed
during an immune response, often because of enhanced transport and growth of
certain cell types. For example, good cellular levels of selenium (Se) are important
for effective antibody production by B-cells as well as for neutrophil chemotaxis
and activity. In general, Se supplementation enhances T-cell responses by upregulation
of the T-cell IL-2 receptor and increases antibody synthesis. Zinc is a component of
proteins involved in signal transduction during T-cell activation and interaction with
B-cells.24 Although copper is essential for optimal immune function, little is known
about the underlying mechanism. Copper deficiency results in decreased T-cell
proliferation, possibly because of suppression of IL-2, and also in reduced numbers
of circulating neutrophils. Iron deficiency leads to reduced neutrophil function,
depression of T-cell numbers with thymic atrophy, defective T-cell proliferative
response, and impaired IL-2 production by lymphocytes.
Amino acids are clearly important during cell growth. Although arginine and
glutamine are generally considered nonessential amino acids, they may be depleted
during an immune response. Nitric oxide (NO) is secreted by macrophages to kill

pathogens, and arginine is the sole substrate for nitric oxide synthase. Glutamine is
a specific fuel for the proliferation of lymphocytes.24a Many sulfur amino acids act
as substrates for acute phase protein and immunoglobulin synthesis. Sulfur amino
acid intake is particularly important for glutathione production, and insufficient levels
will exert a proinflammatory influence.
5.5.2 DIETARY ANTIOXIDANTS

Much of the DNA and tissue damage associated with chronic inflammation occurs
through formation of ROS and reactive nitrogen species (RNS), as well as DNA-
reactive aldehydes, including trans-4-hydroxy-2-nonenal and malondialdehyde from
lipid peroxidation (LPO).25 There is evidence that individuals with chronic inflam-
matory conditions, including the inflammatory bowel diseases, ulcerative colitis and
Crohns disease, show elevated levels of oxidative damage compared to the normal
population.26 Several different ROS and RNS lead to a range of detrimental effects
on different macromolecules. Enhancing the levels of dietary antioxidants (including
free radical scavengers) or enhanced enzymatic cellular defense mechanisms has
been suggested as an important therapeutic strategy for a range of inflammatory
diseases, including lung diseases such as asthma and chronic obstructive pulmonary
disease (COPD).27
Antioxidants that have effective wide-spectrum activity and good bioavailability
and thiols or molecules that have dual antioxidant and anti-inflammatory activity
may not only protect against the direct injurious effects of oxidants but may funda-
mentally alter the underlying inflammatory processes that play an important role in
the pathogenesis of chronic inflammatory lung diseases. Many of the antioxidants
in a normal Western diet are derived from plant foods. Important groups of free
radical scavengers would include ascorbic acid (vitamin C), vitamin E, and caro-
tenoids.25 Although Se is an important component of the immune response, it also
acts as an antioxidant, being a key component of glutathione peroxidase, which
detoxifies the ROS hydroperoxide and hydrogen peroxide.
As well as known micronutrients, a large number of nonnutrients, particularly
polyphenols such as tea catechins and anthocyanins from wine and some fruits and
vegetables, have also been shown to act as powerful antioxidants in vitro.28,29 It is
less clear how well they function as antioxidants in humans,30 although many studies
have demonstrated a link between diets high in polyphenol antioxidants and reduc-
tions in inflammation-related degenerative diseases, including heart disease and
cancer.3133 It is increasingly recognized, however, that many of these antioxidants
have biological activity beyond their abilities to scavenge free radicals, and also
have profound effects on signal transduction.34,35
5.5.3 CELL SIGNALING MOLECULES MEDIATING

INFLAMMATORY RESPONSE
Kundu and Surh36 extensively surveyed the literature for dietary anti-inflammatory
chemicals that disrupted signal transduction, as a rationale for chemoprevention. A
range of nonnutrients or phytochemicals may play an important role. The available
literature largely focuses on various plant polyphenols, including epigalocatechin

galate (EGCG), curcumin, [6]-gingerol and resveratrol, whose effects have been
identified at specific points in the pathways. These compounds have been shown to
inhibit a variety of kinase cascade signal transduction pathways, including protein
kinase C (PKC), I-B kinase (IKK), mitogen-activated protein kinase (MAPK), and
phosphoinositide-3 kinase (PI3K) pathways, resulting in reduced activation of proin-
flammatory transcription factors such as NF-B, AP-1, and -catenin. This has the
downstream effect of reducing the production of proinflammatory enzymes, including
COX-2 and iNOS, and proinflammatory mediators, such as TNF-, IL-8, PGE2,
and NO. Similar effects on cell signaling and inflammation are also reported for
anthocyanins,37 a variety of plant polyphenols common in highly colored foods such
as sweet potatoes.38
5.6 GENEDIET INTERACTIONS IN CHRONIC

INFLAMMATION
The optimal levels of each of the various nutrients or phytochemicals will be
governed by the nature and combination of SNPs in the various genes governing
production of the relevant proteins.
Although companies such as Interleukin Genetics are focusing on pharmaceutical
approaches to modulating the expression of genes such as IL-1, nutrition may actually
provide a practical alternative.8 This will not only be true for delaying the development
and progression of chronic disease, but also for delaying the adverse consequences
associated with aging. Given the number of nutrients affecting the immune response,
a combined approach involving micronutrients and also a variety of lipids may be
appropriate. However, a caution does also need to be sounded. An example is provided
by intervention trials in AD. Walsh and Aisen39 summarized the compelling set of
available published information on a range of in vitro and animal studies suggesting
a causal role for inflammation in development and progression of the disease. This
background provided a strong rationale and justification for clinical trials to test the
effect of anti-inflammatory drugs in this disease. However, such trials, despite testing
a range of agents, have failed to show a consistent protective effect. The same is
likely also to be true for anti-inflammatory dietary approaches.
Selenium provides an example of a nutrient whose desirable levels are impacted
strongly by known genetic polymorphisms. Our own studies suggest that Se levels
may be too low in approximately half of the Auckland population.40 However, for
those individuals with the codon 593 polymorphism in hGPX1, the activity of this
enzyme saturates at lower Se levels than for the rest of the population, and increased
Se does not improve this.21 However, these individuals are highly susceptible to
DNA damage, and this susceptibility may be counteracted by increased Se. They
are also highly susceptible to prostate cancer, and it is not yet known if this risk can
be overcome by supplementation with Se. Conversely, those individuals with the
codon 718 polymorphism in hGPX4 may have decreased prostate cancer risk and
decreased susceptibility to DNA damage. Increased Se may not decrease the risk of
cancer in these individuals, and may actually be harmful (unpublished data; this
laboratory).

Susceptibility to chronic inflammation can occur through prolonged exposures to
environmental pathogens or irritants, or through a combination of variant SNPs in
certain key genes. It is suggested that nutrient combination approaches are essential
if we are to find dietary approaches to overcoming susceptibility to chronic inflam-
mation and its ensuing disease processes. Once the scientific barriers are overcome,
there are still more hurdles. How do we put all this together? What do we tell the
public? And what other work needs to be done? Network and collaborative
approaches are essential if nutrigenomic approaches are to have a significant impact
at the level of public health.
ACKNOWLEDGMENTS
We thank the Auckland Cancer Society, The Cancer Society of New Zealand, and The
Foundation for Research, Science and Technology (New Zealand) for funding support.
KEY READINGS
Ferguson, L. R., Karunasinghe, N., and Philpott, M., Susceptibility to cancer and DNA damage
to lymphocytes: can it be counteracted by dietary selenium?, in Nutritional Biotech-
nology in the Feed and Food Industries: Proceedings of Alltechs Twenty Second
Annual Symposium, Lyons, T.P., Jacques, K.A., and Hower, J.M., Eds., Nottingham
University Press, Nottingham, 2006, p. 143.
Kundu, J. K. and Surh, Y. -J., Breaking the relay in deregulated cellular signal transduction as
a rationale for chemoprevention with anti-inflammatory phytochemicals, Mutat. Res.,
591, 123, 2005.
Philpott, M. and Ferguson, L. R., Immunonutrition and cancer, Mutat. Res., 551, 29, 2004.
Yamada, R. and Ymamoto, K., Recent findings on genes associated with inflammatory disease,
Mutat. Res., 573, 136, 2005.
REFERENCES
1. Dobrovolskaia, M. A. and Kozlov, S. V., Inflammation and cancer: when NF-kappaB
amalgamates the perilous partnership, Curr. Cancer Drug Targets, 5, 325, 2005.
2. Ditsworth, D. and Zong, W. X., NF-kappaB: key mediator of inflammation-associated
cancer, Cancer Biol. Ther., 3, 1214, 2004.
3. Greten, F. R. et al., IKKbeta links inflammation and tumorigenesis in a mouse model
of colitis-associated cancer, Cell, 118, 285, 2004.
4. Luo, J. L. et al., Inhibition of NF-kappaB in cancer cells converts inflammation-
induced tumor growth mediated by TNFalpha to TRAIL-mediated tumor regression,
Cancer Cell, 6, 297, 2004.
5. Pikarsky, E. et al., NF-kappaB functions as a tumour promoter in inflammation-
associated cancer, Nature, 431, 461, 2004.
6. Andreakos, E., Foxwell, B., and Feldmann, M., Is targeting Toll-like receptors and
their signaling pathway a useful therapeutic approach to modulating cytokine-driven
inflammation?, Immunol. Rev., 202, 250, 2004.
7. Yamada, R. and Ymamoto, K., Recent findings on genes associated with inflammatory
disease, Mutat. Res., 573, 136, 2005.
8. Kornman, K. S., Interleukin 1 genetics, inflammatory mechanisms, and nutrigenetic
opportunities to modulate diseases of aging, Am. J. Clin. Nutr., 83, 475S, 2006.
9. Yanagawa, T. et al., CTLA-4 gene polymorphism associated with Graves disease in
a Caucasian population, J. Clin. Endocrinol. Metab., 80, 41, 1995.
10. Marron, M. P. et al., Insulin-dependent diabetes mellitus (IDDM) is associated with
CTLA-4 polymorphisms in multiple ethnic groups, Human Mol. Genet., 6, 1275, 1997.
11. Kristiansen, O. P., Larsen, Z. M., and Pociot, F., CTLA-4 in autoimmune diseasesa
general susceptibility gene to autoimmunity? Genes Immun., 1, 170, 2000.
12. Ogura, Y. et al., A frameshift mutation in NOD2 associated with susceptibility to
Crohns disease, Nature, 411, 603, 2001.
13. Roberts, R. L. et al., Caspase recruitment domain-containing protein 15 mutations in
patients with colorectal cancer, Cancer Res., 66, 2532, 2006.
14. Howell, W. M., Calder, P. C., and Grimble, R. F., Gene polymorphisms, inflammatory
diseases and cancer, Proc. Nutr. Soc., 61, 447, 2002.
15. Chen, X. L. and Kunsch, C., Induction of cytoprotective genes through Nrf2/antiox-
idant response element pathway: a new therapeutic approach for the treatment of
inflammatory diseases, Curr. Pharm. Des., 10, 879, 2004.
16. Walker, D. G. et al., Gene expression changes by amyloid beta peptide-stimulated
human postmortem brain microglia identify activation of multiple inflammatory pro-
cesses, J. Leukoc. Biol., 79, 596, 2006.
17. Mohammad Abdul, H. et al., Mutations in amyloid precursor protein and presenilin-
1 genes increase the basal oxidative stress in murine neuronal cells and lead to
increased sensitivity to oxidative stress mediated by amyloid beta-peptide (1-42), HO
and kainic acid: implications for Alzheimers disease, J. Neurochem., 96, 1322, 2006.
18. Prokunina, L. et al., A regulatory polymorphism in PDCD1 is associated with sus-
ceptibility to systemic lupus erythematosus in humans, Nat. Genet., 32, 666, 2002.
19. Nielsen, C. et al., Association of a putative regulatory polymorphism in PDCD1 with
susceptibility to type 1 diabetes, Tissue Antig., 62, 492, 2003.
20. Hu, Y. J. and Diamond, A. M., Role of glutathione peroxidase 1 in breast cancer: loss
of heterozygosity and allelic differences in the response to selenium, Cancer Res.,
63, 3347, 2003.
21. Ferguson, L. R., Karunasinghe, N., and Philpott, M., Susceptibility to cancer and
DNA damage to lymphocytes: can it be counteracted by dietary selenium?, in Nutri-
tional Biotechnology in the Feed and Food Industries: Proceedings of Alltechs Twenty
Second Annual Symposium, Lyons, T.P., Jacques, K.A., and Hower, J.M., Eds., Not-
tingham University Press, Nottingham, 2006, p. 143.
22. Curran, J. E. et al., Genetic variation in selenoprotein S influences inflammatory
response, Nat. Genet., 37, 1234, 2005.
23. Philpott, M. and Ferguson, L. R., Immunonutrition and cancer, Mutat. Res., 551, 29, 2004.
24. Bao, B. et al., Zinc modulates mRNA levels of cytokines, Am. J. Phys. Endo.
Metab., 285, E1095, 2003.
25. Halliwell, B., Oxidative stress, nutrition and health. Experimental strategies for opti-
mization of nutritional antioxidant intake in humans, Free Rad. Res., 25, 57, 1996.
26. DOdorico, A. et al., Reduced plasma antioxidant concentrations and increased oxi-
dative DNA damage in inflammatory bowel disease, Scand. J. Gastro., 36, 1289, 2001.
27. Rahman, I., Oxidative stress and gene transcription in asthma and chronic obstructive
pulmonary disease: antioxidant therapeutic targets, Curr. Drug Targets Inflamm.
Allergy, 1, 291, 2002.
28. Rice-Evans, C., Plant polyphenols: free radical scavengers or chain-breaking antiox-
idants? Biochem. Soc. Symp., 61, 103, 1995.
29. Dreosti, I. E., Antioxidant polyphenols in tea, cocoa, and wine, Nutrition, 16, 692,
2000.
30. Dragsted, L. O., Antioxidant actions of polyphenols in humans, Int. J. Vit. Nutr. Res.,
73, 112, 2003.
31. Das, D. K. et al., Cardioprotection of red wine: role of polyphenolic antioxidants,
Drugs Exp. Clin. Res., 25, 115, 1999.
32. Ferguson, L. R., Role of plant polyphenols in genomic stability, Mutat. Res., 475,
89, 2001.
33. Bianchini, F. and Vainio, H., Wine and resveratrol: mechanisms of cancer prevention?,
Eur. J. Cancer Prev., 12, 417, 2003.
34. Scalbert, A., Johnson, I. T., and Saltmarsh, M., Polyphenols: antioxidants and beyond,
Am. J. Clin. Nutr., 81, 215S, 2005.
35. Williams, R. J., Spencer, J. P., and Rice-Evans, C., Flavonoids: antioxidants or
signalling molecules? Free Radical Biol. Med., 36, 838, 2004.
36. Kundu, J. K. and Surh, Y. -J., Breaking the relay in deregulated cellular signal
transduction as a rationale for chemoprevention with anti-inflammatory phytochem-
icals, Mutat. Res., 591, 123, 2005.
37. Wang, S. Y. et al., Antioxidant activity in lingonberries (Vaccinium vitis-idaea L.)
and its inhibitory effect on activator protein-1, nuclear factor-kappaB, and mitogen-
activated protein kinases activation, J. Agric. Food Chem., 53, 3156, 2005.
38. Philpott, M. et al., Enhanced coloration reveals high antioxidant potential in new
sweet potato cultivars, J. Sci. Food Agric., 83, 1076, 2003.
39. Walsh, S. and Aisen, P. S., Inflammatory processes and Alzheimers disease, Expert
Rev. Neurother., 4, 793, 2004.
40. Karunasinghe, N. et al., DNA stability and serum selenium levels in a high-risk group
for prostate cancer, Cancer Epidemiol. Biomarkers Prev., 13, 391, 2004.
6 Personalized Prevention
of Type 2 Diabetes
Hans-Georg Joost
CONTENTS
6.1 Introduction: Type 2 Diabetes is a Preventable Disease...............................61

6.2 Pathogenesis of Type 2 Diabetes Data from Animal Models.......................62
6.2.1 Genetics ..............................................................................................62
6.2.2 Nutrition .............................................................................................63
6.2.3 Mechanism of -Cell Failure: Lipotoxicity and Glucotoxicity ........63
6.3 Pathogenesis of Human Type 2 Diabetes: Risk Factors ...............................64
6.3.1 Genes ..................................................................................................64
6.3.2 Anthropometric Parameters ...............................................................65
6.3.3 Dietary Factors ...................................................................................65
6.3.4 Lifestyle: Physical Activity and Smoking .........................................66
6.3.5 Biomarker of Blood Glucose, Insulin Resistance,
and Inflammation ...............................................................................67
6.4 Individual Assessment of the Risk of Diabetes.............................................68
6.5 Future Perspectives: Toward a Personalized Diet in the Prevention
and Therapy of Type 2 Diabetes................................................................... 69
Key Readings...........................................................................................................69
References................................................................................................................70
6.1 INTRODUCTION: TYPE 2 DIABETES

IS A PREVENTABLE DISEASE
Type 2 diabetes mellitus is a complex, polygenic disease with a heterogeneous patho-
physiology, mainly characterized by obesity-associated insulin resistance and a pro-
gressive failure of pancreatic -cells. Its incidence can be lowered by reduction of the
intra-abdominal fat mass and by pharmacological control of postprandial blood glucose
levels. Several phenotypic risk factors and biomarkers (e.g., abdominal obesity, serum
adiponectin, HbA1c, nutritional pattern, and physical activity) are known that deter-
mine the individual disease risk. Such a risk assessment can provide the basis for a
personalized decision as to the quality and intensity of a preventive intervention. So
far, genotyping for the presently known gene variants that confer an increased risk for
type 2 diabetes adds relatively little to the information provided by the phenotypic risk
61
factors. However, a risk assessment based on genetic information is necessary for a

personalized intervention that begins before phenotypic risk factors are detectable.
Furthermore, data from experimental animals indicate that the response to nutritional
parameters, e.g., the dietary fat content, is determined by gene variants that modify
disease risk and progression. Thus, identification of these genes and elucidation of
their role in the human disease will allow a personalized, diet-based intervention.
During the last 30 years, a steady increase in the prevalence of type 2 diabetes has
been observed in industrialized Western countries [1]. In Germany, approximately 5% of
the population suffers from overt type 2 diabetes (fasting hyperglycemia), whereas predi-
abetes (IGT, impaired glucose tolerance) can be detected in 10 to 15% [2]. Thus, it has
been estimated that the prevalence of overt type 2 diabetes in Germany will approach 10%
in 2010. This development is associated with a marked increase in the prevalence of
overweight and obesity. Consequently, the disease is detected in younger adults, adoles-
cents, and even in children who are morbidly obese [3]. Thus, type 2 diabetes and its life-
shortening secondary complications represent a substantial public health problem. The
current trend, if it continues, will have adverse economic consequences for health-care
systems [4,5], and will eventually even reduce the mean life expectancy [6]. Therefore,
an effective prevention strategy is required to stop or even reverse the trend.
Type 2 diabetes develops by interaction of a polygenic basis with exogenous
factors, e.g., nutrition and physical activity. There is ample evidence that a change
in lifestyle leading to a reduction of intra-abdominal fat reduces the incidence of
the disease [7]. Also, a pharmacological intervention with agents that reduce blood
glucose levels, e.g., metformin [8] or the glucosidase inhibitor acarbose [9], reduces
the incidence of diabetes in cohorts with impaired glucose tolerance. Thus, type 2
diabetes can be considered a largely preventable disease.
The decision to start drug therapy for prevention of type 2 diabetes has to be based
on a risk-benefit assessment that takes into account the personal profile of a person
potentially at risk. Indeed, the currently available data already allow a risk assessment
based on phenotypic and, to a limited extent, also on genotypic data. Thus, personalized
prevention of type 2 diabetes is a realistic concept. Furthermore, because its pathophys-
iology and genetics are heterogeneous, it is reasonable to assume that there are subtypes
of individuals at risk who require different (personalized) intervention strategies. Along
these lines, current research tries to identify nutritional factors predisposing for the
disease in association with certain genotypes. Finally, the course of the disease is heter-
ogeneous, and few factors, e.g., blood glucose control, are so far known predicting the
onset of secondary complications. With a better prediction of these complications, per-
sonalized treatment is a realistic perspective for patients affected with type 2 diabetes.
6.2 PATHOGENESIS OF TYPE 2 DIABETES

DATA FROM ANIMAL MODELS
6.2.1 GENETICS
Several strains of rodents exhibit a type 2 diabeteslike hyperglycemia closely resem-
bling the human disease, e.g., the obese mouse strains db/db (monogenic deletion of
the leptin receptor) and NZO (polygenic obesity). In these mouse lines, hyperglycemia
Personalized Prevention of Type 2 Diabetes 63
is associated with morbid obesity and insulin resistance, and reflects a progressive
failure of the pancreatic -cells. The genetic basis of the diabetes can be dissected into
adipogenic and diabetogenic (i.e., causing hyperglycemia) alleles [1012]. Diabetes
requires the presence of both types of predisposing alleles (diabesity). Consequently,
in the absence of diabetogenic alleles, morbid obesity does not produce diabetes (e.g.,
in the ob/ob-mouse or the fa/fa-rat). Lean mouse lines (e.g., NON, SJL, or C57Ks/J)
may carry diabetogenic alleles; these will become active as a diabetogenic background
when the mice become obese either by a single obesity gene (e.g., in the db/db strain)
or after crossbreeding with a polygenic obesity strain such as NZO [1012].
6.2.2 NUTRITION
Most mouse strains are sensitive to a high-fat diet in that they develop obesity and
insulin resistance associated with impaired glucose tolerance [13]. Furthermore, in
diabetes-susceptible mouse strains such as NZO, the development of diabetes can be
accelerated by a high-fat diet [12]. Because diabetes in these mouse models reflects
failure and destruction of insulin-secreting cells, the finding seems to indicate that the
-cell is sensitive against fatty acids and incorporated triglycerides (lipotoxicity; see
later text). However, -cell destruction appears to also depend on the dietary carbo-
hydrates, because substitution of protein [14, 15] or fat [16] for carbohydrates delayed
or prevented the diabetes in db/db or NZO mice, respectively. Interestingly, the effect
of the high-fat diet on the development of diabetes can depend on certain genotypes.
In NZO mice, the effect of a diabetogenic allele on chromosome 5 (tentatively desig-
nated Nob1) depend on the fat content of the diet, whereas the effect of a second QTL
(Nidd/SJL) did not [12]. Such data provide proof of principle for the concept that gene
variants may determine the quality of the response to nutrients. Identification of these
variants in humans will allow the establishment of genotype-specific nutritional rec-
ommendations with regard to dietary fat and diabetes risk.
6.2.3 MECHANISM OF -CELL FAILURE: LIPOTOXICITY

AND GLUCOTOXICITY
The association of diabetes with obesity, and the diabetogenic effect of a high-fat
diet, has lent support to the hypothesis that -cells incorporating excess triglycerides
undergo apoptosis [17]. In mouse models of diabetes and insulin resistance, enhanced
hepatic lipogenesis is a key pathophysiological feature [18,19]. When lipids accu-
mulate in nonadipose tissues during overnutrition, fatty acids enter deleterious
pathways such as formation of ceramide. Ceramide is probably the most damaging
lipid and is a cause of lipoapoptosis [17]. Alternatively, it has been postulated that
elevation of postprandial and postabsorptive blood glucose levels damage the pan-
creatic -cell (glucotoxicity). According to this hypothesis, the -cell is very sensitive
to oxidative damage, or is sensitized by a low antioxidative capacity [20]. It was
suggested that elevated blood glucose concentrations lead to formation of reactive
oxygen radicals [21], which cause loss of essential transcription factors such as
MafA and irreversible cell damage [22]. This hypothesis is supported by numerous
experimental findings: Long-term culture of -cells in the presence of high glucose
concentrations affects their function [23]. In db/db [14,15] or NZO mice [16],
diabetes (defined as histologically assessed -cell damage) can be delayed or pre-

vented by feeding a high-protein or high-fat, carbohydrate-free diet. Thus, carbohy-
drates are essential for -cell destruction in mouse models of diabetes, and lipotoxicity
appears to cooperate with glucotoxicity in this pathogenesis [24].
6.3 PATHOGENESIS OF HUMAN TYPE 2

DIABETES: RISK FACTORS
6.3.1 GENES
Type 2 diabetes is the paradigm of a complex disease that has a genetic basis but
is precipitated by exogenous factors, e.g., lifestyle and diet [1]. The genetic com-
ponent of its pathogenesis appears to be stronger than in type 1 diabetes: in identical
twins, concordance of type 2 is approximately 80% for diabetes and even 96% for
impaired glucose tolerance [25]. The genetic basis of type 2 diabetes is complex; a
combination of numerous diabetogenic alleles appears responsible for development
of the disease [26,27] (Table 6.1).
TABLE 6.1
Anthropometric, Nutritional, Lifestyle, and Genetic Risk Factors
for the Development of Type 2 Diabetes
Risk Factors for the Development of Type 2 Diabetes
Anthropometry Age
Waist circumference
Height
Nutrition Whole grain bread (>50 g/d)
Red meat (>100 g/d)
Sugar-containing beverages (>300 ml/d)
Coffee (150 ml/d)
Moderate alcohol consumption (<40 g/d)
Lifestyle Exercise (sport, hiking, biking, gardening)
Smoking (>20/d)
Gene variants Transcription factor PPAR
Potassium channel KIR 6.2
Transcription factor TCF7L2
Transcription factor HHEX
Zn transporter SLC 30A8
CDKAL1
CDKN2
IGF2BP2
FTO
Note: Arrows indicate increase or decrease of risk with the indicated factors or the rare allele of the
indicated gene. A score based on these factors can be used to assess the individual diabetes risk, and
can guide personalized prevention [61]. Interaction between a genotype and a nutritional variable has
been suggested for the Pro12Ala polymorphism of PPAR and the dietary fat content [62]. Details of
nongenetic risk factors are described in Reference 34, and diabetogenic alleles are reviewed and
described in references 26, 27, 31, and 64.
Considerable efforts have been made to identify genes that cause type 2 diabetes.
In genomewide searches, chromosomal segments (quantitative trait loci, QTL) har-
boring alleles associated with hyperglycemia or other traits related with the disease
were identified. In one of these QTL, positional cloning of the gene responsible for
the effect, the calcium-dependent protease calpain-10, was successful [28]. In the
original cohort, the diabetogenic haplotype of calpain-10 was associated with a
threefold higher diabetes risk. Subsequent studies observed smaller effects, and a
meta-analysis of 26 studies determined a mean effect of the allele of only 20% [29].
In addition, several gene variants associated with an elevated diabetes risk were
identified in numerous case control studies by a candidate gene approach (e.g., the
PPAR Pro12Ala polymorphism and KIR6.2). These studies indicated that the con-
tribution of each variant to the overall diabetes risk is small (20 to 30%), although
statistically significant [26,27,30]. Thus, a single gene can predict only small
increases of the risk for type 2 diabetes. More recently, the transcription factor
TCF7L2 has been identified by positional cloning as a major diabetes gene in an
Icelandic study population, giving rise to a 1.5- to 2.5-fold disease risk [31]. In
addition, six other susceptibility genes (Table 6.1) have recently been identified
in genome-wide association studies [64]. However, it appears reasonable to assume
that many susceptibility genes have not been identified so far. Furthermore, it remains
to be investigated whether the effect of the known susceptibility genes depends on
nutritional or other environmental parameters, as has been demonstrated in mouse
models.
6.3.2 ANTHROPOMETRIC PARAMETERS

Excess adiposity is considered the most important risk factor of type 2 diabetes,
because the anthropometric parameters body weight, body mass index (BMI), and
waist circumference are major predictors of the individual disease risk [3234]. The
risk increase is mainly due to the intra-abdominal and intrahepatic fat depots [35,36],
whereas subcutaneous fat appears not to play a causative role. Thus, waist circum-
ference is the most precise anthropometric risk factor. Intrahepatic triglycerides are
believed to cause insulin resistance and are therefore directly linked with the patho-
physiology of diabetes; even a small reduction significantly lowers the diabetes risk
[37]. The pathophysiological effect of insulin resistance is an increase and prolongation
of postprandial blood glucose excursions, which enhances the exposure of -cells
to toxic glucose concentrations. Furthermore, intra-abdominal and intrahepatic fat
stores might be a marker for the accumulation of triglycerides in -cells, which has
been suggested to produce islet cell failure (lipotoxicity).
6.3.3 DIETARY FACTORS

Data from prospective cohort studies have indicated that the dietary carbohydrates
independent of the body weight significantly modify diabetes risk. A diet rich in
fibers and carbohydrates with a low glycemic index is associated with lower diabetes
risk [38,39]. In contrast, the preferential consumption of carbohydrates with a
higher glycemic index, giving rise to higher postprandial blood glucose excursions,
is associated with a higher diabetes risk. In addition, the diabetes risk is increased
by the consumption of sugar-sweetened soft drinks [40].
In addition to dietary carbohydrates, dietary fat seems to represent a major factor
that determines diabetes risk. The consumption of saturated triglycerides and trans-
fatty acids is associated with an increased, and the consumption of polyunsaturated
fatty acids with a reduced risk [38]. These epidemiological associations may reflect
direct effects of the different types of fat and fatty acids on the pancreatic -cell, or
indirect effects on insulin sensitivity, glucose tolerance and, consequently, on the
postprandial blood glucose excursions. Thus, the data are consistent with a scenario
in which blood glucose excursions, in particular under conditions of insulin resis-
tance and a genetically determined susceptibility of the pancreatic -cell, play a
crucial role in the pathogenesis of diabetes.
In an attempt to identify nutritional patterns rather than single nutrients and food
components associated with an increased diabetes risk, a prudent and a western
dietary pattern were compared [41]. The prudent dietary pattern characterized by a
higher consumption of vegetables, fruit, fish, poultry, and whole grains was associ-
ated with a lower diabetes risk, whereas the Western pattern (higher consumption
of red meat, processed meat, French fries, high-fat dairy products, and refined grains)
was associated with an increased risk (OR 1.6). More recently, the method of reduced
rank regression (RRR) was applied for the first time to identify dietary patterns
associated with a disease risk [42,43]. Using the risk factors serum adiponectin,
HbA1c, HDL, and CRP, a nutritional pattern was identified in incident diabetics
from the EPIC-Potsdam cohort that produced a maximal reduction of these risk
factors. This nutritional pattern consists of a high consumption of fruits, a low
consumption of meat, beer, soft drinks, and white bread, and is associated with an
approximately 80% reduction of the diabetes risk [43]. Similar results were obtained
independently in a Finnish cohort [44].
Moderate alcohol consumption, with the exception of beer [43], has consistently
been observed to lower the diabetes risk [34]. This effect might reflect the inhibitory
effect of alcohol on gluconeogenesis and hepatic glucose output, which could lead
to a reduction of fasting blood glucose levels. A similar effect was observed in a
study with the oral antidiabetic drug metformin, which exerts a blood-sugar-lowering
effect by inhibition of hepatic glucose output [8]. Furthermore, the beneficial effect
of blood sugar control on the incidence of progression of impaired glucose tolerance
to overt diabetes (fasting hyperglycemia) has been shown with the glucosidase
inhibitor acarbose, which blunts postprandial blood glucose excursions [9]. In the
case of beer, the beneficial effect of the alcohol seems to be outweighed by its high
carbohydrate content. Thus, the epidemiologic data as well as the results of different
pharmacological intervention studies are consistent with a scenario in which blood
glucose excursions play a major role in the pathogenesis of -cell failure.
6.3.4 LIFESTYLE: PHYSICAL ACTIVITY AND SMOKING

In addition to genetic, anthropometric, and nutritional parameters, physical activity
is a major factor determining diabetes risk [45]. Prospective cohort studies indicated
that physical activity was significantly lower in diabetics before the onset of their
disease than in healthy controls. Men who watched TV for more than 40 h per week
had an approximately threefold higher diabetes risk than those who spent less than
1 h per week watching TV [46]. Furthermore, a protective effect of exercise as part
of a weight reduction program including diet was observed in an intervention study
[7]. As the intervention resulted in a minor weight reduction (approximately 4 kg)
but a substantial (60%) reduction of diabetes risk, the authors suggested that exercise
exerted an effect that was independent of the weight reduction. This effect may be
due to an increased insulin sensitivity of muscle in response to exercise [47]. It
should be noted, however, that at present there are no data from intervention studies
directly comparing the effect of exercise with that of dietary intervention.
The majority of epidemiologic studies testing the effect of smoking have detected
a moderate increase of the diabetes risk that was reversed after cessation of smoking
[34]. It is noteworthy that the beneficial effect of smoking cessation apparently
outweighs the adverse effect on weight gain. The pathophysiological mechanism of
the effect of smoking is unclear. Nicotine stimulates sympathetic nervous activity
and may therefore stimulate transient elevations of blood glucose; such an effect
would be consistent with the concept that blood sugar excursions play a major role
in the pathophysiology of the disease.
6.3.5 BIOMARKER OF BLOOD GLUCOSE, INSULIN RESISTANCE,

AND INFLAMMATION
During the last few years, several serum factors have been identified that may indicate
an increased risk of diabetes years before the actual onset of the disease. A small
but significant increase of the glycosylated hemoglobin (HbA1c) has consistently
been found to be associated with an increased risk of developing overt diabetes [48].
Type 2 diabetes is preceded by a prediabetic period in which glucose tolerance is
impaired (IGT, impaired glucose tolerance). Fasting blood glucose is still normal,
but increased postprandial blood glucose excursions are responsible for a small
increase in HbA1c levels. There is a probability of 15 to 20% that IGT is converted
to fasting hyperglycemia within the next year. In addition to HbA1c, it has been
shown that a moderate elevation of fasting blood glucose levels predicts an increased
risk of developing diabetes within a period of 6 years [49].
Serum parameters reflecting insulin resistance such as characteristic alterations
of the lipoprotein profile (low HDL cholesterol and elevated triglycerides) and
reduced levels of the cytokine adiponectin are associated with an increased diabetes
risk [50]. Adiponectin is secreted from adipose tissue, and enhances insulin sensi-
tivity of liver, muscle, and adipocyte [51]; its serum levels are reciprocal to body
mass index and intrahepatic triglycerides [52].
Obesity is associated with a massive infiltration of adipose tissue with immune
cells such as macrophages, and it has been suggested that the pathogenesis of type
2 diabetes has an immunologic component in that cytokines released from adipose
tissue play a causal role in the insulin resistance and -cell failure [53]. Consequently,
serum levels of inflammatory cytokines or other biomarkers of inflammatory pro-
cesses such as IL-6 and CRP are elevated long before the onset of overt type 2
diabetes [54,55].
6.4 INDIVIDUAL ASSESSMENT OF THE RISK

OF DIABETES
Convincing evidence indicates that type 2 diabetes is to a large extent a preventable
disease. In persons under the age of 60, with a BMI under 25, and without excess
abdominal fat, the prevalence of the disease is very low. Intervention studies have
demonstrated that weight reduction by lifestyle changes [7] or drug therapy [56],
blood glucose control by oral antidiabetics [8,9], or bariatric surgery [57] signifi-
cantly reduces the incidence of overt type 2 diabetes. Given the life-shortening and
expensive secondary complications of the disease, it appears mandatory to intervene
as early and as effectively as possible. However, the potential negative effects of the
intervention have to be weighed against the benefits, the latter being a function of
the disease risk. At present, there are no guidelines regarding a pharmacological
therapy to reduce the diabetes risk [58], although intervention studies have shown
its efficacy. Also, the decision regarding surgical therapy of obesity (gastric banding,
gastric bypass) is currently based on the severity of the obesity rather on an assess-
ment of the risk of secondary complications. A precise assessment of the individual
diabetes risk would provide a rational basis for any decision as to intervene by drug
or surgical therapy, but could also guide and support efforts to intervene by lifestyle
changes.
The known risk factors described in this chapter allow an accurate assessment
of the risk of developing overt type 2 diabetes within the next 5 to 10 years.
Anthropometric parameters, dietary pattern, and physical activity can be determined
with a simple questionnaire and be used for a simple, orientating risk assessment.
In the next stage, clinical chemistry with parameters such as HbA1c, HDL, CRP,
and adiponectin can be used to improve the precision of the risk assessment. Fur-
thermore, genotyping for the known diabetogenic alleles (of PPAR, calpain-10,
Kir6.1, and TCF7L2) will add to the precision of the assessment in some individuals.
As our knowledge of the diabetogenic genes advances, genotyping will become
more important because it allows a much earlier risk assessment than the anthropo-
metric, lifestyle, and serum parameters.
As a basis for an individualized prevention, simple risk scores such as the
Finnish Diabetes Risk Score have been developed and validated [59]. These risk
scores weigh the parameters age, abdominal obesity, activity, and nutritional
profile. When applied to individuals in other countries, these risk scores appear to
be less accurate: validation of the Finnish Diabetes Risk Score in a German study
population resulted in a lower sensitivity and specificity of the test [60]. Thus,
significant differences appear to exist between countries in the contribution of
individual risk factors, presumably because of cultural and genetic differences.
Consequently, an accurate risk score based on a German study cohort was devel-
oped recently [61]. In addition, specificity and precision of the scores can still be
improved, because they are currently based on a relatively small number of incident
cases. Furthermore, the quality of the scores could be improved by including
known biomarkers such as HbA1c, HDL, CRP, adiponectin, and the known dia-
betogenic haplotypes.
6.5 FUTURE PERSPECTIVES: TOWARD

A PERSONALIZED DIET IN THE PREVENTION
AND THERAPY OF TYPE 2 DIABETES
Current dietary recommendations for the prevention of type 2 diabetes emphasize
weight control, and advocate a healthy diet replacing saturated and trans fats with
unsaturated fats, and refined grain products with whole grains [34]. Based on the
assumption that subgroups with different sensitivity against nutritional variables
(e.g., fat content, fat quality, and glycemic load) and a different response to a dietary
intervention exist, the general recommendations could be personalized with param-
eters predicting the benefit of a particular diet in an individual. At present, there are
few data from human studies indicating that such subgroups can be identified by
genotyping. For instance, in the DPS study, carriers of the 12-Ala allele of PPAR
appeared to benefit more from the intervention (low-fat diet and exercise) with regard
to their diabetes risk than the homozygous carriers of the 12-Pro allele [62]. Thus,
further information regarding the interaction between diabetogenic alleles and nutri-
tional variables may allow a personalized, diet-based prevention of type 2 diabetes.
As with a few other therapies, the treatment of diabetes requires regimens that are
adapted to the individual patient. Of course, the therapeutic regimen is guided by the
metabolic control accomplished as assessed with the parameters blood glucose and
HbA1c. The different regimens of insulin therapy (conventional and intensive insulin
therapy), and their combination with oral antidiabetic drugs are chosen and modified
according to the personal characteristics of the patient. So far, there are almost no
prognostic parameters other than the control of glycemia that could guide the therapy.
Such additional parameters are desirable, because glycemic control is probably not
the only parameter determining the course of secondary complications. At present, the
only other parameter guiding therapy is the BMI, because there is evidence that
suggests overweight diabetics particularly profit from therapy with metformin [63].
For further differentiation of the therapy, complex and expensive studies are required
in which the course of the disease is monitored in patients under different therapeutic
regimens. Furthermore, knowledge of the genetic basis of the disease will help not
only to assess the diabetes risk but might also be useful for differentiation of the
therapy. Implementation of such a genotype-based, personalized approach might still
be some time away but seems entirely realistic, and type 2 diabetes could become a
paradigm of the concept of individualized prevention and therapy of a complex disease.
KEY READINGS
Diamond, J., The double puzzle of diabetes, Nature, 423, 599, 2003.
Hansen, L. and Pedersen, O., Genetics of type 2 diabetes mellitus: status and perspectives,
Diabetes Obes. Metab., 7,122, 2005.
Schulze, M.B. and Hu, F.B., Primary prevention of diabetes: what can be done and how much
can be prevented? Annu. Rev. Public Health, 26, 445, 2005.
Stevenson, C., Barroso, I., and Wareham, N., The genetics of type 2 diabetes. In Nutritional
Genomics, Eds. Brigelius-Floh, R. and Joost, H.-G., Wiley-VCH, Weinheim, 223, 2006.
REFERENCES
1. Diamond, J., The double puzzle of diabetes, Nature, 423, 599, 2003.
2. Hanefeld, M. and Kohler, C., Screening, prevention and early therapy of type 2
diabetics, MMW Fortschr. Med., 146, 28, 31, 2004.
3. Wabitsch, M., Hauner, H., Hertrampf, M., Muche, R., Hay, B., Mayer, H., Kratzer, W.,
Debatin, K.M., and Heinze, E., Type II diabetes mellitus and impaired glucose
regulation in Caucasian children and adolescents with obesity living in Germany, Int.
J. Obes. Relat. Metab. Disord., 28, 307, 2004.
4. Wolf, A.M. and Colditz, G.A., Current estimates of the economic cost of obesity in
the United States, Obes. Res., 6, 97, 1998.
5. Hauner, H., Landgraf, R., Schulze, J., Spranger, J., and Standl, E., Nationales Aktion-
sforum Diabetes mellitus, Prvention des Typ-2-Diabetes mellitus, Positionspapier
des Nationalen Aktionsforum Diabetes Mellitus, Dtsch. Med. Wochenschr., 130, 1053,
2005.
6. Manuel, D.G. and Schultz, S.E., Health-related quality of life and health-adjusted
life expectancy of people with diabetes in Ontario, Canada, 19961997, Diabetes
Care, 27, 407, 2004.
7. Tuomilehto, J., Lindstrom, J., Eriksson, J.G., Valle, T.T., Hamalainen, H., Ilanne-
Parikka, P., Keinanen-Kiukaanniemi, S., Laakso, M., Louheranta, A., Rastas, M.,
Salminen, V., and Uusitupa, M., Finnish Diabetes Prevention Study Group. Prevention
of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired
glucose tolerance, N. Engl. J. Med., 344, 1343, 2001.
8. Knowler, W.C., Barrett-Connor, E., Fowler, S.E., Hamman, R.F., Lachin, J.M.,
Walker, E.A., and Nathan, D.M., Diabetes Prevention Program Research Group.
Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin,
N. Engl. J. Med., 346, 393, 2002.
9. Chiasson, J.L., Josse, R.G., Gomis, R., Hanefeld, M., Karasik, A., and Laakso, M.,
STOP-NIDDM Trial Research Group. Acarbose for prevention of type 2 diabetes
mellitus: the STOP-NIDDM randomised trial, Lancet, 359, 2072, 2002.
10. Leiter, E.H., Reifsnyder, P.C., Flurkey, K., Partke, H.J., Junger, E., and Herberg, L.,
NIDDM genes in mice: deleterious synergism by both parental genomes contributes
to diabetogenic thresholds, Diabetes, 47, 1287, 1998.
11. Plum, L., Kluge, R., Giesen, K., Altmuller, J., Ortlepp, J.R., and Joost, H.G., Type 2
diabetes-like hyperglycemia in a backcross model of NZO and SJL mice: charact-
erization of a susceptibility locus on chromosome 4 and its relation with obesity,
Diabetes, 49, 1590, 2000.
12. Plum, L., Giesen, K., Kluge, R., Junger, E., Linnartz, K., Schurmann, A., Becker, W.,
and Joost, H.G., Characterisation of the mouse diabetes susceptibility locus Nidd/SJL:
islet cell destruction, interaction with the obesity QTL Nob1, and effect of dietary
fat, Diabetologia, 45, 823, 2002.
13. West, D.B., Boozer, C.N., Moody, D.L., and Atkinson, R.L., Dietary obesity in nine
inbred mouse strains, Am. J. Physiol., 262, R1025, 1992.
14. Leiter, E.H., Coleman, D.L., Eisenstein, A.B., and Strack, I., Dietary control of
pathogenesis in C57BL/KsJ db/db diabetes mice, Metabolism, 30, 554, 1981.
15. Leiter, E.H., Coleman, D.L., Ingram, D.K., and Reynolds, M.A., Influence of dietary
carbohydrate on the induction of diabetes in C57BL/KsJ-db/db diabetes mice, J. Nutr.,
113, 184, 1983.
16. Jrgens, H., Schmolz, K., Neschen, S., Ortmann, S., Blher, M., Klaus, S., Tschp,
M., Joost, H.-G., and Schrmann, A., Development of diabetes in morbidly obese
New-Zealand Obese (NZO) mice requires dietary carbohydrates, Evidence for an

essential role of glucose toxicity in -cell destruction, Submitted for publication,
2006.
17. Unger, R.H., Minireview: weapons of lean body mass destruction: the role of ectopic
lipids in the metabolic syndrome, Endocrinology, 144, 5159, 2003.
18. Becker, W., Kluge, R., Kantner, T., Linnartz, K., Korn, M., Tschank, G., Plum, L.,
Giesen, K., and Joost, H.-G., Differential hepatic gene expression in a polygenic
mouse model with insulin resistance and hyperglycemia: Evidence for a combined
transcriptional dysregulation of gluconeogenesis and fatty acid synthesis. J. Mol.
Endocrinol., 32, 195 2004.
19. Lan, H., Rabaglia, M.E., Stoehr, J.P., Nadler, S.T., Schueler, K.L., Zou, F., Yandell,
B.S., and Attie, A.D., Gene expression profiles of nondiabetic and diabetic obese
mice suggest a role of hepatic lipogenic capacity in diabetes susceptibility, Diabetes,
52, 688, 2003.
20. Leahy, J.L., Natural history of beta-cell dysfunction in NIDDM, Diabetes Care, 13,
992, 1990.
21. Brownlee, M., A radical explanation for glucose-induced beta cell dysfunction, J.
Clin. Invest, 112, 1788, 2003.
22. Harmon, J.S., Stein, R., and Robertson, R.P., Oxidative stress-mediated, post-
translational loss of MafA protein as a contributing mechanism to loss of insulin
gene expression in glucotoxic beta cells, J. Biol. Chem., 280, 11107, 2005.
23. Robertson, R.P., Zhang, H.J., Pyzdrowski, K.L., and Walseth, T.F., Preservation of
insulin mRNA levels and insulin secretion in HIT cells by avoidance of chronic
exposure to high glucose concentrations, J. Clin. Invest., 90, 320, 1992.
24. Poitout, V. and Robertson, R.P., Minireview: Secondary beta-cell failure in type 2
diabetesa convergence of glucotoxicity and lipotoxicity, Endocrinology, 143, 339,
2002.
25. Medici, F., Hawa, M., Ianari, A., Pyke, D.A., and Lesile, R.D., Concordance rate for
type II diabetes mellitus in monozygotic twins: actuarial analysis, Diabetologia, 42,
146, 1999.
26. Hansen, L., and Pedersen, O., Genetics of type 2 diabetes mellitus: status and per-
spectives, Diabetes Obes. Metab., 7,122, 2005.
27. Stevenson, C., Barroso, I., and Wareham, N., The genetics of type 2 diabetes. In:
Nutritional Genomics, eds. Brigelius-Floh, R. and Joost, H.-G., Wiley-VCH, Weinheim,
223, 2006.
28. Horikawa, Y., Oda, N., Cox, N.J., Li, X., Orho-Melander, M., Hara, M., Hinokio, Y.,
Lindner, T.H., Mashima, H., Schwarz, P.E., del Bosque-Plata, L., Horikawa, Y., Oda,
Y., Yoshiuchi, I., Colilla, S., Polonsky, K.S., Wei, S., Concannon, P., Iwasaki, N.,
Schulze, J., Baier, L.J., Bogardus, C., Groop, L., Boerwinkle, E., Hanis, C.L., and
Bell, G.I., Genetic variation in the gene encoding calpain-10 is associated with type 2
diabetes mellitus, Nat. Genet., 26, 163, 2000.
29. Song, Y., Niu, T., Manson, J.E., Kwiatkowski, D.J., and Liu, S., Are variants in the
CAPN10 gene related to risk of type 2 diabetes? A quantitative assessment of pop-
ulation and family-based association studies, Am. J. Hum. Genet., 74, 208, 2004.
30. Barroso, I., Luan, J., Middelberg, R.P., Harding, A.H., Franks, P.W., Jakes, R.W.,
Clayton, D., Schafer, A.J., ORahilly, S., and Wareham, N.J., Candidate gene asso-
ciation study in type 2 diabetes indicates a role for genes involved in beta-cell function
as well as insulin action. PloS. Biol., 1, E20, 2003.
31. Grant, S.F., Thorleifsson, G., Reynisdottir, I., Benediktsson, R., Manolescu, A., Sainz,
J., Helgason, A., Stefansson, H., Emilsson, V., Helgadottir, A., Styrkarsdottir, U.,
Magnusson, K.P., Walters, G.B., Palsdottir, E., Jonsdottir, T., Gudmundsdottir, T.,
Gylfason, A., Saemundsdottir, J., Wilensky, R.L., Reilly, M.P., Rader, D.J., Bagger,
Y., Christiansen, C., Gudnason, V., Sigurdsson, G., Thorsteinsdottir, U., Gulcher, J.R.,
Kong, A., and Stefansson, K., Variant of transcription factor 7-like 2 (TCF7L2) gene
confers risk of type 2 diabetes, Nat. Genet., 38, 320, 2006.
32. Colditz, G.A., Willett, W.C., Rotnitzky, A., and Manson, J.E., Weight gain as a risk
factor for clinical diabetes mellitus in women, Ann. Intern. Med., 122, 481, 1995.
33. Wang, Y., Rimm, E.B., Stampfer, M.J., Willett, W.C., and Hu, F.B., Comparison of
abdominal adiposity and overall obesity in predicting risk of type 2 diabetes among
men, Am. J. Clin. Nutr., 81, 555, 2005.
34. Schulze, M.B., and Hu, F.B., Primary prevention of diabetes: What can be done and
how much can be prevented? Annu. Rev. Public Health, 26, 445, 2005.
35. Bajaj, M., Suraamornkul, S., Pratipanawatr, T., Hardies, L.J., Pratipanawatr, W., Glass,
L., Cersosimo, E., Miyazaki, Y., and DeFronzo, R.A., Pioglitazone reduces hepatic
fat content and augments splanchnic glucose uptake in patients with type 2 diabetes.
Diabetes, 52, 1364, 2003.
36. Stumvoll, M., Control of glycaemia: from molecules to men. Minkowski Lecture
2003. Diabetologia, 47, 770, 2004.
37. Petersen, K.F., Dufour, S., Befroy, D., Lehrke, M., Hendler, R.E., and Shulman, G.I.,
Reversal of nonalcoholic hepatic steatosis, hepatic insulin resistance, and hypergly-
cemia by moderate weight reduction in patients with type 2 diabetes, Diabetes, 54,
603, 2005.
38. Hu, F.B., Manson, J.E., Stampfer, M.J., Colditz, G., Liu, S., Solomon, C.G., and
Willett, W.C., Diet, lifestyle, and the risk of type 2 diabetes mellitus in women, N.
Engl. J. Med., 345, 790, 2001.
39. Willett, W., Manson, J., and Liu, S., Glycemic index, glycemic load, and risk of type 2
diabetes, Am. J. Clin. Nutr., 76, 274S, 2002.
40. Schulze, M.B., Manson, J.E., Ludwig, D.S., Colditz, G.A., Stampfer, M.J, Willett,
W.C, and Hu, F.B., Sugar-sweetened beverages, weight gain, and incidence of type 2
diabetes in young and middle-aged women, JAMA, 292, 927, 2004.
41. van Dam, R.M., Rimm, E.B., Willett, W.C., Stampfer, M.J., and Hu, F.B., Dietary
patterns and risk for type 2 diabetes mellitus in U.S. men, Ann. Intern. Med., 5, 136,
201, 2002.
42. Hoffmann, K., Schulze, M.B., Schienkiewitz, A., Nothlings, U., and Boeing, H.. Appli-
cation of a new statistical method to derive dietary patterns in nutritional epidemiology,
Am. J. Epidemiol., 159, 935, 2004.
43. Heidemann, C., Hoffmann, K., Spranger, J., Klipstein-Grobusch, K., Mohlig, M.,
Pfeiffer, A.F., and Boeing, H., A dietary pattern protective against type 2 diabetes in
the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam
Study cohort, Diabetologia, 48, 1126, 2005.
44. Montonen, J., Knekt, P., Harkanen, T., Jarvinen, R., Heliovaara, M., Aromaa, A., and
Reunanen, A., Dietary patterns and the incidence of type 2 diabetes, Am. J. Epidemiol.,
161, 219, 2005.
45. Sigal, R.J., Kenny, G.P., Wasserman, D.H., and Castaneda-Sceppa, C., Physical activity/
exercise and type 2 diabetes, Diabetes Care, 27, 18, 2004.
46. Hu, F.B., Leitzmann, M.F., Stampfer, M.J., Colditz, G.A., Willett, W.C., and Rimm,
E.B., Physical activity and television watching in relation to risk for type 2 diabetes
mellitus in men. Arch. Intern. Med., 161, 1542, 2001.
47. Hawley, J.A., Exercise as a therapeutic intervention for the prevention and treatment
of insulin resistance, Diabetes Metab. Res. Rev., 20, 383, 2004.
48. Khaw, K.T., Wareham, N., Luben, R., Bingham, S., Oakes, S., Welch, A., and Day,
N., Glycated haemoglobin, diabetes, and mortality in men in Norfolk cohort of
European prospective investigation of cancer and nutrition (EPIC-Norfolk), Br. Med.
J., 322, 15, 2001.
49. Lyssenko, V., Almgren, P., Anevski, D., Perfekt, R., Lahti, K., Nissen, M., Isomaa,
B., Forsen, B., Homstrom, N., Saloranta, C., Taskinen, M.R., Groop, L., and Tuomi,
T., Botnia study group, Predictors of and longitudinal changes in insulin sensitivity
and secretion preceding onset of type 2 diabetes, Diabetes, 54, 166, 2005.
50. Spranger, J., Kroke, A., Mhlig, M., Bergmann, M.M., Ristow. M., Boeing, H., and
Pfeiffer, A.F., Adiponectin and protection against type 2 diabetes mellitus, Lancet,
361, 226, 2003, Erratum in: Lancet, 361, 1060, 2003.
51. Trujillo, M.E. and Scherer, P.E., Adiponectin journey from an adipocyte secretory
protein to biomarker of the metabolic syndrome, J. Intern. Med., 257, 167, 2005.
52. Bajaj, M., Suraamornkul, S., Piper, P., Hardies, L.J., Glass, L., Cersosimo, E.,
Pratipanawatr, T., Miyazaki, Y., and DeFronzo, R.A., Decreased plasma adiponectin
concentrations are closely related to hepatic fat content and hepatic insulin resis-
tance in pioglitazone-treated type 2 diabetic patients. J. Clin. Endocrinol. Metab.,
89, 200, 2004.
53. Kolb, H. and Mandrup-Poulsen, T., An immune origin of type 2 diabetes? Diabeto-
logia, 48, 1038, 2005.
54. Spranger, J., Kroke, A., Mohlig, M., Hoffmann, K., Bergmann, M.M., Ristow,
M., Boeing, H., and Pfeiffer, A.F., Inflammatory cytokines and the risk to develop
type 2 diabetes: results of the prospective population-based European Prospective
Investigation into Cancer and Nutrition (EPIC)-Potsdam Study, Diabetes, 52, 812,
2003.
55. Sjoholm, A. and Nystrom, T., Inflammation and the etiology of type 2 diabetes,
Diabetes Metab. Res. Rev., 22, 4, 2006.
56. Torgerson, J.S., Hauptman, J., Boldrin, M.N., and Sjostrom, L., XENical in the
prevention of diabetes in obese subjects (XENDOS) study: a randomized study of
orlistat as an adjunct to lifestyle changes for the prevention of type 2 diabetes in
obese patients, Diabetes Care, 27, 155, 2004.
57. Sugerman, H.J., Wolfe, L.G., Sica, D.A., and Clore, J.N., Diabetes and hypertension
in severe obesity and effects of gastric bypass-induced weight loss, Ann. Surg., 237,
751, 2003.
58. Costa, A., Conget, I., and Gomis, R., Impaired glucose tolerance: is there a case for
pharmacologic intervention?, Treat. Endocrinol., 205, 2002.
59. Lindstrom, J. and Tuomilehto, J., The diabetes risk score: a practical tool to predict
type 2 diabetes risk, Diabetes Care, 26, 725, 2003.
60. Rathmann, W., Martin, S., Haastert, B., Icks, A., Holle, R., Lowel, H., and Giani,
G., KORA Study Group. Performance of screening questionnaires and risk scores
for undiagnosed diabetes: the KORA Survey 2000, Arch. Intern. Med., 165, 436,
2005.
61. Schulze, M.B., Hoffmann, K., Boeing, H., Linseisen, J., Rohrmann, S., Mhlig, M.,
Pfeiffer, A.F.H., Spranger, J., Thamer, C., Hring, H.-U., Fritsche, A., and Joost,
H.-G., An accurate risk score based on anthropometric, dietary and lifestyle factors
to predict the development of type 2 diabetes, Diabetes Care, 30, 510, 2007.
62. Lindi, V.I., Uusitupa, M.I., Lindstrom, J., Louheranta, A., Eriksson, J.G., Valle, T.T.,
Hamalainen, H., Ilanne-Parikka, P., Keinanen-Kiukaanniemi, S., Laakso, M., and
Tuomilehto, J.; Finnish Diabetes Prevention Study. Association of the Pro12Ala
polymorphism in the PPAR-gamma2 gene with 3-year incidence of type 2 diabetes
and body weight change in the Finnish Diabetes Prevention Study, Diabetes, 51,
2581, 2002.
63. U.K. Prospective Diabetes Study (UKPDS) Group: effect of intensive blood-glucose
control with metformin on complications in overweight patients with type 2 diabetes
(UKPDS 34), Lancet, 352, 854, 1998.
64. Zeggini, E., Weedon, M.N., Lindgren, C.M., Frayling, T.M., Elliott, K.S., Lango, H.,
Timpson, N.J., Perry, J.R., Rayner, N.W., Freathy, R.M., Barrett, J.C., Shields, B.,
Morris, A.P., Ellard, S., Groves, C.J., Harries, L.W., Marchini, J.L., Owen, K.R.,
Knight, B., Cardon, L.R., Walker, M., Hitman, G.A., Morris, A.D., Doney, A.S.,
McCarthy, M.L., and Hattersley, A.T., Replication of genome-wide association signals
in U.K. samples reveals risk loci for type 2 diabetes, Science, in press, 2007.
7 Toward Personalized
Nutrition for the
Prevention and Treatment
of Cancer
John C. Mathers
CONTENTS
7.1 Introduction ....................................................................................................75

7.2 Cancer Biology ..............................................................................................76
7.3 Genetic and Environmental Influences on Cancer Risk ...............................77
7.3.1 Genetic Basis of Individual Cancer Risk ..........................................78
7.3.2 DietGene Interactions and Cancer Risk ..........................................80
7.4 Nutritional Factors in Survival after Cancer Diagnosis; Is There
an Opportunity for Personalized Nutrition? ................................................. 82
7.5 Motivation for Behavior Change .................................................................. 83
7.6 Future Perspectives for Personalized Nutrition in the Prevention
and Treatment of Cancer ...............................................................................84
Key Readings...........................................................................................................84
References................................................................................................................85
7.1 INTRODUCTION
Cancers develop because of unrepaired genomic damage, causing aberrant gene expres-
sion that gives the tumor cell and its progeny a competitive advantage. For the majority
of cancers, this genomic damage results from environmental influences, which supports
the idea that if these influences were better understood, a large proportion of the significant,
and growing, global burden of cancer is potentially preventable. There is strong support
from epidemiologic studies that dietary (and lifestyle) factors play a substantial role in
modifying risk for a number of the major cancers including breast, colon, and prostate.
The exciting new information emerging from recent research on what determines personal
cancer risk includes evidence for the influence of (1) individual genotype, (2) dietary
factors on the acquisition and repair of genomic damage, and (3) interactions between
75
polymorphisms in specific genes and food components. It also seems likely that diet will
affect both quality of life and survival after cancer diagnosis. At present, the evidence base
is too limited to justify personalized dietary advice based on knowledge of genotype and
of dietgene interactions in an attempt to reduce cancer risk. However, the limited evidence
available to date is sufficiently promising to encourage research funders and researchers
to adopt this area as a high priority for sustained research. It will also be important to
undertake translational research to discover how best to deliver personalized dietary advice,
products, or other services to elicit the desired behavioral changes.
In the year 2000, 6.2 million people died from cancer, and 4.7 million women
and 5.3 million men were diagnosed with a malignant tumor [1]. Although mortality
rates from cancer are twice as high in developed as in developing countries, the
latter are catching up as the smoking epidemic spreads [1]. By 2020, it is anticipated
that the global burden of cancer will have increased by 50% because of the prevalence
of smoking and other unhealthy lifestyle choices, which are exemplified by the rising
incidence of obesity [1,2]. Better screening and earlier diagnosis combined with
more effective treatments and palliative care will all be important in management
of the growing global cancer burden, but a strong case can be made for a greater
focus on cancer prevention [1,3]. The evidence base for cancer prevention policies
will include understanding of the biology of the disease and of the factors that
contribute to risk for both populations and individuals.
7.2 CANCER BIOLOGY

At its most fundamental, cancer is a genetic disease in the sense that tumors result
from genomic damage that causes aberrant gene expression (Figure 7.1). Loss of
function of tumor suppressor genes (by mutation or epigenetic silencing) and gain of
function by oncogenes are responsible for the hallmark features that characterize
tumor cells [4]. The earliest genomic damage (the initiating event) provides the initiated
Genomic Damage Causes Cancer
Environmental
Factors DNA
Repair
Aberrant
DNA Gene
Damage Expression
Endogenous Apoptosis
Factors
FIGURE 7.1 Tumors arise in (stem) cells that have acquired genomic damage resulting in
abnormal gene expression. Genomic damage can be repaired or damaged cells can be removed
by apoptosis but, when these systems fail, cells that have a competitive advantage (in a
Darwinian sense) may initiate tumorigenesis.
Toward Personalized Nutrition for the Prevention and Treatment of Cancer 77
cell, and its progeny, with a selective advantage. In a Darwinian fashion, further
advantageous genetic and epigenetic events enable the emerging neoplastic clone to
grow through enhanced proliferation, evasion of apoptosis, disregard of differentiation
signals, angiogenesis, and invasion [5]. Mutations of nuclear DNA, which result in
expression of a disabled protein or no protein, are the best-known types of genomic
damage, but mitochondrial DNA mutations, chromosomal abnormalities (loss, rear-
rangement, or duplication), loss of telomere function (which can lead to chromosomal
instability), and aberrant epigenetic marking are also important features of tumors.
Epigenetics describes changes to the genome that result in altered gene expression,
are heritable from one cell generation to the next, but do not involve changes to the
primary sequence. The main epigenetic markings of the genome include (1) posttrans-
lational modifications of the octet of histone proteins around which nuclear DNA is
wrapped, causing alterations in chromatin structure and (2) the pattern of methylation
of cytosine residues in CpG dinucleotides in DNA. Together, these epigenetic marks
are believed to constitute an epigenetic code that regulates gene expression in particular
cells under specific circumstances [6]. Over 20 years ago, Feinberg and Vogelstein
provided early evidence that epigenetic events might be important in carcinogenesis
when they reported that tumors had a lower level of DNA methylation than normal
tissues [7]. It is now clear that epigenetic events play a critical role in most, if not all,
cancers, and that they may be a mechanism through which environmental factors
(including diet) influence stem cell biology and, thus, cancer risk [8,9]. New onco-
genes and tumor suppressor genes whose expression is regulated by epigenetic events
are being discovered, and it has been proposed that epigenetic changes might addict
cancer cells to altered signal-transduction pathways early in tumorigenesis [10,11].
This evidence makes epigenetic markings an increasingly attractive target for identi-
fication of those at enhanced cancer risk and for chemoprevention [9,12,13].
7.3 GENETIC AND ENVIRONMENTAL INFLUENCES

ON CANCER RISK
Studies of twins provide a powerful means of quantifying the contribution made by
genetics to cancer risk. In a landmark analysis of 44,788 pairs of Swedish, Danish,
and Finnish twins, Lichtenstein and colleagues concluded that inherited genetic
factors make a relatively minor contribution (between one fifth and one third see
Table 7.1) to susceptibility to most types of cancer. [14] In this analysis, nonshared,
i.e., individual environmental factors accounted for about 60% of the variance in
risk for most cancer [14]. This concurs with the earlier quantitative epidemiological
analysis by Doll and Peto that concluded that about one third of the variation in
cancer incidence between communities can be accounted for by variation in habitual
diet and a similar proportion by smoking behavior [15]. The importance of eating
patterns and nutrient intake in modifying cancer risk has been reviewed extensively
and forms the basis for recommendations for dietary changes expected to reduce
risk [16,17]. Although the evidence for diet as an etiological factor is much stronger
for some cancer sites, e.g., the colorectum than for others, e.g., the testis, the lack
of convincing evidence should not be interpreted as lack of a dietary effect, given
the paucity of research on the role of diet in the development of many cancers [3,17].
TABLE 7.1
Estimates of the Variance in Cancer Risk Accounted for by Heritable
and Environmental Factors from Studies of Swedish, Danish,
and Finnish Twins
Shared Environmental Nonshared Environmental
Cancer Site Heritable Factors Factors Factors
Stomach 0.28 0.10 0.62

Colorectum 0.35 0.05 0.60
Pancreas 0.36 0 0.64
Lung 0.26 0.12 0.62
Breasta 0.27 0.06 0.67
Ovary 0.22 0 0.78
Prostate 0.42 0 0.58
aData for women only.
Source: From Lichtenstein, P. et al., N Engl J Med, 343, 78, 2000. With permission.
7.3.1 GENETIC BASIS OF INDIVIDUAL CANCER RISK

In most individuals, the genomic damage that causes neoplasia derives from genetic
and epigenetic events experienced by individual somatic (stem) cells. However, for
about 5 to 10% of all cancer cases, germ line defects are responsible for inherited
cancer syndromes such as familial breast cancer (mutations in the BRCA1 and
BRCA2 genes), familial bowel cancer [principally familial adenomatous polyposis
(FAP) caused by mutations in the APC gene and hereditary nonpolyposis colon
cancer (HNPCC) caused by mutations in a DNA mismatch repair gene], and retin-
oblastoma (RB1 gene) and for greatly increased susceptibility to a number of other
specific cancers (see Reference 18 for a review). Studies of these rarer syndromes
have been highly informative about the biology of both inherited and sporadic
tumorigenesis [19]. Although one copy of the tumor suppressor APC gene is mutated
in every cell in the body of patients with a germ line APC mutation, loss of function
of the second copy (by mutation, chromosomal loss, or epigenetic silencing) of this
gatekeeper gene is required to initiate tumor development [19]. In contrast, damage
to only one allele is sufficient to produce gain of function by oncogenes.
In so-called sporadic cancer cases, i.e., those in which there is little evidence of a
familial condition, the genetic defects that are observed in tumors are similar to those
occurring in the familial syndromes, but these defects have been acquired in somatic
tissues [19]. Whereas a series of stochastic events is responsible for the genomic damage
leading to sporadic neoplasia, individual genotype appears to contribute to individual
susceptibility or resistance to tumorigenesis. The hypothesis that polygenic mechanisms
are likely to be responsible for cancer susceptibility has stimulated the search for asso-
ciations between genetic polymorphisms and cancer risk over the last two decades.
Although there is now a large literature on this topic, there are few robust genotype
cancer associations. Houston and Tomlinson carried out a systematic review of 50 studies,
reporting relationships between common variants in 13 genes and bowel cancer risk [20].
They found significant associations in 16 of the 50 studies, but only 3 of these associations
were reported in more than 1 study. After pooling data from a number of studies, they
concluded that significant associations were seen for polymorphisms in three genes
(Table 7.2) [20]. For the I307K polymorphism in APC (which is relatively common in
the Ashkenazi population) and for the variable number tandem repeat polymorphism in
HRAS1, those carrying the unusual versions of the gene had increased incidence of bowel
cancer (Table 7.2). Conversely, those with TT at position 677 in the MTHFR gene had,
on average, about 25% lower risk of bowel cancer. A more recent review by Sharp and
Little of 10 studies involving more than 4000 bowel cancer cases showed that most
studies suggested a lower, but not significantly so, risk for TT individuals [21]. Indeed,
the relative risk of bowel cancer was clearly lower (0.45, 95% confidence interval 0.24
to 0.86) in TT compared with CC individuals only in the study of male physicians
participating in the Physicians Health Study [21]. The authors concluded that the effect
of the MTHFR C677T polymorphism per se is modest and emphasized the lack of
understanding of genegene and of geneenvironment interactions. [21] The MTHFR
gene encodes the enzyme methylenetetrahydrofolate reductase, which catalyzes the rem-
ethylation of homocysteine to methionine. The TT version of the gene results in a protein
with a valine rather than an alanine at amino acid position 222 and which has only about
one third of the normal catalytic activity, because the mutant protein does not bind the
cofactor FAD as strongly as does the wild-type protein [22]. As a consequence, individ-
uals carrying the TT version of MTHFR require higher intakes of folate to maintain
normal circulating concentrations of homocysteine.
Major mapping exercises such as those carried out by the International SNP
Map Working Group and by the International HapMap Consortium have described
the scale and nature of human genetic variability and provide tools for more extensive
genetic association studies [23,24]. As an example, the Cancer Genetic Markers of
Susceptibility Project (http://cgems.cancer.gov/index.asp), which began in 2005 and
is expected to run for 3 years at a cost of $14 million, will use ultra-high-throughput
genotyping to identify genetic variants that increase susceptibility to prostate and
breast cancer. It is possible that such genomewide scans will identify either large
numbers of gene variants each of which makes quite a modest contribution to risk
or a much smaller number of larger-effect variants that, to date, have escaped
detection. Some have urged caution about the potential benefits of such studies,
arguing that (1) they overlook the importance of environmental factors, e.g., diet
and lifestyle and that (2) even if the anticipated genetic variants are discovered, it
is not clear how that knowledge can be translated into clinical benefit [25].
TABLE 7.2
Polymorphisms Associated with Altered Risk of Bowel Cancer
Gene Polymorphism Odds Ratio 95% Confidence Interval
APC I1307K 1.58 1.212.07

HRAS1 Variable number tandem repeat 2.50 1.544.05
MTHFR C677T 0.76 0.620.92
Source: From Houlston, R.S. and Tomlinson, I.P.M., Gastroenterology, 121, 282, 2001. With permission.
7.3.2 DIETGENE INTERACTIONS AND CANCER RISK

There is now ample support for the hypothesis that dietary factors interact with the
genome to modify gene expression and that genotype influences responses to nutri-
ents and, therefore, nutritional needs [26]. Given the evidence from epidemiologic
studies and from animal models that diet can modify cancer risk, dietgene inter-
actions in cancer etiology would be anticipated. Note, however, that much of the
evidence in this area comes from observational epidemiology (case control and
cohort studies), which limits the ability to infer causality.
Genes encoding enzymes catalyzing one carbon transfer reactions and other
folate-related transformations are among the most intensively studied for evidence
of nutrientgene interactions influencing cancer risk. Four out of five studies
reviewed by Sharp and Little provided support for the notion that folate, methionine,
and alcohol intake interacts with MTHFR C677T status to influence bowel cancer
risk [21]. For example, the apparent protection against bowel cancer in men afforded
by carriage of the TT version of MTHFR was highest in those with the highest folate
and methionine intake but was abolished in those who reported consuming five or
more alcoholic drinks per week [27]. Relationships between folate status and
MTHFR genotype have been examined in respect of breast cancer risk in Chinese
women [28]. Although there was no difference in the distribution of MTHFR C677T
genotype among cases and controls, there was a significant inverse association of
breast cancer risk with dietary folate intake for each of the genotypes that appeared
to be stronger for those carrying the TT version of the gene (Table 7.3) [28].
The enzyme manganese superoxide dismutase (MnSOD) is a key component of the
mitochondrial antioxidant defenses, where it dismutates the superoxide anion into oxygen
and H2O2. A polymorphism at codon 16 in the mitochondrial targeting sequence of the
MnSOD gene that results in an alanine (A) rather than a valine (V) appears to enhance
transport of the protein to the mitochondrial matrix and may lead to greater MnSOD
activity [29]. Several years ago, Ambrosone and colleagues reported that premenopausal
women who were homozygous for the A allele and whose intakes of fruits and vegetables
(good dietary sources of antioxidants) were low had increased risk of breast cancer [30].
Independent support for this association was provided subsequently from a study in
Shanghai, China [31]. More recently, a large nested case control study within the
TABLE 7.3
Apparent Interactions between MTHFR C677T Genotype and Folate Intake
on Odds Ratio for Breast Cancer in Chinese Women
Genotype Q4a Q3 Q2 Q1 P for trend
CC 1.00b 1.76 1.75 1.94 0.02

CT 1.16 1.50 1.73 2.17 0.06
TT 0.70 1.66 2.17 2.51 0.003
aQuartiles of dietary folate intake, Q4 = highest intake.
bReference.
Source: From Shrubsole, M.J. et al., Cancer Epidemiol Biomark Prev, 13, 190, 2004. With permission.
Physicians Health Study investigated the association between MnSOD polymorphism

and risk of prostate cancer. An important strength of this study was the fact that antiox-
idant status of the men had been determined prospectively, on average 8 years prior to
cancer diagnosis [32]. Men with high antioxidant scores (based on plasma concentrations
of lycopene, -tocopherol, and selenium) had a 40% lower risk of all prostate cancer
(Ptrend = 0.02) and a 60% lower risk of aggressive prostate cancer (defined as stage C or
D, high-grade tumor (Gleason score 7 to 10) or prostate cancer death during follow up;
Ptrend = 0.002) than those with low antioxidant status [32]. However, there was a signif-
icant interaction between antioxidant score and MnSOD polymorphism that affected
prostate cancer risk. For men carrying one or more V alleles, there was no significant
relationship between antioxidant status and prostate cancer risk. In contrast, men
homozygous for the A allele showed a five and tenfold reduction in total and aggressive
prostate cancer risk, respectively, between those with lowest and highest quartiles of
antioxidant status [32]. Li and colleagues explained their findings by suggesting that the
higher MnSOD activity associated with carriage of the AA allele may result in greater
generation of H2O2 and a consequently greater need for catalase and the selenium-
containing glutathione peroxidase [32]. When antioxidant status is low, the enhanced
production of H2O2 may lead to antioxidant damage and to prostate cancer [32].
The major factors that are hypothesized to contribute to risk of cancer for any
individual are summarized in the simple model known at the Health Pendulum
(Figure 7.2). For each person, the fulcrum of the pendulum is determined by the
Susceptibility Protective
SNP1 SNP1
SNP2 SNP2
Poor Good
Nutrition Nutrition
in utero in utero
Health
Food Food Pendulum
Component 1 Component 2
Physical
Smoking Activity
Time
Health Disease
FIGURE 7.2 The health pendulum. A simple model illustrating how key etiological factors
influence cancer risk. For a given individual, the inherited collection of susceptibility and
protective genes determines the position over the healthdisease continuum from which the
pendulum is suspended. Nutrition in utero and postnatal lifestyle factors interact with genetic
makeup to further modify risk. For most common cancers, risk increases steeply with age.
(Adapted from Mathers, J.C., Br J Nutr, 88 [Suppl. 3], S273, 2002.)
consortium of susceptibility and resistance genes that are inherited from the
individuals parents. Throughout life, the pendulum is able to move being pushed
to the right (further in the direction of cancer) or the left (better health) by a range
of pre- and postnatal factors; i.e., a range of dietary and other factors interact with
genotype to determine risk. A higher-risk lifestyle could include lack of physical
activity, exposure to environmental mutagens, or a poor diet. For most common
cancers, age is a potent carcinogen; i.e., risk rises steeply in old age.
7.4 NUTRITIONAL FACTORS IN SURVIVAL AFTER

CANCER DIAGNOSIS; IS THERE AN OPPORTUNITY
FOR PERSONALIZED NUTRITION?
In comparison with the wealth of literature investigating the role of diet in the
etiology of cancer, much less effort appears to have been directed at determining
whether, and how, dietary patterns can influence survival after cancer diagnosis. As
with cancer etiology, most of the human data come from observational studies, with
few examples of dietary interventions despite the experimental advantages in terms
of timescales, subject targeting, and endpoints afforded by secondary prevention,
quality-of-life, or survival studies.
Higher intakes of vegetables were associated with significantly longer [hazard
ratio 0.75 (95%CI 0.570.99, PTrend = 0.01 for highest vs. lowest third of intake)]
survival after diagnosis with invasive epithelial ovarian cancer in a study of Australian
women [33]. Although there were tendencies for better prognosis (time from diag-
nosis to death) with higher intakes of fruits and vegetables for those diagnosed with
lung cancer in a Danish prospective cohort study, the opposite tendency was observed
for higher consumers of potatoes, and none of the effects was statistically significant
[34]. In a study of 1551 women previously treated for breast cancer, those with
higher plasma total carotenoids concentrations had significantly reduced risk of
breast cancer recurrence (hazard ratio = 0.57, 95% CI, 0.370.89 for highest vs.
lowest quartile) [35]. Women diagnosed with breast cancer had reduced risk of death
from causes other than breast cancer if they adopted a high prudent dietary pattern
(characterized as higher amounts of fruit, vegetables, whole grains, and low-fat dairy
products) [36]. In a recent review, Chan and colleagues concluded that there are
limited data on diet and lifestyle factors that predict survival after prostate cancer
diagnosis [37]. However, higher consumers of fish and tomato sauce appeared to
have some protection against prostate cancer progression [38].
Among the few intervention studies is that by Sandler and colleagues, who used
a double-blind randomized controlled trial to demonstrate that 325 mg aspirin/d
reduced significantly (P = 0.004) the incidence of colorectal adenomas in a group
of 635 patients with previous colorectal cancer [39]. In a much smaller intervention
study of men with prostate cancer and rising prostate-specific antigen (PSA) that
used a placebo-controlled crossover design, a soy-based dietary supplement reduced
the rate of increase in PSA concentration [40].
It would be reasonable to hypothesize that the genotype of the individual patient
might influence prognosis through modulation of both responses to cancer treatment
and responses to diet post diagnosis. There is some evidence that MTHFR genotype
may influence the effectiveness of 5-fluorouracil (5-FU; a common chemotherapy

for bowel cancer) and the severity of side effects, but the studies to date have been
small with conflicting outcomes (reviewed in Reference 41). A recent study of 52
advanced gastric cancer patients suggested that polymorphisms in both glutathione
S-transferase and in thymidylate synthase predicted survival after 5-FU/cisplatin
chemotherapy, and the authors concluded that such genotyping might be used to
select those patients who are likely to benefit from this form of chemotherapy while
sparing others the side effects of the treatment. [42].
It is now clear that individual tumors have distinct molecular portraits that can
be observed as characteristic patterns of gene expression and which are a consequence
of the particular types of genomic damage sustained in their development [43]. This
is the basis for the development of the highly successful monoclonal antibody, Tras-
tuzumab (herceptin), which is targeted against HER2 (human epidermal growth factor
receptor 2) overexpressing breast cancer cells. Two recent major trials have demon-
strated that women with HER2-positive tumors treated with trastuzumab showed
dramatic improvement in a range of clinical endpoints, including time to progression,
response rate, duration of response and, most importantly, survival [44,45]. Such
studies provide proof of principle for genetically targeted cancer therapy, and the
challenge is to discover whether similar targeting can be used to personalize dietary
treatments for those diagnosed with cancer to improve quality of life and survival.
7.5 MOTIVATION FOR BEHAVIOR CHANGE

The big questions for those working in the area of dietgene interactions and cancer
risk are: (1) will genotype-based intervention produce bigger health benefits than
more generic advice and (2) will knowledge of ones genotype promote appropriate
behavior change? In my view, it is much too early to answer the first of these
questions. In the area of cardiovascular disease where understanding of dietgene
interactions is well established, Ordovas has commented recently that it is difficult
to provide genotype-based dietary advice because of conflicting interactions between
different genes [46]. The same is likely to be true for cancer. To date there have
been rather few nutritional intervention studies with cancer as an endpoint [47], and
none that has involved prospective genotyping. Research in the area of cancer
prevention is hampered by the lack of robust surrogate endpoints [48].
It is often assumed that greater understanding of personal risk and of the modifiable
factors that are likely to enhance, or reduce, risk will better motivate appropriate
behavior change. In particular, the increasing availability of relatively cheap genotypic
characterization coupled with information on geneenvironment interactions associ-
ated with altered disease risk provides an opportunity for the development of novel
personalized dietary advice or other personalized products. DNA-based information
might enhance motivation for appropriate behavior change by strengthening belief
(1) in the need for change and (2) that such changes will be beneficial. Alternatively,
such genotype-based information might create a sense of fatalism because of the belief
that genetic risk is immutable [49]. Because of the well-known difficulties in changing
eating and other lifestyle behaviors, it cannot be assumed that provision of genotype-
based information will elicit effective behavior change [50]. Although acknowledging
the potential of genomic profiling to promote a healthy lifestyle, Haga et al. saw little
reason for optimism concerning the potential for genetic test results to motivate behavior
change [51]. The best evidence for the utility of such personalized advice will come
from properly designed, powered, and conducted randomized controlled intervention
trials. However, to my knowledge, no such trials have yet been attempted; such trials
will present formidable design, ethical, and implementation issues.
7.6 FUTURE PERSPECTIVES FOR PERSONALIZED

NUTRITION IN THE PREVENTION
AND TREATMENT OF CANCER
Food choice plays a major role in the etiology of most common cancers and may
do so through influencing the acquisition of genomic damage, which is fundamental
to tumor development. There is fragmentary evidence that individual genotype may
influence cancer risk directly and that this relationship may be modulated by dietary
intake and nutritional status. However, as yet, the evidence base is insufficient to
allow personalization of dietary advice for cancer prevention based on genotypic
knowledge, although phenotypic information, e.g., identification of habitually low
physical activity or raised adiposity, could be used to tailor advice. Advances in this
area will require the development of robust surrogate endpoints that are prerequisites
for genotype-specific dietary intervention studies (with prospective genotyping). It
remains uncertain whether greater understanding of personal risk (based on geno-
type) and of the modifiable factors that are likely to enhance, or reduce, risk will
better motivate appropriate behavior change. This is an important issue for the whole
personalized nutrition field [53], including the development of public health strategies,
and of products or services, that are intended to lower cancer risk.
There appear to be major opportunities to investigate the impact of diet, and of
dietgene interactions, on prognosis after cancer diagnosis. Because of (1) the greater
motivation resulting from diagnosis with a life-threatening disease and (2) the
potentially much greater individual support from health-care professionals following
such a diagnosis, this might be a more fruitful area for eventual, personalization of
dietary advice than primary prevention.
ACKNOWLEDGMENTS
Research in my laboratory on cancer prevention is funded by the Medical Research
Council (G0100496), the Biotechnology and Biological Sciences Research Council
(D20173), and by the Food Standards Agency (N12015, N12016).
KEY READINGS
Feinberg, A.P., Ohlsson, R., and Henikoff, S, The epigenetic progenitor origin of human
cancer, Nat Rev Genet, 7, 21, 2006.
Joost, H.G., Gibney, M.J., Cashman, K.D., Gorman, U., Hesketh, J.E., Mueller, M., van
Ommen, B., Williams, C.M., and Mathers, J.C., Personalized nutrition, Br J Nutr
[in press], 2007.
Mathers, J.C., Nutrition and cancer prevention: diet-gene interactions, Proc Nutr Soc, 62,
605, 2003.
Rock, C.L. et al., Plasma carotenoids and recurrence-free survival in women with a history
of breast cancer, J Clin Oncol, 23, 6631, 2005.
Stewart, B.W. and Kleihues, P., World Cancer Report, IARC Press, Lyon, 2003.
REFERENCES
1. Stewart, B.W. and Kleihues, P., World Cancer Report, IARC Press, Lyon, 2003.
2. Calle, E.E. and Thun, M.J., Obesity and cancer, Oncogene, 23, 6365.
3. Mathers, J.C., Nutrition and cancer prevention: diet-gene interactions, Proc Nutr Soc,
62, 605, 2003.
4. Hanahan, D. and Weinberg, R.A., The hallmarks of cancer, Cell, 100, 57, 2000.
5. Ponder, B.A.J., Cancer genetics, Nature, 411, 336, 2001.
6. Jenuwein, T. and Allis, C.D., Translating the histone code, Science 293, 1074, 2001.
7. Feinberg, A.P. and Vogelstein, B., Hypomethylation distinguishes genes of some
Human cancers from their normal counterparts, Nature, 301, 89, 1983.
8. Jones, P.A. and Baylin, S.B., The fundamental role of epigenetic events in cancer,
Nat Rev Genet, 3, 418, 2002.
9. Feinberg, A.P., Ohlsson, R., and Henikoff, S, The epigenetic progenitor origin of
human cancer, Nat Rev Genet, 7, 21, 2006.
10. Esteller, M., Epigenetics provides a new generation of oncogenes and tumour-
suppressor genes, Br J Cancer, 94, 179, 2006.
11. Baylin, S.B. and Ohm, J.E., Epigenetic gene silencing in cancer a mechanism for
early oncogenic pathway addiction?, Nat Rev Cancer, 6, 107, 2006.
12. Belshaw, N.J. et al., Use of DNA from stools to detect aberrant CpG island methylation of
genes implicated in colorectal cancer. Cancer Epidemiol Biomark Control, 13, 1495, 2004.
13. Mathers, J.C., Reversal of DNA hypomethylation by folic acid supplements: possible
role in colorectal cancer prevention, Gut, 54, 579, 2005.
14. Lichtenstein, P. et al., Environmental and heritable factors in the causation of cancer,
N Engl J Med, 343, 78, 2000.
15. Doll, R. and Peto, R., The causes of cancer: quantitative estimates of avoidable risks
in the USA today, J Natl Cancer Inst, 66, 1191, 1981.
16. World Cancer Research Fund/American Institute of Cancer Research, Food, Nutrition
and the Prevention of Cancer: A Global Perspective, World Cancer Research Fund/
American Institute of Cancer Research, Washington, D.C., 1997.
17. Department of Health, Nutritional Aspects of the Development of Cancer, Report on
Health and Social Subjects No. 48, The Stationery Office, London, 1998.
18. Garber, J.E. and Offit, K., Hereditary cancer predisposition syndromes, J Clin Oncol,
23, 276, 2005.
19. Kinzler, K.W. and Vogelstein, B., Lessons from hereditary colorectal cancer, Cell,
87, 159, 1996.
20. Houlston, R.S. and Tomlinson, I.P.M., Polymorphisms and colorectal cancer risk,
Gastroenterology, 121, 282, 2001.
21. Sharp, L. and Little, J., Polymorphisms in genes involved in folate metabolism and
colorectal neoplasia: a HuGE review, Am J Epidemiol, 159, 423, 2004.
22. Guenther, B.D. et al., The structure and properties of methylenetetrahydrofolate
reductase from Escherichia coli suggest how folate ameliorates human hyperho-
mocysteinemia, Nat Struct Biol, 6, 359, 1999.
23. Sachidanandam, R. et al., A map of human genome sequence variation containing

1.42 million single nucleotide polymorphisms, Nature, 409, 928, 2001.
24. International HapMap Consortium, A haplotype map of the human genome, Nature,
437, 1299, 2005.
25. Baker, S.G. and Kaprio, J., Common susceptibility genes for cancer: search for the
end of the rainbow, BMJ, 332, 1150, 2006.
26. Mathers, J.C., Candidate mechanisms for interactions between nutrients and genes,
in Nutrient-Gene Interactions in Cancer, Choi, S.-W. and Friso, S., Eds., CRC Press
Taylor & Francis, Boca Raton, FL, 2005, chap. 2.
27. Chen, J. et al., A methylenetetrahydrofolate reductase polymorphism and the risk of
colorectal cancer, Cancer Res, 56, 4862, 1996.
28. Shrubsole, M.J. et al., MTHFR polymorphisms, dietary folate intake, and breast
cancer risk: results from the Shanghai Breast Cancer Study, Cancer Epidemiol Biom-
ark Prev, 13, 190, 2004.
29. Sutton, A. et al., The Ala16Val genetic dimorphism modulates the import of human
manganese superoxide dismutate into rat liver mitochondria, Pharmacogenet, 13,
145, 2003.
30. Ambrosone, C.B. et al., Manganese superoxide dismutase (MnSOD) genetic poly-
morphisms, dietary antioxidants, and risk of breast cancer, Cancer Res, 59, 602, 1999.
31. Cai, Q. et al., Genetic polymorphism in the manganese superoxide dismutase gene,
antioxidant intake and breast cancer risk: results from the Shanghai Breast Cancer
Study, Breast Cancer Res, 6, R647, 2004.
32. Li, H. et al., Manganese superoxide dismutase polymorphism, prediagnostic antioxidant
status, and risk of clinical significant prostate cancer, Cancer Res, 65, 2498, 2005.
33. Nagle, C.M. et al., Dietary influences on survival after ovarian cancer, Int J Cancer,
106, 264, 2003.
34. Skuladottir, H. et al., Does intake of fruit and vegetables improve lung cancer sur-
vival?, Lung Cancer, 51, 267, 2006.
35. Rock, C.L. et al., Plasma carotenoids and recurrence-free survival in women with a
history of breast cancer, J Clin Oncol, 23, 6631, 2005.
36. Kroenke, C.H. et al., Dietary patterns and survival after breast cancer diagnosis, J
Clin Oncol, 23, 9295, 2005.
37. Chan, J.M., Gann, P.H., and Giovannucci, E.L., Role of diet in prostate cancer
development and progression, J Clin Oncol, 23, 8152, 2005.
38. Chan, J.M. et al., Diet after diagnosis and the risk of prostate cancer progression,
recurrence, and death (United States), Cancer Causes Control, 17, 199, 2006.
39. Sandler, R.S. et al., A randomised trial of aspirin to prevent colorectal adenomas in
patients with previous colorectal cancer, N Engl J Med, 348, 1939, 2003.
40. Schrder, F.H. et al., Randomised, double-blind, placebo-controlled crossover study
in men with prostate cancer and rising PSA: effectiveness of a dietary supplement,
Eur Urol, 48, 922, 2005.
41. Little, J. et al., Colon cancer and genetic variation in folate metabolism: the clinical
bottom line, J Nutr, 133, 3758S, 2003.
42. Goekkurt, E. et al., Polymorphisms of glutathione S-transferases (GST) and thymidy-
late synthase (TS) novel predictors for response and survival in gastric cancer
patients, Br J Cancer, 94, 281, 2006.
43. Perou, C.M. et al., Molecular portraits of human breast tumours, Nature, 406, 747,
2000.
44. Piccart-Gebhart, M.J. et al., Trastuzumab after adjuvant chemotherapy in HER2-
positive breast cancer, N Engl J Med, 353, 1659, 2005.
45. Romond, E.H. et al., Trastuzumab plus adjuvant chemotherapy for operable HER2-
positive breast cancer, N Engl J Med, 353, 1673, 2005.
46. Ordovas, J.M., Genetic interactions with diet influence the risk of cardiovascular
disease, Am J Clin Nutr, 83, 443S, 2006.
47. Mathers, J.C., Food and cancer prevention: human intervention studies. In Dietary
Anticarcinogens and Antimutagens, Johnson, I.T. and G.R. Fenwick, G.R., Eds., Royal
Society of Chemistry, Cambridge, 2000, chap. 8.1.
48. Sanderson, P. et al., Emerging diet-related surrogate end points for colorectal cancer:
U.K. Food Standards Agency diet and colonic health workshop report, Brit J Nutr,
91, 315, 2004.
49. Marteau, T.M. and Lerman, C., Genetic risk and behavioural change, Br Med J, 322,
1056, 2001.
50. Adamson, A.J. and Mathers, J.C., Effecting dietary change, Proc Nutr Soc, 63, 537,
2004.
51. Haga, S.B., Khoury, M.J., and Burke, W., Genomic profiling to promote a healthy
lifestyle: not ready for prime time, Nat Genet, 34, 347, 2003.
52 Mathers, J.C., Pulses and carcinogenesis: potential for the prevention of colon, breast
and other cancers, Br J Nutr, 88(Suppl. 3), S273, 2002.
53. Joost, H.G., Gibney, M.J., Cashman, K.D., Gorman, U., Hesketh, J.E., Mueller, M.,
van Ommen, B., Williams, C.M., and Mathers, J.C., Personalized nutrition, Br J Nutr
[in press], 2007.
8 Nutrigenomics
and Angiogenesis
in Obesity
Aldona Dembinska-Kiec
CONTENTS
8.1 Introduction ....................................................................................................89

8.2 The Main Principles of Angiogenesis ...........................................................90
8.2.1 Angiogenesis and Adipose Tissue Plasticity .....................................92
8.2.2 Nutrients and Differentiation of Adipose Tissue...............................93
8.2.3 Polyphenols and Other Naturally Occurring Nutrients
in Angiogenesis and Adipogenesis ....................................................95
8.3 Personalized Nutrition and Angiogenesis
in Adipose Tissue Remodeling ......................................................................97
Key Readings...........................................................................................................97
References................................................................................................................98
8.1 INTRODUCTION
Development and maturation/plasticity of adipose tissue vascularity is critical for
the function of this metabolic and the endocrine organ. The proangiogenic factors,
as well as the endogenous inhibitors of angiogenesis, are synthesized locally in
adipose tissue. It has been proved recently that treatment of animals with antiangio-
genic factors (such as anti-VEGF or its receptor antibody) dose dependently, in a
reversible manner, decreases adipose tissue depot and body weight. The involvement
of basal nutrient (such as fatty acids, amino acids, carbohydrates, and polyphenols)
availability and energy supply signals (i.e., caloric restriction) on the nutrient
sensors such as the mammalian target of rapamycin (mTOR) or CHOP-10/gadd
153 and the other pathways regulating adipose tissue morphogenesis are reviewed
on the basis of recent knowledge. Because transcriptomics applied to obesity, as
well as caloric restriction studies in humans, point to the important changes in pro-
and anti-inflammatory status, the essential role of personalized nutrition in regulation
of the angiogenesis-driven remodeling of the stromal vascular fraction of adipose
tissue is discussed.
89
Without appropriate blood supply from the capillary network, tissue cannot
survive; the circulatory system is essential for oxygen and nutrient distribution
between tissues and for the removal of by-products of metabolism. Vasculogenesis
and angiogenesis play an essential role in a number of physiologic and pathologic
events, such as fetal development, vascular and tissue remodeling in ischemia,
inflammation, proliferative diabetic retinopathy, as well as in promotion of solid
tumor growth and invasiveness (1,2).
The development and maturation/plasticity of adipose tissue vascularity is crit-
ical for the function of adipose tissue as the metabolic and the endocrine organ. It
has also been demonstrated that fetal adipocyte development is spatially and tem-
porally related to arteriolar development (3). Recent studies have evidenced that
human endothelial progenitors and preadipocytes express similar patterns of genes
and proteins [cellular markers and the proangiogenic factors such as CD34, AC133,
vascular endothelial growth factor (VEGF) receptors, integrins, serine proteases, and
others], which suggests a possible common origin from the stromal vascular progenitor
cell fraction (SVF) (3,4,5).
In adolescence, the adipose tissue SVF releases only 6% of the main proangio-
genic VEGF amount released by the differentiated adipocytes. Under the stimulus
of insulin, the differentiated adipocytes rather than the stromal vascular fraction are
a source of VEGF released by visceral adipose tissue (6), pointing to the stimulation
of angiogenesis by the growing mass of tissue.
8.2 THE MAIN PRINCIPLES OF ANGIOGENESIS

The multistep process of angiogenesis is organized by a sensitive balance of acti-
vators and inhibitors of recruitment of endothelial cells and their progenitors. It
includes induction of various growth factors, production of proteolytic enzymes to
digest the basement membrane and extracellular matrix, endothelial cell migration,
proliferation and tube formation, and differentiation of pericytes/vascular smooth
muscle with reconstruction of basement membranes. Ischemia, blood shear stress
forces, and cytokine-driven induction of the sequence certain genes are responsible
for the outgrowth and maturation of the new capillary network (1) (Figure 8.1).
There are several proangiogenic growth factors and cytokines induced in
ischemic tissue, such as VEGF, tumor necrosis factor (TNF), basic fibroblast growth
factor (bFGF), interleukin-8, (IL-8), platelet-derived growth factor (PDGF), angio-
poietins, ephrins, proangiogenic vasoconstrictors such as angiotensin II (AngII),
endothelin (End), thromboxan (TxA2), and a plethora of others acting by its specific
receptors, or other angiogenic modulators, such as nitric oxide (NO) (1). VEGF is
the main, most important proangiogenic factor. It exists in several isoforms (VEGF-
A, -B, -C, -D, and -E) that are derived from the alternative splicing of the same
gene. These isoforms transduce their signals to the nucleus mainly through three
receptors. The VEGFR-1(Flt) induces the activation of EC, vessel wall permeability,
EC fenestration, and pathological angiogenesis seen in cancer vasculature (1).
VEGFR-3 is involved in lymphangiogenesis. VEGF2 (Flk/KDR) receptor has thy-
rosine kinase activity, thus regulating EC proliferation (cell cycle) by activation of
extracellular signal-regulated kinase (ERK)1/2 mitogen-activated protein kinases
Nutrigenomics and Angiogenesis in Obesity 91
Inammation Dierentiation
Cytokines: IL-6
TNF- FFA/RA PPAR-//RAR/RXR
Macrophages IL-1 Insulin
MCP-1 cAMP
Others Others
ROS
Adipocyte
Ca++
VSMC SVF
MMPs
VEGF Adipokines:
Angiopoietin Adiponectin, TPA, PAI, TSP
Tie1/2 Leptin, A-II
Others
NO
Angiogenesis
EC VEGF
PDGF EC and VSMC
Angiopoietin
Ephrins
Others
FIGURE 8.1 Multistep process of angiogenesis.
(MAPKs) and p70S6 kinase (S6K), migration of ECs by the stress-activated protein
kinase 2/p38 pathway, and survival (antiapoptotic effect) by Akt activation (7).
VEGF-A induces genes of the main metalloproteinases: MMP-2 and MMP-9, as
well as integrins (VE-cadherin or 3). The IL-8-induced activation of CXCR-2
receptor is necessary for matrix remodeling, adhesion, and EC differentiation during
tubulogenesis. The VEGF gene expression is induced by several factors such as
hypoxia (by the involvement of the hypoxia-inducible-factor, HIF-1), insulin, insulin
growth factors (IGFs), cytokines, as well as by nitric oxide (NO) (1,2,7).
Angiopoietin (Ang)-1 and Ang-2 with their receptors, Tie-1 and Tie-2, are
considered to be the most important partners of VEGF in the maturation of capillary
networks, especially in branching of the sprouts and modifying apoptosis to be not
redundant for capillary-sprouting cells. VEGF signaling is initiated through the
assembly of a multicomponent complex composed of the vascular endothelial (VE)-
cadherin, -catenin, VEGF-R2, and PI3-kinase. This complex then stimulates Akt,
leading to the downstream activation and nuclear translocation of the transcription
factor NF-B, antiapoptotic Bcl-2 protein synthesis, induction of IL-8, and monocyte
chemoattractant protein-1 (MCP-1) gene expression. This is followed by infiltration
of macrophages and tubulogenesis of the growing capillary network (1,7,8). Ephrin
B (EphB2) and its receptor EphB4 expressed on endothelial cell membranes deter-
mine arteriovenous shifting of capillary differentiation (8). Transforming growth
factor (TGF) with tumor necrosis factor (TNF), or platelet-derived growth factor
(PDGF) are induced in ischemic tissue or are released from activated macrophages
or platelets. They establish the lengthening of capillaries by regulating endothelial
cell quiescence, cell-to-cell interactions, and differentiation of fibroblast-like
progenitors to pericytes or vascular smooth muscle cells (9). Nitric oxide (NO)
synthesized by the endothelial nitric oxide synthase (eNOS) is induced by several
proangiogenic factors (including VEGF, TNF, angiotensin II (AII), insulin, leptin,
adiponectin, and others). Nitric oxide plays an essential role in angiogenesis. It
protects endothelial cells against apoptosis by stimulation of Akt, and induction of
Bcl-2 and other antiapoptotic proteins (IAPs). By inhibition of platelet aggregation
and adhesion of inflammatory cells (macrophages, leukocytes, and lymphocytes),
NO inhibits penetration of cells generating pro- and antiangiogenic factors (VEGF,
PDGF, TNF, AII, metalloproteinases, IL-8, IL-1, INF-, etc.). By inhibition of
vascular wall smooth muscle cell (VSMC) proliferation and basement membrane
protein synthesis, NO is responsible for the maturation of the capillary network and
inhibition of pathological remodeling of the vessel wall (1,79).
Simultaneously with the proangiogenic factors, the endogenous inhibitors of
angiogenesis are synthesized locally. Examples are thrombospondins (TSP-1 and
TSP-2) with their receptor CD36, which is also the long-chain fatty acid scavenger-
type receptor, or the other inhibitory products deriving from the protein digestion
such as the angiostatin, endostatin, maspin, and others (1,2,9).
8.2.1 ANGIOGENESIS AND ADIPOSE TISSUE PLASTICITY

Primitive fat cell clusters resembling vascular structures without any or only a few
preadipocytes were observed in fetuses of several species (10). There is also some
evidence for overlap of autocrine/paracrine developmental relationship between the
preendothelial/preadipocyte cells (reviewed by Hausmann and Richardson (3). Both
type of cells express the 3 integrin, release PAI-1 that supports coordination of
angiogenesis and adipogenesis (10). Both the adipose as well as the stromal vascular
(non-fat-containing cell) fraction (SVF) release VEGF; however, adipocytes are the
main source of the proangiogenic VEGF, IL-8 and PAI-1 (but not IL-6), released
from differentiated adipocytes under stimulation with insulin (6). Interestingly,
VEGF induction and release from adipocytes is also activated by factors promoting
lipolysis such as norepinephrine, forskolin, and dibutyryl-cAMP, but not dexametha-
sone (6). Angiopoietin 2 (Ang-2) gene expression, which supports proangiogenic
VEGF activity, was found in preadipocytes as well as ob/ob mice fat tissue under
stimulation with leptin. Leptin and adiponectin exert regulatory effects on angio-
genesis directly or by inducing the infiltration of macrophages (the important source
of VEGF and metalloproteinases, or by activation of endothelial NO release
(1,3,4,10). The long-form leptin receptor is expressed in endothelial cells and adi-
pocytes in human and mouse adipose tissue. Its activation promotes the Erk1/2 and
STAT3 phosphorylation mediating EC proliferation, migration, and survival, and
also (similar to VEGF) increases permeability of endothelium by the formation of
endothelial gaps. Additionally, leptin increases VSMC proliferation and migration
necessary for maturation and remodeling of the vessel wall (3,11).
There is evidence suggesting that angiogenesis is a pivotal factor for adipogenesis.
During fetal development, arteriolar differentiation and synthesis of extracellular
matrix proteins precedes the differentiation of adipocytes (3). The undifferentiated
vessel wall fibroblasts, pericytes, as well as circulating stromal progenitory cells,
may serve as the early preadipocytes, and the adipocyte tissue stromal vascular
fraction (SVF) secrete several factors necessary for angiogenesis and modulation of
the capillary network (4,5,12). Thus, during the postnatal period, VEGF expression
and resulting angiogenesis may precipitate adipogenesis (3). VEGF (as well as leptin)
are induced by ischemia (mediated by HIF-1) or insulin, and the exposure to cold
induces VEGF expression in the brown adipose tissue (1,6,8).
Using several experimental models of genetically modified mice (knockout or
transgenic), it has been demonstrated that the manipulation of the tPA, uPA/PAI-1,
or plasminogen gene expression deeply influences adipose tissue mass and weight
gain of mice fed a high-fat diet. Remodeling of extracellular matrix proteins by
endothelial, adipocyte, or macrophage-derived MMP-2 and MMP-9 is necessary for
angiogenesis and adipogenesis (1,3), and the TIMP-3 deficient mice exhibit
increased adipogenicity during mammary gland involution.
It has been proved recently that treatment of animals with antiangiogenic factors,
(such as anti-VEGF or its receptor antibody) dose dependently, in a reversible
manner, decreases adipose tissue deposition and body weight (13).
8.2.2 NUTRIENTS AND DIFFERENTIATION OF ADIPOSE TISSUE

Manipulation of dietary factors with the aim of decreasing adipose tissue mass
attenuate the inflammatory response in adipose tissue and improve insulin sensi-
tivity. It is an important event for prevention and treatment of metabolic syndrome
and type 2 diabetes. Moloney et al. (14) have demonstrated that that a subgroup of
fatty acids known as conjugated linoleic acid (CLA), in particular, the cis-9, trans-
11-CLA isomer, markedly improves insulin sensitivity and lipid metabolism as well
as inflammatory cytokine biosynthesis by lowering of nuclear P65 levels and decreas-
ing NFB DNA binding.
The developing blood vessels in fat tissue represent a potential target for regu-
lating adipose tissue mass showing metabolism and substrate-fuel buffering proper-
ties. With the microarray method, it has been demonstrated that the expression of
many adipogenic genes (SREBP-1, C/EBP, or PPAR) was significantly decreased
in adipose tissue from obese animals, indicating a process of dedifferentiation, i.e.,
loss of adipocyte phenotype similar to the response seen after treatment of adipose
tissue with TNF. Voros et al. (15) reported that treatment of ob/ob as well as
C57Bl/6 mice on high-fat diet for 15 weeks increases adipo- and angiogenesis in sub-
cutaneous and gonadal fat pads. It was accompanied (dependent on the duration of
feeding) by downregulation of Ang-1 and elevation of TSP-1, pointing to changes in
angiogenesis. In ob/ob mice the placental growth factor (PIGF) and Ang-2 expression
was increased in subcutaneous fat, when TSP-2 in subcutaneous and gonadal fat (15).
The early phase of preadipocyte differentiation is induced by the sequence of
the genes regulated by the two groups of transcriptional factors: the CCAAT/
enhancer binding protein (C/EBP) and peroxisome proliferator-activated receptors
(PPARs). PPAR- and PPAR/ heterodimerize with the other nuclear receptors
(RAR/RXR, LXR, etc.) and regulate the expression of the multiple genes of differ-
entiated adipocytes such as an adipocyte-specific fatty-acid-binding protein (FABP),
FAT/CD36 (scavenger receptor for long-chain FFA and thrombospondin), perilipin,
adipsin, stearoyl-CaA desaturase (SCD-1), GLUT-4, phosphenolpyruvate carboxy-

kinase (PEPECK), and leptin (16). Expression of these genes in progenitor undif-
ferentiated cells (SVF) is connected with inhibition of expression of the genes related
to angiogenesis. Exogenous PPAR ligands cause reduction of VEGF-R1 and VEGF-
R2 expression in endothelial cells, thus inhibiting angiogenesis by inhibition of
endothelial cell migration and proliferation, as well as by the inhibition of the
important-for-angiogenesis infiltration of macrophages. However, the induction of
VEGF gene and protein by PPAR ligands (glitazones) in adipocytes and vascular
smooth muscle cells was reported. Blocking the PPAR pathway by transfecting the
preadipocytes with the PPAR-dominant negative construct abrogates not only adi-
pose tissue differentiation but also reduces angiogenesis (17). Similarly, inhibition
of angiogenesis by blocking of VEGF-R2 activity by the specific antibody inhibited
the preadipocyte differentiation in vivo and in vitro, in spite of the fact that preadi-
pocytes do not express the VEGF-R2 (17). This results in strong support of the
parallel, reciprocal regulation of angio- and adipogenesis in not well recognized but
involving fatty acids and its derivatives paracrine manner (3,11,16,17).
The involvement of basal nutrient availability and energy supply signals (i.e.,
amino acid, glucose sufficiency) on the nutrient sensors such as the mammalian
target of rapamycin (mTOR) or CHOP-10/gadd 153 pathways regulating adipose
tissue morphogenesis remains to be investigated (18). Moreover, the fact that the
transcriptomics applied to obesity as well as the caloric restriction study in humans
points to the important changes in the pro- and anti-inflammatory status, and thus
to the regulation of the angiogenesis regulating factor gene expression in the stromal
vascular fraction of the adipose tissue (19). For example, the Wnt family of proteins,
which are the paracrine growth factors highly expressed in preadipocytes, are the
potent inhibitors of angiogenesis (reviewed in Reference 29). Depending on the local
conditions, Wnt signaling causes precursor cell proliferation, apoptosis, differenti-
ation, or maintenance in undifferentiated status. The inhibition of adipogenesis is
connected with Wnt/fizzled cooperation, which results in downregulation of C/EBP
and PPAR, and inactivation of glycogen synthase kinase 3 (GSK3) by preventing
it from phosphorylating -catenin with subsequent degradation. The stabilization of
-catenin may be responsible for the maintenance of the precursor cells (satellite
cells) in tissues, such as the SVF of adipose tissue. This may be the clue event for
adipose tissue plasticity and adipose tissue accumulation, as well as the jo-jo effect
in patients (1820).
GSK3 modulation by a large number of growth factors (including insulin) and
nutrients is an important regulator of cell differentiation. Disruption of Forkhead box
C2 (FOXC2) gene expression linking cell cycle with glucose sensitivity causes embry-
onic or perinatal death with severe vascular and skeletal defects (20). Also, the sprouty
proteins (Spry), which may both activate or inhibit activity of several receptor tyrosine
kinase (RTK) signaling, are involved in the early progenitor cell fate decision, and
thus also in angiogenic vs. adipogenic differentiation of SVF. Sprouty proteins were
found to suppress the MAPK signaling induced by EGF, VEGF, or PDGF (20). The
importance of polymorphisms of nutrient sensors in metabolic disorders remains to
be established. For that the new scientific approach, i.e., systems biology, is necessary
to understand the nutrientmetabolismgenotype dependence.
8.2.3 POLYPHENOLS AND OTHER NATURALLY OCCURRING

NUTRIENTS IN ANGIOGENESIS AND ADIPOGENESIS
Polyphenols, especially flavonoids, are claimed to act as active components in
prevention of cancer, cardiovascular diseases, or metabolic disorders, including
diabetes. Flavonoids are plant phytochemicals that cannot be synthesized by humans.
The six classes of flavonoids (flavanones, flavones, flavonols, isoflavonoids, antho-
cyanins, and flavans) vary in their structural characteristics. Flavanones occur
predominantly in citrus fruits, flavones in herbs, isoflavonoids in legumes, anthocy-
anins and catechins in fruits, and flavonols in all fruits and vegetables. Grains and
oilseeds, honey, and chocolate have flavonoids; however, the flavonoid content may
be destroyed by food processing. Fresh green tea contains large amounts of catechin
and caffeine polyphenols, whereas resveratrol, myricetin, and quercetin are rich in
grapes, black currants, cranberry, and red wine. Phytic acid [inositol hexaphos-
phate(IP6)] is present in most legumes (corn, soy beans, wheat bran, and nuts).
In adipose tissue, polyphenols (i.e., epigallocatechin gallate) have been shown
to exert a thermogenic effect, suppress adipogenesis by reduction of C/EBP-,
PPAR-, and SREBP-1 gene expression, promote apoptosis of adipocytes, and influ-
ence weight control in rodents as well as in humans (21).
Antiangioprevention and antiangiogenic therapy (preventive/therapeutic inhibi-
tion of tumor angiogenesis) are considered efficient strategies for controlling the
growth and metastasis of solid tumors as well as for other diseases involving patho-
logical angiogenesis such as atherosclerosis, rheumatoid arthritis, psoriasis, and
diabetic retinopathy (1,2,9). The broad spectrum of the chemopreventive and anti-
angiogenic mechanisms of the natural compounds has been suggested to explain
this activity. They include the antioxidant, anti-inflammatory, antiproliferative
(directly by the cell cycle inhibition or by the growth factors and hormone intrac-
ellular signaling inhibition), proapoptotic, immune-enhancing, and modification of
xenobiotic phase I/II enzyme induction properties. In addition to, and independently
of, their antioxidant effects, plant polyphenols enhance the production of vasodilating
factors [NO, prostacyclin, and endothelial hyperpolarizing factor (EDHF)], and
inhibit the synthesis of vasoconstrictory, proangiogenic endothelin-1 (ET-1). The
inhibitory effect on the prostate, breast, colon, lung, and the other tumor angiogenesis
and tumor progression, correlates with the inhibition of metalloproteinases (MMP-
2, MMP-9), as well as inhibition of VEGF and its receptor gene and protein induction
in experimental models. Recent studies indicate that VEGF expression and release
are prevented by red wine polyphenols at concentration as low as 3 mg/l. It has been
also reported that flavones: 3,4 dihydroxyflavone, resveratrol, and luteolin are among
the most potent antiendothelial polyphenols in suppression of cell proliferation with
IC50 of 1.4 to 2 M (22).The effect of polyphenols (including resveratrol, a polyphe-
nolic phytoestrogen found in some grapes and wines, or epigallocatechin-3-gallate,
found in green tea) is mediated by the prevention of the growth factors or ischemia-
induced redox-sensitive activation of polyvalent intracellular signaling, including
Stat3, NFB, PI3-kinase/Akt, p38 MAPK, pathways regulating the endothelial
homeostasis, and intracellular Ca2+ concentration (22,23). Genistein and luteolin,
with regard to its antimitotic activity, inhibits the VEGF-induced phosphorylation
of p70 S6 kinase (S6K), a downstream effector of PI3K responsible for G1 phase

progression of cell cycle, but not the VEGF-induced phosphorylation of the extra-
cellular signaling-regulated kinase 1/2 (ERK1/2) in endothelial cells (23).
Resveratrol and the soybean isoflavones bind and increase the transcriptional
activity of estrogen receptors (ERs) and on endothelial cells. Similar to estradiol
(E2), they rapidly increase eNOS gene expression as well as direct NO release by
increasing eNOS activity. Genistein, daidzeoin, and biochanin A, a soy isoflavone,
inhibit the growth and reduce angiogenesis of transplantable bladder cancer in mice
(2428). Genistein is currently in clinical trials as an angiogenesis inhibitor for the
treatment of breast and prostate cancer (24).
However, clinical evidence to support many of the currently claimed health
benefits of this drugs remains to be confirmed. Despite the advantageous effects of
some of these agents demonstrated in vitro and in experimental animal models, their
mostly dosage-related, possible toxic effects must also be of concern at the same
time (25). The presence of different numbers of -OH moieties on the B-ring of the
flavonols may contribute to their antioxidant activity, as well as to their toxicity and
antiangiogenic activity measured by the proliferation, migration, and the primitive
capillary tubular structure formation (24).
Some studies have also demonstrated pro-oxidant activity of flavonoids. Bleomycin-
dependent DNA damage was accelerated by quercetin and kaempferol (but not
naringin), especially in the presence of Fe+3 and the high concentrations of flavonoid
(23). It has been reported that endothelial response to the red wine polyphenols is
critically dependent on a redox-sensitive mechanism involving the formation of
superoxide anions by a flavin-dependent enzyme (22). In chemoprevention of lung
cancer, as well as antiatherogenic beta-carotene was unexpectedly found to increase
the risk of lung cancer among the high-risk groups of patients.
The beta-carotene-induced expression of genes in endothelial as well as endot-
helial progenitor cells in vitro measured by microarray pointed to the activation of
chemotaxis, homing, decrease of connexin 43, induction of the stress, and xenobiotic
phase I and II metabolizing enzymes groups of genes as well as genes involved in
proangiogenic activity (26).
The inhibition of tumorigenesis by caffeine or tea was shown to be closely
related to the reduction of body fat (25). The mechanism of the flavone-induced
antiadipogenic effect was documented to be connected with the blocking insulin
receptor substrate (IRS) from phosphorylation (thus inhibiting glucose uptake by
decrease of GLUT-4), and with increasing lipolysis by the hormone-sensitive lipase
in adipose tissue. Quercetin, catechin, and kaempferol flavonoids inhibit the terminal
differentiation of preadipocytes by decrease of the C/EBP-, PPAR-, and SREBP-1
gene expression (21). Oral administration of green tea and epigallo-catechin galleate
(EGCG) decreases adipose mass, stimulates thermogenesis and fat oxidation, and
decreases lipemia in animals and humans, but at a rather high dose level. In humans
drinking green tea according to Chinese traditions, serum levels of the most abundant
tea catechin compound EGCG are in the range of 0.1 to 0.3 M (21). The intake of
soy proteins (source of genistein and daidzen) by Chinese and Japanese populations
is considerable higher (30 to 35 g/d) than in Western populations (5 to 10 g/d), and the
concentration of genistein in the urine of humans eating a plant-based diet is 30-fold
higher compared to those consuming a traditional Western diet. Thus, the beneficial
effect may be related also to the ethnic diet.
Interestingly, ferrets receiving the high dose of beta-carotene (the source of
retinoic acid) gained more weight than control animals, which is consistent with the
proadipogenic PPAR/RXR/RAR-induced pathway. No such effect was seen with the
low dose of beta-carotene (27). The size of the subcutaneousinguinal fat depot in
animals treated with the high dose of beta-carotene was significantly higher than
that of animals treated with the low dose and slightly higher than that of controls.
This study also showed that chronic treatment with beta-carotene induced a dose-dependent
hypertrophy of white adipocytes and increased neoangiogenesis in subcutaneous
WAT in all treated ferrets. It is in accordance with the suggested proangiogenic
activity of beta-carotene observed by the gene expression in endothelial cells measured
by microarray (26,27).
Beta-carotene treatment reduced the UCP1 protein levels in the interscapular
BAT as well as in the inguinal and retroperitonal WAT depots, pointing to the
accumulation of WAT-type tissue. Interestingly, rodents with knockout ,-carotene-
15,15-mono-oxygenase (Bcmo1-/-) gene accumulate beta-carotene in adipose tissue
and have increased body fat mass, which is consistent with the previously described
possible proangiogenic and proadipogenic effect of beta-carotene (28).
8.3 PERSONALIZED NUTRITION AND ANGIOGENESIS

IN ADIPOSE TISSUE REMODELING
Diet, with its composition (nutrients) and caloric density, is the pivotal regulator of
angiogenesis, which decides fetal tissue development and tissue remodeling during
the whole period of life. Thus, the individual genotype determining the phenotype also
affects angiogenesis-dependent tissue, including tumor, remodeling (29). Thus, at the
present time polymorphisms of the VEGF, MMP, TPA/PAI, and TNF genes seem to
be of great interest among the markers influencing the magnitude of angiogenesis.
Thus, parallel to mechanistic, the epidemiologic (with the appropriately matched
number of participants and case control) studies aim to prove association between
the nutrient/angiogenesis cooperation in regulation of the accumulation of adipose
tissue mass. The application of quantitative trait locus (QTL) analysis and the more
advanced bioinformatic methods may be helpful for the identification of the gene/
chromosomal localization of the genetic background and genotype/nutrient relation
of metabolic and vascular disease interrelationship (30).
KEY READINGS
Carmeliet, P. Angiogenesis in health and disease. Nat Med 9(2003), 65360.
Hausman, G.J. and Richardson, R.L. Adipose tissue angiogenesis. J Anim Sci 2004; 82,
925934.
Rupnick, M.A., Panigrahy, D., Zhang, C.Y., Dallabrida, S.M., Lowell, B.B., Langer, L., and
Folkman, M.J. Adipose tissue mass can be regulated through the vasculature. Proc
Natl Acad Sci USA 2002; 99, 1073010735.
Voros, G., Maguoi, E., Demeulemeester, D., Clerx, N., Collen, D., and Lijen, H.R. Modulation
of angiogenesis during adipose tissue development in murine models of obesity.
Endocrinology 2005; 146, 45454554.
Wang, S., Yehya, N., Schardt, E.E., Wang, H., Drake, T.A., and Lusis, A.J. Genetic and
genomic analysis of a fat mass trait with complex inheritance reveals marked sex
specificity. PLoS Genetics /www.plosgenetix.org; 2006; 2, 01480153.
REFERENCES
1. Carmeliet, P. Angiogenesis in health and disease. Nat Med 9(2003), 65360.
2. Folkman, J., Merler, E., Abernathy, C., and Williams, G. Isolation of a tumor factor
responsible of angiogenesis. J Exp Med 1971; 133, 275288.
3. Hausman, G.J. and Richardson, R.L. Adipose tissue angiogenesis. J Anim Sci 2004;
82, 925934.
4. Rodriguez, A.M., Elabd, C., Amri, Ez-Z., Alhaud, G., and Dani, C. The human adipose
tissue is a source of multipotent stem cells. Biochemistry 2005; 87, 125128.
5. Prunet-Marcassus, B., Cousin, B., Caton, D., Andre, M., Penicaud, L., and Casteilla,
L. From heterogeneity to plasticity in adipose tissues: site-specific differences. Exp
Cell Res 2006; 312, 727736.
6. Mick, G.J., Wang, X., and McCornick, K. White adipocyte vascular endothelial
growth factor; regulation by insulin. Endocrinology 2002; 143, 948953.
7. Dembinska-Kiec, A., Dulak, J., Partyka, Huk, I., and Malinski, T. VEGF-nitric oxide
reciprocal regulation. Nat Med 1997; 3, 1177.
8. Josko, J. and Mazurek, M. Transcription factors having impact on vascular endothelial
growth factor (VEGF) gene expression in angiogenesis. Med Sci Monit 2004; 10, RA8998.
9. Crandall, D.L., Hausman, G.J., and Kral, J.G. A review on the microcirculation of
adipose tissue: anatomic, metabolic and angiogenic perspectives. Microcirculation
1997; 4, 211232.
10. Park, H.Y., Kwon, H.M., Lim, H.J., Hong, B.K., Lee, J.Y., Park, B.E., Jang, Y., Cho,
S.Y., and Kim, H.S. Potential role of leptin in angiogenesis; Leptin induces endothelial
cell proliferation and expression of metaloproteinases in vivo and in vitro. Exp Mol
Med 2001; 33, 95102.
11. Planat-Benard, V., Silvestre, J.S., Cousin, B., Andre, M., Nibbelink, M., Tamarat, R.,
Clerque, M., Manneville, C., Saillan-Barreau, C., Duriez, M., Tedgui, A., Levy, B.,
Penicaud, L., and Casteilla, L. Plasticity of human adipose lineage cells toward endo-
thelial cell: physiological and therapeutic perspectives. Circulation 2004; 109, 656663.
12. Rechman, J., Tractuev, D., Li, J., Merfeld-Clauss, S., Temm-Grove, C.J., Bovenkerk,
J.E., Pell, C.L., Johnstone, B.H., Considine, R.V., and March, K.L., Secretion of
angiogenic and antiapoptotic factors by human adipose stromal cells. Circulation
2004; 109, 12921298.
13. Rupnick, M.A., Panigrahy, D., Zhang, C.Y., Dallabrida, S.M., Lowell, B.B., Langer,
L., and Folkman, M.J. Adipose tissue mass can be regulated through the vasculature.
Proc Natl Acad Sci USA 2002; 99, 1073010735.
14. Moloney, F., Noone, E., Loscher, C., Gibney, M.J., and Roche, H.M. Cis-9, trans 11
conjugated linoleic acid improves metabolic and molecular markers of insulin sen-
sitivity in adipose tissue and liver. Proc Nutr Soc 2004; 64,58A.
15. Voros, G., Maguoi, E., Demeulemeester, D., Clerx, N., Collen, D., and Lijen, H.R.
Modulation of angiogenesis during adipose tissue development in murine models of
obesity. Endocrinology 2005; 146, 45454554.
16. Rosen, ED. and Spiegelman, B.M. Molecular regulation of adipogenesis. Annu Rev
Cell Dev Biol 2000; 16, 145171.
17. Fukumura, D., Ushiyama, A., Duda, D.D., Xu, L., Tam, J., Chatterjee, V.K.K.,
Garkavtsev, I., and Jain, R.K. Paracrine regulation of angiogenesis and adipocyte
differentiation during in vivo adipogenesis. Circ Res 2003; 93, e88e97.
18. Kim, J.E. and Chen, J. Regulation of peroxisome proliferator-activated receptor-
activity by mammalian target of rapamycin and amino acids in adipogenesis. Diabetes
2004; 53, 27482756.
19. Viguerie, N., Poitou, C., Cancello, R., Stich, V., Clement, K., and Langin, D.
Transcriptomics applied to obesity and caloric restriction. Biochemie 2005; 87,
117123.
20. Novakofski, J. Adipogenesis: usefulness of in vitro and in vivo experimental models.
J Anim Sci 2004; 82, 905915.
21. Dulloo, A.G., Duret, C., Rohrer, D., Girardier, L., Mensi, N., Fathi, M., Chantre, P.,
and Vandermander, J. Efficacy of a green tea extract rich in catechin polyphenols and
caffeine in increasing 24-h energy expenditure and fat oxidation in humans. Am J
Clin Nutr 1999; 70, 10401045.
22. Stoclet, J.C., Chataigneau, T., Ndiaye, M., Oak, M.H., Bedoui, J.E., Chataigneau,
M., and Shini-Kert, V.B. Vascular protection by dietary polyphenols. Eur J Pharmacol
2004; 500, 299313.
23. Bagli, E., Stefaniotou, M., Morbidelli, L., Ziche, M., Psillas, K., Murphy, C., and
Fosis, T. Luteolin inhibits vascular endothelial growth factor-induced angiogenesis;
inhibition of endothelial cell survival and proliferation by targeting phosphatidi-
linositol 3-kinase activity. Cancer Res 2004; 64, 79367946.
24. Cao, Y. and Cao, R. Brakenhielm Antiangiogenic mechanism of diet-derived polyphe-
nols. J Nutr Biochem 2002; 13, 380390.
25. Mennen, L.I., Walker, R., Bennetau-Pelissero, C., and Scalbert, A. Risk and safety
of polyphenol consumption. Am J Clin Nutr 2005; 81, 326S329S.
26. Dembinska-Kiec, A., Polus, A., Kiec-Wilk, B., Grzybowska, J., Mikolajczyk, M.,
Hartwich, J., Razny, U., Szumilas, K., Banas, A., Bodzioch, M., Stachura, J., Dyduch,
G., Laidler, P., Zagajewski, J., Langman, T., and Schmitz, G. Proangiogenic activity
of beta-carotene is coupled with the activation of endothelial cell chemotaxis. Biochim
Biophys Acta May 30, 2005; 1740(2): 22239.
27. Luisa Bonet Murano, I., Morroni, M., Zingaretti, M.C., Oliver, P., Sanchez, J., Fuster,
A., Pico, C., Palou, A., and Cinti, S. Morphology of ferret subcutaneous adipose
tissue after 6-month daily supplementation with oral beta-carotene. Biochim Biophys
Acta May 30, 2005; 1740(2): 30512. Epub November 17, 2004.
28. von Lintig, J., Hessel, S., Eichnger, A., Isken, A., Oberhauser, V., Vogt, K., Goralczyk,
R., and Wyss, A. Analysis of Bcmo1 knock-out mouse model uncovers a role of -
carotene for the regulation of lipid metabolism. In Oxidants and Antioxidants in
Biology. Abstract book of the Oxygen Club of California. Santa Barbara; March
1518, 2006, p 21.
29. Balasubramanian, S.P., Brown, N.J., and Reed, M.W.R. Role of genetic polymor-
phisms in tumor angiogenesis. Br J Cancer 2002; 87, 10571065.
30. Wang, S., Yehya, N., Schardt, E.E., Wang, H., Drake, T.A., and Lusis, A.J. Genetic
and genomic analysis of a fat mass trait with complex inheritance reveals marked
sex specificity. PLoS Genetics /www.plosgenetix.org; 2006; 2, 01480153.
9 Metabolic Programming
during Pregnancy:
Implications for
Personalized Nutrition
Simon C. Langley-Evans
CONTENTS
9.1 Introduction ..................................................................................................101

9.2 Nutritional Programming of Health and Disease........................................102
9.2.1 Evidence from Epidemiologic Studies ............................................102
9.2.2 Evidence from Experimental Studies ..............................................104
9.2.3 Mechanistic Basis of Programming.................................................106
9.3 Programming and the Fetal NutritionGenotype Interaction .....................108
9.3.1 Type 2 Diabetes................................................................................108
9.3.2 Osteoporosis .....................................................................................110
9.4 Future Perspectives for Programming and Personalized Nutrition.............111
Key Readings.........................................................................................................112
References..............................................................................................................113
9.1 INTRODUCTION
This chapter will introduce the concept that nutritional factors before birth impact
disease risk in later life. Evidence from epidemiologic and experimental studies
demonstrates that variation in nutrition during fetal development exerts a program-
ming effect on tissue development and metabolic/physiological status in adult life.
These programming effects have profound implications for the attainment of meta-
bolic phenotypes that promote the development of the metabolic syndrome and
associated disease outcomes. There is growing evidence of interactions between the
genotype and indicators of fetal growth and nutrition in determining this disease
risk phenotype. For example, a well-characterized association between low birth
weight and risk of type 2 diabetes in adult life is dependent on the pro12ala
polymorphism of the peroxisome proliferators activated receptor 2. Similarly, the
established association between the Bsm1 polymorphism of the vitamin D receptor
101
and degenerative disease of bone appears to be modified by fetal growth-related

factors. In considering how this knowledge might be applied to the development of
personalized nutrition, the identification of simple markers of nutritionally pro-
grammed disease phenotypes may be an important adjunct to screening for genotypic
biomarkers. The development of sensitive biomarkers of programmed events using
nutrigenomics approaches should be an important priority for future research in the
metabolic programming field.
9.2 NUTRITIONAL PROGRAMMING OF HEALTH

AND DISEASE
The environment encountered during fetal life and infancy is strongly related to risk
of noncommunicable diseases in adult life. This association of early life factors with
conditions that may not develop for another four decades or more may be explained
in terms of adaptations during critical phases of growth and development that ensure
the maintenance of homeostasis and hence survival when the environment is com-
promised. The means through which events in early life trigger permanent responses
have been described as nutritional or metabolic programming. These terms describe
the process through which a stimulus or insult during a critical window of fetal or
infant development elicits permanent responses that produce long-term changes in
tissue structure or function (Langley-Evans, 2004).
There may be a number of potential insults that initiate such intrauterine adap-
tations. Maternal stress, infection, exposure to toxins, and hypoxia have all been
suggested as agents that may promote long-term physiological change to fetal phys-
iology. For most of the human population, however, the most likely insult will be
variation in nutrient supply during early development. Whatever may drive the
process, programming occurs because of the innate capacity of developing tissues
to adapt to the environment they encounter. For almost all cell types, in all organs,
this plasticity is a characteristic that is present for only a short period before the
time of birth (Gluckman and Hanson, 2004). As a result the prenatal period represents
a unique life-stage in which exposure to suboptimal nutrition may shape the lifelong
capacity of the organism to respond to the environment and, consequently, the long-
term disease risk phenotype.
9.2.1 EVIDENCE FROM EPIDEMIOLOGIC STUDIES

The concept that human disease states may arise, at least in part, through exposure
to adverse programming in utero arose from the findings of a broad range of
epidemiologic studies indicating associations between disease risk and infant char-
acteristics at birth. Many cohort studies have demonstrated that there are inverse
associations between birth weight and risk of coronary heart disease, stroke, hyper-
tension, glucose intolerance, type 2 diabetes, and the metabolic syndrome (Barker,
2004). These relationships are also observed in relation to body proportions at birth.
For example, babies with low ponderal index at birth have significantly greater risk
of insulin resistance and diabetes in adulthood, whereas babies born with a large
head circumference relative to body length appear to be at greater risk of later atopy
Metabolic Programming During Pregnancy 103
(Table 9.1). Low birth weight and disproportion at birth are essentially simple
markers of constrained growth in utero. Such growth impairment is therefore pro-
posed to be central to programming of disease risk, but the effects of intrauterine
growth restriction are also modulated by postnatal growth rates. Thus, the individual
at greatest risk of metabolic syndrome is likely to have been born small but undergone
rapid catch-up growth in infancy, attaining an increased body mass index in adolescence
(Eriksson et al., 2003a).
The relationship between risk of coronary heart disease, non-insulin-dependent
diabetes, and birth weight has been attributed to the effects of variation in maternal
nutritional status on fetal growth. However, this explanation at first glance appears
tenuous as in human populations the association between maternal nutrient intake and
birth weight is difficult to demonstrate (McMillen and Robinson, 2005). It is estimated
that maternal nutrient intake only explains approximately 6% of the variation in birth
weight. In extreme situations, such as the wartime Dutch Hunger Winter and siege of
Leningrad, starvation had only minor effects on birth weight. In well-nourished pop-
ulations, although isolated studies show that intakes of animal protein or sucrose are
related to birth weight and placental weight, the majority of studies find no significant
relationships between maternal diet and weights of either baby or placenta.
Although the lack of clear associations between maternal nutrient intake and
birth weight, or proportions at birth, appears to undermine the nutritional program-
ming hypothesis, there are important studies that directly demonstrate associations
between maternal nutritional status in pregnancy and disease-risk markers in the
resulting offspring. A Jamaican study showed that higher blood pressure in prepu-
bescent boys was predicted by low maternal hemoglobin, weight gain, and triceps
skin fold thicknesses in pregnancy (Godfrey et al., 1994). A study of men born in
Aberdeen in the 1950s found that blood pressure was related to maternal intakes of
animal protein and of carbohydrate (Campbell et al., 1996). Data from the ongoing
TABLE 9.1
Associations between Anthropometric Indices at Birth and Markers
of Disease Risk in Human Populations
Birth Characteristic Associated Disease State Associated Risk Factor
Low birth weight Coronary heart disease, type 2 Obesity, hypertension, insulin
diabetes, asthma, COPD, resistance, poor response to
osteoporosis, depression, vaccination, metabolic syndrome,
schizophrenia, stroke, cataracts raised clotting factors, raised LDL-
cholesterol
Low ponderal indexa Type 2 diabetes, schizophrenia Insulin resistance, hypertension,
hypercortisolism
Reduced abdominal Hypertension, raised LDL-
circumference cholesterol, hyperinsulinemia
Large head circumferenceb Asthma, eczema Raised IgE
a Ponderal index is a marker of relative thinness at birth (birth weight, kg/length, m3).
b Corrected for crown-heel length. COPD = chronic obstructive pulmonary disease.
Project Viva study in the U.S. also appear to support the concept of nutritional
programming. For example, higher maternal calcium intakes in pregnancy result in
lower blood pressure in 6-month-old babies (Gillman et al., 2004).
The inconsistencies in the literature relating to maternal diet and birth weight
almost certainly arise because of the lack of appropriate methodology to assess the
nutrient supply the fetus actually gets. This is partly determined by maternal intake,
but will also depend on maternal stores going in to pregnancy, the metabolic demands
of the mother, and the perfusion of the placenta. These are all problems that are best
addressed by experimental animal models. These have been widely used to investi-
gate the plausibility of the programming concept and to consider the mechanistic
basis of nutritional programming.
9.2.2 EVIDENCE FROM EXPERIMENTAL STUDIES

In addition to the epidemiologic evidence suggesting that human cardiovascular
disease may be programmed by the intrauterine environment, a number of animal
models have been developed to study the process and its underlying mechanisms.
It has been consistently noted in rats, mice, and guinea pigs that fetal exposure to
undernutrition produces elevated blood pressure (Langley-Evans, 2006). For example,
the feeding of a protein-restricted diet in rat pregnancy results in elevations to systolic
blood pressure in the resultant offspring that are of the order of 15 to 30 mm Hg. The
elevation of blood pressure occurs despite the fact that the maternal dietary manip-
ulation is sufficiently mild to produce offspring that are of normal weight at birth.
Blood pressure is elevated from the age of weaning, and the magnitude of increase
depends on the timing of undernutrition during pregnancy (Figure 9.1). Similar
140

120

Systolic BP (mm Hg)
100
80
Male
Female
60
40
20
0
Control LP d0-7 LP d8-14 LP d15-22 LP d0-22
Maternal diet
FIGURE 9.1 Systolic blood pressure in 4-week-old rats exposed to low-protein diet during
specific periods of pregnancy. Undernutrition at any stage of gestation increases later blood
pressure, and these effects persist into adulthood. Figure reproduced from data published in
Langley-Evans et al. (1996). LP = low-protein diet. Full rat gestation is 22 d (d 0 to 7 represents
targeting of protein restriction to early gestation, d 8 to 14 to midgestation, d 15 to 22 to late
gestation). * denotes P < 0.05 compared to control group.
observations in large animal species, such as the sheep, suggest that programming
of cardiovascular function occurs in all mammals.
A broad spectrum of nutritionally programmed disorders have been reported
from animal studies, but in general such experiments uncover blood pressure, glucose
intolerance, insulin resistance, and increased adiposity as common endpoints of
nutritional interventions targeted at pregnancy, or specific periods in pregnancy
(Table 9.2). It is becoming increasingly apparent that in addition to programming
disease states or their direct risk factors (glucose intolerance, adiposity, and hyper-
tension), alterations to the quality or quantity of rodent diets in pregnancy determine
behavioral antecedents of metabolic risk. For example, food restriction and protein
TABLE 9.2
Programming Effects of Maternal Undernutrition in Animal Models
Observed Phenotype Maternal Diet Manipulation
in Offspring in Pregnancy Species
Hypertension Food restriction Rat, sheep, and guinea pig
Low-protein diet Rat and mouse
Uterine artery ligation Rat
High saturated fat Rat
Iron deficiency Rat
Calcium deficiency Rat
Insulin resistance Food restriction Rat and guinea pig

Low-protein diet Rat
Disordered lipid metabolism Low-protein diet Rat
Obesity Postnatal overfeeding Rat

Food restriction Rat, sheep, and guinea pig
Low-protein diet Rat and mouse
High-protein diet Rat
Altered appetite Postnatal overfeeding Rat

Low-protein diet Rat
Bone disorders Low-protein diet Rat
Renal injury Low-protein diet Rat

Food restriction Rat, sheep
Shorter life span Low-protein diet Rat, mouse
Impaired immune function Low-protein diet Rat

Zinc deficiency Mouse
restriction in rat pregnancy have both been shown to impact feeding and locomotor
behavior in the offspring (Vickers et al., 2003).
Working with species that have relatively short life span and gestation allows
evaluation of intergenerational effects and consideration of interactions between
dietary exposures at different developmental stages in pre- and postnatal life. Using
this approach, it has been possible to show that although the combination of under-
nutrition before birth combined with overnutrition in adulthood does not increase
risk of obesity, the introduction of overnutrition in the early postnatal (suckling and
weaning stages) following prenatal restriction can induce extreme central adiposity
(Langley-Evans et al., 2005). Such animal models, although varying in the detailed
approaches adopted, importantly underpin the nutritional programming hypothesis,
demonstrate that programming of cardiovascular changes can occur even in the
absence of fetal growth retardation, and provide useful tools with which to study
the mechanisms involved in nutritional programming.
9.2.3 MECHANISTIC BASIS OF PROGRAMMING

The developmental origins of health and disease hypothesis is a concept that raises
interesting possibilities in terms of the prevention and treatment of disease. One of
the simplest approaches may be to intervene in pregnancy, identifying women most
at risk of bearing children who are programmed for later disease and providing
appropriate advice for dietary change. Alternatively, given appropriate markers to
indicate that adverse programming has occurred, it may be possible to target at-risk
individuals with guidance for nutritional change, or with early pharmaceutical inter-
vention to counteract programmed effects. All of these strategies will require an
understanding of the mechanistic basis of programming.
At the simplest level, nutritional programming in early life is likely to involve
remodeling of key organs and tissues. All tissues develop as a consequence of
proliferation and differentiation of progenitor cell lines. Disruption of these processes
will result in tissues with either a reduced cell number, with altered cell types, or
both of these outcomes. Evidence of this can be seen in the offspring of animals
fed low-protein diets in pregnancy, which exhibit reduced numbers of nephrons in
the kidney, reduced number and vascularization of Islets of Langerhans in the
pancreas, and altered neuronal densities in the hypothalamus (Langley-Evans, 2006).
All of these tissue-remodeling effects could explain many of the observed disease
risk phenotypes due to nutritional programming.
For most species, and in particular for humans, relatively mild variations in mater-
nal nutritional status are unlikely to significantly perturb fetal nutritional status. Release
of maternal stores along with adaptive responses to increase placental blood flow and
nutrient transfer will buffer the fetus against all but the most severe fluctuations in
maternal nutrition. It is difficult, therefore, to see how changes in nutrient intake lead
to the development of major disease in the adult offspring. In the search for mechanisms
that may drive nutritional programming effects, attention has focused on the contri-
bution of relatively subtle shifts in metabolic pathways and on the actions of stress
hormones as vehicles for programming effects. In both cases, the true drivers of
programming are environmentally triggered changes in fetal gene expression.
Variation in nutritional status may act as a physiological stressor. Stress responses

mediated via the hypothalamicpituitaryadrenal axis may impact the developing
fetal tissues, resulting in permanent programmed responses. In most species there
is a gradient in concentrations of active glucocorticoid hormones across the placenta.
For example, in the rat, maternal corticosterone concentrations are between 100-
and 1000-fold greater than in the fetal compartment. This gradient is maintained by
the activity of 11-hydroxysteroid dehydrogenase type 2 (11HSD2), which converts
active glucocorticoids to inactive forms (Seckl and Meaney, 2004). Animal studies
show that 11HSD2 is subject to nutritional regulation. Undernutrition in pregnancy
therefore leads to downregulation of this gatekeeper enzyme in placenta and, con-
sequently, overexposure of the fetal tissues to active glucocorticoids of maternal
origin (Langley-Evans, 2006).
Glucocorticoids are potent regulators of gene expression, and it is proposed that
exposure of the fetal tissues to inappropriate hormone concentrations disrupts the
normal developmental patterns of gene expression and resets homeostatic processes
in the long term through tissue remodeling and modifications to cellcell signaling
pathways. Experiments in which pregnant animals are treated with synthetic gluco-
corticoids that are not substrates for 11HSD2 or with inhibitors of 11HSD2 largely
duplicate the programmed phenotypes that result from variation in maternal nutrition
(Seckl and Meaney, 2004; Langley-Evans, 2006). This strengthens the argument that
at least some of the long-term effects of suboptimal nutrition in pregnancy are
mediated by glucocorticoids.
Gene silencing through processes such as DNA methylation and histone acetyla-
tion may also provide mechanisms through which early life nutrition exerts long-term
programming effects. Waterland and Jirtle (2003) demonstrated using the Agouti
mouse that the expression of transposable elements controlling coat color could be
switched on or off by varying the supply of methyl donors during pregnancy. The
postnatal period also represents a period of plasticity during which dietary or other
environmental stimuli may determine the methylation status of genes and, hence,
their long-term expression (Weaver et al., 2004).
The rat model of maternal low-protein feeding has been one of the most
widely studied and best-characterized experimental protocols used for the study
of nutritional programming (Langley-Evans, 2006). It has been demonstrated
that the deleterious effects of fetal exposure to protein restriction can be, at
least in part, prevented by the addition of folic acid or glycine to the maternal
diet. Folic acid, or more specifically tetrahydrofolate, is the major donor of
methyl groups for DNA methylation, whereas glycine plays a critical role in
folate metabolism during pregnancy. Maternal protein restriction increased
hepatic expression of glucocorticoid receptor (GR) and the peroxisome prolif-
erator-activated receptor in young rats, and these increases were associated
with reduced DNA methylation of these genes (Lillycrop et al., 2005). However,
nutritional modulation of DNA methylation during fetal life does not explain
all aspects of early programming, as some phenotypic effects of maternal
protein restriction are not ameliorated by glycine and folate supplementation,
and some long-term changes in gene expression are not associated with DNA
hypomethylation.
9.3 PROGRAMMING AND THE FETAL

NUTRITIONGENOTYPE INTERACTION
Unfortunately, there are no studies that have been able to consider the interaction
of maternal nutrient intakes or specific nutrient deficiencies in human pregnancy
with known polymorphisms that associate with disease states in adulthood. Thus,
all literature in this field depends on the use of infant anthropometry, either at birth,
or at 1 year of age, as a proxy for the fetal and infant environment. Although weight
at birth or markers of disproportion at birth (for example, a large head circumference
in relation to overall stature) may be seen as rather crude, insensitive indicators of
fetal nutritional status, weight at 1 year provides a reasonable proxy for nutritional
status during infancy, although this will always be confounded by the lack of information
on genetic growth potential.
These caveats notwithstanding, there are a number of interesting studies that
indicate there are strong interactions between the fetal or infant environment and
the genotype. Thus, programming effects can modulate disease risk associated with
a particular genotype and vice versa. The association of low birth weight with disease
may be observed only in individuals carrying particular alleles. Most of the studies
considering these issues have investigated genotype-programming interactions in
relation to either type 2 diabetes or bone health.
9.3.1 TYPE 2 DIABETES

The peroxisome proliferator-activated receptor (PPAR) is a ligand-dependent
transcription factor that is predominantly expressed in adipose tissue and plays an
important role in adipogenesis, fat, and energy metabolism. To date, three isoforms
of PPAR (-1, -2, and -3) have been identified, and each is expressed by different
tissues. PPAR -2 is predominantly expressed in adipose tissue and contains a
polymorphic region in the isoform specific exon B (Eriksson et al., 2002). The Ala
allele of this pro12ala polymorphism has been proposed to have a protective role in
type 2 diabetes, but the association of this allele with improved insulin sensitivity
and lower diabetes risk is somewhat variable and is likely to depend on unidentified
lifestyle factors. Eriksson and colleagues, working with unique cohorts of men and
women born in Helsinki between 1924 and 1933, have assessed the interaction of
the pro12ala polymorphism with markers of prenatal growth in determining adult
insulin sensitivity, cholesterol metabolism, and risk of cardiovascular disease.
These studies demonstrated that the Ala12 allele was associated with signifi-
cantly lower fasting insulin and HOMA-IR index, but that this beneficial effect of
the polymorphism was isolated to individuals who had been of lower weight at birth
(Eriksson et al., 2002). In other words, the detrimental effect of the Pro12Pro
genotype was only seen if weight at birth was low. The genotypefetal environment
interaction worked both ways, and the association between weight at birth and insulin
resistance, which has been so frequently reported in other cohorts (Barker, 2004),
was only seen in individuals with a Pro12Pro genotype (Figure 9.2). The same
genotype was shown by Eriksson et al. (2003b) to associate with total and HDL-
and LDL-cholesterol concentrations in the same subjects. Individuals with the
Pro12Pro genotype were also more likely than individuals with at least one Ala12
PPAR genotype
25
Homa-IR index
20
15
10
5 PP
0 PA or AA
<3.0 3.03.5 >3.5
Birthweight (kg)
FIGURE 9.2 The interaction of genotype for PPAR with birth weight in determining adult
insulin sensitivity. Subjects genotyped according to the Pro12Ala polymorphism in the PPAR
gene were grouped according to weight at birth. The Pro12Pro (PP) variant was associated with
raised HOMA-IR only in subjects in the two lowest birth weight groups (P < 0.005 comparing
PP and PA or AA genotype for birth weight less than 3.0 kg, P = 0.03 for birth weight 3.0 to
3.5 kg). Birth weight was only related to adult HOMA-IR (P = 0.002) in the population subset
with the PP genotype. Figure produced from data reported by Eriksson et al., 2002.
allele to have higher blood pressure and to require hypertensive medication, but once
again this detrimental effect of the genotype was observed only in individuals who
had been of lower weight or shorter stature at birth (Yliharsa et al., 2004).
Overall, these studies of the Pro12Ala polymorphism in PPAR suggest that a
single genotype can give rise to different phenotypes because of variation in early
life experience. This evidence of programmed modulation of the effects of disease-
related alleles is of great importance when considering strategies for personalized
nutrition.
The same considerations are clearly indicated by studies of the interaction between
early life characteristics and the effects of the K121Q polymorphism of plasma cell
glycoprotein 1 (PC1). This is a strong candidate gene for non-insulin-dependent
diabetes, as PC1 is an inhibitor of insulin signaling downstream from the insulin
receptor. The K121Q polymorphism in exon 4 of this gene is associated with
variation in the inhibitory effects of PC1. The 121K variant lowers the inhibitory
action, whereas 121Q enhances the inhibition and is consequently associated with
insulin resistance, raised plasma glucose, and hyperinsulinemia. In the study of
Kubaszek et al. (2004) no variation in insulin concentrations or HOMA-IR index
was noted between individuals with a K121K genotype and those with at least one
121Q allele. However, once grouped by weight at birth, it was apparent that the
121Q allele had adverse effects. Adults that had the lowest weights at birth had
greater insulin concentrations and HOMA-IR index if they carried the 121Q variant.
Most importantly, although there was no variation in prevalence of diabetes or
hypertension that could be attributed solely to PC1 genotype, in individuals who
were of short stature at birth (birth length below 49 cm), the 121Q allele increased
prevalence of hypertension by threefold and type 2 diabetes by twofold. In this study
there was no apparent effect of the genotype until the interaction with a proxy of
fetal growth was considered in the analysis. Thus, without anthropometric data from
birth, genotyping for the PC1 polymorphism would be of little use in the application
of personalized nutritional advice.
Human angiotensin converting enzyme (ACE) has an insertion/deletion (I/D)
polymorphism that appears to impact both circulating and tissue activities of the
enzyme. The insertion is a 287 bp fragment, the presence of which lowers ACE
activity and apparently increases risk of vascular injury in diabetic subjects (Kajantie
et al., 2004). However, the effects of this very common allele (the ID form pre-
dominates in populations, whereas II is the rarest variant) are inconsistent and, as
with the PPAR Pro12Ala polymorphism, this is probably explained by geneenvi-
ronment factors. Kajantie et al. (2004) found that the DD variant was associated
with lower weight, smaller head circumference, and shorter length at birth, imply-
ing that this variant of ACE plays a role in regulation of fetal growth. In adult
life, subjects with the II variant exhibited greater insulin responses to a glucose
tolerance test than those with DD, but this difference was only observed in those
of lower weight at birth. In this case the polymorphic allele may confound the
association between birth weight and disease, because the I allele is itself associ-
ated with shorter gestation and increased weight at birth. However, this study is
a further example of how the effects of a polymorphism on a diabetes risk marker
are modulated by early growth.
9.3.2 OSTEOPOROSIS
Mapping of the genetics of bone metabolism is well advanced, and a number of
putative risk-determining genes have been identified, including the vitamin D receptor
(VDR), the estrogen receptors, and the type 1 collagen A1 gene (Walker-Bone
et al., 2002). Polymorphisms at these loci have been identified, including 22 separate
polymorphisms of VDR, but population studies suggest that these explain only a
small proportion of variation in bone mass. For example, it is estimated that the BB
variant of the Bsm I restriction site in VDR can reduce site-specific bone mineral
density by at most 2%.
The influence of genotype on bone health is clearly modulated by lifestyle and
environmental factors. As with diabetes, any consideration of such factors has to
include some assessment of proxies for early life nutrition and growth. Several
studies have suggested that growth in utero and in the first year of infancy determine
osteoporosis and osteoarthritis (Walker-Bone et al., 2002). In a study of the Bsm 1
polymorphism of VDR, Jordan and colleagues (2005) reported that the B allele
increased severity of osteophytosis (the outgrowth of immature bony processes,
reflecting the presence of degenerative disease) in the lumbar region. In men, but
not women, there was an interaction of the VDR genotype with weight at birth in
determining the severity of the lumbar degeneration. Thus, fetal factors modified
the risk associated with this particular VDR genotype.
The tt variant of the Taq1 polymorphism of VDR is also associated with disease
risk. Individuals with a tt genotype have lower bone mineral density and greater
prevalence of fractures. Adult bone mineral density is strongly a function of the peak
bone mass attained in the third decade of life, and this to some extent depends on
earlier rates of growth. Keen et al. (1997) reported that the tt variant of VDR was
associated with weight at 1 year, suggesting that patterns of skeletal growth even in
the very early years are determined by the VDR genotype. The early environment
may therefore interact with genotype to determine rates of skeletal growth, and the
two factors together program adult bone characteristics.
SNPs have also been identified in the growth hormone gene cluster, which in
addition to the growth hormone (GH1) gene, includes placental lactogen, placental
growth hormone, and the chorionic somatomammotrophin genes. Polymorphisms
of GH1 are associated with variability in bone mineralization. Dennison et al. (2004)
found that over a 4 year period, the loss of bone from the lumbar spine and proximal
femur was greatest in individuals with the 2 allele of the GHV1 and GHV4 loci.
Individuals with these unfavorable allele variants also exhibited blunted growth
hormone secretory profiles. However, these effects depended entirely on an interac-
tion with weight at age 1 year. Individuals in the lowest tertile of infant weight
showed a strong association between GHV1 genotype and rates of bone loss, whereas
in adults who had been heavier infants, GHV1 genotype had no effect on bone loss.
These data show that the environment encountered in infancy can modulate the effect
of GH1 genotype on bone health up to seven decades later.
9.4 FUTURE PERSPECTIVES FOR PROGRAMMING

AND PERSONALIZED NUTRITION
The developmental origins of health and disease hypothesis, and more specifically
the concept that variation in maternal nutritional status can program disease risk,
has important implications for the future management of the health of individuals
and populations. Future goals that arise from research in this area will include
appropriate dietary interventions in pregnancy, early screening and selective treat-
ment of individuals with programmed disease risk, and the development of new
drugs to counteract programming.
With an improved understanding of the life-course basis of disease, it is becoming
clear that disease-causing metabolic disturbances do not just arise as a consequence
of accumulated experience of fetal and postnatal nutrition and the environment.
At the heart of the balance between health and disease lies a complex genetics
programmingmature environment interaction (Figure 9.3). The data from the stud-
ies of the ACE and PPAR polymorphisms described earlier clearly suggest that
certain genotypes may make the impact of programming more powerful. The con-
verse will also be true, as programming may exacerbate the impact of a polymor-
phism, and could just as likely nullify the impact of the polymorphism. Findings of
this nature should lead us to question whether measuring polymorphisms is of any
value in determining personalized nutrition strategies.
In any future development of personalized nutrition strategies, it will be neces-
sary to make assessments of markers of programming, just as effectively as assessing
biomarkers of the genotype. In the programming field we currently lag behind in
this regard as generally the only data that are routinely available are crude anthro-
pometric estimates of fetal growth. The combination of poor intrauterine growth and
childhood catch-up is, however, clearly a warning sign that could already be factored
into personalized nutrition advice for individuals with high-risk genotypes.
Infancy Adult Lifestyle Factors

Growth, nutrition Nutrition, physical activity, environment
The adult disease

risk prole is the
DISEASE-RISK
GENOTYPE product of cumulative
PHENOTYPE
experiences from
conception onward
Fetal Childhood &

Experience Adolescence
Growth, nutrition, Growth, nutrition
hormones, stress lifestyle
FIGURE 9.3 Disease risk in adult life is a product of cumulative experience across the life
course. Geneenvironment interactions occur at all stages of life. The product of this interaction
in early life will determine the strength and nature of such responses in adulthood.
To date the emphasis of the research agenda in the field of metabolic program-
ming has been on demonstrating the biological principle and defining the mechanistic
basis of programming. The time has come to consider the possible nutritional
interventions that might reverse adverse programming effects. To construct appro-
priate personalized nutrition interventions, we require better-quality information
about the characteristics of a programmed disease-risk phenotype. More studies need
to utilize nutrigenomic tools, particularly proteomics and metabolomics, to focus on
the cellular and metabolic consequences of an adverse nutritional environment in
utero or during early infancy. One of the major challenges of applying these
approaches, at least in human studies, will be to separate the effects of early life
programming and the interaction of early life events with the genome from the
effects of lifestyle factors in adult life.
KEY READINGS
Barker, D.J.P. (1998). Mothers, babies and health in later life. Churchill Livingstone. London.
Gluckman, P.D. and Hanson, M.A. (2006). Developmental origins of health and disease.
Cambridge University Press. Cambridge, U.K.
Langley-Evans, S.C. (2004). Fetal Nutrition and Adult Disease: Programming of chronic
disease through fetal exposure to undernutrition. CABI. Wallingford, U.K.
REFERENCES
Barker, D.J.P. (2004). The developmental origins of adult disease. J Am Coll Nutr. 23(6
Suppl.), 588S595S.
Campbell, D.M., Hall, M.H., Barker, D.J., Cross, J., Shiell, A.W., and Godfrey, K.M. (1996).
Diet in pregnancy and the offsprings blood pressure 40 years later. Br J Obstet
Gynaecol. 103, 273280.
Dennison, E.M., Syddall, H.E., Rodriguez, S., Voropanov, A., Day, I.N., and Cooper, C.
(2004). Polymorphism in the growth hormone gene, weight in infancy, and adult bone
mass. J Clin Endocrinol Metab. 89, 48984903.
Eriksson, J.G., Lindi, V., Uusitupa, M., Forsen, T.J., Laakso, M., Osmond, C., and Barker,
D.J. (2002). The effects of the Pro12Ala polymorphism of the peroxisome proliferator-
activated receptor-gamma2 gene on insulin sensitivity and insulin metabolism interact
with size at birth. Diabetes 51, 23212324.
Eriksson, J.G., Forsen, T.J., Osmond, C., and Barker, D.J. (2003a) Pathways of infant and
childhood growth that lead to type 2 diabetes. Diabetes Care. 26, 30063010.
Eriksson, J., Lindi, V., Uusitupa, M., Forsen, T., Laakso, M., Osmond, C., and Barker, D. (2003b).
The effects of the Pro12Ala polymorphism of the PPARgamma-2 gene on lipid metab-
olism interact with body size at birth. Clin Genet. 64, 366370.
Gillman, M.W., Rifas-Shiman, S.L., Kleinman, K.P., Rich-Edwards, J.W., and Lipshultz, S.E.
(2004). Maternal calcium intake and offspring blood pressure. Circulation 110,
19901995.
Gluckman, P.D. and Hanson, M.A. (2004). Living with the past: evolution, development, and
patterns of disease. Science 305, 17331736.
Godfrey, K.M., Forrester, T., Barker, D.J., Jackson, A.A., Landman, J.P., Hall, J.S., Cox, V.,
and Osmond, C. (1994). Maternal nutritional status in pregnancy and blood pressure
in childhood. Br J Obstet Gynaecol. 101, 398403.
Jordan, K.M., Syddall, H., Dennison, E.M., Cooper, C., and Arden, N.K. (2005). Birthweight,
vitamin D receptor gene polymorphism, and risk of lumbar spine osteoarthritis. J
Rheumatol. 32, 678683.
Kajantie, E., Rautanen, A., Kere, J., Andersson, S., Yliharsila, H., Osmond, C., Barker, D.J.,
Forsen, T., and Eriksson, J. (2004). The effects of the ACE gene insertion/deletion
polymorphism on glucose tolerance and insulin secretion in elderly people are modified
by birth weight. J Clin Endocrinol Metab. 89, 57385741.
Keen, R.W., Egger, P., Fall, C., Major, P.J., Lanchbury, J.S., Spector, T.D., and Cooper, C.
(1997). Polymorphisms of the vitamin D receptor, infant growth, and adult bone mass.
Calcif Tissue Int. 60, 233235.
Kubaszek, A., Markkanen, A., Eriksson, J.G., Forsen, T., Osmond, C., Barker, D.J., and
Laakso, M. (2004). The association of the K121Q polymorphism of the plasma cell
glycoprotein-1 gene with type 2 diabetes and hypertension depends on size at birth.
J Clin Endocrinol Metab. 89, 20442047.
Langley-Evans, S.C., Welham, S.J.M., Sherman, R.C., and Jackson, A.A. (1996) Weanling
rats exposed to maternal low protein diets during discrete periods of gestation exhibit
differing severity of hypertension. Clin Sci. 91, 607615.
Langley-Evans, S.C., Bellinger, L., and McMullen, S. (2005). Animal models of program-
ming. Maternal Child Nutr 1, 142148.
Langley-Evans, S.C. (2004). Fetal programming of adult disease: an overview. In Fetal
nutrition and adult disease: Programming of chronic disease through fetal exposure
to undernutrition. Ed. S.C. Langley-Evans. CABI, Wallingford, pp. 120.
Langley-Evans, S.C. (2006). Developmental programming of health and disease. Proc Nutr
Soc. 65, 97105.
Lillycrop, K.A., Phillips, E.S., Jackson, A.A., Hanson, M.A., and Burdge, G.C. (2005). Dietary
protein restriction of pregnant rats induces and folic Acid supplementation prevents
epigenetic modification of hepatic gene expression in the offspring. J Nutr. 135,
382386.
McMillen, I.C. and Robinson, J.S. (2005). Developmental origins of the metabolic syndrome:
prediction, plasticity, and programming. Physiol Rev. 85, 571633.
Seckl, J.R. and Meaney, M.J. (2004). Glucocorticoid programming. Ann NY Acad Sci. 1032,
6384.
Vickers, M.H., Breier, B.H., McCarthy, D., and Gluckman, P.D. (2003). Sedentary behavior
during postnatal life is determined by the prenatal environment and exacerbated by
postnatal hypercaloric nutrition. Am J Physiol Regul Integr Comp Physiol. 285,
R271273.
Walker-Bone, K., Walter, G., and Cooper, C. (2002). Recent developments in the epidemiology
of osteoporosis. Curr Opin Rheumatol.14, 411415.
Waterland, R.A. and Jirtle, R.L. (2003). Transposable elements: targets for early nutritional
effects on epigenetic gene regulation. Mol Cell Biol. 23, 5293300.
Weaver, I.C., Cervoni, N., Champagne, F.A., DAlessio, A.C., Sharma, S., Seckl, J.R., Dymov,
S., Szyf, M., and Meaney, M.J. (2004). Epigenetic programming by maternal behavior.
Nat Neurosci. 7, 847854.
Yliharsila, H., Eriksson, J.G., Forsen, T., Laakso, M., Uusitupa, M., Osmond, C., and Barker,
D.J. (2004). Interactions between peroxisome proliferator-activated receptor-gamma
2 gene polymorphisms and size at birth on blood pressure and the use of antihyper-
tensive medication. J Hypertens. 22, 12831287.
10 Taste as the Gatekeeper

Toshiko Tanaka, Danielle R. Reed,
and Jose M. Ordovas
CONTENTS
10.1 Introduction ..................................................................................................115

10.2 Biology of Taste ...........................................................................................116
10.2.1 Bitter Compounds in Foods and Beverages ....................................116
10.2.2 Phenylthiocarbamide Taste Sensitivity ............................................117
10.2.3 PTC/PROP and Phenotype Associations .........................................118
10.2.4 PTC/PROP Sensitivity and Disease.................................................122
10.3 The Pursuit of the PTC Gene ......................................................................122
10.3.1 Evolution of the PTC Gene .............................................................123
10.3.2 PTC Gene Polymorphisms: GenotypePhenotype Associations ....124
10.3.3 Other Bitter Taste Receptors............................................................124
10.4 Other Taste Modalities .................................................................................125
10.4.1 Sweet and Umami............................................................................125
10.4.2 Sour ..................................................................................................126
10.4.3 Salt....................................................................................................126
10.4.4 Fat.....................................................................................................126
10.5 Genetics of Eating Behavior........................................................................127
10.6 Future Perspective on Personalized Nutrition .............................................129
Acknowledgments..................................................................................................129
Key Readings.........................................................................................................130
References..............................................................................................................130
10.1 INTRODUCTION
Human food preferences and intake are determined by a complex mix of innate,
acquired, and contemporary societal factors. Understanding the individual drivers
that determine what food is consumed and in which quantities, how it is prepared,
and when it is eaten should provide avenues for successful dietary advice and
interventions aimed at decreasing disease risk and achieving healthier aging. This
chapter reviews the current knowledge regarding the genetics of taste with a focus
on bitter taste perception and discusses how genetic variation in taste may be linked
115
to the consumption of foods associated with disease protection and increased well-
being, as well as avoidance of habits associated with increased risk of disease. Better
understanding of the role of taste in dietary behavior may lead to the identification
of genetic markers that can be utilized for the development of personalized dietary
recommendations based on individual taste preferences. Such personalized guide-
lines may improve adherence to dietary recommendations and provide better
preventive strategies for chronic diseases.
In many countries, chronic diseases such as cardiovascular disease and cancer
are the leading causes of mortality. Moreover, the prevalence of conditions such as
metabolic syndrome and obesity are on the rise worldwide. Although these diseases
have complex etiologies, diet is thought to be a major contributing factor. One of
the greatest public health challenges is to alter eating behavior in populations as a
means of preventing chronic diseases. Adherence to dietary recommendations such
as increasing fruit and vegetable consumption proves to be difficult for many people.
Although there are various environmental and physiological factors that determine
individual food selection, taste has been suggested as an important factor in consumer
food choice. There is a wide variety of taste preferences in the population, and part
of this variability is genetically determined [1]. It is of interest to elucidate whether
there are genetic markers of taste that predispose a person to consume a certain type
of diet. Such information would be useful to identify people who are at risk for
diseases based on their taste genotype. The genetic markers may provide the
foundations for tailored recommendations based on their preferences.
10.2 BIOLOGY OF TASTE

There are five modalities of taste that can be detected by most mammals: sweet,
salt, sour, bitter, and umami (the taste of monosodium glutamate) [2]. For our
ancestors, the ability to taste was important to ensure acquisition of nutrients and
for avoidance of noxious substances. In general sweet, salt, and to a certain extent,
umami, drive food consumption, whereas bitter and sour elicit food rejection.
Taste perception begins at the tongue, where specific tastants in food interact
with taste receptors found on taste cells (TCs) [2]. The TCs are specialized epithelial
cells with neuron-like properties, including depolarization, release of neurotransmit-
ters, and formation of synapse to afferent neurons. Approximately 50 to 150 TCs
organize into an onion-like configuration to form a taste bud. Taste buds are found
distributed throughout the tongue in three specialized structures called fungiform,
foliate, and circumvallate papillae. Some taste buds are also located on the soft
palate, uvula, pharynx, larynx, esophagus, and in the epiglottis. Upon binding of the
tastants to its receptors, the TC depolarizes to create an action potential and commu-
nicates either directly or indirectly with afferent neurons. This ultimately leads to the
transmission of signals to the brain.
10.2.1 BITTER COMPOUNDS IN FOODS AND BEVERAGES

Bitter compounds generally elicit food rejection, a behavior critical to avoid ingesting
potentially toxic compounds. Humans can detect various bitter tastes at very low
concentrations, a trait that is presumed to be favorable for survival. However, there
Taste as the Gatekeeper of Personalized Nutrition 117
are some naturally bitter compounds in foods that have been shown to have beneficial
effects on health. Some of these compounds include phenols (found in tea, citrus fruits,
wine, and soy), triterpenes (citrus fruits), and organosulfur compounds (cruciferous
vegetables) [3]. These phytochemicals have been investigated for their putative
anticarcinogenic and antioxidant properties. Various epidemiologic studies have
shown that increased intake of fruit and vegetables is associated with decreased risk
of chronic diseases such as heart disease and cancer. Public health efforts to increase
fruits and vegetable intake have been challenging, and the resistance to increased
consumption may be in part due to the bitter tastes of these foods [4]. Indeed, many
of the healthy vegetables such as cruciferous vegetables are often disliked for their
bitter taste, especially by children [3].
For this reason, it has been hypothesized that variability in bitter taste perception
may influence diet-related health outcomes. Much information investigating the
association between bitter-taste sensitivity and food preferences has centered on the
ability to taste a thiourea compound, phenylthiocarbamide (PTC). The N-C=S group
in PTC is responsible for its bitter taste [5]. Although PTC is not found in nature,
there are chemically related compounds such as isothiocyanates in cruciferous veg-
etables [6]. These bitter compounds often play a protective role for the plants. For
example, glucosinolates are natural pesticides found in cruciferous vegetable and,
when the plant gets damaged, they are hydrolyzed to goitrin, an isothiocyanate [3].
The abundance of bitter compounds in our food supply makes PTC sensitivity an
ideal marker for the study of taste, food preferences, and dietary behaviors.
10.2.2 PHENYLTHIOCARBAMIDE TASTE SENSITIVITY

The focus on the bitter compound PTC came about when A.L Fox discovered that
some people found crystallized PTC to be bitter, whereas others did not [6]. Sub-
sequent studies confirmed this observation and showed that in most populations,
there is a bimodal distribution of taste thresholds dividing people into tasters and
nontasters. Early studies used simple methods to dichotomize subjects, using PTC
crystals or PTC saturated paper [7]. Many modern studies utilize another thiourea,
6-n-propylthiouracil (PROP), because PTC has a slight sulfurous odor. There have
also been advancements in the psychophysical methods, in which subjects are
exposed to varying concentrations of PTC to determine the threshold at which they
can taste the bitterness. Various studies indicate that the thresholds are around 1.0
104 mol/L for tasters and >2.0 104 mol/L for nontasters [6]. Further, it was shown
that there are two types of tasters. Supertasters are extremely sensitive to PROP and
perceive this compound to be more intensely bitter than medium tasters [8].
PTC/PROP sensitivity is modified by sex and age; there is a greater proportion of
tasters among women, and sensitivity declines with age [5]. Anatomically, the tongue
of supertasters has higher densities of fungiform papillae and taste pores. It was
shown that, in general, women have higher densities of fungiform papillae, consistent
with the observed higher frequency of tasters [7]. In most populations, there are
more tasters than nontasters. The frequency of nontasters ranges from about 3% in
western African populations to greater than 40% in some populations in India [5,6].
In North American Caucasians, the frequency of nontasters is estimated to be around
30% [5,9].
10.2.3 PTC/PROP AND PHENOTYPE ASSOCIATIONS

PTC/PROP taster status has been associated with heightened bitter sensitivity and
dislike of bitter foods such as cruciferous vegetables, coffee/caffeine/tea, cheese,
grapefruit juice, and soy products [10 to 18] (Table 10.1). However, PTC/PROP
sensitivity has not shown a consistent pattern of preferences regarding bitter foods
[16]. This may be due to the low acceptance of bitter foods in the general population,
thus minimizing the variability necessary to detect a difference. In addition to bitter
foods and beverages, tasters have been reported to be more sensitive to sweet and
spicy stimuli such as sucrose, saccharin, and capsaicin [19]. There is evidence that
PTC/PROP sensitivity is associated with preferences for macronutrients, namely,
fat. However, the direction of the association has not been consistent. In some studies,
tasters liked foods with higher fat content [17] and disliked these foods in others
[20,21]. There are some reports of no significant difference in fat preferences by
taster status [22,23]. In one study, tasters were able to distinguish between 10 and
40% fat in salad dressing, whereas nontasters could not [19]. However, the taster
did not report preferences for either types of dressing, whereas the nontasters pre-
ferred the high-fat dressing. It has been suggested that tasters are more sensitive to
fat content because of the texture rather than the flavor of fat [19]. Thus, it is unclear
whether ability to taste bitter flavors per se has an effect on preferences for fat
content or whether the greater number of fungiform papillae determines the perceived
pleasantness. Overall, these studies suggest that tasters have a heightened sensitivity
to various tastes and a tendency to perceive these foods as unpleasant.
The connection between taste status and frequency of food intake is not as well
established as food preferences (Table 10.1). Studies showing associations between
PROP/PTC sensitivity and intake of bitter vegetables have been sparse. In one study,
nontasters consumed more cooked turnip and watercress, but no differences were
observed for nine other cruciferous vegetables [24]. In other studies, taste status was
associated with bitter food preferences, and food preferences in turn were associated
with frequency of intake [10,14]. Taste status, however, does not seem to be directly
associated with frequency of food intake [10,14,25,26]. The general low consump-
tion and small variability in intake of bitter foods may negate the ability of small
studies to detect a difference, because of insufficient power. There has been more
convincing evidence of taster status in association with the consumption of macro-
nutrients, specifically for percentage of fat intake [20,27,28]. Using food frequency
questionnaires, taster female college students reported higher intakes of energy from
fat and lower intakes from protein [28]. In another study of children, female tasters
had a lower intake of discretionary fat [20]. A feeding study in which subjects were
given either a high-fat, high-carbohydrate, or mixed meal showed that tasters ate
more fat and less carbohydrates during the mixed meal, but not the others [27].
Further studies are needed to clarify whether taste status is linked with fat intake
and to determine the direction of the associations.
There have been several studies showing an association between PTC/PROP
taste status and alcoholism. Initial case control studies revealed greater frequencies
of nontasters in children with alcoholic family members compared to controls [29].
Taster subjects have been shown to perceive alcohol such as beer to be more bitter
TABLE 10.1
Review of Studies Investigating PTC/PROP Sensitivity and Food Preference/Intake
Subjects: Sex Food Preference and Taster Classification Difference by PTC/PROP Sensitivity
(Age) Sensory Assessment Type of Food Method Sensory or Dietary Intake Reference
146F, 136M Food preference and Twenty-five items that Tasting of PTC saturated No Mattes
(1722 yr) consumption inhibit thyroidal iodine filter paper 1989 [26]
questionnaire uptake, and six bitter foods
17F, 17M Verbal food preference Eight tasted foods, order- Forced choice threshold Sensory: Tasters disliked cheese and liked Anliker
(57 yr) questionnaire and sensory by-choice and 60 food detection and whole milk 1991 [17]
perception checklist suprathreshold method
53F Sensory rating Tasted soy products and Forced choice threshold Sensory: Tasters perceived green tea to be Gayathri
(2045 yr) different dilutions of detection and more bitter and dislike plain tofu, 1997 [15]
green tea suprathreshold method preferred vanilla flavor soymilk
45F, 30M Sensory rating Fats (10 vs. 40% salad Suprathreshold method Sensory: Tasters reported more oral burn Tepper
Taste as the Gatekeeper of Personalized Nutrition
(1851 yr) dressing) from capsaicin and distinguished fat 1997 [19]
capsaicin content
123F Food preference Rated fruits and juices Suprathreshold method Sensory: Tasters disliked naringin and Drewnowski
(2060 yr) questionnaire Sensory testing of naringin grapefruit juice 1997 [18]
Sensory rating solution
118F Sensory and hedonic rating Fifteen Fat/sugar mixtures Forced choice detection Sensory: No Drewnowski
(2040 yr) threshold suprathreshold 1998 [23]
method
159F Food preference checklist One hundred and seventy- Suprathreshold method Sensory: Tasters had lower preference for Drewnowski
(2060 yr) 3-d food record (86 one items from various brussels sprouts, cabbage, spinach, and 1999 [10]
subjects) food categories coffee
Intake: No
(continued)
119
120
TABLE 10.1 (CONTINUED)

Review of Studies Investigating PTC/PROP Sensitivity and Food Preference/Intake
Subjects: Sex Food Preference and Taster Classification Difference by PTC/PROP Sensitivity
(Age) Sensory Assessment Type of Food Method Sensory or Dietary Intake Reference
50F Sensory rating Six bitter foods tasted Suprathreshold method Sensory: Taste perception is associated Kaminski
(2040 yr) FFQ Twenty-two bitter food Tasting PROP saturated with bitterness of food 2000 [14]
items questionnaire paper Intake: No
24F, 22M Food preference survey Eighty-three items Forced choice threshold Sensory: In women sweet and high-fat Duffy
(1840 yr) detection and foods negatively correlated with bitter 2000 [21]
suprathreshold method sensitivity: in men a positive correlation
was found
326F Food preference checklist One hundred and seventy- Suprathreshold method Sensory: Tasters dislike cruciferous, green, Drewnowski
(mean 47.651.7) one food items raw vegetables 2000 [16]
54F Sensory rating Sucrose, caffeine, Suprathreshold method Sensory: Tasters perceived caffeine as Ly
(1830 yr) chocolate Tasting PROP saturated being more bitter and disliked 2001 [12]
paper
92F, 55M Sensory rating Fattiness, saltiness, Suprathreshold method Sensory: No Yackinous
(1736 yr) sweetness of four foods 2001 [22]
85F, 38M Food questionnaire for Forty-three food items Taste recognition threshold Sensory: Tasters rated foods higher than Pasquet
(1959 yr) flavor, cost, effect on representative of Tunisian Suprathreshold method nontasters overall; no differences in bitter 2002 [13]
health diet foods except green tea, which tasters
disliked
33F, 34M Sensory test Tasted ten common Single concentration of Sensory: Tasters disliked broccoli, full-fat Keller
(45 yr) FFQ (child and parent) foods/drinks that are bitter PROP (yes/no) milk, American cheese 2002 [20]
or varying in fat 114 items Taster girls disliked full-fat milk
(FFQ) Intake: Girl tasters had lower intake of
discretionary fat
Personalized Nutrition: Principles and Applications
114F, 69M FFQ Ninety-five items Suprathreshold method Intake: No difference in men Yachinous
(1736 yr) Women tasters obtained lower percentage 2002 [28]
of energy from protein and fruit servings,
higher from fat
Supertasters consumed less green salad
compared to T, MT*
22F,14M Food intake Ad libitum intake of high- Suprathreshold method Intake: Taster had higher fat intake, lower Kamphuis
(mean age 31) fat, high-carb, or mixed CHO intake ad libitum of mixed meal 2003 [27]
meal No difference for high-fat or high-carb
meal
71F, 39M FFQ, food sampling Block 98.1 Suprathreshold method Sensory: Tasters rated sampled bitter Dinehart
(1860 yr) vegetables as most bitter (brussels 2006 [25]
sprouts, asparagus, kale). No association
with perceived sweetness of sweet foods

or from vegetables.
Intake: PROP sensitivity is inversely
associated with vegetable intake; when
Taste as the Gatekeeper of Personalized Nutrition
examining other variables, however, taste

sensitivity was not directly associated
with vegetable intake
*T = taster; MT = medium taster

121
and unpleasant [30]. Taste status has also been linked with smoking habits [31]. In
these studies the proportion of nontasters was greater in smokers compared to
nonsmokers. It has been hypothesized that nontasters have less aversion to the
bitterness of cigarettes and are thus at greater risk to smoke. Together, these data
suggest that PTC/PROP sensitivity may mitigate the development of addictive behav-
iors such as smoking and alcoholism.
10.2.4 PTC/PROP SENSITIVITY AND DISEASE

One of the first diseases that was investigated was goiter, because of the role of PTC
as a thyroid inhibitor [5]. The isothiocyanates found in cruciferous vegetables are
also natural thyroid inhibitors that PTC tasters find bitter. Goiters can be caused by
both over- and underactivity of the thyroid gland. Hypothyroidism can occur in areas
of low iodine, as this mineral is required for the production of thyroid hormones.
In this case, tasters have the advantage, as they would avoid the consumption of
bitter cruciferous vegetables. Conversely, in the case of hyperthyroidism, nontast-
ers would have the advantage. In accordance with these hypotheses, several studies
have found higher prevalence of nontasters in nontoxic goiter patients, and no
association has been observed with diffuse goiter.
Other diseases examined include diabetes, cancer, ulcers, depression, and psy-
chiatric disorders (reviewed in Reference 5). The underlying hypotheses for the
associations include differences in dietary behavior, neurological activity, or
unknown mechanisms. However, after initial reports, subsequent studies failed to
replicate the findings. Therefore, it is difficult to draw conclusions regarding these
associations. Further, many of these studies are small case control studies and the
generalizability of the results may be questionable. Further examination of the
mechanisms underlying the hypothesized associations may provide better under-
standing of the link between PTC/PROP and these diseases.
10.3 THE PURSUIT OF THE PTC GENE

Sensitivity to PTC/PROP is a heritable trait originally proposed to be monogenic
but currently thought to involve several loci. Early linkage studies had shown that
PTC sensitivity is linked with the Kell blood group on chromosome 7 [5]. This was
supported by a study of PROP sensitivity in which significant linkage was found on
chromosome 5p, with a possible linkage on chromosome 7q31 near the KEL locus
[32]. In a subsequent study using PTC sensitivity, there was significant linkage to
chromosome 7q, consistent with the previous studies [33]. Utilizing the results from
these linkage analyses, the genes for bitter taste receptors have been identified by
mining the human genome database. Electrophysiological studies showed that bitter
tastants are mediated through the activation of G-protein-coupled receptors (GPCR).
To identify the bitter receptor genes, segments of the genome linked with bitter taste
perception were screened for novel GPCRs [34,35]. These efforts led to the identi-
fication of a family of bitter taste receptor genes (TAS2R) on chromosomes 5, 7,
and 12. This family of receptors were found to be expressed in taste cells and showed
specific interactions with bitter compounds in vitro [36].
In genomewide linkage analysis, one member of the bitter gene receptor

family, TAS2R38, was shown to be responsible for PTC/PROP sensitivity [37].
Three single-nucleotide polymorphisms (SNPs) in the gene (P49A, A262V, and
V296I) gave rise to five common haplotypes, named in the order of the amino
acids at each SNP. These haplotypes accounted for 55 to 85% of the variance
in PTC sensitivity. Based on associations with taste phenotypes, the taster
haplotype was identified as PAV (proline-alanine-valine) and nontaster haplotype
as AVI (alanine-valine-isoleucine). Functional analysis of the TAS2R38 haplo-
types provided further support for the hypothesis that this gene is the PTC gene.
HEK293 cells transfected with PAV allele receptor displayed elevation of cyto-
solic Ca2+ in a dose-dependent manner, starting at PTC concentrations as low
as 0.1 M [38]. On the other hand, cells expressing the AVI allele receptor did
not respond to PTC at concentrations as high as 1 mM. This result is consistent
with the observation of PAV as the taster and AVI as the nontaster haplotype.
10.3.1 EVOLUTION OF THE PTC GENE

The identification of bitter taste receptor genes and the putative role of bitter taste
perception as a defense system against environmental toxins have drawn interest
from evolutionary biologists. The human TAS2R38 shows 98.2 and 65.5% amino
acid identity with chimpanzee (chTAS2R38) and mouse (Tas2r138) genes. Pheno-
typically, mice do not display a bimodal distribution, and the sensitivity to PTC or
PROP is not correlated as in humans [5]. Thus, in mice, it is speculated that the
Tas2r138 gene is not a receptor for PTC/PROP as in humans, but rather serves as
a receptor for another bitter compound. On the other hand, chimpanzees demonstrate
both taster and nontaster phenotypes. However, the molecular basis of the variation
is different [39]. The polymorphisms that give rise to the human taster haplotypes
are absent in the chimpanzee gene. Instead, a polymorphism in the second position
of the start codon is responsible for the taster phenotypes. This variation results in
a truncated polypeptide that is not responsive to PTC in vitro. These molecular and
phenotypic differences in PTC sensitivity between organisms suggest that this trait
has evolved independently and may be a reflection of the different environments of
the species.
In humans, evidence supports the notion that the PTC/PROP taster phenotype
has been maintained through natural selection rather than by genetic drift [40].
Geneticists have hypothesized that those who are sensitive to bitter tastes will have
a selective advantage because of their ability to avoid toxic foods [5]. However, the
persistence of nontasters in most populations suggests that there are advantages to
being insensitive to bitter compounds [5,40]. This may be explained by the multiple
functions of bitter foods in our environment. Although bitter compounds can be
toxic, many bitter foods have beneficial effects on health. For our ancestors, insen-
sitivity to bitter taste may have provided availability of additional plant-based nutri-
ent sources when food was scarce. For modern humans, it has been proposed that
nontasters may be less prone to reject healthy bitter foods and may choose a diet
with more variety [4].
10.3.2 PTC GENE POLYMORPHISMS: GENOTYPEPHENOTYPE

ASSOCIATIONS
The identification of TAS2R38 has provided the tools to examine whether genetic
measures of PTC sensitivity are linked to dietary behavior and disease status in
humans. In a study of 143 children and their mothers, the P49A polymorphism was
associated with bitter taste sensitivity [41]. However, the variability explained by
the genotype was greater in children than in their mothers, indicating that age may
be a significant modifier. Taste sensitivity was associated with greater preferences
for sweets in children but not adults. In the mothers, ethnicity was the strongest
predictor of sweet preferences, suggesting that cultural influences can override
genetic influences of taste in adults. In another study of 84 healthy adults, TAS2R38
haplotypes were a significant predictor of alcohol consumption [42]. Subjects with
the nontaster haplotype reported higher use of alcohol in comparison to subjects
with the taster haplotype. In a regression analysis, TAS2R38 haplotype and number
of fungiform papillae were both independent determinants of PROP bitterness. When
the bitterness of PROP solution was recorded by TAS2R38 haplotype, the intensity
function was not as distinct as the results using taster groups determined by psy-
chophysical methods. Nevertheless, this study supports the role of TAS2R38 as a
predictor of alcohol consumption.
In a cross-sectional study of older British women, the TAS2R38 haplotype was
not associated with dietary intake as measured by a food frequency questionnaire
[43]. Further, no association was found between this haplotype and CHD risk factors
or CHD. There was marginally greater prevalence of diabetes among the taster
haplotype; however, the significance is unclear, given that multiple associations were
examined. The lack of association in this study may be due to the older age of the
subjects, as PTC sensitivity declines with age. However, this study may also indicate
that this locus is not associated with measurable dietary behaviors in humans. Further
studies are necessary to clarify the association between TAS2R38 haplotype and
dietary intake in both men and women.
These first studies examining TAS2R38 provide tantalizing results for the use of
this gene as a marker for food preferences and, potentially, dietary intake. The lack
of association in the first epidemiologic study may be due to various confounding
factors that were not measured in the study, such as restrictive dietary practices.
Moreover, various cooking methods that cannot be deconstructed by most food
frequency questionnaires can significantly alter the taste qualities of foods. There-
fore, examination of these factors will provide better understanding on the role of
TAS2R38 on dietary behavior. It would be interesting to investigate whether there
are differences in adherence to dietary recommendations based on TAS2R38 haplo-
type. Future studies will most likely provide insight into the applicability of this
marker as a tool to develop targeted dietary recommendations.
10.3.3 OTHER BITTER TASTE RECEPTORS

The PTC gene is one of an estimated ~30 bitter taste receptor genes in humans
[34,35]. There is evidence that other bitter taste receptors are biologically important
and have been linked with alcohol dependence as well as disease risk. For instance,
sequence analysis of TAS2R16 shows that the K172N polymorphism in the gene is
under positive selection based on the observation of high frequency of the evolu-
tionary derived variants in populations worldwide [44]. This result suggests that this
receptor has adapted according to unique human environments. The K172N poly-
morphism lies in the third extracellular loop and functional analysis shows that the
ancestral K172 allele has weaker response to ligands such as salicin, arbitun, and
amygdalin than the evolutionary derived N172 allele. Consumption of foods con-
taining these cyanogenic glycosides has been hypothesized to be protective against
malaria [44]. Consistent with this hypothesis, the K172 allele has been found at
higher frequencies where malaria is endemic. In addition, the K172 has been asso-
ciated with increased risk for alcohol dependence [45]. Because TAS2R38 has also
been associated with alcohol consumption and both TAS2R16 and TAS2R38 are
located on neighboring regions on chromosome 7, it would be interesting to examine
both of these genes in relation to alcohol intake.
In an effort to identify genes associated with myocardial infarction (MI), an associ-
ation study examining 11,053 polymorphisms representing 6,891 genes was conducted
[46]. To identify the variants significantly associated with MI, three case control studies
were analyzed sequentially. In the final analysis, five genes were significantly asso-
ciated with MI, one being TAS2R50 with odds ratio of 1.58 for the susceptible allele.
Interestingly, another gene identified in this study was one of the olfactory receptors,
OR13G1. The authors hypothesized that the association observed for both TAS2R50
and OR13G1 may be through dietary choices. Further studies are needed to confirm
this hypothesis.
With the exception of the known alleles of TAS2R38 and the sensitivity to the
bitterness of PTC and related compounds, there are no known phenotypegenotype
relationships between alleles of the other bitter receptors and the taste perception
of their bitter ligands. However, there are many examples of a wide range of
perceptual sensitivity by human populations to other bitter stimuli. Although PTC
and TAS2R38 are the first example of the effect of genetics on taste perception, the
complement of taste receptors and their many alleles will probably reveal numerous
relationships with bitter taste perception and the motivation to consume bitter foods.
10.4 OTHER TASTE MODALITIES

The discussion thus far has focused on bitter taste; however, there are four other
conventional taste modalities: salty, sour, sweet, and umami and perhaps a fifth
taste, that for fatty acids. We briefly review these other genetic influences on these
taste qualities and how they might contribute to the taste of food. Incorporating all
the taste modalities may improve our understanding of the role of genetic variation
in taste on food consumption and health.
10.4.1 SWEET AND UMAMI

One of the universal experiences in human perception is that sweetness is considered
to be good and desirable, and sugar is well accepted and well tolerated by people
in all human populations studied to date. The initial step in sweet perception is the
binding of the sweetener to the receptor, which is composed of two subunits, TAS1R2
and TAS1R3 (for a review of the discovery and biology of this receptor, see Reference
36). A second subunit (TAS1R1) when paired with TAS1R3, the common partner,
is responsible for the transduction of umami (savory) taste. Alleles of TAS1R3 are
strongly associated with sweet intake in the mouse, and it has been hypothesized
that alleles of human T1Rs may explain part of the variability of the intake of sugars.
These phenotypegenotype studies have yet to be conducted in humans, although
there are many known alleles in these taste receptors, especially the subunit special-
ized to sweet ligands. The umami-specific subunit of the receptor, TAS1R1, also
contains many polymorphisms, and might be related to the sensory blindness to the
savory taste of MSG experienced by a subset of people [47].
10.4.2 SOUR
Individual differences in the ability to perceive sourness exist, but the genetic basis
of this taste quality is unknown. The recent discovery of a key protein in sour taste
in mice provides the foundations to investigate the mechanisms of sour taste trans-
duction and the role of this taste modality on human dietary behavior and health
[48]. Unripe fruit, types of milk and cheese and some vegetables are sour, and people
often confuse sour and bitter, which contributes to the perception of a sour food as
undesirable. Because many nutritious foods are slightly sour, its perception is an
aspect of food selection that is understudied relative to its importance to human
health.
10.4.3 SALT
The perception of saltiness differs from the other taste qualities because differences
among people in sodium threshold and intensity ratings of saltiness are marked, but
are related to immediate food intake rather than to genotype. Although some aspects
of sodium transduction are understood, the search for the full complement of proteins
that participate in this process is still on. It is possible that genetic differences among
people are important, but it could also be that salt perception is labile, and the genetic
differences would only be apparent if all subjects were fed diets equivalent in
saltiness.
10.4.4 FAT
In addition to the basic tastes and their receptor genes, examination of unconventional
taste qualities will likely be important in understanding human diet and health.
Because fat is important to the sensory experience of food, differences in fat taste
perception may influence dietary behavior. Several lines of evidence suggest there
might be genetic determinants of oral fat perception, the most convincing of which
is the observation that fat intake in humans is heritable [1]. There is debate about
whether animals and humans have a sensory system that is specifically tuned to oral
fats, largely because the receptor and transduction mechanisms are not known.
Chewing or tasting fatty foods without swallowing is sufficient to stimulate digestion,
which is evidence that fats are sensed on the tongue [49]. Whether fat perception is
mediated through sensory stimulation of fat texture or a chemosensory detection of

fat taste is controversial. Recently, the glycoprotein CD36 was shown to be
involved in the orosensory detection of dietary lipids in rats and mice [50]. This
protein has been described as a putative receptor for fats, suggesting that taste of
fat may be the sixth taste modality.
There is a human form of the CD36 receptor, but whether it has the same taste
function as in rodents is not known. If this gene is involved in the perception of oral
fats, alleles of this gene might explain individual variation in perception [51]. One
intriguing observation is that the CD36 gene and its alleles in humans have been
studied by investigators who noticed its relationship with diet-induced disease such
as diabetes [52]. It may be that CD36 alleles might change fat perception and intake,
thereby contributing to metabolic diseases. The human homologue of the mouse
CD36 gene is located in the middle of human chromosome 7, and during the last
several years, many alleles of the human CD36 gene have been discovered [51].
Because there is a range in the ability of people to detect fat in foods [53], efficiency
of oral fat receptors may differ among people, therefore influencing preferences and
intake of dietary fat.
10.5 GENETICS OF EATING BEHAVIOR

A principal challenge of studying taste and its relationship to nutrition and health
is translating differences among people in sensory perception of food and drink to
what they actually choose to eat. By introspection we can immediately identify cases
in which a food that is not preferred has been eaten for reasons unrelated to taste,
but rather to social or economic influences. Therefore, it is useful to examine not
only the relationships between taste perception and genotype, but also between
human nutrient selection and genotype. It is the goal that these two fields will
eventually expand to bridge the gap between taste perception and food intake. The
current knowledge on the genetic basis of dietary behavior is reviewed in the
following text.
Family studies have been used to determine the genetic contribution to dietary
behaviors in humans. The heritability estimates for carbohydrates, protein, and fat
range from 30 to 66%, 20 to 70%, and 10 to 56%, respectively [54]. Because
macronutrients are energy yielding, some of the genetic components may include
energy regulation rather than proportion of macronutrients. Heritability estimates
for energy intake in these studies ranged from 11 to 70%. The different data analyses
methods and varied acquisition of dietary intake make the studies difficult to compare
and may explain the wide variability in the heritability estimates. Nevertheless, these
studies support the role of genetics in food consumption. Given the complexity of
human dietary behaviors, with the exception of monogenetic conditions such as with
leptin deficiency, no single gene is likely to have a great impact on food intake.
Although candidate gene association studies of dietary behavior have been
sparse, there have been several positive reports. The majority of the genes reported
to date are those involved in the central control of energy intake including neuropep-
tides and monoamine systems (reviewed in Reference 54). For example, serotonin
is a neurotransmitter that is involved in various physiological processes including
sleep, appetite, mood, thermoregulation, pain perception, hormone secretion, and

sexual behavior. Central and peripheral administration of serotonin decreases food
intake, increases energy expenditure, and drugs that promote serotonergic activity
have an anorectic effect. Serotonin reuptake blockers also increase sweet and bitter
taste perception in humans. These observations collectively make the genes in the
serotonin pathway candidates for influencing dietary behaviors. A SNP in the pro-
moter region of the serotonin receptor 2A (5-HT2A) was associated with total energy
and fat intake in Caucasian populations [55,56]. The minor allele carriers of the
1438G/A SNP consumed fewer calories, and in adolescents the differences in the
energy were driven by fat intake. The major allele homozygotes consumed as much
as 10% more energy in comparison to the minor allele homozygotes. These findings
are interesting in light of the observation of lower frequency of the A allele in
anorexic patients.
In addition to the 5-HT2A gene, polymorphisms in other genes, including the
dopamine transporter, agouti-related protein, and uncoupling proteins, have been
associated with macronutrient intake [54]. These candidate gene association studies
together support the involvement of multiple gene loci on dietary behavior. With the
exception of 5-HT2A, there have been no published reports replicating the observa-
tions in other populations. Replication studies are particularly important for candi-
date gene association studies because of the propensity of false-positive results.
Furthermore, examination of multiple polymorphisms (or haplotypes) rather than a
single SNP may provide better understanding of the significance of these genes as
a marker for dietary behavior. The studies published thus far highlight the complexity
and the diversity of genes that may mediate food intake. Therefore, future studies
investigating the genetics of dietary behavior that consider multitude of genes rather
than focusing on one aspect of dietary behavior may prove to be more fruitful.
One of the drawbacks of the candidate gene approach is that these studies are
limited to current knowledge of the control of feeding behaviors. An alternative
approach for the discovery of genetic contribution of food consumption is to use
the whole-genome scan approach. There have been two genomewide linkage analyses
of macronutrient intake in Mexican Americans, Caucasians, and African Americans.
The first study examined 816 Mexican American participants from the San Antonio
Family Heart Study [57]. Using FFQ developed specifically for this population, total
energy and macronutrient [total protein, total fat, saturated fat (SFA), monosaturated
fat (MUFA), polyunsaturated fat (PUFA), total carbohydrates, and sucrose] intake
were examined. Heritability estimates for these dietary parameters were modest,
ranging from 9% for PUFA to 21% for MUFA. Suggestive linkage was found on
chromosome 2p22 with LOD score of 2.62 for SFA, 2.09 for total fat, 2.22 for total
protein, and 2.05 for MUFA. The proopiomelanocortin (POMC) gene that encodes
for a protein that is the precursor of appetite-regulating hormones was identified as
a candidate gene within this region. However, two polymorphisms on exon 3 were
not associated with saturated fat intake, suggesting that POMC is not the gene
underlying the observed linkage. In the second study, total energy and macronutrient
(fat, carbohydrate, protein, and sucrose) was examined in 794 Caucasians and 198
African American participants of the HERITAGE study [58]. In Caucasians a region
on chromosome 1 (1p22.1-1q22) was linked with total energy and fat intake, 1p22-1p32
and 20q13.3 for total energy intake, and 12q14.1 for total fat intake. In African
Americans, suggestive linkage found at 1q43-44 for sucrose intake, and regions
1p22.3, 6q22.31, and 12q24.21 for fat intake. No specific candidate genes were
examined in the study; therefore, the genes underlying the linkage are yet to be
identified. It is noteworthy that the linkage observed in the studies were specific to
each ethnic group. This may be due to the differences in the methodologies for
dietary assessment or linkage analyses used by the two studies. Or it may be an
indication of different genetic architecture of dietary behavior by ethnicity. These
studies provide clues regarding the regions of the genome that may contribute to
macronutrient intake and may be the initial steps toward identifying novel genes
involved in dietary behavior.
10.6 FUTURE PERSPECTIVE ON PERSONALIZED

NUTRITION
The current nutrigenetic paradigm places an emphasis on interindividual differences
in physiological response to diet. However, one of the biggest challenges for nutrition
practitioners is to change dietary behaviors. Therefore, incorporating genetic markers
that determine predisposition to certain dietary habits may be an important factor
in the development of successful dietary intervention strategies.
In humans, taste is a powerful determinant of food selection. In the specific case
of the bitter taste sensitivity, this may have served in ancestral times as a primary
defense against ingestion of toxic compounds in foods. For modern humans, the
role of bitter taste perception for the avoidance of noxious food is less important.
However, it is thought that individual differences in bitter taste sensitivity may
influence health outcomes through its effect on dietary behaviors. Sensitivity to PTC
and PROP has provided some insight into the role of bitter taste perception on
preferences and consumption of bitter foods. The haplotypes of the TAS2R38 gene
accounts for most of the variability of PTC sensitivity, and several studies suggest
that it may serve as a marker for lifestyle and dietary behavior. However, current
evidence does not yet support a link between genetic variability at this taste candidate
gene, habitual consumption of certain foods, and health-related outcomes. This is
expected when considering the complexity of the large number of taste-related genes
as well as other pathways that modulate eating habits (i.e., serotonin and dopamine).
Therefore, advances in this promising area of translational research will be driven
by a more comprehensive approach to the problem encompassing wide genome
coverage as well as the inclusion of behavioral, cultural, and economic determinants.
It is foreseeable that a better knowledge of the physiology of taste and its genetic
variation in humans will result in an increased ability to carry out successful per-
sonalized dietary recommendations that will be both appealing and successful in
terms of preventing chronic disorders.
ACKNOWLEDGMENTS
We thank Dr. Carol Christensen and Dr. Marcia Levin Pelchat for their generous
contributions in the preparation of this manuscript.
KEY READINGS
Herness, M. S. and Gilbertson, T. A., Cellular mechanisms of taste transduction, Annu Rev
Physiol, 61, 873900, 1999.
Mennella, J. A. et al., Genetic and environmental determinants of bitter perception and sweet
preferences, Pediatrics, 115, e21622, 2005.
Rankinen, T. and Bouchard, C., Genetics of food intake and eating behavior phenotypes in
humans, Annu Rev Nutr, 26, 41334, 2006.
Reed, D. R. et al., Diverse tastes: Genetics of sweet and bitter perception, Physiol Behav, 88,
21526, 2006.
Tepper, B. J., 6-n-Propylthiouracil: a genetic marker for taste, with implications for food
preference and dietary habits, Am J Hum Genet, 63, 12716, 1998.
REFERENCES
1. Reed, D. R. et al., Heritable variation in food preferences and their contribution to
obesity, Behav Genet, 27, 37387, 1997.
2. Herness, M. S. and Gilbertson, T. A., Cellular mechanisms of taste transduction, Annu
Rev Physiol, 61, 873900, 1999.
3. Drewnowski, A. and Gomez-Carneros, C., Bitter taste, phytonutrients, and the con-
sumer: a review, Am J Clin Nutr, 72, 142435, 2000.
4. Drewnowski, A. and Rock, C. L., The influence of genetic taste markers on food
acceptance, Am J Clin Nutr, 62, 50611, 1995.
5. Guo, S. W. and Reed, D. R., The genetics of phenylthiocarbamide perception, Ann
Hum Biol, 28, 11142, 2001.
6. Tepper, B. J., 6-n-Propylthiouracil: a genetic marker for taste, with implications for
food preference and dietary habits, Am J Hum Genet, 63, 12716, 1998.
7. Bartoshuk, L. M., Comparing sensory experiences across individuals: recent psycho-
physical advances illuminate genetic variation in taste perception, Chem Senses, 25,
44760, 2000.
8. Reed, D. R. et al., Propylthiouracil tasting: determination of underlying threshold
distributions using maximum likelihood, Chem Senses, 20, 52933, 1995.
9. Fox, A. L., The relationship between chemical constitution and taste., Proc Natl Acad
Sci, 18, 115120, 1932.
10. Drewnowski, A. et al., Taste and food preferences as predictors of dietary practices
in young women, Public Health Nutr, 2, 5139, 1999.
11. Reed, D., Behavioral modulation of the path from genotype to obesity phenotype,
John Libbey & Company, Paris, 1999, pp. 199208.
12. Ly, A. and Drewnowski, A., PROP (6-n-Propylthiouracil) tasting and sensory responses
to caffeine, sucrose, neohesperidin dihydrochalcone and chocolate, Chem Senses, 26,
417, 2001.
13. Pasquet, P. et al., Relationships between threshold-based PROP sensitivity and food
preferences of Tunisians, Appetite, 39, 16773, 2002.
14. Kaminski, L. C. et al., Young womens food preferences and taste responsiveness to
6-n-propylthiouracil (PROP), Physiol Behav, 68, 6917, 2000.
15. Gayathri Devi, A. et al., Sensory acceptance of Japanese green tea and soy products
is linked to genetic sensitivity to 6-n-propylthiouracil, Nutr Cancer, 29, 14651, 1997.
16. Drewnowski, A. et al., Genetic taste markers and preferences for vegetables and fruit
of female breast care patients, J Am Diet Assoc, 100, 1917, 2000.
17. Anliker, J. A. et al., Childrens food preferences and genetic sensitivity to the bitter
taste of 6-n-propylthiouracil (PROP), Am J Clin Nutr, 54, 31620, 1991.
18. Drewnowski, A. et al., Taste responses to naringin, a flavonoid, and the acceptance
of grapefruit juice are related to genetic sensitivity to 6-n-propylthiouracil, Am J Clin
Nutr, 66, 3917, 1997.
19. Tepper, B. J. and Nurse, R. J., Fat perception is related to PROP taster status, Physiol
Behav, 61, 94954, 1997.
20. Keller, K. L. et al., Genetic taste sensitivity to 6-n-propylthiouracil influences food
preference and reported intake in preschool children, Appetite, 38, 312, 2002.
21. Duffy, V. B. and Bartoshuk, L. M., Food acceptance and genetic variation in taste, J
Am Diet Assoc, 100, 64755, 2000.
22. Yackinous, C. and Guinard, J. X., Relation between PROP taster status and fat
perception, touch, and olfaction, Physiol Behav, 72, 42737, 2001.
23. Drewnowski, A. et al., Genetic sensitivity to 6-n-propylthiouracil and sensory
responses to sugar and fat mixtures, Physiol Behav, 63, 7717, 1998.
24. Jerzsa-Latta, M. et al., Use and perceived attributes of cruciferous vegetables in terms
of genetically-mediated taste sensitivity, Appetite, 15, 12734, 1990.
25. Dinehart, M. E. et al., Bitter taste markers explain variability in vegetable sweetness,
bitterness, and intake, Physiol Behav, 87, 30413, 2006.
26. Mattes, R. and Labov, J., Bitter taste responses to phenylthiocarbamide are not related
to dietary goitrogen intake in human beings, J Am Diet Assoc, 89, 6924, 1989.
27. Kamphuis, M. M. and Westerterp-Plantenga, M. S., PROP sensitivity affects macro-
nutrient selection, Physiol Behav, 79, 16772, 2003.
28. Yackinous, C. A. and Guinard, J. X., Relation between PROP (6-n-propylthiouracil)
taster status, taste anatomy and dietary intake measures for young men and women,
Appetite, 38, 2019, 2002.
29. Pelchat, M. L. and Danowski, S., A possible genetic association between PROP-
tasting and alcoholism, Physiol Behav, 51, 12616, 1992.
30. Intranuovo, L. R. and Powers, A. S., The perceived bitterness of beer and 6-n-
propylthiouracil (PROP) taste sensitivity, Ann NY Acad Sci, 855, 8135, 1998.
31. Enoch, M. A. et al., Does a reduced sensitivity to bitter taste increase the risk of
becoming nicotine addicted?, Addict Behav, 26, 399404, 2001.
32. Reed, D. R. et al., Localization of a gene for bitter-taste perception to human chro-
mosome 5p15, Am J Hum Genet, 64, 147880, 1999.
33. Drayna, D. et al., Genetic analysis of a complex trait in the Utah Genetic Reference
Project: a major locus for PTC taste ability on chromosome 7q and a secondary locus
on chromosome 16p, Hum Genet, 112, 56772, 2003.
34. Adler, E. et al., A novel family of mammalian taste receptors, Cell, 100, 693702, 2000.
35. Matsunami, H. et al., A family of candidate taste receptors in human and mouse,
Nature, 404, 6014, 2000.
36. Reed, D. R. et al., Diverse tastes: Genetics of sweet and bitter perception, Physiol
Behav, 88, 21526, 2006.
37. Kim, U. K. et al., Positional cloning of the human quantitative trait locus underlying
taste sensitivity to phenylthiocarbamide, Science, 299, 12215, 2003.
38. Bufe, B. et al., The molecular basis of individual differences in phenylthiocarbamide
and propylthiouracil bitterness perception, Curr Biol, 15, 3227, 2005.
39. Wooding, S. et al., Independent evolution of bitter-taste sensitivity in humans and
chimpanzees, Nature, 440, 9304, 2006.
40. Wooding, S. et al., Natural selection and molecular evolution in PTC, a bitter-taste
receptor gene, Am J Hum Genet, 74, 63746, 2004.
41. Mennella, J. A. et al., Genetic and environmental determinants of bitter perception

and sweet preferences, Pediatrics, 115, e21622, 2005.
42. Duffy, V. B. et al., Bitter receptor gene (TAS2R38), 6-n-propylthiouracil (PROP)
bitterness and alcohol intake, Alcohol Clin Exp Res, 28, 162937, 2004.
43. Timpson, N. J. et al., TAS2R38 (phenylthiocarbamide) haplotypes, coronary heart
disease traits, and eating behavior in the British Womens Heart and Health Study,
Am J Clin Nutr, 81, 100511, 2005.
44. Soranzo, N. et al., Positive selection on a high-sensitivity allele of the human bitter-
taste receptor TAS2R16, Curr Biol, 15, 125765, 2005.
45. Hinrichs, A. L. et al., Functional variant in a bitter-taste receptor (hTAS2R16) influ-
ences risk of alcohol dependence, Am J Hum Genet, 78, 10311, 2006.
46. Shiffman, D. et al., Identification of four gene variants associated with myocardial
infarction, Am J Hum Genet, 77, 596605, 2005.
47. Lugaz, O. et al., A new specific ageusia: some humans cannot taste L-glutamate,
Chem. Senses, 27, 10515, 2002.
48. Huang, A. L. et al., The cells and logic for mammalian sour taste detection, Nature,
442, 9348, 2006.
49. Mattes, R. D., Fat taste and lipid metabolism in humans, Physiol Behav, 86, 6917,
2005.
50. Laugerette, F. et al., CD36 involvement in orosensory detection of dietary lipids,
spontaneous fat preference, and digestive secretions, J Clin Invest, 115, 317784,
2005.
51. Ma, X. et al., A common haplotype at the CD36 locus is associated with high free
fatty acid levels and increased cardiovascular risk in Caucasians, Hum Mol Genet,
13, 2197205, 2004.
52. Corpeleijn, E. et al., Direct association of a promoter polymorphism in the CD36/FAT
fatty acid transporter gene with Type 2 diabetes mellitus and insulin resistance, Diabet
Med, 23, 90711, 2006.
53. Reed, D. R. et al., Sensory and metabolic influences on fat intake, Elsevier Applied
Science, London, New York, 1992, pp. 117137.
54. Rankinen, T. and Bouchard, C., Genetics of food intake and eating behavior pheno-
types in humans, Annu Rev Nutr, 26, 41334, 2006.
55. Aubert, R. et al., 5-HT2A receptor gene polymorphism is associated with food and
alcohol intake in obese people, Int J Obes Relat Metab Disord, 24, 9204, 2000.
56. Herbeth, B. et al., Polymorphism of the 5-HT2A receptor gene and food intakes in
children and adolescents: the Stanislas Family Study, Am J Clin Nutr, 82, 46770,
2005.
57. Cai, G. et al., Quantitative trait locus determining dietary macronutrient intakes is
located on human chromosome 2p22, Am J Clin Nutr, 80, 14104, 2004.
58. Collaku, A. et al., A genome-wide linkage scan for dietary energy and nutrient intakes:
the Health, Risk Factors, Exercise Training, and Genetics (HERITAGE) Family Study,
Am J Clin Nutr, 79, 8816, 2004.
11 Personalized Nutrition
and Public Health
Pieter vant Veer, Edith J.M. Feskens,
and Ellen Kampman
CONTENTS
11.1 Introduction ..................................................................................................133

11.2 Chronic Disease Etiology ............................................................................135
11.2.1 Nutrition and Environment ..............................................................135
11.2.2 Genetic Factors and Their Interplay with Diet ...............................136
11.3 Examples of the Status Quo: Carcinogenesis
and Diabetes.................................................................................................137
11.3.1 Diet, Genes, and Colorectal Cancer ................................................137
11.3.2 Diet, Genes, and Diabetes................................................................138
11.4 Prevention Paradox: Sick Individuals and Sick Populations ......................139
11.4.1 Prevention Paradox ..........................................................................139
11.4.2 GeneEnvironment Interactions: Individual
and Population Health......................................................................140
11.4.3 Additive and Synergistic Effects .....................................................141
11.5 Discussion Framework for PN.....................................................................142
11.5.1 Context .............................................................................................142
11.5.2 Genomics..........................................................................................143
11.5.3 Tailoring ...........................................................................................144
11.6 Future Perspectives for PN ..........................................................................145
Acknowledgment ...................................................................................................145
Key Readings.........................................................................................................146
References..............................................................................................................146
11.1 INTRODUCTION
Personalized nutrition (PN) addresses the idea that optimal nutrition differs between
individuals as a result of their genetic makeup. One hundred years after Mendel and
50 years after Watson and Crick, the current deciphering of the human genome holds
the promise that knowledge of the genetic makeup and the omic technologies will
133
enable us to provide highly specific dietary advice fitting the nutritional requirements
of individuals.1
Nutritional requirements are traditionally based on physiological needs, e.g., for
growth and development, to maintain nutritional status, to minimize disease risk, or
to optimize life expectancy. Based on this biomedical approach, nutritional require-
ments and recommendations are provided for energy intake and nutrient needs for
specific populations according to age, sex, physical activity, health status, life cycle,
etc. Beyond these variations in requirements at the group level, variability in nutri-
tional requirements has been denoted as natural biological variation, stemming
from genetic variation, uncontrolled environmental factors, and measurement error.
Thus, effects of dietary factors may at least in part depend on inherited
individual susceptibility, allowing the strength of the diethealth equation to differ
between subjects. In this framework, biological variation in classical risk factors
such as blood pressure and blood lipids serve as phenotypical intermediates of
dietgene interactions.
When we apply this biomedical approach of PN to disease prevention and health
promotion, it enters the domain of public health, defined as the science and art of
preventing disease, prolonging life and promoting health through organized efforts
of society.2 To enhance public health, nutritional requirements and recommenda-
tions are translated into food patterns, which, at least implicitly, incorporate national
and local traditions and cultural issues, consumer acceptance, economic interests,
and feasibility. These dietary guidelines complete the sequence from requirements
to recommendations and, subsequently, to guidelines, and the process provides a
basis for evidence-based nutritional policy.
In the context of public health, PN can fulfill a role in effective implementation
of dietary guidelines at the individual level, i.e., in dietetics and health promotion.
Here, the challenges are similar to those of clinical nutritionists and dieticians
treating specific distortions in blood pressure, lipid metabolism, etc.; both in a
clinical and in a public health context, population nutrition recommendations
have to be translated to an individualized dietary optimum in everyday food
choice matching the lifestyle of the individual.
Thus, although the biomedical concept of PN is scientifically challenging and
may help to further uncover and define physiological requirements, it also has an
independent meaning in the domain of public health, where existing and future
nutritional requirements are applied to populations (public health) and individuals
(PN) aiming at disease prevention and health care.
In this contribution, we will delineate the link between PN and public health.
We will do this from the viewpoint of epidemiology. We will first address the role
of diet (and lifestyle) in chronic disease epidemiology at the global level and then
clarify the basic concept of dietgene interaction. Subsequently, developments in
the status quo on research in nutrient/dietgene interactions will be summarized for
colorectal carcinogenesis and diabetes, representing two pathological processes of
major importance to population health. Then, the so-called prevention paradox and
its implications for PN will be illustrated and knowledge gaps will be touched upon.
Finally, elements for a discussion framework on PN are proposed, and future pros-
pects are delineated.
Personalized Nutrition and Public Health 135
11.2 CHRONIC DISEASE ETIOLOGY

11.2.1 NUTRITION AND ENVIRONMENT
During the past century, improvements in environmental factors such as water supply,
sanitation, housing, immunization, and health care have resulted in a major increase
in life expectancy, accompanied by successive replacement of infectious diseases
by health outcomes such as trauma (because of increased use of private transporta-
tion), and age-related cardiovascular disease and malignancies, with overweight and
diabetes as intermediary health states. This epidemiologic transition has accompa-
nied a nutrition transition, first focusing on food security (energy needs), followed
by food safety and food distribution resulting in increased workforce and economic
development, culminating in attention to food quality and dietary advice (food and
nutrition policy). This evolution has mainly taken place in the Western countries,
and is still occurring in developing economies. These subsequent stages of devel-
opment require increasingly higher levels of societal infrastructures, and they are
clearly visible cross-sectionally at the global level.
What are the main avoidable risk factors? The World Health Report3 quantified
attributable risks of death for the most important exposures: high blood pressure
and high serum cholesterol are leading, and known to have direct dietary determi-
nants; next, overweight, physical inactivity, and low consumption of fruits and
vegetables also indicate an unbalanced diet and lifestyle. Alcohol and tobacco are
addictive lifestyle factors. These exposures are much more important in Western
economies compared to developing countries, but undernutrition in regard to zinc,
vitamin A, and iron is almost unique for the latter.
The young are especially vulnerable to undernutrition, as is apparent from
survival curves of developing countries. But in developed economies the risks of
a Western lifestyle are becoming visible only after prolonged, lifelong exposure
(which also implies that the risks are intrinsically low). The dramatic improvements
in life expectancy and physical well-being during the past century resulted from
gradual changes in dietary and environmental factors. Still, about one third of age-
related cardiovascular and malignant diseases in developed economies remain diet
related. Thus, now that major environmental risk factors have been controlled and
have resulted in higher life expectancy, genetic variability between individuals
within a population will become a relatively more important factor in explaining
the remaining variability in life expectancy. International frameworks and consor-
tiums such as the European Nutrigenomics Organisation, NuGO, are now decipher-
ing the human genome and developing omics technologies, leading to deeper
understanding of dietgene interactions, potentially prolonging healthy life span
and preventing disease.
The opportunities provided by the genomics revolution must be viewed against
their relevance for individuals and the societies they are living in. For a rational
approach to PN and public health, an important question is, to what molecular level
we can break down human health and physiology into highly specific molecular
cellular nutrition-related processes in order to provide scientific proof, without
arriving at infinitely small subpopulations that lack any predictive meaning for
(groups of) individuals and in our societies.
11.2.2 GENETIC FACTORS AND THEIR INTERPLAY WITH DIET

Monogenetic traits, with high attributable risks have long served as a model to
understand geneenvironment interactions. Phenylketonuria (PKU) is a classical
example. Phenylalanine (Phe) is a necessary nutrient, but in the absence of a normal
phenylalanine hydroxylase enzyme (PAH), Phe metabolites accumulate and eventually
lead to mental retardation. The frequency of homozygous deficient subjects is on
the order of 1/10,000, which resembles the disease risk because of the ubiquitous
exposure to Phe via dietary protein. Because the adverse effect of the nutrient is
limited to the population that harbors the genetic variant, all cases can be prevented
by adapting the diet in this specific group.
In epidemiologic terms, the PKU example basically envisages geneenvironment
interactions as the effects resulting from the interplay between a combination of two
factors: Phe and PAH, both either present (+) or absent (), yielding four possible
combinations (++, +, +, ) in which only the combination Phe+ (applicable to
whole population) with PAH (applicable to 1/10,000) exhibits a large increased
risk of disease. Because of the ubiquitous exposure to Phe and the rare PAH defect,
PKU is perceived as an inborn single-gene disease, although the successes of dietary
treatment illustrate that it is in fact a highly specific geneenvironment interaction.
Note that in this example of geneenvironment interaction, the risk of disease in
Phe+/PAH subjects is extremely high, whereas the other three combinations carry
no risk. Also note that there are many different PAH polymorphisms leading to a
similar distortion in Phe metabolism and subsequent mental retardation.
In epidemiology, it was initially hoped that relatively small population-based
case control or cohort studies could suffice to detect such strong interactions as in
the PKU example. It was stipulated that the weak associations for certain dietary
risk factors might be the average of strong associations in certain (relatively small)
subgroups of the population, with no association being present in the remainder. To
account for the low gene frequency, genetic epidemiologists also developed efficient
study designs making use of the genetic relationships between family members and
corresponding data analysis techniques. Apart from shared lifestyle factors, it
appeared that some syndromes have a major genetic component. In the field of CVD,
these are, for example, familial hypercholesterolemia, and the hereditary forms of
arrhythmias such as the long QT syndrome, or Liddles syndrome in hypertension.
In the field of diabetes, they are for example, maturity-onset diabetes of the young
(MODY), and in the field of cancer examples are LiFraumeni Syndrome and
hereditary non-polyposis colorectal cancer (HNPCC). The genes involved in these
familial syndromes are responsible for only a small percentage (say, 5 to 10% at
most) of the total burden of these diseases. These groups have very high risks of
developing the disease and, hence, they develop it at a young age. Nevertheless,
about 10 to 20% of the population has some type of positive familial history for a
specific chronic disease with relative risks on the order of 2 to 3, depending on
definitions of familiar risk and the disease as such; thus, about 20% of the case load
involves some familial background. The next section provides an overview of the
status quo of both types of population-based and family-based studies in the etiology
of colorectal cancer and diabetes.
11.3 EXAMPLES OF THE STATUS QUO:

CARCINOGENESIS AND DIABETES
11.3.1 DIET, GENES, AND COLORECTAL CANCER
Already in 1968, MacMahon suggested that environmental factors may influence
cancer risk through the genetic mechanism of mutation, whereas genetic mechanisms
may modify the impact of the environment.4 More recently, genetic mutations have
been described in the etiology of cancer, such as mutations in the BRCA1 and
BRCA2 genes in hereditary breast cancer, and mutations in mismatch repair genes,
e.g., MLH1 and MSH2, in the HNPCC syndrome. Mutations in these genes result
in a 40 to 80% lifetime risk of developing disease. Such alterations are rare, however,
and taken together account for a very small proportion of cancer cases, probably
not more than 1 to 5%. Because highly penetrant cancer genes in frequently occurring
types of cancer have not been identified thus far, it seems likely that cancer results
from biological interaction of both genetic and environmental factors such as diet.
Lifestyle factors, such as dietary habits, smoking, and alcohol consumption are
considered to explain more than 30% of cancer incidence worldwide.5 A common
cancer that is thought to be predominantly influenced by lifestyle factors is colorectal
cancer (CRC). It is the second most common cancer in the Western world, although
it is relatively uncommon in most African, Asian, and South American countries.
The marked international variation in CRC incidence, the rapid increase in incidence
in families moving from a low- to high-CRC incidence country (migrant studies)
and the results from association studies imply that lifestyle factors contribute impor-
tantly to the etiology of this disease. It is estimated that more than 50% of CRC
may be preventable by reducing alcohol intake, avoidance of weight gain, more
physical activity, and decreasing (processed) meat consumption, while increasing
the consumption of plant-derived foods.5
Relative risks (RRs) observed for dietary factors and CRC in association studies
are, however, relatively small (RR < 2). Two decades ago, inherited susceptibility
to diet-related carcinogens or specific intakes of nutrients, vitamins, and nutrients
provided a promising explanation for these weak associations and inconsistent find-
ings in association studies. Research began to focus on more common genetic
alterations that by themselves may not substantially impact disease risk but that in
concert with environmental exposures may lead to development of cancer.
Although human experimental studies do show interesting individual variation in
response to specific foods and nutrients,6 the results of most observational studies on
inherited susceptibility to dietary factors in cancer risk appeared to be disappointing.7
Observed dietgene associations are weak and inconsistent. So far, many individual
studies have been designed to examine the main effects of individual factors and do
not have adequate power to examine interactions. Often, only one or two single SNPs
are studied simultaneously, the functionalities of which are not known. Thus, these
studies are not equipped to take complex dietgene and genegene interactions into
account. Also, the focus on SNPs as markers of inherited variation may be insufficient
to capture the complexity of cancer, given that it ignores other genetic alterations (loss
of heterozygosity, epigenetic silencing, somatic mutations, etc.).
11.3.2 DIET, GENES, AND DIABETES

The incidence of type 2 diabetes (T2DM) has risen rapidly during the past decades
together with its major risk factor, obesity. According to the CDC in Atlanta (U.S.),
the age-adjusted prevalence of diagnosed diabetes increased, for example, 81% for
men and 59% for women from 1980 to 2004. Nowadays, the U.S. prevalence rate
amounts to 5.1%, and the lifetime risk for U.S. inhabitants is 33%. Similar trends
occur in other Western countries. The WHO has estimated that because of both
demographic changes and increases in obesity, about 230 billion people will suffer
from diabetes in 2030. The increasing prevalence illustrates that genes cannot be
the rate-limiting factor, because the genetic background of the population cannot
change that much so quickly. However, T2DM remains a good example of dietgene
interaction: already in 1927, Elliott P. Joslin stated that heredity loads the canon
but obesity and other stresses pull the trigger.
The heritability of T2DM in twin studies has been estimated to range from 25% to
up to 80%.8 Having a sibling with T2DM increases the risk about threefold. Indeed,
McCarthy and his group pointed out that this sets the upper limit for the combined
effects of all susceptibility variants.9 Linkage studies have revealed QTLs associated
with either T2DM or related continuous traits such as glucose levels and insulin levels
(higher in prediabetes, indicating insulin resistance). So far about 17 QTLs on 12 different
chromosomes have been identified.10 Next, metabolic studies gave rise to a long list of
candidate genes, such as glucose transporters, insulin-receptor-related factors, sul-
fonurea receptors, glycogen synthase, PPAR-gamma and related cofactors, but also
lipid-metabolism-related factors, beta-2 and beta3 adrenergic receptors, and many obe-
sity related factors as well.9 Interestingly, but perhaps not surprisingly, the products of
some of these candidate genes are diabetes drug targets. For example, the SUR1/KIR6.2
gene locus (ABCD8/KCNJ11 on chromosome 11p) plays a role in the membrane
potential of the beta cell. Before the genetics was even fully disclosed, the sulfonureas
were designed to stimulate insulin secretion in T2DM. The antidiabetic thiazolidinedi-
ones (TZDs) are PPAR-gamma agonists and work as an insulin sensitizer. Again, PPAR-
gamma is an important candidate gene, for which the scientific evidence is so compelling
that the Pro12Ala variant is associated with T2DM. Calpain 10 (CAPN10) is an inter-
esting candidate gene as it is one of the few susceptibility genes that has been found
by genomewide screening of families.11 Finally, very recently, a new surprising gene
has been discovered, TCF7L2, transcription factor 7-like 2,12 of which the mechanism
and targets for treatment or prevention are now being investigated.
Few studies so far have investigated the topic of dietgene interactions in T2DM.
Researchers of the Finnish Diabetes Prevention Study have been active in this field
and their findings have been summarized recently.13 PPAR-gamma was one of the
first genes studied. Carriers of the low-risk 12Ala allele did not develop diabetes
when they were part of the lifestyle intervention group (diet and physical activity),
whereas there was an increasing trend for T2DM with 12Ala in the control group.
This difference was accompanied by differences in weight loss. However, these
findings also demonstrate the premature state of this type of research, as the high
diabetes risk for 12Ala is contrary to what is generally found. An English group,
for example, noticed that when the dietary P:S ratio is low, BMI and fasting insulin
levels in Ala carriers are higher than that in Pro/Pro homozygotes, but when the P:S
ratio is high, the opposite is seen.14 In fact, this would suggest that on a low dietary
P:S ratio diabetes risk is larger in Ala carriers, and on a high P:S ratio diabetes risk
is larger in Pro/Pro homozygotes. As fatty acids are natural ligands for PPAR-gamma,
these findings await confirmation and further study; also, a lack of power and other
methodological issues make interpretation difficult.
11.4 PREVENTION PARADOX: SICK INDIVIDUALS

AND SICK POPULATIONS
11.4.1 PREVENTION PARADOX
In 1985, Sir Geoffrey Rose pointed at the so-called prevention paradox. He stressed
the importance of distinguishing sick individuals and sick populations.15 Essentially, the
prevention paradox implies that individual benefits and population benefits do not nec-
essarily coincide and that cost efficiency must be included in the equation to find an
optimal prevention strategy. With respect to the role of genes and environmental factors
such as nutrition, Rose indicated that in populations with homogeneous exposures, the
classical epidemiologic studies will fail to detect the role of environmental factors and
they will only identify markers of susceptibility. With respect to nutrition, the major
differences are between populations on the world scale, as illustrated by the epidemio-
logic transition; apart from socioeconomic gradients, the variation in nutritional expo-
sures within populations is limited and the (relative) risks of nutritional factors are modest
or weak indeed. Thus, markers of susceptibility become relatively more prominent in
terms of the (relative) risks. Furthermore, Rose argues, strategies in disease control are
the high-risk approach, which seeks to protect susceptible individuals (e.g., by PN)
and the population approach, which seeks to control the causes of incidence.
Population strategies aim to shift the distribution of exposure as a whole. For
instance, industry-mediated approaches (with or without legislative incentives) that
control the distribution of foods over the population have proved effective and cost-
effective. Apart from the classical example of John Snow,16 who controlled water
supply and eliminated cholera in London, other examples include fluoridated drink-
ing water, iodized salt, and elimination of trans fatty acids; these approaches benefit
to public health, whereas the risk reduction in any individual subject is of minor
importance. Mass-media approaches also aim at reaching the population as a whole
and changing lifestyles and dietary patterns; this, however, has not been very effec-
tive and is suspected to reach the well-educated socioeconomic strata of society,
who already have relatively good health and low disease risks.17
In line with medical tradition, nutritional advice is also provided to individuals
at high risk. For instance, food industries are also targeting lifestyle and food choice
and offer functional foods (with or without health claims) specifically formulated
for high-risk groups of the population or those motivated to improve their quality
of life. Along these lines, it has been suggested that knowledge of differential risks
resulting from predisposing metabolic genetic traits can be useful in behavior
modification programs and chemoprevention strategies that will have the maximum
impact.18 Even apart from cost-efficiency issues, it remains to be seen if such high-
risk approaches followed by behavioral modification and personalized nutrition are
as effective as the whole-population approach.
11.4.2 GENEENVIRONMENT INTERACTIONS:

INDIVIDUAL AND POPULATION HEALTH
Now that data from association studies on nutrients, genes, and their interaction have
become available from epidemiologic studies, individual and population benefits can
be used to derive the case load from each of the components. We express the strength
of associations as the relative risk RR, and we use RRN for nutritional factors, RRG
for the genetic factor, and RRNG for the effect of their joint exposure. The population
attributable risk (PAR) describes the (relative) case load owing to the joint distribu-
tion of nutritional and genetic factors relevant to public health; the PAR depends on
both the RRs mentioned and the proportion of the population Pij in each of the
population segments. Rather than providing formulas, we prefer to visualize the case
load as a function of the prevalence of the genetic factor and presence/absence of
nutritional exposure.
As an example, RRs and prevalences of the genetic variants are taken from a
case control study on aflatoxins, polymorphisms in aflatoxin-metabolizing genes,
and the risk of liver cancer in Sudan.19 In this study, consumption of aflatoxin-
contaminated peanut butter was taken as a marker of exposure to aflatoxins. Genes
considered were GSTM1, GSTT1, and EPHX. The RRs obtained from this study
are summarized in Figure 11.1. The y-axis depicts the RRs and the x-axis the
6
GSTM1 GSTT1 EPHX
7%
5
RR
2
40% 17% 54%
46% 4%
15%
1
27% 23% 20% 35%
12%
0
Present Null Present Null Wild-type Mutant
Genotype Frequency
FIGURE 11.1 Graphical representation of population attributable risk [PAR (%), surface
area] as a function of relative risk (RR) and genotype frequency. Data were derived from a
case control study on liver cancer as related to aflatoxin-contaminated peanut butter and
genotypes of GSTM1, GSTT1, and EPHX, conducted in Sudan.
proportion of the population harboring the genotypes considered. Thus, for each
genotype, the case load is graphically represented by the surface given by the product
of the prevalence Pij and the RR of the nutritional exposure, here aflatoxin-
contaminated peanut butter (this is the case because we have used a median split
for aflatoxin exposure). For the three genes considered, aflatoxin reduction in the
whole population would prevent about 60% of all cases, independent of genotype
(we neglect the fact that the PARs for exposure to aflatoxin are not identical because
of weak associations between genotype and exposure and rounding errors). For the
highly prevalent GSTM1 and GSTT1 null polymorphisms, a substantial part of
the case load derives from the exposed and genetically nonsusceptible subgroup.
For the variant of the EPHX gene, only one tenth of all exposure-related cases are
derived from the category of the genetically susceptible.
This example illustrates that any single screening test, even when the mutant
genotype is highly prevalent (GSTs), tends to provide relatively small PARs for
exposure, whereas major opportunities to prevent disease in large segments of the
population are not seized. The RRs do not differ greatly according to genotype, and
the rare frequency of the high-risk genotype renders the case load due to the
interaction relatively small. As illustrated in the previous section, published
dietgene interactions in obesity, diabetes, and cardiovascular and malignant disease
all suggest that this example of liver cancer reflects a general phenomenon.
11.4.3 ADDITIVE AND SYNERGISTIC EFFECTS

The way in which genetic and dietary factors act jointly is relevant to individual vs.
population health and depends on the RRs and Pijs. This joint action may be additive
or synergistic (or antagonistic).
When genes and diet work additively, the risks of exposure and genetic back-
ground tend to add, i.e., RRNG = 1 + (RRN 1) + (RRG 1). An additive model might
apply when the etiological pathways in which the gene and the dietary factor are
involved are completely disjoint. In fact, in such cases, dietgene interaction is not
really an issue: the nutritional factor adds the same risk independent of the genotype
and, consequently, individuals with the mutant and those with the wild-type genes
would reduce risk by the same extent by dietary modification.
In all other cases, nutritional and genetic risks act in a synergistic or antagonistic
way. In synergistic interaction, the effect of genes and diet together exceeds their
separate effects; in the case of antagonism, it is less. In synergism, individuals would
benefit from screening, but the full potential of nutritional modification would not
be met, because a large part of the case load will still derive from the exposed group
with the low-risk genotype. In epidemiology, logistic regression models and Cox
Proportional Hazards models tend to be used for estimating the independent risks
of two or more factors, which implicitly assumes synergy of the form RRNG = RRN
* RRG. Of course, evaluation of this assumption by using statistical interaction terms
may provide evidence for deviations from the multiplicative model, leading to
additive or supramultiplicative effects.
When the prevalence of the high-risk genotype is small, RRNG must largely
exceed RRN * RRG to justify screening. When the risk of exposure is almost entirely
restricted to the genetic variant (as in the PKU example), i.e., RRN = 1, the case load
resulting from exposure and the genetic factor coincide. In this specific and uncom-
mon case of geneenvironment interaction, dietary modification (if possible) can
reduce the risk among the genetically susceptible. When RRN exceeds 1, the case
load emerging from the relative large size of the less susceptible genetic subpop-
ulation does not justify an individualized approach that only targets this subgroup,
unless both the prevalence of the high-risk genotype is large and RRNG is large as
well.
The foregoing observations become highly relevant if results on different genes
are considered simultaneously. The study on liver cancer could provide a hypothetical
example: One could speculate that if the genes mentioned are independent and
if the RRs are multiplicative that the RR of liver cancer for exposure to aflatoxin
contaminated peanut butter could rise to RR = 2.7 * 2.3 * 7.7 = 48, which would
be applicable to a population segment of 0.44 * 0.41 * 0.04 = 0.007, resulting in a
very low PAR. Thus, when several genes are addressed simultaneously, the RRs and
individual benefits can be large. However, the higher the RR, the smaller the pop-
ulation segment it applies to. Finally, any of the data in this example should be used
with caution, both because of uncertainties due to sampling and because of uncer-
tainties with respect to independence of the polymorphisms and the biochemical
pathways involved.
Thus, for combinations of specific dietgene interactions, the biological inter-
action (synergism or antagonism) of genes and diet is superimposed on background
risk. Joint effects can be quantified using theoretical models, but we still have to
learn what type of model applies most commonly to geneenvironment interactions.
It is highly questionable whether elucidation of biochemical pathways can contribute
to valid inference on additivity, multiplicativity, or other synergistic/antagonistic
effects; at the same time, high-throughput mass approaches that simultaneously
include a large number of gene variants and dietary factors in epidemiologic research
are often incapable of yielding sufficient evidence for specific genenutrient inter-
actions. Uncertainty on individual risks is a necessary consequence and will limit
the possibilities for PN advice.
11.5 DISCUSSION FRAMEWORK FOR PN

11.5.1 CONTEXT
Discussions on possibilities and limitations of PN would benefit from a framework
that distinguishes the context in which PN is applied (clinical, prevention), the
research angle (genomics), and issues relevant to individual advice (tailoring).
For effective treatment of clinically sick and very-high-risk individuals, drug
treatment is indicated. Genetic factors are gradually being incorporated in clinical
settings to optimize treatment according to prognostic indicators. In the preventive
domain, risks are much lower, metabolic aberrations are relatively mild and inter-
ventions in lifestyle are implemented via, e.g., dietetic advice, the basic form of PN.
The objective is to successfully alter habitual dietary intake and thereby reduce risk
factors of disease (blood lipids, blood pressure, overweight, etc.) within the individ-
uals range of possibilities. Apart from compliance issues (also relevant to drug
treatment), such individual advice can indeed lead to direct changes in dietary intake,
but demonstrating effects on habitual diet, nutritional status, risk factors of disease,
and long-term disease risk is increasingly difficult.
In contrast to effective individual treatment, an important issue in prevention is
effective communication. Mass-media approaches to low-risk populations are known
for poor efficacy, conceptually similar to poor compliance in drug treatment. Group
approaches (e.g., at schools) to communicate dietary recommendations are considered
more effective, and the idea is that tailored dietary advice based on individual
genomics data would be even more effective. This notion is questionable, however.
As we can see in the practice of clinical dietetics, dietary advice incorporates dietary
assessment, provides options for dietary change, helps to set priorities, is practical,
and monitors changes, provides feedback on dietary intake (and exposure/health
markers). For the individual, the trade-off is between lowering a high risk at the
expense of giving up a habitual diet. For PN the healthy individual will make a
trade-off between (unsure) future risk reduction and rewards of (positive) feedback
on the one hand and direct losses of well-being related to changing habitual patterns
of food preparation, psychological aspects, taste, and social aspects of eating. Direct
changes in quality of life must be balanced against uncertain future longevity.
(See Chapter 13 and Chapter 14.)
11.5.2 GENOMICS
Research on dietgene interactions and genomics research are in their infancy. So far,
the results of dietgene studies have not been equivocal. Several methodological issues
may account for this, such as the need for large study samples, a consistent and well-
measured phenotype, high-quality information on dietary exposures, etc. There
certainly is an enormous growth potential for research, and we are beginning to see
its potency in explaining variation of nutrient needs and risks between individuals.
In this chapter, we focused on genetic data; information on human disease risks
as related to transcriptomics, proteomics, and metabolomics is not available nowa-
days. The genetic blueprint can be regarded as a somewhat static trait that tells us
about the potential of our body, but differently so for different individuals and results
are still scarce. Whole-genome scans employing thousands of SNPs will soon
become available and indicate new relevant loci, with possibly surprising findings.
This may enable the search for underlying pathways and study of the simultaneous
effects of SNP-patterns and dietary modification. However, statistical methods and
common scientific principles are still being developed. For instance, the etiological
distance between SNPs and a clinical disease is long, and dilution occurs. Shortening
this distance by reducing the phenotype from individual to organ, tissue, cellular,
or molecular levels tends to increase the strength of scientific proof on specific
interactions; at the same time, however, zooming in on molecular laboratory research
excludes higher-level interactions and increases uncertainty due to other factors,
thereby tempering the generalizability to individual subjects and populations. In
addition to SNPs, future results of transcriptomics, proteomics, and metabolomics
may provide powerful noninvasive techniques to screen populations and monitor the
effect of dietary interventions. For instance, the role of folate in DNA methylation and
other reversible epigenetic processes may be examined using these techniques on blood
samples from a large number of individuals, taking inherited differences in folate

metabolism into account. However, the current enthusiasm about scientific promises
in genomics research should not result in overestimating the upper limit of genediet
interactions on public health. For instance, for T2DM a relative risk of about 3 is
provided by family history studies and similar figures apply to cancer etiology; as
illustrated, extreme high-risk groups are clinically important, but they tend not to
be of major public health importance.
An intrinsic scientific problem in genomics research is that genes and diets play
their role in many biochemical pathways at the same time. Each gene may have many
polymorphisms/SNPs relating to (slightly) different mRNAs, proteins, and metabolite
patterns. Studying all functions under controlled conditions and assessing their asso-
ciated risks is virtually impossible. Thus, there is a need for integrative markers, i.e.,
phenotypic markers reflecting pathophysiologically relevant distortions in metabolic
pathways, attributable to key regulatory genes of that pathway or to genetic variation
in the enzymes involved. This calls for identification of pathways starting from known
classical risk factors of disease or of being healthy in old age and using the metabolites,
proteins, transcriptome, and genetics as explanatory factors. In conclusion, even if the
scientific hopes will be realized, it will take a long time, before the results are ready
for use in individual advice and the benefits for population health remain limited.
11.5.3 TAILORING
As mentioned earlier, balancing direct gains in well-being against future risk is now-
adays restricted to clinical settings with high and intermediate risks. However, ICT
technologies can make tailored advice available to a larger proportion of the population
in a cost-effective way. A personal balanced diet, however, should not be based on
nutrient composition alone, but it should also include dietary assessment (as the starting
point to identify options for change), energy balance (physical activity), lifestyle factors
(e.g., smoking and drinking), and medical risk factors (BMI, blood lipids, blood
pressure, etc.), whereas genomic markers might be added in the distant future.
Nevertheless, PN targets the individual, and remains a high-risk-group approach, rather
than the population-based approach advocated by Sir Geoffrey Rose.
The concept of PN also calls for defining health as more than the absence of
disease, also including personal well-being (as WHO does). If this is ignored, it
may lead to adding years to life rather than life to years. In a clinical high-risk
context, risk reductions are paid for not only economically, but also personally by
accepting side effects that adversely affect well-being in a physical (surgery), mental
(pain, suffering), or social sense (hospitalization). The price individuals want to pay
for reducing moderate or low risks are logically much less, but they are taken into
account in tailored advice to individual patients. If well informed, individuals can
decide their own personal balance of long-term risks (with attached values) and
direct benefits in the domain of well-being (with their attached values). In this view,
personal is based on individual values rather than genomic risks.
In practice, it appears that some people are willing to pay large amounts of
money for genetic testing and tailored advice, even though the advice provided is
quite similar to generic guidelines; however, individual risk cannot be addressed
with reasonable accuracy and precision, leaving the predicted risks associated with
high uncertainty. Thus, similar to the widespread use of supplements and the increas-
ing use of functional foods, the PN approach may preferentially reach the worried,
healthy, well-educated, and health-literate population segments that already have the
highest life expectancy. This way, PN may contribute to widening the socioeconomic
health gradient in populations. This calls for careful balancing of the population-
targeted policies against the PN approach.
11.6 FUTURE PERSPECTIVES FOR PN

The idea that genomics research can lead to PN is theoretically sound, but has a
very long way to go and is currently a bridge too far. Data are available on a
limited number of specific dietgene interactions, but they are not promising for
public health. Moreover, it is not clear what these data imply for polygenic and
multifactorial interactions in disease etiology. Moreover, a theoretical framework is
lacking that could help to reconstruct individual and population risks from the
laboratory-based pieces of evidence of dietgene interactions. And even when large-
scale human data using genomic technology and systems biology become available,
large uncertainties would remain when scientific data on risks among groups of
subjects are applied to the individual.
Nevertheless, at some point in time, systems biology, individual, and population
health might converge for some highly relevant genediet interactions. Requirements
and technologies for PN may evolve parallel in time, probably starting from dietetic
counseling practices. In the meantime, research on genomics and tailoring must be
accounted for, as the idea and rough prototypes are already on the market. This
stresses the need to further explore the possibilities, limitations, and implications of
full-scale implementation, simultaneously by scientists, stakeholders, and the public.
Nevertheless, if PN is realized, the prevention paradox remains, i.e., individual
benefits and population benefits do not necessarily coincide, unless direct benefits
in subjective well-being are included in the health concept. At least the time-scale
of the risks, cost-efficiency, and personal well-being must be included into the
equation to find the optimum balance. For now, the omics are not relevant to public
health, as proteomics and metabolomics are expensive, and statistical techniques for
interpretation are still in development. For the time being, family history probably
remains the most useful tool as an easy and efficient summary measure of a large
variety of highly specific genetic factors.
Applications of PN await convincing scientific evidence, and it is possible that
such evidence may never materialize. This should be taken as a challenge for a
serious debate between molecular sciences, social sciences, and public health pro-
fessionals to balance the health and well-being of individuals and populations with
respect to PN.
ACKNOWLEDGMENT
The work underlying this chapter has been funded by a grant of the Wageningen
University Board (VIVRE project MyFood, 20042005).
KEY READINGS
Brennan, P., Gene-environment interaction and aetiology of cancer: what does it mean and
how can we measure it?, Carcinogenesis 23(3), 3817, 2002.
Mutch, D.M., Wahli, W., and Williamson, G., Nutrigenomics and nutrigenetics: the emerging
faces of nutrition, FASEB J. 19(12), 160216, 2005.
Ordovas, J.M. and Corella, D., Nutritional genomics, Annu Rev Genomics Hum Genet, 5,
71118, 2004.
Perera, F. P., Environment and cancer: who are susceptible?, Science 278(5340), 106873, 1997.
Rose, G., Sick individuals and sick populations, Int J Epidemiol 14(1), 328, 1985.
REFERENCES
1. Mutch, D.M., Wahli, W., and Williamson, G., Nutrigenomics and nutrigenetics: the
emerging faces of nutrition, FASEB J. 19(12), 160216, 2005.
2. Report of the committee of inquiry into the future development of the public health
function and community medicine. HMSO, London, 1988.
3. The World Health Report 2002 Reducing risks, promoting healthy lifestyle WHO,
2002.
4. MacMahon, B., Gene-environment interactions in human disease, J Psychiatr Res 6,
393242, 1968.
5. WCRF, Food, nutrition and the prevention of cancer: a global perspective WCRF/
American Institute for Cancer Research, Washington, D.C., 1997.
6. Lampe, J. W. and Peterson, S., Brassica, biotransformation and cancer risk: genetic
polymorphisms alter the preventive effects of cruciferous vegetables, J Nutr 132(10),
29914, 2002.
7. Brennan, P., Gene-environment interaction and aetiology of cancer: what does it mean
and how can we measure it?, Carcinogenesis 23(3), 3817, 2002.
8. Beck-Nielsen, H., Vaag, A., Poulsen, P., and Gaster, M., Metabolic and genetic
influence on glucose metabolism in type 2 diabetic subjects experiences from
relatives and twin studies, Best Pract Res Clin Endocrinol Metab 17(3), 44567, 2003.
9. McCarthy, M. I. and Zeggini, E., Genetics of type 2 diabetes, Curr Diab Rep 6(2),
14754, 2006.
10. Florez, J. C., Hirschhorn, J., and Altshuler, D., The inherited basis of diabetes mellitus:
implications for the genetic analysis of complex traits, Annu Rev Genomics Hum
Genet 4, 25791, 2003.
11. Sreenan, S. K., Zhou, Y. P., Otani, K., Hansen, P. A., Currie, K. P., Pan, C. Y., Lee,
J. P., Ostrega, D. M., Pugh, W., Horikawa, Y., Cox, N. J., Hanis, C. L., Burant, C. F.,
Fox, A. P., Bell, G. I., and Polonsky, K. S., Calpains play a role in insulin secretion
and action, Diabetes 50(9), 201320, 2001.
12. Grant, S. F., Thorleifsson, G., Reynisdottir, I., Benediktsson, R., Manolescu, A., Sainz,
J., Helgason, A., Stefansson, H., Emilsson, V., Helgadottir, A., Styrkarsdottir, U.,
Magnusson, K. P., Walters, G. B., Palsdottir, E., Jonsdottir, T., Gudmundsdottir, T.,
Gylfason, A., Saemundsdottir, J., Wilensky, R. L., Reilly, M. P., Rader, D. J., Bagger, Y.,
Christiansen, C., Gudnason, V., Sigurdsson, G., Thorsteinsdottir, U., Gulcher, J. R.,
Kong, A., and Stefansson, K., Variant of transcription factor 7-like 2 (TCF7L2) gene
confers risk of type 2 diabetes, Nat Genet 38(3), 3203, 2006.
13. Uusitupa, M., Gene-diet interaction in relation to the prevention of obesity and type
2 diabetes: evidence from the Finnish Diabetes Prevention Study, Nutr Metab
Cardiovasc Dis 15(3), 22533, 2005.
14. Luan, J., Browne, P. O., Harding, A. H., Halsall, D. J., ORahilly, S., Chatterjee, V. K.,
and Wareham, N. J., Evidence for gene-nutrient interaction at the PPARgamma locus,
Diabetes 50(3), 6869, 2001.
15. Rose, G., Sick individuals and sick populations, Int J Epidemiol 14(1), 328, 1985.
16. Smith, G.D, Behind the broad street pump: aetiology, epidemiology and prevention
of cholera in mid-19th century Britain, Int J Epidemiol 31(5), 920932, 2002.
17. van der Pal-de Bruin, K.M., de Walle, H.E., de Rover, C.M., Jeeninga, W., Cornel,
M.C., de Jong-van den Berg, L.T., Buitendijk, S.E., and Paulussen, T.G.. Influence
of educational level on determinants of folic acid use, Paediatr Perinat Epidemiol
17(3), 25663, 2003.
18. Perera, F. P., Environment and cancer: who are susceptible?, Science 278(5340),
106873, 1997.
19. Tiemersma, E. W., Omer, R. E., Bunschoten, A., vant Veer, P., Kok, F. J., Idris, M. O.,
Kadaru, A. M., Fedail, S. S., and Kampman, E., Role of genetic polymorphism of
glutathione-S-transferase T1 and microsomal epoxide hydrolase in aflatoxin-associated
hepatocellular carcinoma, Cancer Epidemiol Biomarkers Prev 10(7), 78591, 2001.
Section II
Personalized Nutrition and
Stakeholders in Society
12 Imminent Applications
of Nutrigenomics:
A Stakeholder Analysis
J.T. Winkler
CONTENTS
12.1 Introduction ..................................................................................................151

12.2 Scientists.......................................................................................................152
12.3 Timeless Issues behind Current Controversies............................................153
12.4 Food Companies ..........................................................................................154
12.4.1 Personalized Marketing....................................................................155
12.4.2 The Counterpart to Promise.............................................................155
12.4.3 Foods vs. Drugs, Science vs. Law...................................................156
12.5 Consumers ....................................................................................................157
12.5.1 To Regulate or Not to Regulate? .....................................................158
12.5.2 Influences on Acceptability..............................................................158
12.5.3 Familiarization with Genetic Testing...............................................159
12.6 Competitive Athletes ....................................................................................160
12.6.1 Gene Transfer: Therapy and Doping ...........................................161
12.6.2 Optimizing Nutrition........................................................................162
12.6.3 Ethics: Con and Pro .........................................................................162
12.7 Health-Care Providers..................................................................................163
12.7.1 Offloading Care................................................................................163
12.7.2 Personalized Nutrition as Self-Care ................................................164
12.7.3 Policy Options..................................................................................164
12.7.4 Nutritional Compliance....................................................................165
12.8 Future Perspectives ......................................................................................166
Key Readings.........................................................................................................166
References..............................................................................................................167
12.1 INTRODUCTION
The promise of nutrigenomics is grand; the practicalities are daunting. Many scien-
tists researching the frontiers of the field believe the complexities are such that useful
applications are decades away. But, for better or for worse, scientists do not always
151
determine what happens in the world. Other groups have interests in nutrigenomics.
For them, benefits are in prospect genetic understanding and personalized diets,
better health and enhanced capacities, improved foods and targeted marketing, lower
costs, and higher revenues. Rewards such as these may inspire a more optimistic
interpretation of the scientific evidence. And provide incentives to immediate action.
This chapter identifies probable actions of key interest groups whose interplay will
shape the future of nutrigenomics. The intent is description, not prescription. The
perspective is intentionally agnostic, to describe not what is desirable, but what is
likely, for good or for ill. The method is stakeholder analysis. Five groups will be
considered: scientists, food companies, consumers, competitive athletes, and health-
care providers. Their behavior will raise issues on which policy decisions will have
to be taken, sooner or later. For all these parties, the discovery of inherited suscep-
tibilities and capabilities is a prerequisite for whatever actions they want to take.
Hence, this chapter is as much about genetic testing as it is about nutritional
interventions.
12.2 SCIENTISTS
Most people active in nutrigenomics at present are scientists, conducting research
in genomics, transcriptomics, proteomics, metabolomics, and associated disci-
plines. In this field, as everywhere else, differing views coexist even among the
cognoscenti.
Nonetheless, the view of numerous leaders is that, because of the multiple genes,
multiple polymorphisms, and multiple metabolic pathways involved, unraveling
genetic influences on diet-related diseases will take a long time. We are in the stage
of basic science and likely to remain so for many years. With the present state of
knowledge, speculation regarding practical applications of nutrigenomics would be
premature.
Thus, for example, Ben van Ommen, Coordinator of the NuGO program: I
think it is way too early to give dietary advice based on your genes.1 Similarly,
Sian Astley, U.K. coordinator of NuGO, feels there is no way that nutrigenomics
is ready yet to give individual advice or help health policy.2 Not in our lifetimes,
says optimist Bruce German.3 Never, feels Larry Parnell, because it is technically
impossible to cover all possible genegene permutations.4 Prudence is most vividly
expressed by Jose Ordovas: What we are seeing now is the very, very beginning.
Nutrigenomics has established proof of concept, but thats it. To commercialize the
science at this point is almost like throwing darts in the dark.5
Even within the scientific community, however, others urge immediate use. Most
obvious are scientists in companies selling genetic tests, such as Sciona and Genelex.
Among clinical geneticists, Aubrey Milunsky condenses the positive case into the
title of his book, Your Genetic Destiny: Know your Genes, Secure your Health, Save
your Life.6 For others, moral as well as practical arguments encourage incremental
application of nutrigenomic knowledge as soon as it emerges. Diet-related diseases
are so widespread, responsible for so much mortality, morbidity and misery, that it
would be criminally negligent to continue prescribing ineffective general diets
when personalized nutrition is possible.7
Imminent Applications of Nutrigenomics: A Stakeholder Analysis 153
12.3 TIMELESS ISSUES BEHIND CURRENT

CONTROVERSIES
Between Nutrigenomics Now! and Never! lie issues in the philosophy of
science. Among them is the classic: how much certainty is needed before initiating
practical action? In translation: what percentage of the variance in disease rates
needs to be explained by a set of genes before therapeutic interventions are
justified?
That question has practical immediacy because nutrigenetic tests are not static,
either useful or inadequate, once and for all. Rather, they are developed incrementally
as evidence accumulates. Scionas first test, for example, assessed 7 genes, then 19,
with further expansions anticipated.
Another fundamental issue for nutrigenomics has surfaced in ironic form: should
the use of tests be determined by the quality of available responses to the information
they produce? Public health screening programs are sometimes postponed because
no effective remedies exist or the cost of screening exceeds the population benefits
of subsequent treatment.
In contrast, one criticism of nutrigenetic testing, following research by the U.S.
Government Accountability Office, was that it only produced commonsense
recommendations.8 Many nutritionists would be flattered and surprised that contem-
porary dietary advice is now common sense. The usual criticism of nutrition
science is that some of its therapies are unproven, or only erratically effective,
because the mechanisms of action are incompletely understood. To which the stan-
dard rebuttal has been: they may do good and probably will do no harm. Such benign
insouciance will no longer suffice. With the advent of nutrigenetic testing, the
standards of proof for nutritional therapies will rise. So too will the specificity of
nutritional advice provided to the tested.
Indeed, one of the greatest promises of nutrigenomics is to identify the mecha-
nisms of nutritional action, by monitoring genetic influences on metabolism, through
genetic and other biomarkers along pathways. This provides an alternative to the
orthodoxy of double-blinded randomized controlled trials, which are impossible with
foods. Meanwhile, a prolonged debate is in prospect about the accuracy, acceptability,
and advisability of nutrigenomic tests. It will include the scientist vs. scientist
squabbles that enrage nonspecialists who want certainty. And exonerate those who
would change nothing.
However, a related issue may hasten the introduction of tests. Expectations
are rising for genetic controls on subjects in nutritional trials. For example, two
issues in the relationship between salt and hypertension are the extent and
strength of inherited salt sensitivity. Those convinced that genetic influences
on blood pressure are strong (including the salt and savory foods industries) feel
the absence of genetic controls in earlier studies undermines the positive asso-
ciations they produced. Researchers may henceforth feel it judicious to incor-
porate genetic controls, however imperfect, using whatever tests are available
at the time.
Underlying all these issues is a common question: when is the science sufficient?
The correlative question raised by stakeholder analysis is: who decides?
12.4 FOOD COMPANIES

That grand promise of nutrigenomics to identify inherited susceptibilities to diet-
related diseases, then provide personalized nutritional advice creates commercial
opportunities. There is a large prospective market for targeted foods, products
designed for genetically susceptible people. This is the nutritional analogue to
pharmacogenetics, producing drugs tailored to patients inherited characteristics.
Targeted foods should be understood broadly. The category includes not just
manufactured products, but also nutritional supplements, herbal remedies, and
enhanced versions of fresh produce. Some fresh vegetables, fruits, eggs, meats, and
fish containing higher levels of beneficial components may be produced by conven-
tional breeding and feeding, others by genetic modification. Current examples of
both approaches are provided by brassicas with higher levels of glucosinolates,
offering a measure of protection against some cancers. Nutritionally modified plants
(in which nutrient profiles are altered rather than agronomic properties) are not, at
present, conceived as part of nutrigenomics, although the concept could expand.
The point, for present purposes, is simply that the commercial response to nutrige-
nomics will include a wide variety of targeted products.
Some will be niche products, such as glutenfree foods for celiacs. Some may
become mass-market products, where the critical genetic variations are widespread
in a population. Candidates, as nutrigenomics develops, might include genetic influ-
ences on insulin resistance, leading to new versions of traditional foods formulated
with lower glycemic loads, relevant to obesity and diabetes. That topical example
suggests targeted foods might also appeal to people suffering from the same problem
without the inherited susceptibility, expanding the market. They are an advanced
version of a 20th century industry development functional foods, with added
genetics. In fact, opportunities may be greater still. Nutrigenomics involves a two-way
relationship. Nutrient intakes influence gene expression, and expressed or silenced
genes influence nutrient metabolism.
Current practice is, by default, to adapt food intakes to inherited susceptibilities,
through both reductions and additions. Those with familial hypercholesterolemia,
for example, are urged to cut saturated fat intake and eat more cardioprotective
foods.9 People inheriting galactosemia exclude milk products. Those suffering from
cystic fibrosis supplement with fat-soluble vitamins.
As nutrigenomics develops, however, we may expect more targeted products
designed to influence gene expression/silencing. Some of these genes may influence
diseases that are not today considered diet-related. Targeted foods could become
a large category. If so, such products would arrive on the back of an already strong
trend. To a degree unimaginable 30 years ago, foods are promoted for their nutritional
strengths, on health platforms. No longer just less bad (low-fat, reduced salt,
sugar-free), high-tech fortified foods now promise positive benefits. Phytosterol-
enriched margarines to lower cholesterol are the best-documented example.
At an abstract level, it is a form of progress that health is becoming a popular
basis for choosing foods, alongside taste, price, and convenience. But assessing the
nutrient profiles of products and linking them to personal genetic profiles is complex.
Consumers will need help.
Dietary advice will become a routine part of genetic counseling. Conversely,

dieticians will have to incorporate genetics into their repertoire. Nutrition education
will be transformed, once one-size-fits-all recommendations are no longer plausible.
Food companies will naturally offer pointed guidance, and they are much bigger
nutrition educators than governments. Ahead lies a public contest for share of mind.
12.4.1 PERSONALIZED MARKETING

Genetic tests could also transform another trend in the food industry, toward direct
marketing. One clich in advertising is that 50% of it is wasted, but we do not
know which half. Much mass-market advertising is showered on people who will
never be interested in the products being promoted.
Advertisers have long attempted to reduce this wastage, to concentrate marketing
on serious potential customers, a process also known as targeting. Early efforts were
primitive, based on media audience profiles, demographic characteristics, and attitudinal
orientations. With widespread adoption of bar coding and supermarket loyalty cards,
advertisers extract individual buying patterns from electronic-point-of-sale data, identi-
fying established customers to encourage repeat purchases. Prospective buyers are now
picked out from interests expressed though Internet search patterns. In contemporary
marketing jargon, this is establishing a personalized relationship with the customer.
In future, individuals genetic test data could facilitate the ultimate personal
relationship, genetic marketing. When, eventually, nutrigenomics verifies individ-
ual susceptibilities and nutritional needs, it will simultaneously identify potential
customers for targeted foods. In a sporting metaphor popular among marketers, this
creates the chance to go one-on-one with the customer in the kitchen.
Persuading consumers to share their genetic data, even bits of it, will probably
provoke a larger and longer version of the competition between formula and baby
food manufacturers to sign up new mothers for supplies while they are still in
maternity wards. Inducements may be necessary, for institutions as well as individuals.
Economics enters the equation in another way. From a marketing perspective,
targeted foods are archetypal value-added products. And the value they add beyond
conventional foods is a particularly important one, health. In one version of market
logic, better products justify higher prices.
In sum, nutrigenomics will stimulate new foods, new ways of selling them, and
enlarged margins. These are powerful temptations to rapid applications of nutrige-
nomic research. So powerful that some food companies may force the pace, by
making specific genetic tests directly accessible to consumers at subsidized rates,
or indeed, free, a loss leader to recruit customers for new foods (see Chapter 14).
Personalized nutrition will bring with it personalized marketing. Targeted foods will
be promoted though targeted advertising. But, of course, this brave new commercial
world will not be problem free.
12.4.2 THE COUNTERPART TO PROMISE

Those who buy targeted foods, as opposed to those who sell them, may view higher
prices differently, as a levy tacked onto products that are alleged to do people good,
a health premium. Some economists will agree with them, because that pricing
strategy creates economic problems.
Some people, made aware of their nutrigenetic susceptibilities, will become
anxious about their future. In distress, some will feel they need to grasp whatever
nutritional remedies are available, whatever the cost. In the language of economics,
they become price insensitive; they have inelastic demand for those targeted
foods that are relevant to their risks. This is the exploitation of anxiety. In many
countries, commercial appeals to fear are prohibited by advertising regulations. But
genetic susceptibility may be presented more subtly, fright disguised as fate (see
Chapter 15).
Not only is individual psychology at issue, but macro health economics. Diet-
related diseases are overwhelmingly diseases of poverty. With nutritional problems,
as with much else, the poor suffer more. If targeted foods are permanently positioned
at higher prices, then the people who need them most will be least able to acquire
them. The public health promise of nutrigenomics would then not be realized (see
Chapter 11).
But a different economic calculus leads to an intermediate position. Even for
poor people, absolute price is not always the concern; sometimes they seek value
for money. Targeted foods, if genuinely effective, would not only diminish disease
risks and symptoms, but also associated costs. These include outgoings on care, but
equally important, income foregone because of hindrances to employment. There
may be net benefits, even if prices are high. Whether targeted foods really represent
value for money, however, is conditional on another fundamental issue, namely, their
efficacy. Will personalized nutrition be effective in coping with inherited, diet-related
diseases? Even the most cautious researchers are optimistic on that point, or they
would not be working in the field. But decades of nutrigenomic research may still
not suffice, because this is as much a legal as a scientific issue.
12.4.3 FOODS VS. DRUGS, SCIENCE VS. LAW

At present, the law says unequivocally that foods do not prevent, treat, or cure
disease. In the U.S., foods cannot even ameliorate it. Or, more precisely, no one
may claim that they do. Such medicinal claims are reserved for licensed drugs.
That is one of the few points on which food laws around the world actually agree.
The prohibition has always been scientifically irrational. Modern nutritional
science actually began with the demonstration that vitamins can prevent and cure
deficiency diseases. That, of course, is still true. For example, U.K. dietary recom-
mendations state that the essential and undisputed roles of vitamin C (ascorbic acid)
are to prevent scurvy [and] 10 mg/d [is] also sufficient to cure clinical signs
of scurvy .10 So say scientists. But food manufacturers may not make the same
statement about products containing vitamin C, no matter how enriched. These laws
were, in fact, designed to staunch the gushing hyperbole of patent medicines. They
did that job effectively. But foods were caught in the same net. Nutrients are treated
the same as Victorian elixirs. Still, today, even the most nutrient-rich foods may not
be presented as being prophylactic or therapeutic.
Meanwhile, most developed countries have resorted to fudge. A compromise
emerged that, although companies may not claim that foods reduce disease, they
may claim that foods reduce risk factors for disease. Companies may be required
at some stage to present the scientific evidence behind their health claims, a
surrogate assessment of efficacy. The U.S. has adopted a particularly complex variant
of this approach, with four grades of qualified claims, depending on the quality
of the evidence, with even preliminary and unpersuasive science allowed at the
bottom rung.
Nutrigenomics has the potential to end this hotchpotch. When mechanisms of
action are eventually proved, it will offer new and stronger forms of evidence, beyond
normal nutritional arguments, about foods beneficial actions against disease. That
should, logically, undermine the legal distinction between foods and drugs.
However, doctors and most other health professionals receive little training in
nutrition. They are often suspicious of dietary interventions, citing the absence of
randomized controlled trials. This is part rationality, part protective mechanism
against alternative therapies. Furthermore, interests have built up to defend the
status quo, especially in pharmaceutical companies. Changing the law will be a
lengthy, perhaps interminable, process. Legal pragmatism has already resisted sci-
ence for more than a century and may continue to do so.
In the interim, possibly a long interim, nutrigenomic tests will be offered to
people with varying degrees of completeness and certainty. Targeted foods will come
wrapped in imprecise statements about what they do and how well. To paraphrase
Eliot, this is the way nutrigenomics may begin, not with a bang but a whimper.11
But food companies will strive for a more explosive beginning. They will actively
promote both tests and products. Advertising may be as persuasive in implying the
efficacy of targeted foods as it has been about many other goods. Consumers may
respond more to these promises than to the scruples of scientists and lawyers.
12.5 CONSUMERS
Compared to the corporates, consumers have a more intense as well as a more
personal interest in nutrigenomics. For some, their life depends on it. For others,
only their quality of life hangs in the balance, but that is motivation enough for
most. At present, relatively few consumers know about nutrigenomics, so they do
not yet constitute a coherent interest group. But targeted commercial promotions
to the vulnerable will spread information quickly. And with it, hope.
Sick consumers sometimes employ a different rationality from scientists, defined
by their context. In the early days of HIV/AIDS drugs, sufferers refused to wait for
the conclusion of long randomized controlled trials. Their personal end might arrive
before scientists had proof. The rigor of three-stage clinical trials was seen, not
as good scientific practice, but as delay, or worse, denial of treatment.
Seriously ill patients are increasingly demanding access to experimental, unap-
proved drugs.12 The controversy over Herceptin for breast cancer is the latest example
of consumer impatience with scientific method and regulatory due process. Sellers
of remedies often encourage, support, and even finance, such demanding consumers.
Elected politicians may respond to their appeals. They did in both the cases cited.
Diet-related diseases are usually chronic, lacking the life-and-death drama of these
examples. But the popular perception of genes-as-destiny may lend nutrigenomics
an urgency that nutrition has seldom acquired, outside famines. Consumers are
already more accepting of nutritional remedies than doctors and regulators. Nutrige-
nomics will reinforce their preferences.
However, both the health and commercial benefits of targeted foods depend on
consumers willingness to share their nutrigenetic data with others, not just health
professionals, but also food companies. Who will have access to genetic profiles?
12.5.1 TO REGULATE OR NOT TO REGULATE?

The initial public debate in many countries, led by consumer and patient organiza-
tions, suggests widespread reluctance, despite inducements. Attention has focused
on possible misuses of data, especially by insurance companies and employers. The
concern is potential discrimination against the genetically susceptible.
Advocates urge measures to protect privacy, forbidding compulsory testing and
restricting access to the results of voluntary tests. Many call for more regulation
of tests, laboratories, biobanks, direct-to-consumer sales, promotional claims, prod-
ucts linked to test results, nutritional supplements, fortified foods, and more.
Historically, this is an unusual development. Normally, regulation trundles along
belatedly after new technologies, like a rubbish van after a parade, sweeping up the
mess. Those concerned about the new genetics are trying to preempt problems.
This strategy contrasts with another new technology that emerged around the
same time, the Internet. For all its many virtues, the Internet is accused of facilitating
numerous kinds of fraud, identity theft, mass disruptions through viruses, breaches
of security and privacy through hackers, stimulating pornography and pedophilia,
and not least for nutritionists, contributing to the surge in childhood obesity. Yet
preventing regulation of the Internet is, for many, a point of principle. When Internet
service providers accept some restrictions in China, this is interpreted as capitulation
to power in pursuit of money.
Preemptive regulation is potentially a significant development in consumer
protection. In part, it reflects the particular sensitivity many people feel about all
things genetic, because of associations with eugenics and enduring battles with
religion. Among its several purposes, this book is intended as a corrective to those
apprehensions.
However, in many other areas of modern life, developed societies are in a
deregulatory phase. Regulation is often presented as red tape, a burden on
enterprise. In some countries, deregulation is explicit policy. In the EU, all legislative
proposals on foods must produce a regulatory impact analysis.
At its grandest, we have here a conflict between two approaches to organizing
modern society. Practically, we are in for a lengthy series of acrimonious regulatory
battles about genetic testing and its unintended consequences.
12.5.2 INFLUENCES ON ACCEPTABILITY

But even within the field of genetic testing, nutrigenomics is different. Testing for
diet-related susceptibilities is not like testing for Huntingtons disease, the subject
of much attitudinal research. For some inherited diseases, testing means disclosure
of an unavoidable fate, even a death foretold.
Nutrigenomic testing has two distinct attributes that make it exceptional, and
probably more acceptable. First, a positive response is always possible, changing
diet. Second, that response is under the individuals control. These are intrinsic
advantages of nutrigenomic testing, compared to other types of screening. Here,
within economic constraints, knowledge empowers individual action.
But extrinsic factors will also influence the acceptability of these tests. Those
who are not just at risk, but already ill with a diet-related disease, may be more
insistent than resistant about testing. As with the aforementioned drugs, some will
be more concerned about short-term relief from problems than long-term invasions
of privacy. With nutrigenomics, as with other subjects, the views of consumer
organizations are not always those of individual consumers.
The terms and conditions on which nutrigenomic tests are presented to the public
will further shape uptake. Tests subsidized by industry would certainly be cheaper
than those available individually from specialist testing companies. And perhaps
more accessible, too, if distributed through local outlets, even unconventional ones
such as pharmacies and supermarkets. Lower prices and easier availability would
widen the customer base and help counter the allegation that nutrigenomics tests
are merely a luxury for the worried well. They would become a resource for the
demonstrably sick as well as those at risk.
Moreover, the agreement of consumers to share results with third parties, includ-
ing food companies, might be a condition for providing subsidized tests. To those
worried about privacy, such requirements will seem extortionate. But concerned
consumers might be willing to share information with those they believe, rightly or
wrongly, could help them, especially as initial tests are likely to be specific to genes
for defined problems, rather than expensive whole-genome scans that could allow
later fishing expeditions for other information.
Decisions on sharing personal data will not be taken by individuals in isolation,
but within an evolving social context. Nutrigenomics is arriving at a time when
skepticism toward conventional medicine is growing. Alternative or complementary
therapies are increasingly popular, including dietary responses such as supplements,
functional foods, and herbal remedies, as well as simply healthier eating. Nutrige-
nomics again reinforces an existing trend.
But it is already clear that nutrigenomics will be the focus of a public debate in
its own right, amplified in popular media, about whether it is endangering or empow-
ering.13 The outcome is by no means certain. In Europe, the opponents of the new
genetics in all its forms have had considerable success. Other parts of the world
are more enthusiastic.
12.5.3 FAMILIARIZATION WITH GENETIC TESTING

Underneath the fireworks of public polemic, however, a long-term process of cultural
familiarization is under way, with both genetic testing and positive applications of
its results. DNA tests add a new dimension to the already widespread interest in
genealogy. DNA evidence, to identify murderers and rapists, often after years of
conventional investigative failure, is already a staple of mainstream crime reporting
and TV police dramas. A more compassionate use is to exonerate the victims of
miscarriages of justice.14
Proceeding slowly and incrementally is preimplantation genetic diagnosis of

embryos (PGD), to avoid disabling inherited diseases. At present, it is used only for
couples attempting in vitro fertilization. In time, in utero testing may become pos-
sible, permissible, and popular.
Just entering public awareness is genetic testing for what is gently described as
paternity discrepancy, to determine whether the apparent father of a child is
actually its biological father. The results, surprising to many, may foreshadow wider
use of such tests.
The major instrument of familiarization, however, is likely to be pharmacoge-
netics. Genes influence the ability to metabolize pharmaceuticals, just as with nutri-
ents. As a result, prescription drugs are in many cases, often in most cases, either
ineffective or produce harmful side effects.15 In theory, genetic testing could sub-
stantially reduce both waste and iatrogenesis. If such tests became routine, or indeed
required, before the administration of drugs, the mass of the population would rapidly
become accustomed to them in a helpful context.
Many practical obstacles lie ahead in this scenario. But considerable financial
pressures are driving it forward. For health-care providers, pharmacogenetics prom-
ises to make treatments more effective, reduce drug bills, and cut the cost of treating
side effects. For pharmaceutical companies, it should make their products more
effective, preempt damage litigation and facilitate development of new drugs. Both
sides have interests in applying pharmacogenetics widely and quickly.
None of these developments directly involves nutrigenomics. But increasing famil-
iarity with genetic tests in various forms and perceptible benefits from their use may
in time influence public acceptance of testing, including nutrigenomic testing.
In sum, it is possible that nutrigenomic tests, however imperfect, may actually
be popular with consumers, whatever scientists may think and, hence, targeted foods
too. The pull of demand reinforces the push of supply.
12.6 COMPETITIVE ATHLETES

Most of the discussion within nutrigenomics (and within this chapter, so far) has
concentrated on identifying inherited susceptibility to disease, and then doing some-
thing about it. But, that is only half the potential. The other half is about promoting
well-being as well as preventing disease, facilitating the good as well as preempting
the bad.
That prospect is already subject to a complementary debate. In its strongest
form, the issue is human enhancement, the idea that human capabilities can be
improved through a variety of means, including genetic and nutritional.16 It is a large
subject, beyond the broadest definition of nutrigenomics. For many, it seems a
futuristic fantasy or nightmare, like Huxleys Brave New World, the fictional precursor
of current concerns about designer babies.
In fact, one form of human enhancement is an immediate prospect, possible
gene doping of athletes at the 2008 Olympics in Beijing.
Gene doping is the thin but sharp end of a large wedge. It is introduced here to
illustrate additional issues that nutrigenomics will face in future. Hence, among the
stakeholders are elite athletes, as well as those who advise, coach, promote, and
otherwise minister to them. More important for public health, this category also
includes the much larger numbers who aspire to become leading, or at least improved,
sportsmen and women, the millions who train for marathons, as well as the handful
who win them. Hope is as vulnerable to exploitation as fear.
12.6.1 GENE TRANSFER: THERAPY AND DOPING

Gene doping emerged as a public issue following announcement of experimental
gene therapy to increase muscle bulk in elderly patients with muscular dystrophy.
The researchers then reported receiving inquiries, not just from clinicians, but from
coaches wanting a quick fix to build up players.17 Gene doping is a dramatic
pejorative but, as this anecdote indicates, it is really just one specific application of
gene therapy and, hence, extendable; indeed, it already is being extended. Others
are working on the same lines, identifying candidate genes that influence, for exam-
ple, oxygen uptake, heart efficiency, fast muscle fiber responses, and blood flow.
Different target tissues are appropriate for different sports. The goals include improving
speed and stamina, as well as strength. The focus is designer adults rather than
designer babies.
In a parallel with nutrigenomics generally, many specialists feel gene transfer
is at an early stage and much research is needed before it can be routinely applied.
But it is widely recognized that athletes are not so much interested in research, but
outcomes. They will jump on a product or a technique or a strategy that they perceive
to work, on the basis of no research at all.18
Prudence suggests gene doping is imminent. The World Anti-Doping Agency
(WADA) has already convened two conferences on preventing it. They concluded
the benefits are real and that it is feasible now. Initially, detection prospects seemed
poor. Since then, at least eleven research projects have been commissioned, so
WADA is now optimistic that it is ahead of potential sporting users.19
In practice, applications of genetics in sport, including nutrigenomics, will be
introduced in stages. Initially, tests will identify individuals with the potential to
become outstanding athletes. They have the capacity to improve, not only through
high-tech interventions, but also with conventional training and sports diets, by
maximizing metabolic performance.
The first such test is already on the market, for variants of the ACTN3 gene,
which allegedly separate sprinting sheep from long-distance goats.20 Many more
such tests will become available, with variable predictive value, but nonetheless
appealing to the ambitious, and to their parents. Instead of identifying disease
susceptibility, these tests indicate performance predisposition. People are selected,
not for treatment but enhancement.
Influencing gene expression is a more advanced stage. Expressing genes for
IGF-I and silencing genes for myostatin are prime candidates for muscle enlarge-
ment. Pharmaceutical triggers are one route. Eventually, nutritional influences on
gene expression/silencing might offer a truly invisible form of manipulation, becom-
ing the cheating of choice: personalized nutrition for peak performance.
Simultaneously, using transcriptomics and proteomics to identify variations in
gene expression is one of the most promising forms of detection for the full gene
doping on the horizon.
12.6.2 OPTIMIZING NUTRITION

Introducing nutrigenomics into sport would extend another trend. Nutrition has long
since moved beyond its original concerns with deficiency diseases and minimum
daily requirements. Many now conceptualize the science as striving for optimal
nutrition, improving nutritional status in all phases of the life cycle; if not absolute
well-being, at least better-being is the goal. Optimal nutrition has already been
embraced by athletes: eat to compete. At a simple level, this logic is manifest in
champions who foreswear fast food, dieticians advising at sports clubs, and health
food shops selling the cornucopia of nutritional supplements that have become a
routine part of competitive sport.
The most sophisticated example is the popular diet book, The Zone, which argues
that human capabilities are optimized by maintaining nutrients within a narrow
range, especially those influencing eicosanoids.21 The approach was originally devel-
oped to improve the performance of a swimming team.
Adding nutrigenomics to the repertoire of sports nutrition will require new
technical skills, but no leap in imagination. No imagination is needed either to
envisage ethical issues that will erupt.
12.6.3 ETHICS: CON AND PRO

The heart of the matter is genetic manipulation of human nature. This is not just
about sporting prowess, but also other physical traits, including the cosmetic. Nor
is it limited to the physical, but encompasses mental development, psychology, and
behavior. Enhancement is not just for athletes; it could involve everyone. Improving
health, the initial focus of nutrigenomics, is only part of the emerging agenda.
That prospect horrifies some, raising moral issues about life, about playing God.
For others, it reopens the most notorious episode of the 20th century: enhancement
as the acceptable face of eugenics. At its core is always some form of social selection,
maybe not race, but something unpleasant. The alternative name for the process,
trans-humanism, is even worse, with echoes of Frankenstein.
Current disputes over abortion, stem cells, and PGD prefigure the controversies
ahead. These are some of the most passionately disputed social issues of the age.
The debates over human enhancement will be of that intensity, and more, especially
if they begin in the context of sporting nationalism.
Less familiar is the articulate affirmation of gene doping.22 The new genetics
introduces a new realism. Successful athletes have always had a genetic advantage.
We used to call it talent, now we can prove it is genes. Sports stars are natural
mutants. Manipulating genes of competitors is thus not cheating, it is actually fair
play, leveling the genetic playing field. Further, if gene transfers are acceptable for
the wasting elderly, why not for the promising young? If it is all right to manipulate
physical capabilities, why not other talents? Why not, indeed, any traits people find
desirable?
The future may include premium enhancement packages available to consum-
ers to improve athletic performance. They would present parents with a moral/
economic dilemma, whether to invest in their childrens sporting futures. This is not
scientific fantasy or commercial rip-off. It is actually one of the scenarios for 2025
being used now by the Department of Trade and Industry in the U.K. as the basis
for public consultations on the ethics of science. If enhancement for sport arrives,
purchasable packages for additional traits and talents will follow. They would put
prices on the qualities people value most. They would create a market in superiority.
Ethicists will see gene transfer as raising fundamental questions about humanity.
For the more worldly, it exacerbates eternal arguments about inequality. For others,
this is just a new technical option for giving people what they have always wanted,
but were previously unable to obtain.
Whatever the merits of arguments for and against, gene doping in popular sports
would rapidly increase public familiarity with genetic testing and beneficial
responses. In the long term, gene therapies for otherwise incurable diseases may be
the more compassionate face of genetic modifications.
The point here is not to endorse any of these arguments, pro or con. It is merely
to anticipate the imminent disputes, vicious but profound, about human enhancement
in all its forms, including the nutrigenomic. Gene doping at the Olympics may fire
the starting gun.
12.7 HEALTH-CARE PROVIDERS

Health-care providers everywhere, whether funded by taxation or insurance, are
under financial pressure. Demand for health care is rising, and so are the costs of
providing it, but budgets are not increasing proportionately. Providers respond var-
iously, but basic strategies include cutting costs, restricting demand, and rationing
provision.
One common component in cost control is transferring a proportion of bills to
patients, in the form of charges, copayments, deductibles, or cost recovery schemes.
A recurrent strategy for restricting demand involves limiting services to essentials,
excluding nonessential treatments. Dentistry is a frequent casualty. Demand is
thereby transferred away from the normal provider. Patients must buy services from
someone else, or go without.
12.7.1 OFFLOADING CARE

The radical solution, however, is to transfer treatment to the patient. It is never
presented so bluntly, of course. Euphemisms abound. Patients are invited to assume
more responsibility for their health, and to increase engagement in their own
care. The U.K. government commissioned a banker to produce two reports on the
fully engaged patient, those who monitor and treat their own conditions.23,24
Programs followed to develop self-managing patients, coping with minor illnesses
and chronic conditions or, in complex cases, becoming expert patients, and better
yet patient entrepreneurs.25
Patients relationship with health-care providers is redefined as partnership,
increasing lay involvement, to the point of shared decision making; shared care
and costs, too. The role of the National Health Service is empowerment, not
nannying, said the British prime minister, a double-edged slash at his countrys
welfare state.26
In the market-oriented U.S., the language is different, but the goal is the same.
The strategy is to shift the incentives to people having ownership of their own
health and therefore health care, as opposed to thinking some third party pays the
bills.27 In the misty past, when our proverbs were coined, sages spoke more plainly,
Patient, heal thyself. Self-care is the generic name for this approach. On balance,
over time, it may or may not prove better for patients. But its sprouting in the 21st
century has more to do with finance than therapeutics.
12.7.2 PERSONALIZED NUTRITION AS SELF-CARE

The relevance here is that personalized nutrition fits perfectly within the paradigm of
self-care. It facilitates all three of the strategic transfers: costs, demand, and treatments.
Genetic tests identify susceptibilities, for which the afflicted themselves take corrective
action, changing their diets. To health-care providers, nutrigenomics offers help with
both demand and supply. It can reduce demand in two ways. If effective, it prevents
some diet-related diseases. If not, it deflects professional consultations into self-care.
Conventional screening programs are often viewed ambivalently by health min-
isters and managers. They identify formerly invisible problems, stimulating more
treatments and additional costs. Nutrigenomic testing has opposite results. It shunts
therapy onto patients, requiring fewer treatments and less expenditure.
Personalized nutrition also reduces costs in more fundamental ways. Appropriate
remedies change from expensive drugs to cheaper foods. To reduce cholesterol, for
example, take not statins, but phytosterol-fortified margarines. For hypertension, not
ACE inhibitors, but weight loss. Statins and ACE inhibitors are, in many countries,
two of the most commonly prescribed drugs. Even more importantly, although
providers pay for much of the drugs, food is bought by consumers. This is not cost
sharing, this is transferring the whole bill, passing the buck. For all these reasons,
health-care providers will probably develop positive assessments of nutrigenomics.
12.7.3 POLICY OPTIONS

This appreciation could manifest itself in different policies. Regulation, the priority
for genetic skeptics, is but one of several options, and, for nutrigenomics, not the
most likely. The easiest alternative is passive acceptance of commercial initiatives
in testing and targeted foods, with no serious attempts at regulation. Why should
the state or insurance companies control access to nutrigenomic tests? Why not
open access for consumers? Self-care legitimizes laissez faire.
Alternatively, providers might actively engage with nutrigenomics, themselves
offering selected tests for common or expensive diet-related diseases. Expanded
clinical genetics services could cover some nutrigenomic testing and counseling.
Savings could be large.
In between, benign facilitation would encourage cooperation among interested
parties. For example, the U.S. Food and Drug Administration (FDA) recently reached
an understanding with the pharmaceutical industry, whereby companies will share
their pharmacogenetic test results with the government and in return receive regu-
latory immunity. The FDA will not use the data in drug approval processes, but
will gain valuable information.28 This is co-optation rather than control.
This is an intimation of arrangements in nutrigenomics. In principle, even the

limited genetic profiles produced by the initial testers would be useful for research,
public health, and individual treatment. Health-care providers would value them for
good diagnostic reasons as well as simple economics. Testers in turn would be keen
to recover the costs of subsidized tests and to gain legitimacy amid controversy. All
parties would favor an exchange, on terms and conditions to be negotiated.
12.7.4 NUTRITIONAL COMPLIANCE

The implications for individuals are mixed. One of the hopes of personalized nutri-
tion is that it will increase peoples motivation to eat healthily. Population dietary
recommendations are widely disregarded, despite much claimed conformity. People
may comply more readily with recommendations designed personally for them. So
far, there is little evidence to support this scenario, but logic and anecdote make it
plausible (see Chapter 13).
The policy question ahead is whether increased compliance will remain just an
aspiration or become an enforceable obligation. What expectations will be placed on the
individual to act on the results of nutrigenomic tests, to follow their personalized diet?
One template already exists. Parents of children with spina bifida or celiac
disease are expected to control their diets strictly and, further, to train children to
restrain their own intakes as they mature. Even optimists recognize that it requires
considerable effort to enforce such regimes, which leads to emotional stress in
families, and which takes its toll on the parents and the child.29 However, failure
to do so constitutes bad parenting, provoking interventions to protect children. These
are extreme cases. But will analogous expectations gradually be applied, in some
degree, to other genetically influenced dietary susceptibilities?
Already we see hints of social control. Increased insurance charges for some diet-
related conditions are established, tolerated, and paid. Restrictive conditions on access
to care is a further turn of the screw. For example, in the U.K., obesity drugs are reserved
for the very overweight, who must also demonstrate a successful commitment to dieting.
A more sophisticated version of conditional care is being introduced in the U.S.
Personal responsibility requirements are attached to publicly funded health programs
for the poor.30 Again, the logic is expressed in economic terms. It gives people rewards
and incentives to improve their health, and punishments if they do not.
Applicants must sign a pledge to attend health improvement programs, have
routine screenings, keep appointments, and take medicines. If they do, they are
entitled to enhanced benefits. If they refuse, or fail to keep their undertakings, they
are reduced to basic services.
Such experiments are controversial and may not endure.31,32 But the underlying
logic is particularly applicable to diet-related diseases. Changing food consumption
is essential for their cure, and nutrigenomic tests provide a technical basis for
personalized rules about diet. Eating appropriately for your genes could become a
prerequisite for care. Nutrigenomics will certainly create new norms. The issue then
becomes: will compulsion be applied to ensure compliance? If so, how much?
In sum, health-care providers will probably embrace personalized nutrition. To
others, pharmaceutical companies as well as patients, that may feel like a squeeze.

There are many unresolved issues with personalized nutrition, concerning both tests
and subsequent interventions. Much time and research will be needed to settle them.
Scientists are right to be cautious, technically speaking. But they are lined up against
other stakeholders with interests in rapid application of the developing nutrigenom-
ics. Significantly, all have more resources to pursue their goals than scientists, and
they all also have the prospect of increased revenues or substantial savings if they
succeed. For them, nutrigenomics is an investment.
Food companies have the wherewithal to develop new products, to market
targeted foods, to subsidize genetic tests, and to induce data sharing. Health-care
providers have the purchasing power to conduct or buy in nutrigenomic testing,
leading to substantial reductions in their drug bills and treatment costs. Most con-
sumers are not rich, but their combined spending power could provide the economic
base for both tests and products. These days, even elite athletes have more resources
than scientists.
In the earlier sections on these four stakeholder groups, there have been two
common themes. One is that their interests in nutrigenomics are all reinforcing
existing trends toward functional foods, alternative therapies, optimal nutrition,
and self-care. They are swimming with the tide.
Nonetheless, second, nutrigenomics will be controversial, provoking public debates
about unproven science, exploiting the vulnerable, protecting consumers, and inade-
quate health services, with the grandest of them all, GM humans, just over the horizon.
Genetics has always been special specially controversial. It began with polemics
between Darwinism and religion that persist to this day, followed by the 19th century
eugenics of criminality, the extensions to Social Darwinism, the 20th century eugenics
of race, and the postwar emergence of sociobiology and ethology. Since the mapping of
the human genome, the controversies have multiplied to include GM crops, intelligent
design, genetic testing, in vitro fertilization, and genetic databanks, among others.
Nutrigenomics will not escape controversy. In the years ahead, there will be both
scientific advance and mudslinging. It is potentially a major development, socially,
economically, and politically, as well as scientifically, and for public as well as indi-
vidual health. Understandably, people will feel strongly about it, one way or another.
At the end, this analysis returns where it began. It has been intentionally agnostic.
It has not sought to decide who is right or what is better. It has drawn attention to
policy issues that will arise, without predicting or prescribing how they might be
resolved. The conclusion is not whether personalized nutrition is desirable or not,
simply that it is likely to become a reality, and soon.
KEY READINGS
Miller, P. and Wilsdon, J., Better Humans? The politics of human enhancement and life
extension, Demos, London, 2006.
Milunsky, A., Your Genetic Destiny: Know Your Genes, Secure Your Health, Save Your Life,
Perseus, Cambridge, 2001.
Sweeney, H.L., Gene doping, Scientific American, July 2004, p. 37.
U.S. Government Accountability Office, Nutrigenetic Testing, GA-06-977T, Washington,

D.C., 2006.
Wanless, D., Securing Our Future Health: Taking a Long-Term View, HM Treasury, London,
2002.
REFERENCES
1. Cited in, Pray, L.A., Dieting for the genome generation, The Scientist, 19, 14, 2005.
2. Astley, S., Nutrigenomics, presented at Demos Symposium, You eat what you are,
London, April 20, 2005.
3. German, B., Nutritional phenotypes in the age of metabalomics, presented at NuGO
Week Symposium, Pisa, September 1013, 2005.
4. Parnell, L., The complexity of nutritional disorders presents a bioinformatic challenge
to genotyping, presented at NuGo Week Symposium, Pisa, September 1013, 2005.
5. Cited in Pray, L.A., op cit.
6. Milunsky, A., Your Genetic Destiny: Know Your Genes, Secure Your Health, Save
Your Life, Perseus, Cambridge, 2001.
7. Gibney, M., Public communication at NuGO Week Symposium, Pisa, September
1013, 2005.
8. Kutz, G., Nutrigenetic testing: tests purchased from four web sites mislead consumer,
U.S. Government Accountability Office, GAO-06-977T, Washington, D.C. 2006.
9. Heart U.K., www.heartuk.org.uk.
10. Department of Health, Dietary Reference Values for Food Energy and Nutrients for
the United Kingdom, HMSO, London 1991.
11. Eliot, T.S., The hollow men, in Collected Poems 19091962, Harcourt Brace Jovanovich,
New York, 1991.
12. Groopman, J., The right to a trial, The New Yorker, December 18, 2006, p. 40.
13. Trivedi, B., Hungry genes?, New Scientist, 193, No. 2587, 34, 2006.
14. The true state of C.S.I. justice, New York Times, January 29, 2007.
15. Connor, S., Glaxo chief: our drugs do not work on most patients, The Independent,
December 8, 2003.
16. Miller, P. and Wilsdon, J. (Eds.), Better Humans? The politics of human enhancement
and life extension, Demos, London, 2006.
17. Sokolove, M., The lab animal, New York Times Magazine, January 18, 2004, p. 28.
18. Hamilton, B., Human enhancement technologies, uncorrected evidence, House of
Commons Science and Technology Committee, October 25, 2006.
19. World Anti-Doping Agency, www.wada-ama.org.
20. Yang, N. et al., ACTN3 genotype is associated with human elite athletic performance,
Am J Hum Genet, 73, 627, 2003.
21. Sears, B., The Zone, Regan Books, New York, 1995.
22. Miah, A., Genetically Modified Athletes: Biomedical Ethics, Gene Doping and Sport,
Routledge, London, 2004.
23. Wanless, D. Securing Our Future Health: Taking a Long Term View, HM Treasury,
London, 2002.
24. Wanless, D., Securing Good Health for the Whole Population, HM Treasury, London,
2004.
25. Cayton, H., Patients as entrepreneurs: who is in charge of change?, in Andersson, E.,
Tritter, J., and Wilson, R. (Eds.), Healthy Democracy, The National Centre for
Involvement, London, 2006.
26. Donnelly, L., PM calls for empowerment instead of nannying, Health Serv J, 12,
August 3, 2006.
27. McKinnell, H.A., cited in Berenson, A., An insiders critique of the health care system,
New York Times, June 4, 2005.
28. Food and Drug Administration, Pharmacogenomic data submissions, final guidance,
Issue 22, March 2005.
29. Milunsky, op cit.
30. Eckholm, E., Medicaid plan prods patients towards health, New York Times, December
1, 2006.
31. Steinbrook, R., Imposing personal responsibility for health, New Eng J Med, 355,
353, 2006.
32. Bishop, G. and Brodkey, A., Personal responsibility and physician responsibility:
West Virginias Medicaid plan, New Eng J Med, 355, 356.
13 The Personal Factor

in Nutrition
Communication
Laura I. Bouwman, Maria A. Koelen,
and Gerrit J. Hiddink
CONTENTS
13.1 Introduction ..................................................................................................169

13.2 The Personal Factor in Theory: Personal Relevance ..................................171
13.2.1 Creating Awareness about Health Communication.........................171
13.2.2 Threat Appraisal in Health Communication....................................172
13.2.2.1 Perceived Severity.............................................................172
13.2.2.2 Perceived Vulnerability.....................................................173
13.2.3 Coping Appraisal in Health Communication ..................................174
13.2.3.1 Perceived Effectiveness
of the Recommended Action ............................................174
13.2.3.2 Perceived Self-Efficacy.....................................................174
13.2.3.3 Cost-Benefit Evaluation....................................................175
13.3 Influence of Information about Genes, Nutrition, and Health
on Personal Relevance .................................................................................176
Key Readings.........................................................................................................181
References..............................................................................................................181
13.1 INTRODUCTION
Information that is personalized to take into account a targeted individuals char-
acteristics and situation is more effective in influencing that persons health behav-
ior than general information. The central factor for effectiveness is perceived
personal relevance. In most theoretical models of behavior change, concepts relat-
ing to this personal factor are included, such as perceived personal risk, effective-
ness of recommended actions, self-efficacy, and cost-benefit evaluation. In the field
of nutrition, communication can be tailored to an individuals food-related needs,
habits, preferences, and interests shaped by his or her physical and social
169
circumstances. When the messages are also delivered in a location and at a time
desired by the individual, we can speak of personalized nutrition communication.
The innovative field of nutritional genomics (nutrigenomics) is expected to lead to
more insights into the interaction between diet, genes, proteins and metabolites,
and health. This will hopefully lead to evidence-based strategies for the develop-
ment of stronger health messages that will influence both the perceived severity
and effectiveness of the actions recommended in nutrition communication. Nutrige-
nomics aims to use genetic testing to assess personal vulnerability to the devel-
opment of nutrition-related illnesses. Such a test could contribute to perceived
personal vulnerability and to the perceived effectiveness of recommended actions,
thereby influencing the perceived personal relevance of healthy eating. Next to a
personalized assessment, the availability of a personalized solution by means of
a diet, product, or nutrient that helps prevent nutrition-related diseases is a prereq-
uisite for the concept of personalized nutrition to become integrated in health
behavior change strategies.
Many causes of premature death and illness are preventable, or at least
postponable, at the level of individual behavior. As individuals, if we did not
smoke, exercised more, and ate less saturated fat and more fruits and vegetables,
we would probably be healthier. In the last decades, a lot of effort has been put
into improving dietary habits through nutrition communication. However, it has
not been effective in changing the behavior of populations or individuals: in most
European countries, actual consumption is not in line with basic recommendations
for healthy nutrition. Although consumers know what they should be doing, diets
still contain too much energy, saturated fat, sugar and salt, and insufficient veg-
etables, fruits, and fish. Dietary habits are important determinants of health
because unhealthy eating, coupled with poor lifestyle choices, increases the risk
of disease such as obesity, diabetes, cardiovascular disease, and cancer. The
growing incidence of diet-related diseases accentuates the need for innovative
approaches that motivate people to eat healthily [13]. A promising approach is
personalization of nutrition communication. Reviews on health interventions [4,5]
and research on the effect of personalization [68] have shown that information
that is personalized to a targeted individuals characteristics and situation is more
effective in influencing that persons health behavior than general information.
Central to this chapter is perceived personal relevance, because personalized
nutrition communication that is not perceived as being relevant to the individual
will not induce motivation to eat healthily in the long term. We discuss how
personal relevance is integrated in communication, fear arousing communication,
and health behavior change theory. In most theoretical models, concepts relating
to this personal factor are included, such as selective perception, perceived per-
sonal risk, effectiveness of recommended actions, and self-efficacy [914]. The
innovative field of nutritional genomics is expected to lead to more insight into
the interaction between diet, genes, protein, and metabolites, and health [15]. The
possible influence of this innovation on perceived personal relevance of nutrition
communication is discussed in Section 13.3. Finally, we discuss some of the issues
surrounding the personalization of nutrition communication and topics for future
research.
The Personal Factor in Nutrition Communication 171
13.2 THE PERSONAL FACTOR IN THEORY:

PERSONAL RELEVANCE
In communication, fear-arousing communication, and health behavior change theory,
concepts include personal factors in several stages of the behavior change process:
creating awareness, the threat appraisal, and the coping appraisal. Weinsteins precau-
tion adoption model identifies a series of steps preceding the taking of preventive
action, the first three of which relate to awareness [16]:
Realize that a specific risk exists

Acknowledge that the risk is significant and can affect people
Recognize that one is personally vulnerable to the risk
13.2.1 CREATING AWARENESS ABOUT HEALTH COMMUNICATION

Every day, we are confronted with an enormous number of messages. Nutrition
information is provided through mass-media channels, on product labels, bill-
boards, on the Internet and, in specific cases, through schools or by health pro-
fessionals. It is impossible to pay attention to all of these messages: we have to
be selective. People have a tendency to expose themselves to information that is
consistent with their own attitudes and opinions [9]. This process is also known
as selective perception. Attention is only given to information that is perceived as
being somehow personally relevant; this means that exposure does not automati-
cally elicit attention. However, what personally relevant means is not defined
specifically in the literature. Batra and Ray [17] define relevant as stronger and
providing topics that are of interest or valued by individuals. McGuire [18]
describes relevant as interesting. Personal relevance can be described as con-
sistent with personal attitudes and opinions. Sherif et al.s social judgment theory
[19] also states that people tend to accept ideas that agree with their own personal
views. Messages that are not in line with their personal latitude (range) of accept-
able options are ignored or dismissed. Another important factor relating to message
content is the level of involvement. According to the elaboration likelihood model
[10], the persuasive impact of a message can be central or peripheral. The key
variable is involvement: the extent to which an individual is motivated and able
to think about the position advocated in the message (issue-relevant thinking).
When involvement is high, elaboration is also high. Elaboration involves cognitive
processes such as evaluation, recall, critical judgment, and inference and occurs
through the central persuasive route. Changes are stronger when induced through
the central route. An issue that becomes more personally relevant to a recipient
will increase his or her motivation to engage in thoughtful consideration and action
[10]. Communication that contains information opposed to personal beliefs, atti-
tudes, or opinions induces uncomfortable feelings. People tend to reduce those
feelings by avoiding dissonant information. This process is known as cognitive
dissonance [20]. A person who strongly believes that healthy food tastes bad will
ignore information that aims to persuade him or her to eat healthily by saying that
healthy food is tasty. Research has shown that many people have misperceptions
about their personal food intake. They rate their food intake as healthy and
therefore do not consider nutrition communication on healthy eating as being
personally relevant [2124].
13.2.2 THREAT APPRAISAL IN HEALTH COMMUNICATION

In between creating awareness and the behavior change process, feelings of personal
risk will depend on the perceived severity of the consequences for an individuals
health, and the effectiveness and costs of preventive behavior. Well-known models
of preventive health behavior such as the health belief model [13], theory of planned
behavior [25], and protection motivation theory [12] contain the concepts of per-
ceived severity and vulnerability as influencing factors on motivation to change
behavior. These models assume that people are able to adequately assess the risk to
themselves associated with their behavior.
13.2.2.1 Perceived Severity
In most communication about nutrition, the messages are about the consequences
of unhealthy eating. The theory of planned behavior, protection motivation theory,
and health belief model all rely on severity as the influencing factor for perceived
personal relevance. Fear appeals are often made to spell out the severity of
nutrition-related diseases such as diabetes and cardiovascular diseases, but the
fact that these are outside the experience of most people may explain why the
appeal is not effective. Gleicher and Petty [26] and Liberman and Chaiken [27]
state that fear arousal can induce two different coping strategies: either acting as
a motivator to induce intensive (and accurate) message processing or inducing
defense motivation, both temporarily. Defense motivation is most likely to occur
when a health threat is both severe and personally relevant because personal
beliefs are being threatened. According to the heuristic-systematic model [27],
the processing goal of defense-motivated people is to confirm the validity of a
particular attitudinal position (I am eating healthily) and to disconfirm the validity
of others (your eating choices place you at risk). Defense-motivated people will
process information selectively in the way that best supports their own beliefs
(see also: selective perception). Risk perception research has raised questions
about the assumptions of most models in preventive health behavior, which is,
as stated earlier, that people are able to adequately assess the risk to themselves
associated with their behavior. Risk assessment is a complex process influenced
by several factors that interfere with accurate assessment of personal risk. The
catastrophic effect, controllability, reversibility, and whether the risk is taken on
a voluntary basis or not influence risk perception and, thereby, fear arousal. For
instance, perceived risks of unhealthy lifestyles (voluntary) are known to be lower
than perceived risks of new technology (nonvoluntary) [28]. Thus, people tend
to have misperceptions about their personal behavior, depending on the context
in which the risk information is presented, the way the risk is being described,
and their personal and cultural characteristics [11]. Furthermore, estimation of
personal risks tends to be biased. Many people overestimate small probabilities
(plane crashes) and underestimate large probabilities (heart disease). Risks that
are cognitively available through personal experience or intense media coverage
tend to be overestimated. This bias process is related to Tversky and Kahnemans
[29] availability heuristic and refers to peoples tendency to judge an event as
more probable to the extent that it is more easily pictured or recalled. The lack
of knowledge about the specific relationship between food intake and individual
risk of disease may interfere with the perceived severity of nutrition communi-
cation and contribute to misperceptions of personal risk.
13.2.2.2 Perceived Vulnerability
In protection motivation theory and the health belief model, perceived vulnera-
bility or susceptibility refers to the subjective risk of acquiring an illness if no
countermeasures are taken [30]. In combination with high perceived severity,
perceived vulnerability is known to build fear and increase the personal relevance
of messages. Research shows that people are quite aware of the relative risk of
specific activities or behavior, but this tends to change when personal risk needs
to be assessed. For instance, smokers accept the association between smoking
cigarettes and disease but do not believe themselves to be personally at risk [31].
This is referred to as unrealistic optimism from Weinsteins [32] paper that
focused on comparative risks in health risk perception. Van der Pligt [11]
describes six causes of unrealistic optimism that can lead to perceived personal
invulnerability, perceived control, egocentric bias, personal experience, stereo-
typical or prototypical judgment, self-esteem maintenance, and coping strategies.
Risks judged to be under personal control tend to induce feelings of optimism
[33]. People generally know more about their own protective behavior than about
that of others, causing an egocentric bias that can generate optimism. They also
tend to focus on their own risk-reducing behavior and are less aware of personal
behavior that increases risk. Personal experience with a risk tends to be relatively
vivid and can decrease unrealistic optimism. Stereotypical or prototypical judg-
ment is a relatively extreme image people have of high-risk groups, which is
unlikely to fit with their self-image, thereby increasing optimism. Generally,
people tend to rate their own actions, lifestyle, and personality as better than
that of others: this is known as self-esteem maintenance or enhancement. The
last factor that influences unrealistic optimism relates to coping strategies. Con-
ditions of high stress or threat can induce denial, thereby reducing emotional
distress but also reducing the likelihood of preventive actions or their success.
In general nutrition communication, vulnerability is addressed without reference
to the personal factor. Recent research on the stage model of processing of fear-
arousing communications developed by Das et al. [34] concluded that unless
individuals can be persuaded of their vulnerability to health risk, they are unlikely
to take protective action [35]. Through face-to-face consultation, vulnerability
to nutrition-related disease can be made more personally relevant, based on
assessment of the individuals lifestyle (e.g., calories), physical parameters
(e.g., blood pressure), and environmental circumstances (e.g., sedentary work).
As in relation to communication on severity, uncertainties about whether
unhealthy eating will actually lead to illness, also known as probabilistic

outcomes, interfere with the strength of the messages and, thereby, with perceived
vulnerability [36].
13.2.3 COPING APPRAISAL IN HEALTH COMMUNICATION

In the following section, personal factors in models of preventive health behavior
that influence the coping appraisal, such as perceived effectiveness of the recom-
mended action or response efficacy, perceived self-efficacy, and cost-benefit evalu-
ation, are discussed.
13.2.3.1 Perceived Effectiveness of the Recommended Action
The appraisal of recommended actions in terms of being effective in reducing or

avoiding health risks is included in several theories. Having undertaken an exten-
sive review, Sutton concludes that increasing communication on the efficacy of
the recommended action strengthens the individuals intentions to adopt that
action [37]. In protection motivation theory, motivation to engage in the recom-
mended behavior is also codependent on the appraisal of both response efficacy
and self-efficacy [12]. In the health belief model, the effectiveness of a recom-
mended action is a function of the perceived extent to which preventive behavior
will reduce the threat (perceived benefits) and the perceived negative aspects of
a preventive behavior (perceived barriers) [13]. Recommendations in most nutri-
tion messages are generic, or sometimes tailored to specific life stages such as
childhood or pregnancy. Perceived effectiveness of the recommended action at
the individual level can therefore be low. As during the threat appraisal, uncer-
tainties about the effectiveness of recommended actions (or probabilistic out-
comes) can interfere with perceived effectiveness of recommended actions. For
instance, a healthy diet is not necessarily a safeguard against the development of
cardiovascular disease [30].
13.2.3.2 Perceived Self-Efficacy
In Banduras [14] social learning theory (later called cognitive learning theory),
the central concept relating to personal factors is self-efficacy. It describes a
cognitive state of taking control in which people believe they are capable of
carrying out the specific behavior and can help create and control their environment
in doing so. This concept of reciprocal determinism of behavior and environment
is associated with concepts of self-management and self-control and is influenced
by several processes, such as direct experience. It is also influenced by the storing
and processing of complex information in cognitive operations that facilitate
anticipation of the consequences of actions, represents goals, and weighs evidence
from different sources to assess personal capabilities. This leads to a situation-
specific self-appraisal that induces feelings of confidence or insecurity about
behavior in new, unpredictable, or stressful situations. Self-efficacy is, then, the
perception of ones own capacity to successfully organize and implement new
behavior largely based on experience with similar actions and situations encountered
or observed in the past, also called performance history [38]. Self-efficacy is also
influenced by indirect or vicarious learning experience gained by observing others
(modeling), such as a parent, teacher, or television personality who seems to enjoy
a specific behavior. It is assumed that people learn more from models that are
competent, attractive, likable, admired, and loved. Also, similarity to/empathy with
the observer is known to influence learning [30]. Another influence on self-efficacy
stemming from others is verbal persuasion. Strong persuasive messages from a
respected, trusted person, such as a dietician, can have a positive influence on
feelings of self-efficacy. Besides influencing behavior, self-efficacy affects thought
patterns and emotional reactions, thereby inducing or reducing feelings of anxiety
or coping ability. It is linked to specific skills and, by personalizing communica-
tion, attention can be paid to an individuals feelings of self-efficacy. Communi-
cation about what to eat can, for instance, be matched with an individuals level
of cooking skills. Perceived behavioral control in the theory of planned behavior
is closely related to self-efficacy and refers to the fact that people can have positive
attitudes toward certain behavior but simply lack the resources to carry it out. In
protection motivation theory, the coping appraisal will be positively influenced
by response efficacy, the equivalent of effectiveness of recommended action and
self-efficacy.
13.2.3.3 Cost-Benefit Evaluation
The aforementioned theories on health behavior also include an evaluation of the

material and immaterial costs and benefits of changing behavior in line with the
recommended action. Those perceived benefits and costs are anticipated or
expected, but not yet realized. The cost-benefit evaluation is integrated in both the
threat appraisal (severity of health costs) and the coping appraisal (effectiveness/
health benefits). At the same time, other consequences relating to physical, mental,
social, and economic values will also be evaluated. Healthy meals that are not
appreciated by certain members of the family can raise issues at meal times. These
social costs of healthy eating can be perceived to be high and have a negative
influence on the cost-benefit evaluation. In the theory of planned behavior, the behav-
ioral beliefs reflect beliefs about the consequences of indulging in the behavior.
Together with the evaluation of those consequences, attitudes toward certain
behaviors are influenced. The perceived barriers in the health belief model refer
to the perceived negative aspects of a particular recommended behavior, such as
financial and social costs, and the efforts required to carry out the behavior. In
protection motivation theory, these are referred to as response costs. If these costs
are perceived to be too high, the personal relevance of the recommended action
can be perceived as low [30]. Figure 13.1 represents important personal factors
in the early stages of the behavior change process. In Table 13.1, an overview is
presented of the discussed concepts. Table 13.2 provides an overview of the
factors discussed relating to personal relevance and interfering concepts in stages
of behavior change. Table 13.3 contains the personal factors in the discussed
theory.
Stage 1: Unaware of
Awareness
Stage 2: Unengaged
Perceived personal
vulnerability
Stage 3: Deciding
about acting
Perceived severity
Perceived eectiveness of
the recommended behavior
Cost-benet evaluation
Stage 4: Decided Stage 5: Decided to

against
Stage 6: Acting
Stage 7: Maintenance
FIGURE 13.1 Personal factors in different stages of behavior change. (From Weinstein, N.D.,
Health Psychol., 7, 355, 1988. With permission.)
13.3 INFLUENCE OF INFORMATION ABOUT GENES,

NUTRITION, AND HEALTH ON PERSONAL
RELEVANCE
It has already been acknowledged that individual variability affects individual dietary
and nutrient requirements, nutritional status, and, hence, health. Therefore, recommen-
dations on nutrient intake vary according to age, sex, and ethnicity [39]. The relatively
TABLE 13.1
Personal Factors Influencing Perceived Personal
Relevance
Concepts Reference
Optimistic bias/unrealistic optimism Weinstein, 1980

Selective perception Sears and Freedman, 1971
Defense motivation Liberman and Chaiken, 1992
Probabilistic outcomes Zimbardo and Leippe, 1991
Misperception/bias of personal risks van der Pligt, 1996
TABLE 13.2
Factors Relating to Personal Relevance and Interfering Concepts in Stages
of Behavior Change
Awareness Threat Appraisal Coping Appraisal
Personal relevance Personal risk Perceived severity Perceived efficacy of

perception Perceived vulnerability the recommended
action
Cost-benefit
evaluation
Perceived self-efficacy
Interfering concepts Selective perception Defense motivation Probabilistic outcomes
Cognitive dissonance Optimistic bias: Fatalism
Optimistic bias unrealistic optimism
TABLE 13.3
Concepts Relating to Personal Factors in Discussed Theories
Concept Related to Personal
Theories Factors Reference
Heuristic-systematic model Defense motivation Liberman and Chaiken, 1992

Social judgment/involvement Involvement Sherif, Sherif, and Nebergall,
theory Latitude of acceptance 1965
Elaboration likelihood model Involvement Petty and Cacioppo, 1986
Cognitive dissonance theory Cognitive dissonance Festinger, 1957
Social learning theory/social Self-efficacy Bandura, 1982
cognitive theory
Theory of planned behavior Behavioral beliefs Ajzen and Madden, 1986
Outcome expectancy
Perceived behavioral control
Protection motivation theory Perceived severity Rogers, 1983
Perceived vulnerability
Intrinsic/extrinsic rewards
Response efficacy
Self-efficacy
Response costs
Health belief model Perceived severity Janz and Becker, 1984
Perceived susceptibility
Perceived benefits and barriers
Precaution adoption model Perceived awareness Weinstein, 1988
Stage model of processing fear- Perceived vulnerability Das et al., 2003
arousing communications Perceived effectiveness of
recommended actions
new science of nutrigenomics examines the response of our genes, proteins, and
metabolism to different foods. Nutrigenomics is expected to lead to evidence-based
dietary intervention strategies for maintaining, and perhaps restoring, health and fitness
and preventing diet-related disease. It is expected that, in the long term, nutrigenomic
technologies will be used to determine how our body responds to foods that affect our
long-term health [15]. Nutrigenomics is also targeting the assessment of personal
vulnerability to the development of nutrition-related illnesses through genetic testing.
Next to a personalized assessment, the availability of a personalized solution by
means of a diet, product, or nutrient that helps to prevent nutrition-related diseases is
a prerequisite for the concept of personalized nutrition to become integrated in health
behavior change strategies. The assumption is that individuals will be able to use this
information to reduce their risk of common diseases such as heart disease, diabetes,
and obesity or to improve overall health and well-being. But not much is known yet
about how individuals will actually use the information and whether it will contribute
more to behavior change than the information currently supplied [4043].
The most promising contribution to behavior change may lie in the reduction of
uncertainties on a general and personal level, thereby reducing the influence of the
interfering concept of probabilistic outcomes. The expected insights into the relation-
ship between genes, nutrition, and health may provide a stronger base for designing
clearer health messages about severity and effectiveness of the recommended actions.
On a personal level, advice based on genetic testing can provide insight into individual
vulnerability to nutrition-related illnesses and into the effectiveness of preventive
strategies, thereby strengthening messages targeted at perceived personal vulnerability
and perceived efficacy of recommended actions. Also, beliefs about the effectiveness
of a treatment recommendation based on genotypic information could be strengthened.
From research, it is known that tests offering great certainty of results (clinical validity),
with available treatment and prevention options, are more readily undertaken [44].
Another potentially positive influence on perceived personal relevance is the avoidance
of optimistic bias that leads to feelings of invulnerability. Uncertainties as to whether
or not an individual is at risk of developing nutrition-related disease can be influenced
by the results of genetic testing.
However, information on individual genetic makeup can also have undesired
effects on motivation to change behavior. It is known that, when fear appeals become
too strong, some people will react defensively; this leads to inaction. Also, higher
susceptibility to developing nutrition-related disease can induce feelings of fatalism,
thus decreasing motivation to change. Given the common perception that genetic
risks are immutable, motivation to change behavior may be decreased by weakening
beliefs that changing behavior will reduce risk. Perceived self-efficacy could also
be negatively influenced by weakening the belief in the ability to change behavior:
Its in my genes so I cant change it.
Further research has to be undertaken to gain more insight into how people will
include information on genetic makeup in the process of behavior change and
whether it will either enhance or decrease motivation. In Table 13.4, an overview is
presented of the possible contribution of innovations to perceived personal relevance
of nutrition communication in respect of creating awareness, the threat appraisal,
and the coping appraisal.
TABLE 13.4
Possible Contribution of Information on NutritionGenesHealth to
Perceived Personal Relevance of Nutrition Communication in Stages
of the Behavior Process
Awareness Threat Appraisal Coping Appraisal
Contribution of information Reduce Strengthen severity messages

on nutrition geneshealth probabilistic Strengthen effectiveness of
outcomes recommended actions
Contribution of genetic Increase accuracy of vulnerability Increase accuracy
testing assessment of recommended
actions

The main conclusion to be drawn is that not enough is known yet about the impact
of personalized nutrition interventions with respect to both reach and effect on
behavior. The possible negative influence of information about the relationship
between genes and health on perceived self-efficacy, as discussed in this chapter, is
a point of concern. Current evidence does not suggest that providing people with
DNA-derived information about risks to their health increases their motivation to
change behavior beyond that achieved with nongenetic information [40,45]. In the
authors view, the question as to whether and how the inclusion of information on
genes and health will influence perceived personal relevance of nutrition information,
and thereby affect the motivation to eat healthily, needs to be the central focus in
further research. Such insights could contribute to the development of more effective
health communication. Figure 13.2 presents an amusing depiction of the dilemma
faced.
The authors suggest that more effort needs to be put into understanding factors
that influence personal eating style and are therefore perceived as personally
relevant in nutrition communication. Eating style is an important, relatively
constant characteristic that reflects individual beliefs and behavior concerning pro-
duction, distribution, and consumption of food. It is often based on the notion that
a certain diet offers specific individual benefits or causes harm and is constructed
in the context of daily life. Research should start with exploring whether and how
genes are currently represented in personal eating style.
A last point of discussion is the fact that personalized nutrition communication
based on information from nutrigenomics will only contribute to health and well-
being if end users are sufficiently motivated and enabled to follow up personalized
recommendations on food intake. But the empowerment of individuals to improve
their food intake depends not only on individual behavior but also on the interaction
with the legal, physical, and social environment. Providing this stimulating environ-
ment in which the healthy choice is the easy choice is partly the responsibility of
many: for instance, the government, for the right regulation; health professionals
I have the gene, so what can I do? I have the gene, so I eat healthily
FIGURE 13.2 What will be the effect of information on genes and health?
and education offices, for services, information, and social support; and industry,
for products. Views on how nutrigenomics-based personalized nutrition communi-
cation will impact individuals and society need to be exchanged among all actors
concerned to ensure its legitimate and successful introduction. As with other new
technologies, personalized nutrition will entail benefits and risks and may change
social structures, culture, norms, and values. These will best be addressed by the
people that will be confronted with personalized nutrition in their daily life. Early
involvement in the development and implementation process of innovations influ-
ences personal commitment to those innovations. The WHO also recently stated that
capacity building through partnerships is an important strategy to promote health.
Partnerships are important for bringing together diversity in expertise, skills, and
resources for more effective health outcomes [46]. However, partnerships can only
be successful if participants share visions about goals, leadership, and the necessary
investment of each participant. Most often, this does not reflect reality.
A first step toward an open dialogue to create partnerships on personalized nutrition
was taken at the round table discussion at the conference of the European Nutrige-
nomics Organisation in November 2005. The views of representatives from different
scientific disciplines, industry, and government were collected about who should be
involved in a dialogue and what topics should be on the agenda. Although the discus-
sion was very lively, it was clear that its content remained scattered, leaving many
topics touched upon unexplored in depth. The reactions of the participants were limited
to their own specific interest, and the discussion did not elaborate further on specific
topics. Further action is needed to facilitate extensive and fruitful dialogue about
relevant topic such as dissemination of knowledge, practical relevance of scientific
insights, and social-ethical issues such as expected high costs of applications.
KEY READINGS
Frosch, D.L., Mello, P., and Lerman C., Behavioral consequences of testing for obesity risk,
Cancer Epidemiol. Biomarkers. Prev., 14(6), 1485.
Khoury, M.J., Genetics and genomics in practice: the continuum from genetic disease to
genetic information in health and disease, Genet. Med., 5, 261.
Lerman, C. et. al., Genetic testing: psychosocial aspects and implications, J. Consult. Clin.
Psychol., 70(3), 784, 2002.
WHO (2005) The Bangkok Charter for Health Promotion in a Globalized World. Available
at http://www.who.int/healthpromotion.
REFERENCES
1. Department of Health, Choosing Health: making healthy choices easier, Public Health
White Paper; Command paper CM 6374, U.K., 2004.
2. Ministry of Health, Welfare and Sports, Living longer in good health also a
question of a healthy lifestyle. Dutch Ministry of Health, Welfare and Sports, The
Hague, 2004.
3. World Health Organization, The World Health Report 2002: reducing risks, promoting
healthy life, World Health Organization, Geneva, 2002.
4. Contento, I.R. et al., The effectiveness of nutrition education and implications for
nutrition education policy, programs and research: a review of research, J. Nutr. Educ.,
27, 277, 1995.
5. Contento, I.R., Randell, J.S., and Basch, C.E., Review and analysis of evaluation
measures used in nutrition education intervention research, J. Nutr. Educ. Behav., 34,
2, 2002.
6. Curry, S.J. et. al., A Randomized trial of self-help materials, personalized feedback,
and telephone counselling with non-volunteer smokers, J. Consult. Clin. Psychol.,
63(6), 1005, 2005.
7. Brug, J. et al., The impact of computer-tailored feedback on fat, fruit and vegetable
intake, Health Educ. Behav., 25, 517, 1998.
8. Kreuter, M.W. and Stretcher, V.J., Do tailored behaviour change messages enhance
the effectiveness of health risk appraisals? Results from a randomized trial, Health
Educ. Res., 11, 97, 1996.
9. Sears, D.O. and Freedman J.L., Selective exposure to information: a critical review,
in The Process and Effects of Mass Communication, Schramm, W. and Roberts, D.F.,
Eds., University of Illinois Press, Urbana, 1971, p. 209.
10. Petty, E.P. and Cacioppo, J.T., Communication and Persuasion: Central and Periph-
eral Routes to Attitude Change, Springer-Verlag, New York, 1986.
11. Van der Pligt, J., Risk perception and self-protective behaviour, Eur. Psychol., 1(1),
34, 1996.
12. Rogers, R.W., Cognitive and psychological processes in fear appeals and attitude
change: a revised theory of protection motivation, in Social Psychology: A Source
Book, Cacioppo, J.T. and Petty, R.E., Eds., Guilford Press, New York, 1983, p. 153.
13. Janz, N. and Becker, M.H., The health belief model: a decade later, Health Educ. Q.,
11, 1, 1984.
14. Bandura, A., Self-efficacy mechanism in human agency, Am. Psychol., 37,122, 1982.
15. Mller, M. and Kersten, S., Nutrigenomics: goals and strategies, Nat. Rev., 4, 315, 2003.
16. Weinstein, N.D., The precaution adoption process, Health Psychol., 7, 355, 1988.
17. Batra, R. and Ray, M.L., Advertising situations: the implications of differential
involvement and accompanying affect responses, in Information Processing Research
in Advertising, Harris, R.J., Ed., Lawrence Earlbaum Associates Publishers, London,
1983, p. 127.
18. McGuire, W.J., Attitudes and attitude change, in Handbook of Social Psychology,
Lindzey, G. and Aronson, E., Eds., Random House, New York, 1985, p. 233.
19. Sherif, C.W., Sherif, M., and Nebergall, R.E., Attitude and Attitude Change: The
Social Judgement-Involvement Approach. W.B. Saunders, Philadelphia, PA, 1965.
20. Festinger, L., A Theory of Cognitive Dissonance. Stanford University Press, Stanford,
CA,1957.
21. Brug, J., Hospers, H., and Kok, G.J., Differences in psychosocial factors and fat con-
sumption between stages of change for fat reduction, Psychol. Health, 12, 719, 1997.
22. Brug, J. et al., Self-rated dietary fat intake: association with objective assessment of
fat, psychosocial factors and intention to change, J. Nutr. Educ., 26, 218, 1994.
23. Glanz, K., Brug, J., and Van Assema, P., Are awareness of dietary fat intake and
actual fat consumption associated? A Dutch-American comparison, Eur. J. Clin. Nutr.,
51, 542, 1997.
24. Lechner, L., Brug, J., and De Vries, H., Misconceptions of fruit and vegetable
consumption: differences between objective and subjective estimation of intake, J.
Nutr. Educ., 29, 313, 1997.
25. Ajzen, I. and Madden, T.J., Prediction of goal directed behaviour: attitudes, intentions
and perceived behavioural control, J. Exp. Soc. Psychol., 22, 453, 1986.
26. Gleicher, F. and Petty, R.E., Expectations of reassurance influence the nature of fear-
stimulated attitude change, J. Exp. Soc. Psychol., 28, 86, 1992.
27. Liberman, A. and Chaiken, S., Defensive processing of personally relevant health
messages, Pers. Soc. Psychol. Bull., 18, 669, 1992.
28. Koelen, M.A. and Lyklema, S., (in Dutch) De consument en perceptie van voed-
selveiligheid, in Ons Eten Gemeten. van Kreijl, C.F. and Knaap, A.G.A.C., Eds.,
Bohn, Stafleu, Van Loghum, RIVM, Bilthoven, the Netherlands, 2004, chap. 8.
29. Tversky, A. and Kahneman, D., Judgement under uncertainty: heuristics and biases,
Science, 185, 1127, 1974.
30. Koelen, M.A. and Van den Ban, A., Health Education and Health Promotion,
Wageningen Academic Publishers, Wageningen, the Netherlands, 2004.
31. Pechacek, T. and Danaher, B., How and why people quit smoking: a cognitive-
behavioural analyses, in Cognitive Behavioural Intervention: Theory, Research and
Procedures, Kendall, P. and Hollon, S., Eds., Academic Press, New York, 1979, p. 389.
32. Weinstein, N.D., Unrealistic optimism about future life events, J. Pers. Soc. Psychol.,
39, 806, 1980.
33. Otten, W. and Van der Pligt, J., Risk and behaviour: the mediating role of risk
appraisal, Acta Psychol., 80, 325, 1992.
34. Das, E.H.H.J., de Wit, J.B.F., and Stroebe, W., Fear appeals motivate acceptance of
action recommendations: evidence for a positive bias in the processing of persuasive
messages, PSPB, 29, 650, 2003.
35. De Hoog, N., Stroebe, W., and de Wit, J.B.F., The impact of fear appeals on processing
and acceptance of action recommendations, PSPB, 31(1), 24, 2005.
36. Zimbardo, P.G. and Leippe, M.R., The Psychology of Attitude Change and Social
Influence, McGraw-Hill, Boston, MA, 1991.
37. Sutton, S.R., Fear arousing communications: a critical examination of theory and
research, in Social Psychology and Behavioural Medicine, Eiser, J.R., Ed., John Wiley
& Sons, Chichester, 1982, chap. 13.
38. Green, L.W. and Kreuter, M.W., Health Program Planning: An Educational and
Ecological Approach, 4th ed., McGraw-Hill, New York, 2005.
39. Darnton-Hill, I.I., Margetts, B., and Deckelbaum, R., Public health nutrition and
genetics: implications for nutrition policy and promotion, Proc. Nutr. Soc., 63, 173,
2004.
40. Marteau, T.M. and Lerman C., Genetic risk and behavioural change, BMJ, 322, 1056,
2001.
41. Haga, S.B., Khoury, M.J., and Burke, W., Genomic profiling to promote a healthy
lifestyle: not ready for prime time yet, Commentary, Nat. Genet., 34(4), 347, 2003.
42. McCain, M. and Schmid, G., From Nutrigenomics Science to Personalized Nutrition:
the market in 2010. Institute for the Future, Silicon Valley, California, 2003.
43. Massoud, M. et al., The Future of Nutrition: Consumers Engage with Science, Institute
for the Future, Silicon Valley, California, 2001.
44. Marteau, T.M. and Croyle, R.T., The new genetics: psychological responses to genetic
testing, BMJ, 316, 693, 1998.
45. Lerman, C. et. al., Genetic testing: psychosocial aspects and implications, J. Consult.
Clin. Psychol., 70(3), 784, 2002.
46. World Health Organisation, The Bangkok Charter for Health Promotion in a Globalized
World, Available at http://www.who.int/healthpromotion, 2005, accessed June 2006.
14 A Marketing and Consumer

Behavior Perspective
on Personalized Nutrition
Hans C.M. van Trijp and Amber Ronteltap
CONTENTS
14.1 Introduction ..................................................................................................185

14.2 Personalization and Customization..............................................................187
14.3 Process Models for Personalization.............................................................190
14.4 Costs and Benefits Involved in Personalization ..........................................192
14.4.1 Consumer Costs and Benefits at the Outcome Level......................193
14.4.2 Consumer Costs and Benefits at the Interactive Process Level......193
14.4.3 Value to the Company......................................................................194
14.5 Consumer Evaluation of Personalization ....................................................195
14.6 Operational Dimensions of Personalized Nutrition ....................................196
14.7 Critical Success Factors from a Marketing Point of View .........................198
14.8 Conclusion....................................................................................................201
14.9 Future Perspectives on Personalized Nutrition............................................201
Key Readings.........................................................................................................202
References..............................................................................................................202
14.1 INTRODUCTION
Adjusting marketing offerings to the identified needs and wants of consumers is at
the heart of the marketing philosophy (e.g., Kotler and Keller, 2006). Essentially,
marketing aims to achieve an exchange of values for the mutual benefit of both the
customer and the supplier. What the customer receives in terms of need satisfaction
from products or services and gives in terms of monetary and nonmonetary sacrifices
to obtain the product is the mirror image of what the supply chain provides in terms
of products and services, and receives in terms of money to satisfy its financial
objectives of the value exchange. This process benefits both parties: the customer
benefits from superior need satisfaction through consumption, whereas the suppliers
benefits are financial means for growth and business continuity. When there are
many customers and multiple suppliers, the processes of segmenting, targeting, and
positioning become crucial. In other words, companies attempt to understand
185
heterogeneity in consumer preferences, identify promising market segments, and

position their marketing offerings (products and services) for maximum benefit. In
strategic marketing, this is known as the STP process: segmenting, targeting, and
positioning.
Historically, how this adjustment of marketing offerings to consumer needs has
been achieved has depended largely on the market situation and the relative power
of consumers and suppliers in the market economy. In the early days, many of the
products were literally tailor-made. The industrial revolution changed the
situation considerably. Economies of scale became a fact, and it became possible to
produce products economically on a large scale. The key concerns were efficiency
and the price of products, and the degree of personalization is probably best exem-
plified by Henry Ford, who in the 1920s stated that: Any customer can have a car
painted any color that he wants so long as it is black. Later, when markets became
more abundant, supply more differentiated, and consumers exhibited greater pur-
chasing power and more sophisticated articulation of their needs, market segmen-
tation and product positioning became more crucial in the marketing process.
Companies began to produce and market differentiated products targeted at identified
market segments.
In the late 20th century this process reached a new level (Sheth, Sisodia, and
Sharma, 2000; Wind and Rangaswamy, 2001; Pine and Gilmore, 1999). Companies
now realize that the crucial objective of marketing strategy is customer value, that
demand is heterogeneous and fragmented, that consumer segments are diminishing
in size, and that the crucial element in competitiveness and marketing success has
become the fit with the individual consumer. The variety of product and service is
important, but as Pine, Peppers, and Rogers (1995: 103) state: Customers, whether
consumers or businesses, do not want more choices. They want exactly what they
want when, where and how they want it and technology now makes it possible
for companies to give it to them. A key concept in this trend is personalization.
Personalization is about building consumer loyalty by building a meaningful one-
to-one relationship; by understanding the needs of each individual and helping satisfy
a goal that efficiently and knowledgeably addresses each individuals needs in a
given context (Riecken, 2000: 27). So, personalization is about reducing the infor-
mation gap between consumers and the supply chain (information asymmetry)
beyond the level of traditional segmentation. It implies a degree of interactivity
between supply chains and consumers, such that information about the consumer
becomes available in an actionable format and that the fit of marketing offerings
with those needs becomes more explicit. Personalization is central to the firms
customer relationship management (CRM) and to customer involvement, such as
through consumer intimacy strategy (e.g., Treacy and Wiersema, 1993).
Personalization has pervaded many different fields, most prominently that of digital
products and services, where it is relatively easy to obtain and analyze consumer input
interactively and adjust the marketing offerings accordingly. The field of nutrition, too,
is becoming increasingly affected by personalization, both in terms of information
provision and in terms of products tailored on the basis of consumer information (e.g.,
Bouwman et al., 2005; Brug, Oenema, and Campbell, 2003; Brug, Campbell, and van
Assema, 1999). The recent advances in nutrigenomics have given extra impetus to the
A Marketing and Consumer Behavior Perspective on Personalized Nutrition 187
field of personalized nutrition. The Institute of the Future defines personalized nutrition
as the application by individuals of their knowledge of nutrigenomics to their everyday
decisions about nutrition (Cain and Schmid, 2003: 2). However, we believe that
personalization goes beyond the field of genetics and includes other information related
to personal preferences as well, such as perceived state of health, tastes, values, and
other relevant segmentation variables.
Personalization in the field of human nutrition provides great potential to all
stakeholders, including the consumer, the firm, and those concerned with public health.
The process of personalization provides consumers with the opportunity to find an
offer that is better tailored to their specific needs. This might be in the form of
nutritional information, personal advice, and even customized food products. In this
way, personalization may also help consumers to reduce their search and evaluation
costs in product choice behavior. For the firm, personalization may help to distinguish
it from competitors, build customer relationships, and increase customer loyalty. If
well executed and substantiated in nutritional science, personalized nutrition may
contribute to the improvement of public health. However, there are benefits and costs
associated with personalization both to the individual consumer and to the company.
In this chapter we will address personalization from a consumer and marketing
perspective. First, we will delineate the definition of personalization and distinguish
it from customization. Then, we will review process models of the personalization
process and use these models to identify and discuss the costs and benefits involved
in personalization. Next, we will focus on consumer evaluation of personalization,
before we turn to the operational dimensions of personalized nutrition. Lastly, we
will discuss critical success factors from a marketing perspective, turn to conclusions,
and look at the outlook for personalized nutrition.
14.2 PERSONALIZATION AND CUSTOMIZATION

Although personalization has quite a long history, particularly in the area of Informa-
tion and Communication Technology (ICT), there is still a lack of consensus about its
definition (Vesanen and Raulas, 2006). Also, personalization has been approached
from many different perspectives (Riecken, 2000). Table 14.1 provides an overview
of various definitions of personalization as they appear in literature. Some definitions
focus on personalization as a capability of the company, whereas others (e.g., Peppers
and Rogers at www.1to1.com; Riecken, 2000) define it in terms of its implementation
in marketing execution. However, the majority of the definitions define personalization
as the process of adjusting the marketing offerings (products, services, or information)
to the identified needs of customers. Several key elements can be identified from the
variety of definitions (Adomivicius and Tuzhilin, 2005).
Personalization involves the following:
1. Tailoring certain offerings, such as information (e.g., Web pages), ser-

vices, personalized product and service recommendations (e.g., books,
CDs, and vacations), and E-commerce interactions
2. Made available by a certain provider, such as E-commerce Web sites, or
a food company
TABLE 14.1
Definitions of Personalization
Reference Definition
Peppers and Rogers One-to-one marketing based on the idea of an enterprise knowing its
(www.1to1.com) customer. Through interactions with a customer, the enterprise can learn
how a customer wants to be treated and can then treat that customer
differently from other customers.
Hagen (1999) The ability to provide content and services tailored to individuals based
on knowledge about their preferences and behavior.
Personalization consortium The use of technology and consumer information to tailor electronic
(2005) commerce interactions between a business and each individual customer.
Using information either previously obtained or provided in real time
about the customer, the exchange between the parties is altered to fit
that customers stated needs, as well as the need perceived by the
business based on the available customer information.
Dyche (2002) The capacity to customize customer communication based on knowledge
preferences and behaviors at the time of interaction with the customer.
Peppers and Rogers (1997) The process of using a customers information to deliver a targeted solution
to that consumer is known as personalization or one-to-one marketing.
Riecken (2000) Personalization is about building customer loyalty by building a
meaningful one-to-one relationship; by understanding the needs of each
individual and helping satisfy a goal that efficiently and knowledgeably
addresses each individuals needs in a given context.
Hanson (2000) A specialized form of product differentiation in which a solution is
tailored to a specific individual.
Peppers, Rogers, and Dorf Customizing some feature of a product or service so that the customer
(1999) enjoys more convenience, lower cost, and some other benefit.
Allen, et al. (2001) Company-driven individualization of customer Web experience.
Imhoff, Loftis, and Geiger Personalization is the ability of a company to recognize and treat its
(2001) customers as individuals through personal messaging, targeted banner
ads, special offers on bills, or other personal transactions.
Wind and Rangaswamy Personalization can be initiated by the customer (i.e., customizing the
(2001) look and contents of a Web site) or by the firm (i.e., individualized
offering, greeting customer by name).
Coner (2003) Personalization is performed by the company and is based on a match of
categorized content to profiled users.
Roberts (2003) The process of preparing and individualizing communication for a
specific person based on stated or implied preferences.
Riemer and Totz (2003) To match one objects nature to one subjects needs. More precisely: to
customize products, services, content, communication, etc., to the needs
of single consumers or customer groups.
Bonnett (2001) Personalization involves a process of gathering user information during
interaction with the user, which is then used to provide appropriate
assistance or services, tailor-made to the users needs.
Source: Adapted from Vesanen, J. and Raulas, M., Journal of Interactive Marketing, 20(1), 520, 2006
and Adomavicius, G. and Tuzhilin, A., Communications of the ACM, 48(10), 8390, 2005.
3. To the needs of their customers, such as consumers or Web site clientele

4. Based on knowledge about those customers, which may be explicit or
implicit knowledge, preferences, or behaviors
5. Through an iterative process in which the customer often directly or
indirectly acts as codesigner
6. With certain profit goals in mind, which can be simple (improving brows-
ing and shopping experience, such as by presenting only content relevant
to the consumer) to much more complex (e.g., building long-term rela-
tionships with the customer)
Personalization thus requires knowledge about the consumer and his or her specific
needs, and adjusting offerings to fit those needs. Some offerings, such as services
(which are by nature often developed in interaction with the consumer) and infor-
mation can be adjusted to consumer needs relatively easily. However, adjusting
physical products to individual consumer needs entails substantially increasing the
complexity of the supply chain, with potential additional costs as a consequence.
This is the field of customization.
Sometimes referred to as product personalization (Riemer and Totz, 2003),
mass customization is the field of research that focuses on using flexible manufac-
turing processes to customize physical products to the needs and preferences of
consumers with mass production (or near mass production) efficiency (Piller and
Mller, 2004). Personalization and mass customization are different but closely
related terms. Wind and Rangaswamy (2001) differentiate between the two terms,
arguing that personalization is the process located on the consumer side of the
marketing spectrum, whereas mass customization is the process on the operational
product side. Much of the mass customization literature originates from the literature
on flexible manufacturing. In customization, consumers can alter or even create
products that contain precisely those attributes that the individual consumer specifies
(Godek, 2002). In other words, the marketing offerings are being adjusted by, or
for, the user very close to the moment of purchase or consumption. An example
would be companies such as Dell and Gateway, which have adopted flexible and
responsive manufacturing systems that create products to meet the needs of individ-
ual consumers upon their request (Pine, 1993). In contrast, personalization does not
presuppose on-the-spot personalized production. Usually based on directly elicited
or indirectly inferred consumer preferences, the firm recommends those products
from an existing range that provides the best fit with those preferences (Godek,
Yates, and Yoon, 2002). An example of personalization would be Amazon.com,
which recommends alternatives on the basis of consumer-expressed preferences
(customers who bought this item also bought ). In conclusion, customization
goes one step beyond personalization, in that the customer is an active codesigner
of the product (Bonnett, 2001). Mass customization in the context of nutrition
involves a large number of complex supply chain issues that are beyond the scope
of this chapter. We will only discuss its important implications for the marketing of
personalized nutrition (see Section 14.7).
Personalization implies an interaction between the customer and the supply chain.
Simonson (2005) discusses this interactive process and emphasizes that customer
satisfaction is not only enhanced through delivering products with superior fit to the
customers individual preferences, but also that through self-assessment, the interactive
process itself can help shape the customer preferences, particularly when customers
do not yet have stable preferences. In this way, personalization may be a means to
offer customers what they want, often even before they know that they want it (i.e.,
latent needs). Others (Fotheringham and Owen, 2000) have argued that although
interaction is important, personalization is not restricted to synchronous interaction.
Personalization may be based on previously collected consumer information. Also,
personalization issues are dominant in but not restricted to online interactions
(Murthi and Sarkar, 2004) but can be realized at all user interfaces (Riecken, 2000).
14.3 PROCESS MODELS FOR PERSONALIZATION

In the personalization literature (see Table 14.2), process models have been proposed
that describe personalization as a number of consecutive and interlinked steps
(Vesanen and Raulas, 2006). These models differ, depending on the specific fields
they originate from, and tend to be biased toward the marketing side rather than
the consumer evaluation side of the personalization process. Several authors
TABLE 14.2
Several Proposed Process Models for Personalization
Adomavicius and Murthi and Sarkar Vesanen and
Tuzhilin (2005) (2003) Pierrakos et al. (2003) Raulas (2006)
Understanding the Learning about Collecting data Customer

consumer consumer Interactions
Collecting data preferences Customer data
Creating consumer
profiles
Delivering personalized Matching offerings to Preprocessing data Processing
offering customers Discovering patterns Consumer profile
Matching Postprocessing knowledge Customization
Delivery and Personalization Marketing output
presentation Delivery
Measuring impact of Evaluating the Customer
personalization learning and
Adjusting personalization matching processes Interaction
strategy
Reports Customer data

Processing
Customer profile
Source: Summarized in Vesanen, J. and Raulas, M., Journal of Interactive Marketing, 20(1), 520, 2006.
With permission.
(Steckel et al., 2005; Piller, 2005) have argued for more research on consumer
evaluation, as successful personalization depends on an enduring learning relation-
ship between the customer and the firm. Hence, the existing process models as included
in Table 14.2 should be conceived of as cycles rather than linear processes, with their
success depending on the ability to generate repeated interaction with the customer.
Process models by Murthi and Sarkar (2003) and Adomavivius and Tuzhilin
(2005) separate the personalization process into three stages: (1) understanding and
learning about customers preferences, (2) delivering personalized offerings to cus-
tomers, and (3) evaluating the learning and matching process. Pierrakos et al. (2003)
describe the personalization process as having four basic features, namely: (1) the
two-way nature of the communication system, (2) the level of response control that
each party has in the communication process, (3) the personalization in the com-
munication process, and (4) the use and involvement of database technology. How-
ever, the specific tasks they distinguish in the personalization process again relate
to the understanding (data collection), personalization (data preprocessing, pattern
discovery, knowledge postprocessing), and evaluation (reports) substages.
Vesanen and Raulas (2006) synthesize these process models by distinguishing
between four objects (customer, customer data, customer profile, and marketing
output) and four operations (interactions, processing of information, customization,
and delivery) as the key variables that together define the marketing process with
individual consumers.
The Vesanen and Raulas (2006) model defines the personalization process as a
loop, with the customer as the starting point. Through interactions with a customer,
relevant data are collected, such as expressed preferences (e.g., through question-
naires), Web site/purchasing behavior, or other more objective measurements such
as blood parameters and genetic typing. This customer information is processed
into relevant customer profiles that serve as a basis for the differentiation and
segmentation of customers. Sophisticated techniques such as data mining and neural
networks are increasingly being used for this purpose. The customer profiles are
used to generate personalized marketing output, such as tailored communication
material, personal advice, or personalized products. The delivery stage describes
how (e.g., through which channel) the personalized marketing offerings reach the
customer and will bring about a response from the customer as a new interaction,
resulting in new customer data. This closes the loop in the Vesanen and Raulas
(2006) model, as this new customer information will be processed into customer
profiles as a new input into the customization process.
At a more general level, the proposed stages can be classified in terms of three
responsibility domains: (1) consumer, (2) consumerfirm interaction, and (3) firm.
The consumer domain refers to those stages that involve consumer actions, whereas
the consumerfirm interaction domain comprises the processes by which the supply
chain and the consumer interact through various interfaces. Finally, the firm domain
includes the value-creation process undertaken by the supply chain on the basis of
the consumers personal information.
Ronteltap et al. (2006) have developed a process model specifically for the area
of personalized nutrition, which defines specific stages in each of the three respon-
sibility domains within the personalization process: the consumer, the interaction,
7. Usage 3. Data
handling
6. Delivery
8. Evaluation 4. Design
2. Communication
1. Information 5. Production
sharing
Consumer Domain Interaction Company Domain

Domain
FIGURE 14.1 The structure of a personalized recommendation system.
and the firm. The model (see Figure 14.1) describes the personalization process in
eight different stages (see also Steckel et al., 2005; Wendel et al., 2006). During the
first stage of the exchange process, consumers make available certain personal, and
possibly sensitive, information (e.g., current health condition) to the supply chain
(stage one: information sharing). In the next stage, this information passes through
a physical interface (e.g., service desk), a digital interface (e.g., electronic question-
naires), or a combination (stage two: communication). Stage three of the exchange
process (data handling) pertains to the receiver of the information (i.e., the supply
chain) that will transform the personal information into a personalized solution on
the basis of a decision model (stage four: design). Subsequently, the supply chain
will create or select personalized recommendations (information or lifestyle or
product advice) that address the needs of a particular client (stage five: production).
In stage six (delivery), this personalized advice, involving different types of infor-
mation possibly combined with products or services, is communicated and distrib-
uted via a user interface (e.g., e-mail) and received by the consumer. After acting
upon the recommendation (stage seven: usage), the consumer may evaluate the
added value of the recommendations and assess any personal benefit obtained from
the interaction (stage eight: evaluation). This evaluation will then serve as input to
the next cycle and the decision about whether or not to repeat the interaction.
By emphasizing the three domains, the model illustrates the joint responsibility
of the consumer and the firm in establishing the personalized interaction. Whether
such interactions occur depends largely on the perceived costs and benefits from the
personalization process, which we will review next. In Section 14.5 and Section 14.6
we will specifically focus on the consumer evaluation of personalized nutrition.
14.4 COSTS AND BENEFITS INVOLVED

IN PERSONALIZATION
To assess the marketing feasibility of personalization requires an analysis of the
perceived costs and benefits for the actors in the personalization process. Some authors
(e.g., Piller and Mller, 2004) argue that from a marketing perspective, the consumer
decision to engage in the personalized advice process (on nutrition or another matter)
is basically the result of a simple equation: if the (expected) returns exceed the
(expected) costs, the likelihood that customers will engage in this option will increase.
The decision of suppliers to provide personalized advice follows the same logic: only
if the expected returns exceed the expected cost will the firm offer personalized advice.
The costs to the company are the costs of collecting personalized consumer information
and differentiation (e.g., delivering personalized advice and the costs of resources
required to do so). The returns are the price premium that can be charged for individ-
ualized options, the relationship and customer loyalty that individualization may build,
and the enhancement of corporate or brand image. The key question is how to define
the costs and returns at the consumer level. Although the firms costs and benefits can
be expressed in monetary units, the consumer value is more complex and involves
psychological and ethical factors, too (Karat et al., 2003). At the consumer level, costs
and benefits are usually (e.g., Piller and Mller, 2004; Simonson, 2005; Karat et al.,
2003) differentiated into those emerging from the outcomes of the personalization
process and those emerging from the interactive process itself.
14.4.1 CONSUMER COSTS AND BENEFITS AT THE OUTCOME LEVEL

Consumer benefits of personalized offerings arise from the increased utility of
products and services that better fit personal needs compared to the best standard
option attainable. In terms of personalized nutrition, the benefits would be repre-
sented by the added contribution to personal health and the simplification of the
choice process. Provided that the information is simple and trustworthy, personalized
advice can reduce confusion and the costs of sifting through the amount of nutrition
information available.
The costs factor would constitute the costs of what the consumer would have to
compromise in terms of taste and convenience value and the price premium the
consumer would have to pay for personalized offerings compared to standard offer-
ings (Piller and Muller, 2004).
14.4.2 CONSUMER COSTS AND BENEFITS AT THE INTERACTIVE PROCESS LEVEL

Consumer costs and benefits at the interactive process level are much more diverse
and are often psychological rather than economic in nature. Consumers may derive
value from codesigning the marketing offerings (Piller and Muller, 2004), i.e., being
totally immersed in the interaction process itself (Csikszentmihalyi, 1990), successfully
fulfilling the codesign task (Dellaert and Stremersch, 2005; Franke and Piller, 2004),
and appreciating the presentation format or context (Simonson, 2005). Consumers may
also experience symbolic benefits from the process of codesign, such as pride of
authorship, sheer enjoyment, and a sense of creativity in task accomplishment (Piller,
2005). The interactive process itself can help consumers construct their preferences,
adding to a sense of self-knowledge (Simonson, 2005). Further, personalization
in nutritional advice can facilitate the empowerment of the individual consumer
(e.g., Bouwman et al., 2005) in terms of access to information, improved decision
making, and ultimately to exercise control over genetic destiny (Chadwick, 2004).
However, the interactive process also involves perceived costs on the part of
the consumer. These costs primarily constitute the cost of disclosing personal
information (Karat et al., 2003); here, privacy issues become pertinent. This is
particularly problematic in the case of genetic information, as the information is
hard-wired and stable over time, may be relevant to many different parties (e.g.,
insurance companies), and it is uncertain what applications the information may
yield in the future (Chadwick, 2004). As a result, consumers may be reluctant to
accede to their genetic information being stored, as in future it may be used against
them (Chadwick, 2004; Economist, 2005). As a psychological cost, consumers
may experience a lack of freedom and may feel that marketers are trespassing on
their personal preferences; or they may interpret personalization as attempts to
persuade and manipulate (Simonson, 2005). Also, the active involvement in the
codesign task may result in the consumer feeling psychologically at a disadvantage,
because the manufacturer possesses much more information. The consumer has
few options available, other than trusting that the manufacturer will propose the
best personalized option.
14.4.3 VALUE TO THE COMPANY

Value to the company can be expressed in terms of the costs and benefits associated
with the gathering of personal information from customers and the costs and benefits
obtained from marketing personal solutions (Karat et al., 2003). Piller and Mller
(2004) distinguish between the differentiation, cost, and relationship aspects of
personalization. Offering products and services tailored exactly to the customers
needs gives the firm a competitive advantage, takes the offerings out of the
commodity offerings, and, if recognized by customers, may generate consumer
preference for these differentiated offerings. The cost aspect of personalization
depends on the extra costs the firm incurs in the production and marketing of
personalized solutions in relation to the price premium that can be charged for the
personalized offerings. The relationship aspect of personalization and customization
arises from a customer bonding with the company and the increase in customer
loyalty arising from that (Simonson, 2005). Importantly also, this enduring learning
relationship with the customer may result in specific information that is often hard
to capture by conventional methods of market research. Building on this difficult
to obtain information may in turn deter competitors.
Riemer and Totz (2003) elaborate on the companys economic motivation of
personalization in relation to customer retention, which may be the result of the
consumers technological (e.g., incompatibility of technological systems), contrac-
tual (contract periods), or psychological obligations (e.g., brand preferences caused
by satisfaction) to the company. Each of these obligations constitutes a so-called
lock-in effect, caused by switching costs, which from a customer perspective
are defined as any cost associated with the migration to a new supplier, vendor, or
service provider. The costs of switching customer relationships can be subdivided
into direct switching costs, opportunity costs, and sunk costs. Consumers experience
direct switching costs because they find it more difficult to compare personalized
products from rival companies and because personalized products increase their
emotional obligations to the firm. Once a customer has entered into a relationship
on the basis of personalized information, he or she is confronted with direct switching

costs, and with opportunity costs associated with the probability of losing the
advantageous effects of the current relationships. Finally, sunk costs (although eco-
nomically irrelevant) associated with consumers irreversible prior investments in
establishing the current relationship are likely to reduce consumers willingness to
invest in a new relationship (Riemer and Totz, 2003). As a result, successful per-
sonalized relationships based on trust in the supplier might drive customer satis-
faction, trust and investment and increase switching costs preventing customers
from defecting (Riemer and Totz, 2003: 3839).
14.5 CONSUMER EVALUATION OF PERSONALIZATION

As identified earlier, an area less well understood (Steckel et al., 2005; Piller, 2005;
Piller and Mller, 2004) is how consumers evaluate and choose consumerfirm
interaction mechanisms as complex as those in personalized nutrition. Two fields of
research that may be informative here are the technology acceptance model (e.g.,
Davis, Bagozzi, and Warshaw, 1989) and the self-service technology acceptance
model (e.g., Dabholkar and Bagozzi, 2002), both building on Rogers seminal work
on the adoption and diffusion of technological innovations. Rogers (2003) diffusion
of innovations theory states that the speed of adoption of innovations can be partly
explained by consumer characteristics (such as innovativeness) and the perceived
characteristics of the innovation: (1) relative advantage to the user, (2) compatibility
with existing habits and values, (3) perceived complexity of the innovation, (4) ability
to try out the innovation on a small scale, and (5) the visibility of results of applying
the innovation. Rogers theory has enjoyed wide application in the marketing and
innovation literature. In a meta-analysis of applications, Tornatzky and Klein (1982)
show that three of the perceived innovation characteristics (relative advantage, com-
plexity, and compatibility) are indeed consistently related to adoption (see also Meuter
et al., 2005). The research tradition on consumer adoption of information technology
in the working place has subsequently emphasized the so-called Technology Accep-
tance Model (Davis, 1989; Venkatesh, Morris, Davis, and Davis, 2003), which posits
that ease of use and perceived usefulness are the key determinants of adoption of
information technology. The model has also been applied in consumer adoption of
self-services in the supermarket environment (e.g., Dabholkar and Bagozzi, 2002),
where it has been shown that perceived enjoyment in using the system adds to
predictive validity of the adoption models over and above ease of use and usefulness.
In a first systematic study on consumer evaluation of personalized nutrition sys-
tems, Wendel et al. (2006) extended the Technology Acceptance Model with consumer
perceptions of privacy or safety (see Figure 14.1). They showed that usefulness,
enjoyment, and privacy protection are the key perceived benefits on which consumers
base their evaluation of personalized recommendation systems, together accounting
for 80% of the variance in consumer preferences. They also show that these perceptions
of benefits almost completely mediate the effects of personalized nutrition features on
consumer preferences. It thus seems that variants of the Technology Acceptance Model
are useful for understanding consumer evaluations of personalized nutrition, with
usefulness, enjoyment, and privacy protection as key benefits.
14.6 OPERATIONAL DIMENSIONS OF PERSONALIZED

NUTRITION
To our knowledge, no studies to date have elicited consumer responses to different
operationalizations of the personalized nutrition process. In this section we will
summarize the results from two working papers (Wendel et al., 2006; Ronteltap and
Van Trijp, 2007) on this issue. Both studies are empirical tests of the model shown
in Figure 14.1. In both studies, each stage of the model was operationalized at three
levels (see Table 14.3). Based on these levels, systematically varied scenarios were
TABLE 14.3
Factors and Factor Levels Applied
in the Scenario-Rating Tasks
1. Information sharing
a. Blood composition
b. DNA/genetic makeup
c. Food consumption habits
2. Communication
a. Through fitness club
b. Through hospital
c. Through family practitioner
3. Data handling
a. Fully anonymous
b. Patient and family practitioner
c. Available to commercial food company
4. Design
a. Commercial company
b. Insurance company
c. Government nutrition center
5. Production
a. At ingredient level
b. Food product groups
c. Special branded products
6. Delivery
a. Through e-mail
b. Through fitness club
c. Through family practitioner
7. Usage
a. Within existing meal patterns
b. Add special products to diet
c. Prepare own adjusted meals
8. Evaluation
a. No feedback for verification
b. Feedback for verification optional
c. Feedback for verification mandatory
Box 14.1 Consumers Evaluation of Alternative Operationalizations

You go to your family practitioner [fitness club] to deliver information on your

food habits [DNA information].
The personal information you provide will be to known only to you and your
family practitioner [shared with commercial companies].
On the basis of your personal information, the Nutrition Center [an insurance
company] will compile personalized nutritional advice, which you will receive
through your family practitioner [fitness club].
The personalized nutritional advice is in the form of product groups [branded
food products] that you should eat more often or avoid. It is put together in
such a way that you can easily implement it in your regular diet [you have to
prepare separate meals, with special products].
After 3 months, you will probably already feel better, and this system gives you
the opportunity to verify the improvements in your health status [does not
provide feedback].
Note: The most (italics) and least [in brackets] preferred scenarios are shown.
developed as different combinations of levels of each of the eight stages (see

Figure 14.1). These scenarios (see Box 14.1 for an example) were then evaluated
by consumers in a conjoint task (Ronteltap and Van Trijp, 2007) and a scenario-
rating task (Wendel et al., 2006). Consumer evaluation measures included overall
preference and intended use (Ronteltap and Van Trijp, 2006), and the perceived
benefits ease of use, usefulness, privacy protection, and enjoyment (Wendel et al.,
2006). As both of these studies have been reported elsewhere in detail, only the main
results will be discussed here.
Together, the two studies confirm that:
The personalization process can indeed be conceived of as consisting of three

domains: a consumer domain, an interactive domain, and a firm domain.
Efforts in each of these domains contribute to overall consumer evaluation.
Each of the eight stages of the personalization process (see Figure 14.1)
contributes to the overall evaluation of personalized nutrition offerings,
lending support to the personalization process model.
What consumers are most suspicious of are commercial applications in
terms of the type of offerings (i.e., branded food products), the ownership
of the database technology (to translate the consumer profile into a per-
sonalized diet), the sharing of personal information with commercial
companies, and the route to interaction (e.g., through fitness clubs).
Consumers prefer to receive personalized nutritional advice through estab-
lished health institutions, particularly the family practitioner.
Consumers prefer to be able to choose the specificity of nutritional advice

(preferring advice at the level of ingredients and product groups, rather
than about specific brands and products) and optional (rather than man-
datory or no) feedback.
Consumers prefer solutions that they can easily incorporate into their daily
lives.
The effects of design features on consumer attitude and intentions appear
fully mediated by the perceptual dimensions of ease of use, usefulness,
enjoyment, and privacy (Wendel et al., 2006).
As an illustration of the conjoint task results, in Box 14.1 we show the two scenarios
from the Ronteltap and Van Trijp (2006) study that consumers thought were the
most (italics) and least (between brackets) desirable scenarios for personalized
nutritional advice.
Overall, the results as presented in Box 14.1 show that there are still a number
of hurdles to be overcome before personalized nutrition can be applied commercially
in marketing. In the next section, we will discuss the critical success factors from
a marketing point of view.
14.7 CRITICAL SUCCESS FACTORS FROM A MARKETING

POINT OF VIEW
From a marketing point of view, personalized gene-based nutrition implies that
food companies would focus on smaller and smaller consumer segments that are
homogeneous in terms of genetic profiles and associated nutritional needs (e.g.,
Jiang, 2000). To be justifiable as a strategic marketing tool, any segmentation
needs to conform to a set of specific criteria: (1) measurability, (2) substantiality,
(3) accessibility, (4) responsiveness, and (5) actionability (Kotler and Keller, 2006).
In this section we assess the feasibility of personalized gene-based nutrition in
terms of these criteria. Segments need to be measurable to the extent that individual
consumers can accurately and consistently be assigned to segments at justifiable
cost. In the case of genetics-based segmentation, this would mean that it should
be relatively easy to accurately identify which consumers belong to which genetic
segment. Also, segments should be substantive, in the sense that the segments
identified should be large enough to justify separate marketing programs. Segments
should be accessible, in the sense that targeted marketing programs could be
developed that effectively reach the segment of consumers. Segments should be
responsive, in the sense that different segments respond differently to the differen-
tiated marketing mixes, as otherwise it would not be profitable to develop these
separate marketing programs. Finally, the segments should be actionable, in the
sense that it is both possible and advantageous to target different segments with
different marketing programs. Much of the marketing potential of personalized
nutrition will depend on further and as yet unknown scientific progress in the areas
of nutrigenetics and nutrigenomics. However, the five criteria for effective seg-
mentation provide a useful tool for reflecting on the marketing potential of per-
sonalized nutrition.
In terms of measurability, genetics-based personalized nutrition has the advantage

that genotype is a segmentation variable that is stable over time. However, the
measures are not easily obtained, because of consumer resistance to divulging per-
sonal genetic information to third parties, particularly when these are not in the public
health domain. Uncertainties exist about how this information will be used now and
in the future. For food companies, it is unclear how they can obtain this information
with sufficient accuracy and in large enough quantity. In terms of substantiveness of
the segments, the key question is whether homogenous consumer groups of sufficient
size can be identified to justify targeted marketing. In other words, if these are very
small groups of consumers, it will not be feasible for the mainstream food industry
to develop targeted marketing programs, although it may be a relevant niche market
with marketing opportunity for entrepreneurial small- and medium-sized companies.
However, such small segments are probably catered to more efficiently and effectively
through the medical segment than through the commercial segment of food compa-
nies. In terms of accessibility, given that genetic constitution is highly idiosyncratic
and not overtly visible, it will be necessary for food companies to reach the segment
members through an individualized approach. Consumer reluctance to share this
sensitive genetic information directly with commercial food companies complicates
such one-to-one marketing interactions with the individual consumer.
The greatest opportunity for marketing personalized nutrition assuming that
once consumers experience the health benefits of personalized nutrition they are
likely to actively seek the personalized foods that fit their specific preferences is
responsiveness to personalized offerings. However, most of the health benefits are
long term and not easily discernible by the consumer. Hence, it is crucial for
personalized nutrition to have demonstrable short-term effectiveness: for example,
inducing a change in consumer-relevant biomarkers (such as lowering cholesterol
level), or an improved feeling of well-being. Actionability touches on the distinction
between personalization and mass customization. There is great potential for giving
personalized advice to consumers about which existing products to choose or to
avoid. However, the situation becomes much more complex when the challenge is
to develop and market personalized food products. This would require customizing
product offerings to match smaller and smaller segments of consumers, which in
turn requires substantial adjustments to the existing food supply chains that are
currently based on large-scale production. One possibility would be to develop
specialized foods for particular genotype segments; whether this is feasible depends
largely on the willingness of this segment to pay a premium price that will cover
the extra costs involved in production, distribution, and marketing. Another option
would be to add specific nutrients and ingredients to products in the very last step
in the production chain, rather like a coffee-vending machine, which allows the
consumer to add certain ingredients to the basic product in-store to make it fit with
his or her personal nutritional needs. However, this would bring the application of
personalized nutrition very close to that of supplements and medicine, and a key
challenge would be to meaningfully differentiate the personalized food option from
supplements and medicine. Further complicating the marketing of personalized
nutrition are the regulations regarding nutrition and health claims (note that in the
European Union, such regulations have recently come into force). It is expected that
genetics-based products will focus either on maintaining health or on reducing risk

of disease (or maybe even curing disease). Although claims to this effect are now
permitted under the new European Union regulations, they are strictly regulated on
the basis of scientific substantiation. The high costs of this scientific substantiation
will be borne by a relatively small market segment, because the alternative is to
make genomics knowledge publicly available, which would substantially reduce the
competitive advantage of the institutions that create knowledge.
Overall, we believe that current developments in nutrigenomics augur well for
improving our understanding of the relation between food, health, and disease, and
individual differences therein. There is potential for personalized nutrition advice
for those segments of consumers willing to make their genetic information available
in order to obtain such advice. Some commercial companies such as Sciona
(www.mycellf.com) are already practicing this particular application of nutrigenom-
ics. Provided that these applications are scientifically sound and consumers are
willing to share their genetic information, this may be one route for personalized
nutrition to explore. But note that the Sciona business model relies heavily on the
marketing of diagnostic kits, not on customized food products. Also, we believe that
further advances in the science of nutrigenomics may be helpful in the marketing
of food products whose selling point is the absence of certain ingredients and
possibly nutrients (guaranteed free from) as is currently already the case in aller-
gen-free foods. If nutrigenetics provides evidence for sufficiently large consumer
segments, this may be an opportunity for the mainstream food industry; otherwise,
gene-based personalized nutrition may be a niche market relevant for smaller food
companies. However, at present there are too many factors that hinder the commer-
cial application of nutrigenetics-based practices for marketing food products on the
basis of their content of specific combinations of ingredients and nutrients. It is
unlikely that in the near future the benefits arising from customer loyalty and
willingness to pay will outweigh the costs associated with personalized nutrition.
The consumer segments are likely to be small and the additional production, distri-
bution, and marketing costs high. For most larger food companies, this will deviate
too much from their current business models to justify the effort. There may be
potential for small specialist companies, provided that they can get the distribution
channel right. However, such applications are more likely to be closer to the medical
and supplement businesses than to the food business. In the short term, the added
value of nutrigenomics is more likely to emerge from the better insight into the
fundamental processes underlying health and disease than in the application of
marketing for personalized diets. From consumers great confidence in and reliance
on their family practitioner, we further see that it is this communication channel
that has most potential. We therefore believe that genetics-based insights can be best
implemented through the existing health systems, with a central role being played
by the family practitioner. Our consumer research suggests that commercial appli-
cations through fitness and health centers are still a step too far for most consumers.
Much of the existing segmentation and targeting in the food industry is being
conducted on the basis of phenotype information rather than genotype information.
Market potential should therefore be assessed by examining the added value of gen-
otype segmentation over and above existing phenotype segmentation. For example, it
is possible to identify a relevant segment of consumers at risk based on their current

blood cholesterol level, on specific lifestyle features (age, lifestyle, job stress, etc.), or
based on their genotype information. Blood cholesterol levels are relatively easy to
measure and specific psychographic and lifestyle factors are an accessible feature to
which communication to the target audience can be tailored. This raises the question
of what added value genotype information related to blood cholesterol levels will have
for marketing purposes. Clearly, genotype information has the potential for identifying
individuals at risk much earlier in their lives, before the phenotype with the high
cholesterol level has developed, and this may be important for preventive purposes.
However, the crucial factor is the reliability of genotype information for predicting
later development of high cholesterol levels, and the extent to which the effect of the
genotype can be attenuated by environmental factors such as nutrition. Currently, the
predictive accuracy seems to be low, especially for diseases such as diabetes and
obesity that involve multiple genes. Finally, regarding actionability in terms of the
development of specific dietary advice and tailored products, the question arises as to
whether the additional genotype information would lead to relevant adjustments in the
current dietary recommendations about the consumption of macro- and micronutrients.
14.8 CONCLUSION
Personalized nutrition based on insights from nutrigenomics is a relatively new
concept. Building on marketing literature about personalization and effective market
segmentation, in this chapter we have reviewed the marketing potential of person-
alized nutrition. Although personalization has revolutionized the marketing approach
in many fields (see Wind, 2001; Wind and Rangaswamy, 2001 and Riemer and Totz,
2003 for more details), we have specifically focused on the potential of personalization
in the food domain. We have argued that personalized nutrition based on nutrige-
netics may provide potential for personalized nutritional advice with the purpose of
assisting consumers to make better choices from the existing product supply. In
terms of personalized food products, we have argued that currently there is limited
marketing potential for the mainstream food industry, because the consumer seg-
ments are likely to be small and the additional costs involved in production, distri-
bution, and marketing are large. However, niche markets of sufficient size and
purchasing power might develop that can profitably be catered to by smaller and
specialized food companies. Another potentially successful application of nutrige-
nomics knowledge in the short term may be foods personalized on the basis of the
absence of certain ingredients and nutrients.
14.9 FUTURE PERSPECTIVES ON PERSONALIZED

NUTRITION
Despite these constraints to personalized nutrition becoming a large-scale marketing
opportunity, we believe that it has potential to add to public health in the more
medically oriented domain. Again, this depends on the advances in the science of
nutrigenomics, as complex states of health and disease cannot currently be predicted
from genetics. Health-care professionals such as the family practitioner, a highly
trusted source of health information, could be important gatekeepers in this process

of advising consumers about personalized nutrition.
This chapter has highlighted that only a limited amount of scientific marketing
and consumer research has been done on personalization in the food domain. There
is clearly a need for more research in this area, not just from a marketing and
consumer behavior perspective, but also more broadly at the interface between
biochemical and nutrition science on the one hand and social sciences such as
communication, ethics and marketing, and consumer behavior on the other. Only
such a multidisciplinary approach will ensure that consumer triggers and reservations
can be identified up front and serve as input to the further development of nutrige-
nomics practice to make it a substantive success both commercially and in terms of
public health.
KEY READINGS
Adomavicius, G. and Tuzhilin, A., Personalisation technologies: a process-oriented perspec-
tive, Communications of the ACM, 48(10), 8390, 2005.
Murthi, B.P.S. and Sarkar, S., The role of the management sciences in research on personal-
ization. Management Science, 49(10), 13441362, 2003.
Piller, F.T., Mass customization: reflections on the state of the concept. The International
Journal of Flexible Manufacturing Systems 16: 313334, 2005.
Vesanen, J. and Raulas, M., Building bridges for personalization: a process model for mar-
keting. Journal of Interactive Marketing, 20(1), 520, 2006.
Wind, J. and Rangaswamy, A., Customerization: the next revolution in mass customization.
Journal of Interactive Marketing, 15(1), 1332, 2001.
REFERENCES
Adomavicius, G. and Tuzhilin, A., Personalisation technologies: a process-oriented perspec-
tive, Communications of the ACM, 48(10), 8390, 2005.
Bonnett, M., Personalization of Web services: opportunities and challenges. Ariadne 28 (June),
www.ariadne.ac.uk, 2001.
Bouwman, L.I., Hiddink, G.J., Koelen, M.A., Korthals, M., Vant Veer, P., and Van Woerkum,
D., Personalized nutrition communication through ICT application: how to overcome
the gap between potential effectiveness and reality. European Journal of Clinical
Nutrition 59(Suppl. 1), S108S116, 2005.
Brug, J., Campbell, M., and Van Assema, P., The application and impact of computer-generated
personalized nutrition education: a review of the literature. Patient Education and
Counseling 36, 145156, 1999.
Brug, J., Oenema, A., and Campbell, M., Past, present and future of computer-tailored nutrition
education. American Journal of Clinical Nutrition 77(Suppl.), 1028S1034S, 2003.
Cain, M. and Schmid, G. From nutrigenomic science to personalize nutrition. The market in
2010. Palo Alto (CA): Institute for the Future. New Consumer New Genetics Program,
report SR793 (www.ifth.org), 2003.
Chadwick, R., Nutrigenomics, individualism and public health. Proceedings of the Nutritional
Society 6, 161166, 2004.
Csikszentmilhalyi, M., Flow: The Psychology of Optimal Experience. New York: Harper and
Row, 1990.
Dabholkar, P.A. and Bagozzi, R.P., An attitudinal model of technology-based self-service:

moderating effects of consumer traits and situational factors. Journal of the Academy
of Marketing Science, 30(3), 184201, 2002.
Davis, F.D., Perceived usefulness, perceived ease of use and user acceptance of information
technology. MIS Quarterly 13(3), 145158, 1989.
Davis, F.D., Bagozzi, R.P., and Warshaw, P.R., User acceptance of computer technology:
a comparison of two theoretical models. Management Science 35(8), 9821002,
1989.
Dellaert, B.G.C. and Stremersch, S., Marketing mass-customized products: striking a bal-
ance between utility and complexity. Journal of Marketing Research 42(May),
219227, 2005.
Fotheringham, M.J. and Owen, N., Introduction Interactive health communication in
preventive medicine. American Journal of Preventive Medicine 19(2), 111112, 2000.
Franke, N. and Piller, F.T., Value creation by toolkits for user innovation and design: the case
of the watch market. Journal of Product Innovation Management, 21: 405415, 2004.
Godek, J., Personalization and customization: implications for consumer decision making and
behavior. Advances in Consumer Research 29, 155157, 2002.
Godek, J, Yates, J.G., and Yoon, Y., Customization and personalization: the influence of
perceived control and perceived capability on product evaluations. Advances in Con-
sumer Research 29, 155157, 2002.
Jiang, P., Segment-based mass customization: an exploration of a new conceptual marketing
framework. Internet Research: Electronic Networking Applications and Policy, 10(3),
215226, 2000.
Karat, C.M., Brodie, C., Karat, J., Vergo, J., and Alpert, S.R., Personalizing the user experience
on ibm.com. IBM Systems Journal, 42(4), 686701, 2003.
Kotler, P. and Keller, K.L., Marketing Management: analysis, planning, implementation and
control, 12th ed. Englewood Cliffs, NJ: Prentice Hall, 2006.
Meuter, M.L., Bitner, M.J., Ostrom, A.L., and Brown, S.W., Choosing among alternative
service delivery modes: an investigation of customer trial of self-service technologies.
Journal of Marketing 69(April), 6183, 2005.
Murthi, B.P.S. and Sarkar, S., The role of the management sciences in research on personal-
ization. Management Science, 49(10), 13441362, 2003.
Pierrakos, D., Paliouras, G., Papatheodorou, C., and Spyropoulos, C.D., Web usage mining
as a tool for personalization: a survey. User modeling and User-adapted interaction,
13: 311372, 2003.
Piller, F.T., Mass customization: reflections on the state of the concept. The International
Journal of Flexible Manufacturing Systems 16: 313334, 2005.
Piller, F.T. and Mller, M., A new marketing approach to mass customisation. International
Journal of Computer Integrated Manufacturing, 17(7), 583593, 2004.
Pine, B.J., Customizing for the new consumer. Fortune 128(December 16), 45, 1993.
Pine, B.J. and Gilmore, J.H., The Experience Economy. Boston, MA: Harvard Business School
Press, 1999.
Pine, B.J., Peppers, D., and Rogers, M., Do you want to keep your customers forever? Harvard
Business Review 73(MarchApril), 103114, 1995.
Riecken, D., Personalized views of personalization. Communications of the ACM 43(8),
2728, 2000.
Riemer, K. and Totz, C., The many faces of personalization: an integrative economic overview
of mass customization and personalization. In Tseng, M.M. and Piller, F.T. (Eds.),
The customer centric enterprise: advances in mass customization and personaliza-
tion. Berlin: Springer, 2003, pp. 3550.
Rogers, E.M., Diffusion of innovations, 5th ed. New York: Free Press, 2003.
Ronteltap, A. and Van Trijp, H.C.M. (2007). Consumer preferences for personalised nutritional
advice: a conjoint analysis study. Wageningen University; Marketing and Consumer
Behaviour Group. Working Paper (in preparation).
Ronteltap, A., Wendel, B.S., Van Trijp, H.C.M., and Dellaert, B.G.C., Marketing van producten
die passen bij de genetische structuur van het individu [Marketing of products that
fit the genetic predisposition of the individual]. Jaarboek van de Nederlandse Verenig-
ing van Marketing Onderzoekers, 2006, pp. 127141.
Simonson, I., Determinants of customers responses to customized offers: conceptual frame-
work and research propositions. Journal of Marketing 69(January), 3245, 2005.
Sheth, J., Sisodia, R., and Sharma, A., The antecedents and consequences of customer-centric
marketing, Journal of the Academy of Marketing Sciences, 28(1), 5566, 2000.
Steckel, J.H., Winer, R.S., Bucklin, R., Delleart, B.G.C., Dreze, X., Haubl, G., Jap, S.D.,
Little, J.D.C., Meyvis, T., Montgomery, A.L., and Rangaswamy, A., Choice in inter-
active environments. Marketing Letters 16(3/4), 309320, 2005.
Tornatzky, L.G. and Klein, K.J., Innovation characteristics and innovation-adoption imple-
mentation: a meta-analysis of findings. IEEE Transactions on Engineering Manage-
ment 29(1), 2845, 1982.
Treacy, M. and Wiersema, F. (1993), Customer intimacy and other value disciplines. Harvard
Business Review 71(1), 8493.
Venkatesh, V., Morris, M.G., Davis, G.B., and Davis, F.D., User acceptance of information
technology: toward a unified view. MIS Quarterly 27(3), 425478, 2003.
Vesanen, J. and Raulas, M., Building bridges for personalization: a process model for market-
ing. Journal of Interactive Marketing, 20(1), 520, 2006.
Wendel, S., Ronteltap, A., Dellaert, B.G.C., and Van Trijp, H.C.M., CustomerFirm Interaction
in the creation and delivery of personalized recommendations. Working paper, 2006.
Wind, J., The challenge of customerization in financial services. Communications of the ACM,
44(6), 3944, 2001.
Wind, J. and Rangaswamy, A., Customerization: the next revolution in mass customization.
Journal of Interactive Marketing, 15(1), 1332, 2001.
15 U.S. Consumer Attitudes

toward Personalized
Nutrition
David B. Schmidt, Christy White,
Wendy Reinhardt Kapsak, Josh Conway,
and Elizabeth Baily
CONTENTS
15.1 Methods........................................................................................................205
15.2 Results ..........................................................................................................206
15.2.1 Genomics: General Attitudes and Perceptions ................................206
15.2.1.1 Awareness and Interest Regarding Genomics..................206
15.3 The Opportunity for Personalized Nutrition ...............................................209
15.3.1 Key Findings on Consumers Perceptions
Regarding Health and Nutrition ......................................................209
15.3.1.1 Interest in and Favorability toward
Personalized Nutrition ......................................................209
15.3.1.2 Applications for Personalized Nutrition ............................210
15.4 Discussion ....................................................................................................212
15.4.1 Awareness, Interest, and Barriers Regarding Genomics .................212
15.4.2 Awareness, Interest, and Applications
for Personalized Nutrition................................................................215
15.5 Conclusion and Principles for Effective Communication
Related to Personalized Nutrition................................................................216
Key Readings.........................................................................................................218
References..............................................................................................................219
15.1 METHODS
Since 2004, the Cogent Syndicated Genomics Attitudes and Trends (CGAT) survey
has been annually taking the publics pulse regarding the issue of genomics. CGAT
explores public awareness, understanding, favorability, and interest in genomics and
its applications, including nutrigenomics and pharmacogenomics. This chapter
205
describes data from the CGAT survey, collected in January 2006, via a Web-based
205-question survey of more than 1000 Americans. Using both balancing and
weighting techniques, the sample was designed to reflect the U.S. population in
terms of key demographics, including education, age, income, and gender.
Data in this chapter are also from the International Food Information Council
(IFIC) survey, Consumer Attitudes toward Functional Foods/Foods for Health. This
survey measures consumer awareness of, and interest in, functional foods/foods for
health as well as the application of genomics to nutrition and personal health, a science
otherwise known as nutrigenomics or personalized nutrition. This trending survey
was first commissioned in 1998, with subsequent surveys conducted in 2000 and 2002;
some survey questions gauge trends in consumer attitudes toward nutrition and health.
Questions regarding consumer perceptions of personalized nutrition were added in
2005. The quantitative Web-based survey was fielded in May 2005 with more than
1000 randomly selected U.S. adults, aged 18 years and older. The findings were
weighted by education, age, and ethnicity, reflecting the 2003 U.S. population census
estimate, to allow the findings to be representative of the American public as a whole.
The findings from both surveys rely primarily on univariate analyses and cross-
tabulations. The data were analyzed according to the demographic profile of the
population, including primary characteristics (e.g., age, income and gender), health-
based characteristics (e.g., health history, and current health status), and in some
cases, attitudinal characteristics (e.g., receptivity to medical advances).
15.2 RESULTS
15.2.1 GENOMICS: GENERAL ATTITUDES AND PERCEPTIONS
15.2.1.1 Awareness and Interest Regarding Genomics
Although three fourths (76%) of the American public say they have heard or read
something about genomics (defined as using individual genetic information to under-
stand and optimize health), a closer analysis indicates that this broad awareness is
quite limited in depth. First, only a very small number (6%) of Americans say they
have heard a lot about genomics, whereas half (49%) say they have heard only
a little bit. Further, when those who say they have heard something about genomics
were asked to describe what they have heard, about one fourth (27%) were unable
to recall any details, and another one third (36%) mention having heard only about
the basic connection between genes and health (Chart 15.1).
Still, there is a small group of Americans who understand the concept of genomics
and some of its practical implications. Specifically, of the approximately 75% of
Americans who have heard something about genomics, only about 15% say they
have heard that genes can be altered for health benefits and about 5% have heard
that genomics can aid in early detection of disease or that genomics can be used to
help customize health-related treatments.
Although knowledge is limited, more than half (52%) say they favor the idea
of using individual genetic information to understand and optimize health. Similarly,
nearly half (46%) of Americans say they are interested in using their personal genetic
information to understand and optimize their health. This includes using their personal
U.S. Consumer Attitudes toward Personalized Nutrition 207
A lot
6%
A fair Nothing
amount 24%
21%
A little bit
49%
CHART 15.1 U.S. consumers genomics awareness. n = 1015.
genetic information to provide nutrition and diet-related recommendations (i.e.,

nutrigenomics or personalized nutrition). Findings specific to personalized nutrition
will be explored later in this chapter.
Consumer interest in using genomics is highest when it can be connected to
identifiable health benefits. For example, three fourths (72%) of the public are
interested in using genomics to reduce risk of specific diseases. Furthermore, less
than half (46%) of Americans say they would be even more interested if the test
results could indicate they were certain to get the disease. In contrast, about one
fourth (27%) of Americans would be more interested in using genomics if the test
results were able to tell them that they carry the relevant gene but not whether they
will develop a disease. The public is particularly interested in using genomics to
target diseases that are treatable.
Privacy. Among the most interesting findings is the number of people who are
concerned about privacy issues. Two thirds (66%) of the American public is
concerned about how personal genetic information would be stored and who would
have access to that information. One third (33%) of the public say their concern
about privacy would actually prevent them from having a genetic test. A potential
reason for this may lie in the fact that 36% of the public sees genetic tests as being
more sensitive, and thus more personal, than other kinds of medical tests.
Regarding information misuse specifically, about half of Americans say they are
concerned that their DNA sample may be used for tests other than ones [they] have
authorized. More generally, the public is worried about misuse by insurance
companies, government, and employers (51%).
Insurance discrimination. Almost two thirds (64%) of the public agree with state-
ments regarding insurance discrimination, including: insurance companies will do
everything possible to use genetic information to deny health coverage and insurance
companies will use information to deny coverage for drugs people need if their genetic
profile indicates a low chance of responding. This may be why more than half (56%)
of the American public say they would be less interested in using genomic services if
their test results would become part of their medical record and therefore be obtainable
by insurance companies. Conversely, the public does want insurance companies to

pay for genetic tests. Almost half (48%) indicated they would be less interested if their
health insurance company would not cover the cost of genetic testing.
Government discrimination. A similar dynamic exists with regard to the govern-
ment and its role in genomics. First, about two thirds (65%) say they are concerned
that the government might obtain unauthorized access to their personal genetic infor-
mation, and only 2% say they would want the government to receive a copy of the
results of any genetic test they might choose to conduct. However, Americans are
looking to the government to protect their privacy, as three fourths (72%) of the public
believe the federal government should establish specific laws and regulations to
protect the privacy of genetic information. That said, 70% of Americans do not know
whether laws to protect the privacy of their genetic information currently exist, and
of those who are aware of existing laws, 60% believe the public needs more protection.
Clear and strict protections could alleviate concerns, given that about 60% of Amer-
icans say they would be more interested in genomics if they were assured by law
that no one could access [their] DNA information without [their] consent.
Employer discrimination. Lastly, Americans are concerned about employers
accessing their genetic information. More than half (54%) of Americans say they
are concerned that their employer would gain unauthorized access to their personal
genetic information. Furthermore, only 1% say they would want their employer to
receive a copy of the results of any genetic test they might choose to receive.
Moral issues. Another challenge facing genomics is public concern about moral
issues. According to the research, there is a segment of the population that holds
deep moral reservations about genomics. In this case, this group is a minority. One
fourth (25%) of the public agree that the recent advances surrounding the use of
genetic testing make me uncomfortable from a moral standpoint. Slightly less (21%)
agree that meddling with our genes and DNA is trying to play God [and that]
scientists, researchers, and doctors should stay out of it altogether.
The CGAT survey shows two trends that have emerged that may be important
to consider. First, although the number of Americans who have moral reservations
is low, this cohort has grown slightly over the past year (see Chart 15.2).
Second, about one third (33%) of the American public has not yet developed a
firm moral stand on these issues; that is, they neither agree nor disagree with these
2006 2005 2006 2005

Top 2 Box Top 2 Box Bot. 2 Box Bot. 2 Box
(Agree) (Agree) (Disagree) (Disagree)
We shouldnt meddle with our genes 21% 16% 46% 50%
Genetic testing should be stopped 19% 15% 47% 55%
Genetic testing makes me morally

25% 23% 42% 46%
uncomfortable
CHART 15.2 Moral reservations about genetic testing. n = 1015 (2006), 1018 (2005).
statements noted in Chart 15.2. Perhaps speaking to both moral and privacy concerns,
we see that almost half (46%) of all Americans agree with the statement, I am torn
on the use of genetics. The potential benefits are incredible, but the potential for
misuse is considerable.
Emotional uncertainty. The third challenge facing genomics is the emotional
burden that genetic test results can bring. For example, about one fourth (26%) of
the public agree that knowing their genetic profile is too great a responsibility
because it impacts [them], [their] spouse, and ultimately [their] children. Similarly,
one third (32%) agree that it would be too depressing to know [they were] going
to get a disease, particularly if there is nothing [they] can do about it. This finding
reinforces the consumer appeal of genomics when it addresses treatable diseases
and actionable preemptive choices.
15.3 THE OPPORTUNITY FOR PERSONALIZED

NUTRITION
15.3.1 KEY FINDINGS ON CONSUMERS PERCEPTIONS REGARDING
HEALTH AND NUTRITION
Survey findings indicate that most Americans (94%) believe that they have some
control over their own health. In addition, the majority of consumers (69%) believe
food and nutrition play a great role in maintaining or improving overall health.
Interestingly, more consumers (90%) said in 2005 that family health history plays
a moderate to great role in maintaining and improving health than in previous
years (82% in 2002, 80% in 2000, and 85% in 1998). This finding likely indicates
consumers increasing awareness of how their unique genetic makeup, inherited
from their parents, may affect their future health status.
The majority (88%) of American consumers agree that certain foods have health
benefits that go beyond basic nutrition and may reduce the risk of disease or other health
concerns. Consumer interest in learning more about foods for better health remains high.
Specifically, more than eight out of 10 (83%) Americans are interested in learning more
about foods that have health benefits that go beyond basic nutrition and may reduce the
risk of disease or promote better health. Those who are very interested in learning
more about healthful foods are those who feel that food and nutrition play a great role
in overall health (45% vs. 40% of those who say it plays a moderate role, and 22%
of those who stated no role); are more likely to be vitamin/supplement users (49% vs.
27% of nonusers); consumers with a college degree or higher (49 percent vs. 38 percent
of those with some college or less); and consumers aged 35 to 54 (48% vs. 41% of those
aged 18 to 34 and 35% of those who are 55 years and older).
15.3.1.1 Interest in and Favorability toward

Personalized Nutrition
More Americans have heard or read about using individual genetic information in
the context of nutrition and diet-related recommendations (42%) than in any other
context, including prescription drugs (36%), over-the-counter medications (29%),
and beauty and cosmetic products (19%). In fact, nearly half (47%) of all Americans
say they have heard or read something about personalized nutrition (defined as using
genetic information to make informed dietary choices).
When asked what they thought about various terms to describe the practice of
using genetic information to develop nutrition and diet-related recommendations,
Americans overwhelmingly preferred either personalized nutrition (70% like a
lot or a little) or individualized nutrition (68% like a lot or a little) over
terms such as nutrigenomics (19% like a lot or a little).
As with genomics in general, slightly more than half (51%) of the American
public hold a highly favorable view of personalized nutrition. Among the 48% of
Americans who hold an unfavorable opinion of personalized nutrition, most indicate
that their favorability is low due to a lack of knowledge about the science. Whereas
20% cite lack of knowledge directly, most do so indirectly, saying they are not sure
how they could benefit from personalized nutrition (17%) or dont see a need for
the science (14%). Others say they fail to see how genomics would be superior to
their doctors opinion or to good-old common sense. A small number (14%) say
they are not favorable at this point in time because the unknown consequences of
personalized nutrition may be harmful, whereas others have more immediate con-
cerns such as the perceived high cost of this technology (4%).
Interestingly, given Americans fears of discrimination when it comes to genom-
ics in general, only about 10% say their lack of favorability toward personalized
nutrition is attributable to concerns regarding misuse of genetic testing information.
When the topic turns away from general favorability toward using this technology
for a personal benefit, just less than half (42%) of the American public express a
strong level of interest in using genetic information to receive personalized, diet-related
recommendations. Only one fourth (27%) of the American public characterizes their
interest in personalized nutrition as low.
15.3.1.2 Applications for Personalized Nutrition
Desired benefits of nutrigenomics. Although Americans find numerous potential

benefits of personalized nutrition to be appealing, distinct preferences for specific
applications are apparent. Specifically, more holistic, preventive health benefits rank
highest, with 63% of consumers interested in disease prevention and 61% interested
in overall wellness. Singular, less general health benefits, such as mental alertness
(58%), stamina (53%), weight loss (52%), and vitamin/mineral optimization (48%)
fall into a second tier, but are still appealing to a large number of people about
half of all Americans. Using personalized nutrition for appearance benefits ranks
lowest of all the areas tested (40% interested). An overwhelming percentage (86%)
of Americans favor receiving personal nutrition recommendations that provide guid-
ance on both additions to and subtractions from their diet.
Products of interest. As shown in Chart 15.3, when it comes to the specific types
of products Americans would be most interested in having tailored to their specific
dietary needs, vitamins and supplements top the list. Interest in fortified foods and
natural foods appeals to 50% of the American public, whereas only about 25 to 38%
are interested in specific food products, including beverages and meal replacement bars.
Vitamins and supplements

Nutrient fortified foods
Whole/natural foods
Beverages
Foods using biotechnology
to add nutrients
Sports or meal replacement bars
0 10 20 30 40 50 60
CHART 15.3 Interest in genetically tailored products. n = 1015.
The power of the physician as guide and gatekeeper. Nearly 90% of Americans
say they would consult their doctor when deciding to pursue a genomics test to
facilitate personalized nutrition and diet-related recommendations, and almost a third
(31%) of all Americans would only pursue testing after receiving a doctors specific
recommendation.
The role of the doctor goes beyond deciding whether or not to have a test. Once
Americans have decided to take a genetic test for the purpose of receiving person-
alized nutrition recommendations, 67% cite the doctors office as the most desirable
location for administering such a test. The majority of Americans (70%) are also
most likely to indicate that they would want to receive their test results in a face-
to-face consultation similar to how many patients receive their medical test
information today.
The power of the individual and other sources in decision making. Although
the doctors opinion is undeniably important, more than half (57%) of Americans
say they will make their own decision after consulting their doctor, and an addi-
tional 10% say they will not consult their doctor at all. This means that a majority
of Americans have some measure of confidence in their own ability to decide whether
or not to pursue genetic testing.
Although secondary to physicians (70%), significant numbers of consumers rated
nurses and physician assistants (58%), and dietitians (57%) as highly credible
sources for information on using genetic information for personalized nutrition
information. Genetic counselors (48%) were also considered among the highly
credible sources. The inclusion of health associations as a credible source (52%)
demonstrates that consumers may not necessarily need a face-to-face, intimate
history for a high level of trust. Pharmacists (43%) were considered only slightly
less credible than genetic counselors. Moderately credible resources also included
university or other health newsletters (42%) and government agencies (37%). Other
sources, ranging from news media to product labels and insurers, received relatively
low ratings when it came to being a credible source on personalized nutrition (see
Chart 15.4 for the remaining figures).
Testing locations and results delivery. This research has shown the doctors office
as Americans most desirable testing location. However, a sizeable number of Americans
MDs
Nurses/PAs
Dietitians
Health associations
Genetic counselors
Pharmacists
University or other
health newsletters
Government agencies
Food product labels
Dietary supplement labels
Consumer groups, mass media,
Web, insurance cos.
0 10 20 30 40 50 60 70 80
CHART 15.4 Personalized nutrition resource credibility. n = 1015.
say they would be interested in testing at home (40%), perhaps because of the privacy
it provides. In fact, at-home testing slightly edged out the offices of two other medical
professionals genetic counselors (39%) and dietitians/nutritionists (31%). Testing
locations that are largely unacceptable to consumers include health spas, fitness
centers, personal trainers office, and health food stores (each with 8%).
Furthermore, although face-to-face delivery of test results tops the list (70%),
not far behind are paper results (60%), which suggest that, given the right approach,
home testing could be widely accepted. Americans were less interested in their
willingness to receive their results through electronic media 35% were willing
to receive results via a secured Web site, 32% by e-mail, and 28% by disk. A phone
call with a consultant was the least preferred means of receiving test results (26%).
15.4 DISCUSSION
15.4.1 AWARENESS, INTEREST, AND BARRIERS REGARDING
GENOMICS
Research results show that consumers ideal scenario for genomics-related applica-
tions is one in which the genetic marker (or test) provides a high level of certainty
that one would actually experience a specific disease for which there are known
remedies. Given these high expectations, additional consumer education will likely
need to address the many factors that can impact disease risk, including genetics as
well as the environment (see Figure 15.1).
Barriers to adoption. Despite broad awareness, a positive perception, and general
interest in genomics and its applications, there are some formidable challenges facing
genomics. If left unchecked, the challenges have the potential of limiting widespread
usage of genomics, or even becoming the flashpoint for a negative shift in public
Reduce Risk of
Specific Diseases
Ensure
Condition Can Be
Treated
Identify Diseases
Certain to
Experience
FIGURE 15.1 Consumers are most interested in certain knowledge of treatable diseases.
opinion. The research showed there are three barriers of specific concern for genomic
testing. These concerns include: privacy, moral issues, and perceived emotional
drawbacks.
Almost two thirds of the American public were concerned about issues related
to privacy, which included insurance, government, and employer discrimination.
Given this finding, it is likely that a regulatory and statutory framework that takes
these concerns into account will be debated before widespread adoption of genomics
testing occurs. In the U.S., governing bodies are already beginning to look at genetic
testing and some of the challenges to come. In the scientific and academic commu-
nity, the Institute of Medicine held its first-ever meeting on nutrigenomics, sponsored
by the National Institutes of Healths Office of Dietary Supplements, the National
Cancer Institute, and the U.S. Department of Agricultures Agriculture Research
Service, which looked at the role of various stakeholders and their impact on this
emerging area of science.1 Additionally, in the U.S., the Department of Health and
Human Services has formed the Secretarys Advisory Committee on Genetics,
Health, and Society (SACGHS) to provide a forum for experts to discuss ethical,
social, legal, and medical issues related to innovations and applications in genetic
technology. The SACGHS also helps inform other government agencies about spe-
cific questions related to genetics and makes recommendations to the secretary on
such issues.2 For more than a decade, members of Congress have also been working
to pass the Genetic Information Nondiscrimination Act (GINA) to address the
growing concern surrounding the misuse of genetic information in insurance and
employment decisions.
Around the globe similar discussions have also begun. In Canada, Genome
Canada commissioned a project to compare different ethical approaches to public
involvement in governing genomics.3 The U.K. also has a commission in place that
provides the government with advice about genomics. This commission has a par-
ticular focus on ethical, legal, and social issues related to human genetics.4
Genomics is entering a crucial phase in which scientific advances are blossom-
ing, while simultaneously a sizable minority of the public is struggling to understand
how these advances relate to their deeply held convictions. Given that about one
third of the American public have not yet developed a firm moral stand on issues
related to genetic testing, these fence sitters may be highly influenced by messaging
regarding genomics either pro or con. If a substantial number of these individuals
are persuaded and join the concerned minority, the total group holding moral objections
would form a near majority. In contrast, messaging that promotes the responsible use
of genetic information as well as the implications for public (and personal) health may
swing the balance more heavily toward positive viewpoints (see Figure 15.2).
Emotional concerns held by Americans may, in part, signal recognition of the
significance or even value of genomics. However, there is also evidence that
the emotional burden for some may be too great, enough to preclude interest for
a sizable minority. Providers of genomic testing and products may be able to pursue
strategies to lessen this burden, especially through professionals who provide testing
coupled with consultation.
More research has been conducted to assess consumer acceptance and reactions
to genetic testing than in the area of personalized nutrition. One area, which this
research did not assess, was the behavioral consequences that could be associated
with receiving a genetic test. One study, involving healthy individuals, asked par-
ticipants to imagine that they were at increased risk for obesity based on the results
of a genetic test. This preliminary study found that a positive test (i.e., showing that
they were at increased risk of becoming obese) was a motivator for more healthful
st
tere
& In
ility
orab
Fav
Moral
Issues
Emotional
Consequences
Privacy Concerns
FIGURE 15.2 Favorability and interest in genomics outweigh concerns.

dietary behavior.5 This finding calls attention to the importance of providing moti-
vational or behavioral information along with a genetic test to help consumers
implement dietary or lifestyle changes.
15.4.2 AWARENESS, INTEREST, AND APPLICATIONS

FOR PERSONALIZED NUTRITION
Of the various genomic applications, the public is most aware of personalized

nutrition. In addition, the public considers personalized nutrition to be aimed at
general health, or as a more tailored diagnostic tool. Opportunities present them-
selves as the concept of using genetic information to provide personalized diet
recommendations was favorably received by the majority of Americans, who are
interested in learning more.
Terminology related to personalized nutrition. Research findings show that the
majority of Americans preferred the terms personalized nutrition and individu-
alized nutrition to other terms such as nutrigenomics. Consumer insights from
previous IFIC focus groups indicated that whereas the former terms are seen as
focusing on one person (speaking to me or about me), consumers found the
latter term too technical and difficult to understand. Even more worrying, many
found this term to be scary. Consumer communications that take into account such
terminology preferences may be more effective in building acceptance of this emerging
field.
Applications of personalized nutrition. In the case of pharmacogenomics, the
CGAT survey showed that the public is most interested in identifying specific
medical issues, as well as providing directional means to treat or prevent those issues.
In contrast, personalized nutrition is held to a different set of success standards,
according to which individuals are less concerned with specific treatments and more
interested in general health and wellness outcomes. This difference in expectations
may illustrate the publics belief in the functional limits of food and nutrition with
regard to health. It also may be that the public is unaware of how specific genetic
conditions or predispositions can be influenced by altering lifestyle choices, includ-
ing the consumption of nutrients or food components over time (as opposed to taking
prescription medications). In either event, it is again made clear that the public would
benefit greatly from having a better general understanding of the interplay between
genes and dietary choices in determining ones present and future health. Given the
breadth of nutrition and lifestyle modifications that these tests may encompass,
dietetic professionals are particularly poised to play a primary role in helping to
meet this need.6
Relaying genetic information to consumers. At present, though awareness is high
for personalized nutrition, the lack of familiarity with potential benefits, specific
products, and the testing process would likely drive consumers to consult their trusted
medical professionals for guidance regarding this emerging approach to personal
health. Genetic testing, and by extension personalized nutrition, appear to be heavily
rooted in the medical realm, and Americans are approaching them as they would
any other medical need or concern.
It is not surprising, then, that for many, the doctors office would be a prelim-
inary step to pursuing any kind of genetic-based dietary approach. Thus, although
efforts to communicate the general benefits and practical knowledge of personal-
ized nutrition to the consumer are critical, so too is providing the latest develop-
ments and research in the field of genomics and its nutrition applications to the
primary care physicians whose advice and knowledge would be sought by most
individuals. This might, in part, help to alleviate the emotional burden that a lay
reading of the genetic test results might produce. Such an unfamiliar medical test
could potentially suggest a multitude of recommendations and changes. Creating
an interdisciplinary partnership among health professionals could help mitigate
the responsibility of one health professional being responsible for administering
a genetic test, analyzing the results, explaining the results to a client, making
recommendations on how to implement behavioral changes based on the results
of the test, and following up with the client to determine how his or her life has
been affected by the test results.
Considering the high levels of credibility and influence assigned to health-care
professionals beside physicians, face-to-face result consultations may well provide
an acceptable level of trust, particularly if performed by a dietitian/nutritionist or a
pharmacist. In studies completed with consumers and genetic testing, it appears
there are a multitude of factors that may influence an individuals decision to undergo
genetic testing. Research has shown a decrease in genetic testing for women at risk
of developing breast cancer after they spoke with a genetic counselor. However, in
situations where a genetic test was offered without genetic counseling, people were
more likely to undergo the genetic testing.7 Partnerships and information sharing
will be necessary to ensure that all those who work with genetic tests are providing
the most accurate information to consumers and are equipped to answer any ques-
tions or concerns related to the genetic test being offered.
As public knowledge of and familiarity with the field grows, so too might con-
sumers willingness to test and use the information in the privacy of ones own home.
One option could be to deliver paper results with an official seal of a trusted health
association. However, business case studies have already shown that even with the
best intentions of maximizing the benefits and minimizing the risks of a new science
and technology, not all innovations will prove socially acceptable. Ideally, science,
ethics, and the law working in concert could lay the foundation needed to ensure
optimal social outcomes and public acceptance of this new technology.8
15.5 CONCLUSION AND PRINCIPLES FOR EFFECTIVE

COMMUNICATION RELATED TO PERSONALIZED
NUTRITION
Communicating the promise of personalized nutrition can be exciting and almost
intoxicating for professionals working in the field. However, it is important to temper
this enthusiasm with the reality of the pace of scientific discovery. Other chapters
highlight some of the foremost international experts on nutrigenomics, and their
research has uncovered fascinating implications for diet and gene expression. Yet,
the scientific community would be the first to emphasize that there is more to discover
about which genes are most important to our health and which may be impacted by
nutrition. It is important to continue to communicate what we know about the
promise of personalized nutrition, while at the same time communicating the limi-
tations of that knowledge. Reports of scientific advances in this field should be
accompanied by enough context to communicate the pieces of the puzzle that are
still missing. Although the science continues to emerge, effective communication
can help us move further down that road and avoid the obstacles that misinformation
can create.
There are both challenges and opportunities in communicating the potential
health benefits of foods and food components and how they ultimately may be
associated with the practice of personalized nutrition. Overall, awareness of
genomics is broad and interest is high, especially in regard to definitive tests for
treatable medical conditions. At the same time, few adults have sufficient knowl-
edge regarding genomics, leaving many unaware of the potential benefits to per-
sonal health or worse, vulnerable to misinformation. Moreover, the American
public has a number of latent concerns that could quiet enthusiasm or even turn
the tide against the entire field of genomics. According to the CGAT survey
conducted by Cogent Research as well as the IFIC survey, Consumer Attitudes
toward Functional Foods/Foods for Health, much of the American public is inter-
ested in using their personal genetic information to optimize their overall health
through personalized nutrition. Communications about personalized nutrition
should focus on balancing consumers expectations with the reality of the science
by not overpromising on potential outcomes. As applications of personalized
nutrition become more widespread, it will be important to communicate both
consumer benefits as well as address concerns. To build support, stakeholders,
such as government, industry, health professionals, and others, need to work
together to communicate effectively with each other and the public to increase
knowledge of the benefits associated with personalized nutrition.
Consumers are primed for personalized messages about foods that provide benefits
beyond basic nutrition and how to incorporate these foods into their diet. Because of
this research finding and the fact that most Americans feel hopeful and curious about
emerging science, a door is open for communicators to deliver more information that
increases the understanding of personalized nutrition and helps consumers enjoy
health-promoting foods as part of an overall healthful lifestyle. Keeping the consumer
in mind, and adherence to the following communication principles, will be vital to
consumer acceptance and success of personalized nutrition.

The line of communication is not always a straight path. Americans acquire health
and nutrition information from numerous sources. It is important for everyone in
the communication chain to provide consistent and scientifically accurate informa-
tion to consumers as they begin to apply this new technology.
To aid in this process, the IFIC Foundation partnered with the Institute of Food
Technologists (IFT) to develop the Guidelines for Communicating the Emerging Sci-
ence of Dietary Components for Health. These guidelines provide communicators with
a checklist to help enhance the publics understanding of emerging science and the
role it plays in overall health.9 Communicators, ranging from health professionals,
educators, scientists, scientific journal editors, government officials, and journalists,
should consider these points when translating how the latest research about food and
nutrition, including personalized nutrition, could change what is on the publics plate:
Serve up plain talk about food and health.

State that scientific research is evolutionary, not revolutionary.
Carefully craft communications.
Make messages meaningful.
Cite study specifics.
Point out the peer-review process as a key measure of a studys objectivity.
Consider the full facts when assessing a studys objectivity.
One of the best ways to improve the publics understanding of personalized nutrition
is to put partnerships in place that provide clear and consistent communication
messages. Multiple communicators conveying the same messages increase the like-
lihood that those messages will resonate with consumers. It is particularly important
that these partnerships be created while the science is still developing and less
information has been imparted to consumers. These partnerships can help to ensure
that the first information consumers receive, related to the applications and benefits
of personalized nutrition, is the most accurate information.
KEY READINGS
Borra, S., Kelly, L., Tuttle, M., and Neville, K. Developing actionable dietary guidance
messages: dietary fat as a case study. J Am Diet Assoc. 2001; 101: 678684.
International Food Information Council. 2005 Functional Foods/Foods for Health Executive
Summary. July 2006. http://www.ific.org/research/funcfoodsres05.cfm. Accessed on
January 22, 2007.
International Food Information Council Foundation. Tools for Effective Communication:
Beginning a New Conversation with Consumers. http://www.ific.org/tools/intro.cfm.
Accessed on January 22, 2007.
Kaput, J. et al. The case for strategic international alliances to harness nutritional genomics
for public and personal health. Br J Nutr. 2005; 94: 623632.
Sanderson, S.C., Wardle, J., Jarvis, M., and Humphries, S. Public interest in genetic testing
for susceptibility to heart disease and cancer: a population-based survey in the U.K.
Prev Med. 2004; 39: 458464.
U.S. Department of Health and Human Services, National Institutes of Health, National
Cancer Institute, Making Health Communication Programs Work, Bethesda, MD,
2001.
REFERENCES
1. Institute of Medicine. Nutrigenomics and Beyond: Informing the Future. Nutrige-
nomics Workshop Presentations July 12, 2006. www.iom.edu/CMS/3788/31286/
33150/35223.aspx. Accessed on December 18, 2006.
2. Department of Health and Human Services. Secretarys Advisory Committee on
Genetics, Health, and Society. http://www4.od.nih.gov/oba/sacghs/sacghsdocuments.
html. Accessed on December 28, 2006.
3. Burgess, M. Starting on the right foot: public consultation to inform issue definition
in genome policy. Electronic Working Paper Series. W. Maurice Young Centre for
Applied Ethics, University of British Columbia. 2003.
4. Human Genetics Commission, Human Genetics Commission Fifth Report from April
2005 to March 2006. www.hgc.gov.uk/UploadDocs/DocPub/Document/Final%20PDF.
pdf. Accessed on December 19, 2006.
5. Dominick, L., Frosch, P.M., and Caryn, L. Behavioral consequences of testing for
obesity risk. Cancer Epidemiol Biomarkers Prev. 2005; 14(6): 14851489.
6. Debusk, R., Fogarty, C.P., Ordovas, J.M., and Kornman, K.S. Nutritional genomics
in practice: where do we begin? J Am Diet Assoc. 2005; 105: 589598.
7. Bowen, D.J., Battuello, K.M., and Raats, M. Marketing genetic tests. Health Educ
Behav. 2005; 32(5): 676685.
8. Kaput, J. and Rodriquez, R. Nutrients and norms: ethical issues in nutritional genomics
in Nutrition Genomics Discovering the Path to Personalized Nutrition. Wiley, 2006,
pp. 419434.
9. International Food Information Council Foundation, Institute of Food Technologists.
Guidelines for communicating the emerging science of dietary components for health.
2005. www.ific.org/nutrition/functional/guidelines/guidelinesfulldoc.cfm. Accessed
on December 19, 2006.
16 Ethics of Personalized
Nutrition
Michiel Korthals
CONTENTS
16.1 Introduction ..................................................................................................221

16.2 Opening the Black Box of Personalized Nutrition .....................................223
16.3 Two Ethical Perspectives .............................................................................223
16.4 Three Main Areas of Ethical Concern.........................................................225
16.4.1 The Relationship between Food and Health (Drugs)......................225
16.4.2 The Second Main Area of Ethical Concern:
The Relationship between Personalized and Public
Health Food Policies ........................................................................228
16.4.3 Third Main Area of Ethical Concern: Personalized
Nutrition for the Consumer .............................................................230
Key Readings.........................................................................................................233
References..............................................................................................................233
16.1 INTRODUCTION
Personalized nutrition focuses on the value of health in choosing food, and prejudges
an individualizing policy instead of a public health nutrition policy; it is faced with
many uncertainties and is difficult to reconcile with research outcomes. Personalized
nutrition raises many issues of ethical concern. The three main issues are mostly
approached from a utilitarianist or rights-based perspective of individual autonomy
(informed consent or informed choice) that aims at protecting individuals against
harmful interventions. However, it is also fruitful to develop a more collective and
proactive ethical perspective, because of the collective consequences of individual
choices. The first ethical issue is that of the relationship between food and health
(drugs), which is at present subject to rather strict regulation. Owing to the rise of
nutrigenomics, new gray zones emerge, for instance, when food becomes a preven-
tive drug or when drugs are used by nonpatients, as in the case of enhancement
medicine. There are no compelling arguments against new developments, but we
are in need of debate about coping with the ethical impacts. The second issue is that
of the relations between personalized and public health nutrition, which requires
221
new responsibilities of both consumers and producers in incorporating health con-

cerns into pluralist food styles. The third one covers ethical issues of making use
of personalized nutrition and requires consultations between consumers and profes-
sionals. Moreover, independent gatekeepers are necessary to safeguard the indepen-
dence of testing, marketing, communicating, and providing the services.
In the 1980s, nutrigenomics started with the idea of improving general diets of
inhabitants of Western countries. By enhancing the quality of food such as common
crops and meat products, the goal was to raise the quality of public health. It was
the time of vitamin-enriched food and sugar beets without sweets, or, as the article
in Time of June 1999 said, guilt-free, nondigestible sugar: Sweet Taste of Success.
Genetic engineers in Holland claim they have finally come up with the holy grail
of the food industry: a plant that provides sugar and can be used to make nonfattening
and guilt-free cakes. Unlike normal sugar beet, whose content is easily digested, a
new form has been created that can make fructans, sugars that come in long molecule
chains that are more difficult for the body to absorb [1]. Even 8 years later, this
guilt-free sugar beet still does not exist, so the promise of the engineers was made
a bit too early.
Nevertheless, very soon after the promising start on the level of public health
(compare Reference 2), nutrigenomics made a more individualizing turn toward
personalized nutrition, with the idea that genetic profiles differ very much, and that
the genetic profile of individual consumers should determine what they eat. Person-
alized nutrition has, in a certain sense, already a long history as part of public health
nutrition policies, particularly during industrialization because of the health condi-
tion of the urban labor force. General practitioners made food charts that signaled
food deficiencies [3]. Food diseases were proved to be connected with bad food and
undernourishment, and good food could cure these diseases.
However, nowadays personalized nutrition nutrigenomics is driven by individual
prevention, not the cure, of food-related risks such as vitamin deficiencies, alcoholism,
obesity, allergies, and food-related diseases such as diabetes mellitus type 2, hyper-
tension, gallstones, osteoporosis, problems of lipid metabolism, and cancer. Over-
weight is, in particular, a case in point that is said to be preventable by individualized
nutrition. The World Health Organization (WHO) wrote a report in 2004 in which
it was stated that in 2020, 50% of the world population is expected to become
overweight. For instance, in the U.S., in general, 15 to 20% of the adults in some
states are obese, and in some even more than 25% of the inhabitants (BMI > 30)
are obese. The personal factors that determine overweight and obesity are the
imbalance between energy intake and expenditure because of high intake of saturated
fat and sugar; but inactivity (computer and television), social economic factors such
as urbanization, educational level of parents, and the lifestyle of parents do play a
role as well [4,5,6].
In this chapter I will first analyze the main structure of personalized nutrition,
which covers fours aspects: laying bare general relationships between genetic pro-
files, organisms, and environment such as food intake and lifestyle; testing and
screening of individuals to understand their vulnerabilities; giving personal nutrition
advice to individuals and, lastly, producing innovative food products. I will next
describe two ethical perspectives that in literature on food and health are mostly
Ethics of Personalized Nutrition 223
mentioned, i.e., the utilitarianist perspective of individual autonomy (informed con-

sent or informed choice) and the rights-based (Kantian) and communicative per-
spective that develops a more collective ethical outlook emphasizing the collective
consequences of individual choices. I will subsequently show that personalized
nutrition focuses on the value of health in food choice, and seems to prejudge an
individualizing policy instead of a public health nutrition policy; this focus results
in three main areas of ethical concern. The first is the relationship between food and
health (or food and drugs), the second is the relations between personalized and
public health nutrition, and the third issue covers items that emerge in making use
of personalized nutrition. I will finish with future prospects of personalized nutrition
and its ethics.
16.2 OPENING THE BLACK BOX OF PERSONALIZED

NUTRITION
The first assumption of personalized nutrition is that food is predominantly not seen
as positive, having pleasurable features or even as neutral, but as having impacts on
human health in the long run in preventing or even accelerating risks for diseases
such as cardiovascular diseases, cancer, and allergies. The second main idea is that
people should choose their diet on the basis of their genetic profile and that diet-
induced diseases should be detected as early as possible and should be prevented
by a change in diet.
Personalized nutrition covers four aspects (see also Reference 7). The first aspect
is the complex relationship between the genetic makeup of individuals and their
food intake. Second, the screening and testing of an individual to find out in how
far he or she runs any risk. To this aspect belong questions such as, how to interpret
the chances and how to share this information with your relatives, who are possibly
affected as well. Another question is the privacy of the screening and test results.
Third, there is the personalized advice to eat in a certain way, or more broadly, to
follow a healthy life style that this same person gets from some professional. The
fourth aspect covers the products (foods or drugs) that are the innovative fruits of
nutrigenomics research. These products again can have multiple effects, and it is
often unclear in how far these are taken into account. Long-term health effects of
products are mostly not tested, and therefore not known.
16.3 TWO ETHICAL PERSPECTIVES

Everyone wants to be healthy, but many of us differ in what role health should play
in life, what the meaning of health is, and also refuse to act in healthy ways. What
bearing, if any, should our food, health, and life choices have on the ethics and
policy of food production, food regulation, clinical practice, and health policy? What
do personal freedom, responsibility, and solidarity mean here? Should risk-takers
have the same claim on scarce resources, such as organs for transplant, as those
whose plight is owing to no choices of their own?
With respect to the issues at hand, we can discern at least two ethical perspectives
that are rather common now in medical ethics and can shed some light on nutrigenomics
oriented toward personalized nutrition. The first one emphasizes the protection of
consumer/citizens by viewing all ethical issues from the point of view of informed
consent or informed choice, i.e., from the point of view of individual freedom and
autonomy, which seeks to limit social powers that threaten to transgress the bound-
aries of privacy and of private freedom. Although issues of justice, i.e., the socially
fair distribution of resources plays a role here, the idea is primarily that social powers
should not be allowed to dominate individual consumers. Patients and consumers
should be enabled to protect themselves against powers that invade their personal
responsibility, by giving them rights [8]. In the alternative model, the emphasis is
on coordination and interaction between consumers, producers, and other stakeholders,
which means that individual consent and choice can play a role, but in the social
context of interaction and social learning, and of distributing responsibilities according
to issues and needs. Consultations and deliberations with stakeholders (as a matter
of fact, including consumers) to pinpoint agreement and disagreement play an
important role here. The aim is not to protect patients and consumers against
professional powers; patients and consumers are invited and encouraged to take part
in consultations and are given the opportunity to tell their narratives, interests, and
anxieties.
In both models, the concept of autonomy plays a pivotal role; however, in the
first as individual autonomy and, in the second, as social or collective autonomy.
The first concept of individual autonomy (and protection) is more connected to
utilitarianism, which emphasizes individual utility and happiness as the most important
criterion for the outcome of ethical decisions. The concept of personal autonomy lies
at the heart of the utilitarian doctrine of, for example, John Stuart Mill (18061873)[9]:
The only freedom which deserves the name is that of pursuing our own good in our
own way, so long as we do not attempt to deprive others of theirs, or impede their
efforts to obtain it. Each is the proper guardian of his own health, whether bodily, or
mental or spiritual [10].
As long as one can do what one thinks is in favor of ones own health and as long
as one does not harm others, then actions are ethically seen allowed. Harm toward
others is the category that marks the distinction between personal autonomy and
social life, and harm is the indicator that something is ethically wrong. This means
that individuals first and foremost should be protected from the harmful actions of
others or should be enabled to protect themselves from those harmful actions.
Of the two alternative models, the second one, of social autonomy, is directly
linked to the German philosopher Immanuel Kant (17241804) [11], the foremost
thinker of autonomy. Kant thinks more positively than Mills about the relation
between individual and society, in terms of enrichment and humanization and not
in terms of harm. Kant argues:
Laziness and cowardice are the reasons why such a large part of humanity, even long
after nature has liberated it from foreign control (naturaliter maiorennes), is still happy
to remain infantile during its entire life, making it so easy for others to act as its keeper.
It is so easy to be infantile. If I have a book that is wisdom for me, a therapist or
preacher who serves as my conscience, a doctor who prescribes my diet, then I do not
need to worry about these myself. I do not need to think, as long as I am willing to
pay [11] (my italics).
As food choices are a component of ones autonomy, consumers should determine

their own food (diet); as a consequence, the food markets and food professionals
should follow suit, and they should act as our keeper. Only paying for your food,
without acting according to your autonomy, is a sign of infantility, or of having no
control over your mouth. Autonomy not only encompasses what goes out of the
mouth but also what goes into the mouth.
Kant gives some more interesting clues on what autonomous food choice
means elsewhere [12]. In general, he considers humans to be subject to both Nature
and Reason. Man has neither only capacities from nature nor capacities from reason
but from both, where nature stands for passions and sensual experience and reason
for transcending nature by using the reasoning faculty in knowing, willing, and
appreciating (judging). For instance, enjoying art and food socially means transcend-
ing nature by assessing something that is given and subsequently structured accord-
ing to the judgmental structure that you share with other rational beings. This type
of (communicative) assessing is what autonomy means, so autonomy has nothing
to do with informed consent or choosing or doing what you want. Human beings
are involved in a development toward standing on their own feet, although hesitating
and stumbling. Taste has, according to Kant, the extraordinary feature that it stim-
ulates solidarity in enjoying. Enjoying food is related to culture and society. But
there is more. A good meal enjoyed over a long period with congenial people is an
occasion where experience and reason are united and which can be repeated again
and again. Through having common meals, reason acts upon the experiences,
because they stimulate solidarity and humanity. Meals are a stimulus to come
together and share your thoughts with other rational beings. Kant takes into account
that we do not have the same taste for food: some like this and others like that.
However, he introduces the concept of comparative universality to make clear that
on closer inspection, by experiencing and talking, many of our tastes overlap.
Comparative universality is a term that refers to the narrative overlapping of our
food preferences. It recognizes both that taste is personal, which you have to per-
sonally recognize in food, and that tastes have some commonality, proved by the
fact that we enjoy the meal together [12, p. 567].
16.4 THREE MAIN AREAS OF ETHICAL CONCERN

On the basis of these ethical perspectives one can discern at least three main areas
of ethical issues connected with nutrigenomics-based personalized nutrition. With
these two perspectives in mind, we can tackle the three main areas of ethical concern
connected with nutrigenomics.
16.4.1 THE RELATIONSHIP BETWEEN FOOD AND HEALTH (DRUGS)

The first area of ethical concern covers the relationship between food and health or
between food and drugs, and addresses all the aspects of personalized nutrition
mentioned earlier. Although there has always been in the past some overlap between
food and health (see the case of food in curing diseases), in general, the distinction
between the two up to now has been rather strict. The most important distinctions
between food and health (drugs) are located in the dimensions of personal lives,
culture, politics, organizations, and market forces. First, food is for everyone, drugs
only for the ill; food should not have too severe negative effects, because of its
lifelong intake, and drugs sometimes have (acceptable) negative effects because of
their temporary intake; food choices involve multiple reasons, whereas drugs are
chosen for only one reason (to be cured); foods are freely available for all, and drugs
only on prescription; and finally, food is not tested for efficacy, but only for safety,
whereas drugs are tested for both efficacy and safety. When food is indeed becoming
a drug in a preventive way, it is possible for persons to get side effects caused by
their food choice; they are expected to choose their food only for health reasons;
they should consult their physicians before eating and have to learn new types of
behavior that redistribute responsibilities, duties, and rights in a fundamental way.
People will have to deal with the blurring of these important cultural, political,
personal, social, and scientific distinctions and their ethical implications [14,15]
(Table 16.1).
People have a complex set of considerations governing their food choice, which
are different from the considerations that determine their use of drugs. At least four
factors determine our food preferences, in general: genetic factors, social and cultural
factors, bodily history, and personal historical factors [17]. The genetic factors,
depending on genetic makeup, are generally not changeable; they determine taste
in general (like bitter and sweet) and, partially, taste development. The social and
cultural factors consist of determining organizational structures and values, such as
family relationships and health or nationalistic values. The bodily history influences
tongue and mouth, but also nose and ear, which are part and parcel of food choices.
When one is tasting food, one is also tasting ones own body. Finally, there are
TABLE 16.1
Some Differences between Food and Drugs
Food Drugs
1. Side effects Not accepted Accepted

2. For whom? All Groups of patients
3. Choice Multiple reasons (lifestyle) Health reasons
4. Time taken Over lifetime During illnesses
5. Delivery Freely available Over-the-counter/prescription
6. Activities Shopping, talking, Pill taking
preparing, sharing
7. Testing Safety, not efficacy Safety and efficacy
8. Production Food companies, farmers (open Drug companies: closed lab
production systems)
9. Responsibility Governments and food sector Physicians, drug companies,
governments
10. Trust [16] Low trust in sector Sector trusted
personal historical factors, such as the family history and how it is internalized.
Moreover, when actually having a meal, additional factors do play a role as well,
such as the present conditions of the meal, the present condition of the body, and
the actual pleasure and displeasure in tasting. Gustatory taste and food choice are
strongly intentional and cognitive: it matters what one eats and to know what one
eats. They comprise the visual, the smell (olfactory), tongue senses (gustatory), and
the auditory (e.g., crunchy or not). The fact that taste is cognitive implies that taste
can be learned, and can develop in a certain direction. A yellowish substance with
a rotting smell turns suddenly into a delicious cheese upon tasting it, preferable with
nice and knowledgeable people.
More generally, in the domains of food and health culture, politics, science, and
society, the most important distinctions between food and drugs are that food is for
everyone, drugs only for the diseased; that food should not have negative effects,
because of its lifelong intake, and drugs sometimes have (acceptable) negative effects
because of their temporary intake; that food choices are motivated by multiple
reasons (see earlier text) and drugs by only one (to be cured or protected); that food
is freely available and drugs only on prescription available, and that food is not
tested for efficacy, but only for safety.
However, from an ethical point of view, objections against the identification of
food and health (drugs) can be heard; for example, food should not be seen as a
drug and society is not a hospital, meaning that health should not be the all-
determining value in food choices. For instance, Crawford [18] warns of the medi-
calization of daily life, which leads to a narrowing of perspective. As an observer
notices: For some decades the doctorpatient relationship has been the central
concern of medical ethics. This focus has marginalized public health issues by
concentrating on individual patients and individual practitioners, and thereby on one
aspect of medical structures in the richer parts of the world [19]. The British Food
Ethics Council views the exclusive orientation on health in food choices as a trans-
formation of society into a hospital [20], which is in the case of prevention of obesity
not desirable and not necessary as well, because physical activity is a good recom-
mendation as well [21]. Petrini [22], the founder of Slowfood, and a fierce defender
of the social and cultural aspects of enjoying food, quotes Madame Sevigny approv-
ingly, Health is enjoying the other enjoyments. When the other enjoyments are
taken away, we live longer, but we lose our health.
However, because of the development of the food sciences, traditional truths
regarding healthy food collapse: the health value of wine, chocolate, vegetable oils,
and milk are reevaluated. Food intake should correspond with changes in lifestyles:
What to eat in the new situation of sedentary work and other changing social
contexts?
The exact boundaries between food and health (drugs) are unclear, and there
have always some blurred boundaries (the ancient Greek thinker Hippocrates said:
Let food be your medicine![23]. Moreover, we are facing now two new gray areas:
food as drugs and drugs as enhancements, and we cannot eliminate them. This means
that the relationship between food and drugs needs to be reconsidered: the refusal
to reconsider the traditional distinction between food and drugs looks like sheer
dogmatism. Even with Kant [11] in hand, one could argue thus: There is a difference
TABLE 16.2
New Gray Zones between Food and Drugs
Gray Zone: Food as
Drugs and Drugs for
Nonpatients (modafinil,
Ritalin, enhancement
Food Drugs drugs)
1. Side effects Not accepted Accepted ??

2. For whom? Everybody Groups of patients Quasi patients
3. Choice Multiple reasons Health reasons Health prevention,
enhancement
4. Time taken Over lifetime During illnesses Preventive against
susceptibilities; over
lifetime
5. Delivery Freely available Over the counter Both
6. Testing Only safety, not efficacy Safety and efficacy ?
7. Activities Shopping, talking, preparing, Taking pills Testing, getting advice,
sharing shopping
8. Production Food companies, farmers Drug companies: Mixed
(open production systems) Closed lab drugs in greenhouses/
pharmadrugs
9. Responsibility Governments and food sector Physicians, drug ??
companies,
governments
10. Trust Sector not highly trusted Sector trusted ?
between obeying a physician and taking into account the advice of experts on food.
Moreover, the Eurobarometer [16] shows that the association of food with health is
in the EU only made by one person in five; the idea that society should transform
into a hospital is not very realistic. As long as health remains one of the consider-
ations in food choice and supply, and the multiple functions of food stay in place,
there is some reason to reconsider the intricate relationship between the two [14].
When personal nutrition on a modest scale is implemented, then the first area of
ethical concern can be overcome, is our final assessment. This allows for reflection
on the new gray zones between food and health when personalized nutrition on a
modest scale is implemented in certain fields. This reflection can begin by analyzing
the second and third main areas of ethical concern (Table 16.2).
16.4.2 THE SECOND MAIN AREA OF ETHICAL CONCERN:

THE RELATIONSHIP BETWEEN PERSONALIZED AND PUBLIC
HEALTH FOOD POLICIES
The second main area of ethical concern is the relationship between personalized
or individualized nutrition and social or public nutrition (and health). Many argue
on the basis of ethical and empirical arguments that a population or public health
nutrition and public health policy should have precedence over an individualist health
approach for reasons of efficacy [24] or for ethical reasons [25]. Moreover, they
argue that personal nutrition and health care have collective effects. When most rich
people in the Western world only want personal nutrition and care, the danger
emerges that the public nutrition and health system will seriously be damaged as a
result of a kind of personal/collective dilemma [18]. We will run the risk that the
food and medical problems of the poor in the West and the South will be neglected
because the richer people all will choose according to their own personal nutrition
and health projects. Because of these enormous collective effects, personalized
nutrition is not only a personal decision up to the individual, but affects the whole
population. Social responsibility seems to require that the whole population be
considered. For instance, it is a well-known fact that socioeconomic status (SES)
and health expectancy are strongly linked. The rich always have more opportunities
to look after their health than the poor and have therefore a longer life expectancy.
International public health (for example, World Health Organization [4]) aims at
improving health on a populational level. How far should personal responsibility for
health be taken into account in setting the agenda for national and global public
health nutrition policies, and what role has collective responsibility to play?
The answer of many food scientists and food industrialists to this question is
that personalized nutrition and its concomitants, such as a gene passport, comprising
the results of screening and testing, are means of individual empowerment and should
replace public health nutrition policies [26,27]. Moreover, insurance companies and
many governments are of the opinion that consumers should take more responsibility
for their health and not rely too much on governmental health regulations. They are
arguing: Youre responsible to do what is healthy and call this self-control and
autonomy (see the Blair government paper [28]).
However, some food companies are rather hesitant to enter the market of per-
sonalized nutrition. In an adjacent field, that of personalized medicine, which is
already more developed, it is very often the case that personalized medicines are
not commercially viable. Personalized medicine means that companies will have to
be content with only a (small) share of the market and cannot cover all the (potential)
patients. As a result, few new pharmagenomic drugs are filed for license in the U.S.
According to the FDA [29], the number of new drugs and biological applications
submitted to it has declined significantly. Moreover, the U.K.-based Public Health
Association [25] argues in its reaction to the Blair governments policies [28] that
public health should be the main target of governmental policies. This association
argues that in the long run the poorer parts of the population, and health itself, will
lose in a health and food system in which individual responsibility is the main
organizing principle [31]. In the same vein, Lang [24] also makes a strong case for
the primacy of public vs. individualized health nutrition politics. Targeting whole
populations provide governments with better chances of public health success,
whereas targeting at risk individuals could be socially divisive [24, p. 110]. But
in the end he concedes that individual tendencies [24, p. 120] do have to play a
role in a public health system.
Moreover, from an ethical point of view, one could argue with Kant [11] that
personalized nutrition in the sense of a gene passport or chart means disempowerment
and not enrichment of social autonomy, because one makes oneself dependent on a
physician. Only as a socially autonomous member of a rational community can
personalized nutrition be given a try, and the issues boil down to the balanced
relationship between autonomy and solidarity, not an either/or relationship [14].
Concluding, there are some arguments in favor of personalized nutrition, as long
as it does not displace public health nutrition [30]. Also, from a utilitarian standpoint,
there is place for personalized nutrition as long as individual responsibility can
defray the costs.
16.4.3 THIRD MAIN AREA OF ETHICAL CONCERN: PERSONALIZED

NUTRITION FOR THE CONSUMER
Granted that there is indeed a role for personalized nutrition in health, we are faced
with ethical issues concerning the development of personalized nutrition, such as the
protection of privacy (input, use, and sharing of databases) and misuse of information
(as in the example on the DNA pill hype; this concerns the second and third aspect
of personalized nutrition); the regulation of health claims (regarding the fourth aspect
of personalized nutrition, i.e., the development of foods); and the balance of personal
and social responsibility (regarding the second, third, and fourth aspects). Also, issues
related to communication on personalized services and products such as claims, right
to know and not to know, and informed consent need to be discussed. Preparation
for optimal consumer protection brings up issues related to intellectual property and
equal access. On the level of food choice, expression of identity with food and the
functions of food other than for staying healthy will with personalized nutrition
undergo dramatic changes that will challenge established ethical routines [31]. The
notion of personal responsibility for health is vulnerable to intentional manipulation
by self-interested (commercially interested) parties [36,37].
The scientific peculiarities of personalized nutrition and nutrigenomics, in par-
ticular, the huge complexities of the interactions between genome, genetic profiles,
and environments, such as diets and lifestyles, reduce severely the possibility of
reproducing experiments in other locations, because food and lifestyles differ so
much and are difficult to exclude [32,33]. It means that validation and falsification
of scientific relationships are confronted with extremely difficult and expensive
research designs. Nutrigenomics is not about a single relationship between a single
gene and its expression, but about the reactive genome, contextualizing at all levels.
Genes influence the expression of other genes in proteins and the organism and vice
versa, so the genome is not one, stable, all-determining commando post. Food has
multiple effects on the genetic profile, and is dependent on the food choices, which
in turn are influenced by lifestyle factors. It is not only that risks abound, such as
the now established connection between smoking and increased chance of lung
cancer, but also that ambiguities are quite normal; for example, two glasses of wine
give a higher risk of breast cancer but a lower risk of cardiovascular disease.
However, so far this evidence, which has led to some widely promoted dietary
recommendations, has come from mainly observational epidemiologic studies. The
complexity of multifactorial diseases and other interwoven complex causal relations
between genetic factors, omics factors, and the environment (lifestyle, social context)
confronts ethics with the necessity of transforming itself from a taken for granted
discipline to a searching and experimenting discipline. Ethics will need to incor-
porate these uncertainties and ambiguities into the use of nutrigenomics services
and products in the daily life of consumers, in adequate policy measures and in
ethical acceptable research agendas for research institutes.
An ethics for consumers can draw inspiration from Kants adage, dont obey
your doctor [11], which in a certain sense is right, but does not assist in establishing
the new necessary connections between food and health (drugs). The traditional
ethical concepts such as informed choice and individual responsibility are not
adequate for this new task [10]. When the collective implications of the individual
choice for personalized nutrition are so huge, more collective ways of opinion and
decision making should be tried in which all stakeholders have to readjust their
opinions and interests according to the ongoing discussion. This implies that the
ethics of protection of the individual against the apparently mighty powers of state
and professionals is not appropriate; more desirable is an ethics of participation in
continuous discussion on the intricacies of nutrigenomics.
Moreover, consumers need an ethics of dealing with uncertainties in the sense
of identifying and selecting between important, major, minor, and unimportant
choices. There are no general tools or guidelines to deal with that process of selection,
but there are general procedures such as consultations, deliberations, and exchange
of war stories and anecdotes. The main purpose of these procedures is to discern
commonalities and particularities in your own life and that of other affected indi-
viduals (Kants comparative universality! [12]). Some ethical support can be given
by Putnams distinction between commonsense doubt, meaning the selecting of more
and less certain cases, and philosophical doubt, i.e., the radical denial of all certain-
ties, which is most of the time not meaningful [34]:
We have to remind ourselves of the distinction between common sense doubt and
philosophical doubt. Finishing in believing in something is not really a human possi-
bility. Criticism cannot be a reason for universal skepticism. The fact that sometimes
we are wrong is not a reason to really doubt every particular conviction.
The reason that informed choice is not a good ethical concept in this context has to
do with the fact that it does not say anything about selecting the more certain and
less certain recommendations and the incorporation of well-established health con-
siderations into ones diet and personal health. The needed ethics of cooperation
between consumers and professionals and of sharing life experiences implies that
consumer groups should try to bridge the gap between production and consumption
of food and new technologies and to incorporate the new health considerations into
their food style together with the main stakeholders such as governments and food
industry [35].
With respect to the ethics for governments and their food policy, the uncertainties
in the gray zone between food and health are best tackled not by prohibiting the gray
zone but by regulating these on a social and technological level (coevolution) [14].
Governments should organize the research agenda of nutrigenomics in a democratic
way, not exclusively oriented toward personalized nutrition, but oriented toward the
prevention of common illnesses, common conditions, and chronic diseases. There is

a role to play for public health nutrigenomics, directed at general risk profiles.
Governments should regulate the three sectors, food, drugs and the newly emerging
gray zone, improve research ethics and medical testing of health foods, and allow only
affordable health foods. Its main aim in nutrition policy should be the encouragement
of the formal and informal ties between pleasurable eating and fundamental health.
Health should be a secondary goal in eating, and consumers and their associations
should be empowered as stakeholders (most trusted according to Reference 16: 33%)
and be given a voice not only downstream of nutrigenomics (with respect to the end
products) but also upstream of the nutrigenomics research agenda. To tackle the issue
of easy expectations and hype, it is necessary to establish independent gatekeepers to
safeguard the independency of testing, marketing, communicating, and providing the
services of nutrigenomics. Trust of citizen-consumers in the food and health sector
can be best maintained by the establishment and maintenance of independent bodies
fixing and monitoring rules with respect to claims.
With respect to ethics for scientists, scientists are in need of codes of practice
governing forward ethically acceptable research agendas and validating nutrigenom-
ics scientific claims. Research focusing on common illnesses should not be disrupted,
as also research on affordable and common drugs and food. Making healthy food
accessible for everyone remains an important challenge for nutrigenomics.
Of course, these three areas overlap with each other; how far food should be
connected with health considerations, or even identified with it, is both an issue for
collective consideration as well as a personal affair. The issue of responsibility is
also to be considered in these three main areas. But some issues (such as those of
the last area) demand more individualistic concepts, such as informed consent and
autonomy and protection, and others more collective ethical concerns such as trust
and justice.

Nutrigenomics has still an uncertain future, and a lot depends on the interaction of
research developments and priorities, prospects for profits, and ethical concerns. An
important first issue to be settled for the future of nutrigenomics is the relationship
between food and health (drugs). Although there are good reasons to be careful and
to make this distinction, there is room for taking into account new connections
between food and health; this gray zone is in need of regulation and requires
additional attention on the part of consumers and producers. Attention is in particular
needed so that health does not socially and personally function as the only motive
behind food choices, and health policies should not move in that direction. Second,
the issue of the relation of public health vs. personalized nutrition requires a balance
between solidarity and individual autonomy. In many societies, one is motivated to
look for such a balance, and only processes of continuous consultation and deliber-
ation can find out what this implies. It is unmistakably the case that a more personal
appeal to customers of food and medicine to incorporate recent health standards into
their food styles can have some effect. From this perspective, it seems ethically
desirable that consumers learn to cope with the hype [36,37], the uncertainties, and
relative certainties of this new approach to identifying and selecting their chief
vulnerabilities and be able to incorporate these in their narrative life plans. This
implies not only room for informed choice, but also for consultations and deliber-
ations on these uncertainties. The ethics of these issues is still in its infancy, because
until now the ethics of food (and of health) has had a strong defensive, and not
proactive outlook. However, from a public health point of view, the potentialities of
nutrigenomics in improving the health of the population, in particular of the least
well off, should be developed as well. Governmental policies and research institutes
should allow their research agendas to be shaped not only by personalized nutrition
but by common nutritional issues as well.
KEY READINGS
Castle, D., Cheryl Cline, Abdallah S. Daar, Peter A. Singer, and Charoula Tsamis, 2006, Science,
Society and the Supermarket : The Opportunities and Challenges of Nutrigenomics, Wiley.
Food Ethics Council, 2005, Getting Personal: Shifting Responsibilities for Health, Brighton.
Korthals, M., 2004, Before Dinner: Philosophy and Ethics of Food, Springer, Berlin.
Korthals, M. (Ed.), Genomics and Obesity, Dordrecht, Springer, forthcoming.
Mller, M. and Kersten, S., 2003, Nutrigenomics: goals and strategies, Nature Reviews
Genetics, 4, 4, 315322.
REFERENCES
1. TIMES, Sweet Taste of Success, June 24, 1999.
2. Della Penna, D., Nutritional genomics: manipulating plant micronutrients to improve
human health, Science 285: 375379, 1999.
3. Porter, R., The Greatest Benefit of Mankind, Norton, New York, 1999.
4. WHO (2004): Global strategy on diet, physical activity and health. Fifty-seventh World
Health Assembly. WHA57.17. May 22, 2004; WHO (2005): The Bangkok Charter for
health Promotion in a Globalized World. WHO, August 2005. Available on www.who.int.
5. Friedman, J.F., Modern science versus the stigma of obesity. Nature Medicine, 10(6):
563569, 2004.
6. Roth, J., Qiang, X. et al., The obesity pandemic: where have we been and where are
we going? Obesity Research 12(Suppl.): 88S100S, 2004.
7. Kaput, J. and Rodriguez, R.L., Nutritional genomics: the next frontier in the postge-
nomic era. Physiological Genomics. 16, 166177. PMID: 14726599, 2004.
8. Chadwick, R., Nutrigenomics, individualism and public health. Proceedings of the
Nutrition Society, 63(1): 1616, 2004.
9. For a general discussion on utilitarianism and other ethical perspectives. Singer, P.,
Practical Ethics, Cambridge University Press, Cambridge, 1979; 2nd ed., 1993.
10. Mill, J.S., On Liberty, Oxford, Oxford University Press, 1859, 1978.
11. Kant, I., What is enlightenment?, Guyer, P. Ed. The Cambridge Companion to Kant.
Cambridge University Press, 1992. Translation from Was ist Aufklrung? 1785.
12. Kant, I., Anthropologie in pragmatischer Hinsicht (Anthropology from Pragmatic
Point of View) (1798/1790), idem, Kants Werke, Darmstadt, 1975.
13. ibidem, p. 567.
14. Korthals, M., Before Dinner, Dordrecht: Springer, 2004.
15. Castle, D., Cline, C., Daar, A.S., Singer, P.A., and Tsamis C., Science, Society and the
Supermarket : The Opportunities and Challenges of Nutrigenomics, Wiley, 2006.
16. Eurobarometer, 2005, Risk issues, http://ec.europa.eu/public_opinion/archives/ebs/
ebs_238_en.pdf.
17. Korsmeyer, C., Making Sense of Taste, Cornell University Press,1999; Korsmeyer
doesnt take into account the personal history of tasting in listing factors that deter-
mine food preferences, p. 82, 98.
18. Crawford, R., Healthism and the medicalization of every day. International Journal
of Health Services, Vol. 10, No. 3, 1980.
19. ONeill, Onora, Public health or clinical ethics: thinking beyond borders, Ethics and
International Affairs, 16, 2002.
20. Food Ethics Council, Getting personal: shifting responsibilities for health, Brighton, 2005.
21. Hillsdon, M., Foster, C., Naidoo, B., and Crombie, H., The effectiveness of public
health interventions for increasing physical activity among adults: a review of reviews.
Evidence briefing of the NHS Health Development Agency. www.hda.nhs.uk/evidence.
2004.
22. Petrini, The Pleasures of Slow Food: Celebrating Authentic Traditions, Flavors and
Recipes, Chronicle Books, 2002.
23. Foucault, M., History of Sexuality, Vintage, Part 2, 1990.
24. Lang, T. and Heasman, M., 2005, Foodwars, London: Earthscan.
25. UKPHA, Public Health Association: Choosing Health or Losing Health? 2003.
26. German, J.B. and H.J. Watzke, Personalizing foods for health and delight. Compre-
hensive Reviews in Food Science and Food Safety 3: 145151, 2004.
27. Koelen, M.A. and Lindstrm, B., Making healthy choices the easy choices: the role
of empowerment. EJCN. 59, Suppl. 1, S10S16, 2005.
28. U.K. Department of Health, Choosing Health, London, 2004.
29. Darnton-Hill, I., Margetts, B., and Deckelbaum, R., Public health nutrition and genet-
ics: implications for nutrition policy and promotion, Proceedings of the Nutrition
Society, Vol. 63, No. 1, February 2004, pp. 173185(13).
30. Khoury, M.J. and Mensah, G.A., Genomics and the prevention and control of common
chronic diseases: emerging priorities for public health action. Prevention of Chronic
Diseases (serial online) April 2005. Available from http://www.cdc.gov/ pcd/issues/
2005/apr/05_0011.htm.
31. Cain, M. and Schmid G., From nutrigenomics science to personalized nutrition: The
market in 2010. The Institute for the Future, CAL.; Castle (2003). Clinical challenges
posed by new biotechnology: the case of nutrigenomics. Post-Graduate Medical
Journal, 79: 6566, 2003.
32. Sinha, G., the diet that fits analyzing metabolism for personalized nutrition,
Scientific American 292(3): 22 March 2005; DeCamp, M. and Sugarman, J. (2004).
Ethics and research assessing the relative roles of genes and the environment.
Accountability in Research: Policies and Quality Assurance 11: 161182.
33. Tate, T.K. and Goldstein, D.B., Will tomorrow s medicines work for everyone?
Nature Genetics 36(11): S34S42, 2004.
34. Putnam, H., Pragmatism: An Open Question, Oxford: Blackwell, pp. 5781, 1995.
35. Liakopoulos, M. and Schroeder, D., Trust and functional foods. New products, old
issues, Poiesis & Praxis: International Journal of Technology Assessment and Ethics
of Science, Vol. 2, No. 1, 4152, 2003.
36. Meek, J., Public misled by gene test hype. In The Guardian, March 12, 2002.
37. BBC-News. (July 17, 2005). DNA test for diabetes and obesity. Health. Retrieved
November 7, 2005.
Section III
The Future of Personalized
Nutrition
17 International Efforts
on Nutrigenomic Health
for Individuals in the
Global Community
Jim Kaput
CONTENTS
17.1 Introduction ..................................................................................................238

17.2 The Scientific Diversity Challenges ............................................................239
17.2.1 Chemicals in Foods..........................................................................239
17.2.2 Genetic Diversity .............................................................................240
17.2.3 Complexity of Health and Disease Processes .................................242
17.2.4 Complexity of Data Sets..................................................................242
17.3 The Changing Shape of Biomedical Research............................................243
17.4 The Road Map for Harnessing Nutrigenomics for Personal
and Public Health.........................................................................................249
17.4.1 The Status of the Nutrigenomics Society 2006 .........................249
17.4.1.1 Promoting, Organizing, and Communicating Interest .....250
17.4.1.2 Best Practices....................................................................250
17.4.1.3 Global Biological Databases ............................................251
17.5 The Extraordinary Challenges of Involving Scientists
and the Public in Developing Countries......................................................251
17.5.1 A Humanitarian Need also Exists ...................................................252
17.5.2 Funding Nutrigenomic Research in Developing Countries ............252
17.5.3 Issues Facing Collaborative Research
in Developing Countries: The Ethiopian Example .........................253
Acknowledgments..................................................................................................255
Key Readings.........................................................................................................255
References..............................................................................................................255
237
17.1 INTRODUCTION
Nutrigenomics is the study of how constituents of the diet interact with genes,
and their products, to alter phenotype and, conversely, how genes and their products
metabolize these constituents into nutrients, antinutrients, and bioactive com-
pounds. The data and results from nutrigenomics research are predicted to provide
individuals with knowledge for optimizing nutrient intakes with the goal of pre-
venting or delaying the onset of diseases. The research and application path to
personalized nutrition faces the challenges of the chemical complexity of food,
genetic heterogeneity of human populations, intricacies of biological processes,
and the need for new methods of analyzing high-dimensional data sets generated
by omics technologies. With the success of the Human Genome and HapMap
projects, many scientific disciplines are beginning to develop best practices and
share data sets for understanding biological processes. A number of large-scale
collaborative programs focusing on nutrientgene interactions have been initiated
within the past 3 years. Nutrigenomic researchers individually, in centers, and in
multi-institutional programs have recognized the need for developing best prac-
tices, fostering international collaborations, and sharing data sets. The scientific
and humanitarian need for forming a network of networks and the initial steps on
that pathway are reviewed in this chapter.
The excitement caused by the unparalleled international collaborations to ana-
lyze the human genome (1,2) and its variability (3) created lofty expectations in the
medical and pharmaceutical sectors, food industries, media, and informed public.
Information obtained with genetic and emerging omic tests are predicted to person-
alize medical treatments (4), drug selection (5), and nutritional advice (69), and
perhaps lead to the development of healthier processed foods targeted to individuals
(10,11). Although many believe these goals will be realized, expectations have given
way to the sobering recognition of the challenges of understanding genetic causes
of complex traits such as chronic diseases and the role of environment in regulating
expression of genetic makeup.
Results from genetic studies attempting to associate a gene or its variants to a
disease or specific metabolic phenotype illustrate the challenges: many published
genedisease associations could not be replicated in subsequent studies [e.g.,
(12,13)]. This led one accomplished epidemiologist to publish a report entitled Why
Most Published Research Findings Are False (14), identifying experimental designs
where sample sizes lacked appropriate statistical power, control groups were not
appropriately matched to cases, study participants had differing genetic admixtures
(populations stratification), and data that were overinterpreted (among others, see
References 15 to 18). Researchers working at the intersection of nutrition and
genetics added that dietgene interactions are major contributors to the control of
gene expression and could influence associations among genesphenotype and
dietary intake (7,1921). Reaching the goal of personalized nutrition will require an
understanding of the scientific challenges of identifying nutrientgene interactions
involved in health or disease development and realistic solutions to surmount those
challenges.
International Efforts on Nutrigenomic Health 239
17.2 THE SCIENTIFIC DIVERSITY CHALLENGES

17.2.1 CHEMICALS IN FOODS
Humans have adapted to almost every nutritional environment on earth, including
northern tundra, equatorial, and high-altitude ecosystems. Based on this adaptability,
one could conclude that as long as minimum amounts of energy and micronutrients
were consumed, the most significant contribution to health and disease susceptibility
would be ones genetic makeup, However, environment, and specifically nutrients,
alters the expression of genetic makeup leading to promotion of health or the
development of chronic diseases (7). The most obvious example of the importance
of environment on disease is the rise in the incidence of obesity and its complications
(http://www.cdc.gov/nccdphp/dnpa/obesity/) over the past 20 years. This time span
is not long enough for genetic changes to cause such rapid weight gains in individ-
uals. Rather, changes in diet and physical activity habits have produced the epidemic
of obesity, type 2 diabetes (T2DM), and their complications. This conclusion is
supported by traditional epidemiologic and immigration studies (e.g., Reference 22)
that associate environment with chronic diseases. However, the underlying molecular
mechanisms cannot be ascertained by these statistical associations.
Numerous in vitro, animal, and human studies have demonstrated that chemicals
classified as macronutrients, such as certain fatty acids, and micronutrients such as
genistein, sterols, and vitamins regulate gene expression directly (reviewed in
Reference 23) or through changes in signal transduction pathways (7,24). Hence,
for the specific chronic disease of obesity, overconsumption of calories (25) and
subsets or specific bioactive chemicals (e.g., Reference 26) alter the regulation of
genes that cause increased weight deposition in genetically susceptible individuals.
In addition to chemicals present in unprocessed foods, our diets are now greatly
influenced by modern food production methods. Humans have been able to extract
and recombine food chemicals for mass markets only since the onset of the industrial
revolution. Designing new foods was done primarily for taste and economic pur-
poses. Hence, it is not surprising that our current national and international food
systems have not produced the most healthful foods. Nevertheless, modern food
production provides enough calories and most macro- and micronutrients, at least
for a large proportion of the worlds population.
Regardless of the type of food, identifying the specific pathways and the effect
of different amounts of these bioactive chemicals is a critical part of understanding
of how diet maintains health or leads to disease in each individual. For more complete
analyses of genedisease (or phenotype) associations, studies must assess food intake
as well as genetic differences. Quantifying nutrient intake in epidemiologic studies
is challenging because free-living humans do not regard daily life as a science
experiment where the amount and type of food is accurately and precisely recorded.
It is generally accepted in the nutrition community that food frequency questionnaires
(FFQs) for assessing intakes are less than ideal (27). Nevertheless, FFQs are likely
to be the most convenient assessment for large-scale studies. Quantitative biological
measurement tools for nutrient intakes need to be developed for more accurately
assessing nutrient intake. (28,29).
Converting food intake to nutrients is done using food composition databases.

The most complete is the USDA Food Composition Database (http://www.nal.usda.
gov/fnic/foodcomp/Data/SR18/sr18.html). However, plant species, similar to other
organisms, respond to their environments, which may result in different nutrient
contents depending on the genotype of the plant and its growth conditions (30).
Because a full nutrient profile costs about $2000 (per sample in 2002, from
Reference 31) and it is typical to average six samples, a ranking scheme (31) will
be needed to prioritize analyses of foods. Progress in developing national and
international food composition databases is being made, as noted by the International
Life Sciences Institute (ILSI) Crop Composition Database (http://www.cropcompo-
sition.org/), the International Network of Food Data Systems (http://www.fao.org/
infoods/directory_ en.stm), and the European Food Information Resource Network
(EuroFIRhttp://www. eurofir.net/).
An added layer of complexity occurs when comparing across populations. Cul-
tural differences in food manufacturing, preparation, and eating customs differ not
only among nations and ethnic groups, but also among different religions (e.g.,
Reference 32). Personal choices such as various vegetarian practices will also con-
found assessments of nutrient intakes. Other aspects of culture and personal habits
[e.g., sleep time and continuity (33,34), activity levels (3538), and other factors
(reviewed in Reference 39)] are all known to alter expression of genetic information.
Individual studies have successfully included one or more of these variables, but to
date, no study has been published with complete environmental analyses. The devel-
opment of admixture mapping (40), a genetic method that uses differences among
ancestral groups to identify chromosomal regions involved in disease, provides an
example of using comparative methods to identify causal factors in complex systems.
Similar comparative analyses might be developed for nutrient components involved
in health and disease.
17.2.2 GENETIC DIVERSITY

Genetic analyses showed that Africans have the most DNA sequence variations
compared to other populations, a result consistent with the hypothesis that modern
humans evolved on that continent (41,42). During the stepwise migrations from east
Africa and subsequent population centers, migrating groups carried subsets of the
genetic diversity to each new location. Rapid growth of these isolated populations
produced populations that share 99.9% of genomic sequences. Examination of the
remaining 0.1% indicated that there is a 12 to 14% difference between geographically
distinct populations (42) for example, between Asia and Europe, whereas the
majority of human variation (estimated range of 86 to 88%) occurs within a geo-
graphic population (Europe, for example).
Food availability and other factors contributed pressures for selecting and main-
taining specific gene variants in new environments. Lactase persistence is a prime
example of this natural selection process. In many adult humans, severe gastrointes-
tinal discomfort and diarrhea result when milk is ingested because adults cannot
metabolize lactose. Certain populations, such as Europeans and many of their descen-
dents, have the ability to drink milk because the gene encoding lactase is expressed
in adults. A polymorphism in the promoter [C-13910T (43)] of the lactase phlorizin
hydrolase gene (LCH) that arose as a mutation around 9000 years ago is responsible
for the altered expression pattern of the gene. Individuals not descended from the
founding European population retain the normal metabolism of being unable to
metabolize lactose as adults: less than 5% of the individuals in southeast Asia can
safely drink milk (7). Certain populations in Africa also exhibit lactase persistence
even though the African and European populations have been separated for at least
~35,000 years. DNA sequence analyses showed that a large percentage of Africans
in certain populations have a G-14010C polymorphism 5 to the LCH gene. This
polymorphism arose about ~7000 years ago (44). The maintenance of these LCH
promoter variants may confer selective advantages that include improved nutrition,
prevention of dehydration, and improved calcium absorption in environments that
are nutrient poor. The independent selection and maintenance of separate mutations
in African and European populations are a prime example of convergent natural
selection (44). Different alleles conferring the same phenotype (lactase persistence)
have practical consequences for both pharmacogenomics and nutritional genomics:
gene variantphenotype associations may be population specific because members
of different ancestral groups may have different polymorphisms affecting the same
gene.
The implications of this genetic similarity and diversity are also significant for
nutrigenomic researchers because the frequency of gene variants differs among
populations. Perhaps as importantly, gene variants and their products (either RNA
or DNA) interact with other genes and their products in pathways and networks.
Such genegene interactions [or epistasis (4547)] work at the protein and enzyme
[and probably iRNA (48)] levels to buffer biochemical reactions (49,50). The
concept of buffering is important because the presence or absence of a particular
gene variant (i.e., SNP) is not an all or none phenomenon. For example, the activity
of a gene A variant may be affected by a second gene B (through its protein product)
that is inherited separately (i.e., that is, on an unlinked chromosomal region). The
result of epistasis is that a single SNP is not deterministic but expresses itself within
the background of the individuals genome (that is, the total set of all variations in
their DNA). Hence, SNP analyses of individual genes must be accompanied by
analyses of genomic variations, an analytical approach that has yet to be developed.
A specific example is illustrative: a specific group of SNPs (called a haplotype,
and in this case, HapK) in leukotriene A4 hydrolase (LTA4H) was associated with
increased risk of myocardial infarction (MI). However, the risk of MI was significantly
greater in African Americans who carried the HapK haplotype derived from Europe
(51) compared to the presence of the HapK in a European genetic background. The
genetic explanation is that SNPs in LTA4H affect the activity of that enzyme product
and its interaction with other genes whose variants are found more frequently in
Africans. These results demonstrate the concept that a SNP or haplotype is context
specific (52); that is, a SNP or haplotype may contribute different amounts to disease
risk depending on ancestral background because of epistatic (genegene) interactions.
Because some of the genes that cause disease are influenced by diet, the same nutrients
may differentially affect individuals with different ancestral backgrounds. Researchers
have used self-described ethnicities as proxies for genetic architecture, but skin color
is a poor substitute for genetic analyses (53,54).
17.2.3 COMPLEXITY OF HEALTH AND DISEASE PROCESSES

Chronic disease is often thought of as an all or none phenomenon one is either
healthy or has the symptoms of some disease. This dichotomous view of health and
disease does not accurately describe biological processes. Complex traits, such as
the chronic diseases obesity, diabetes, cardio- or cerebrovascular diseases, Alzheimers,
and certain cancers are caused by the contribution of many genes interacting with
multiple environmental factors (reviewed in references 20 and 55). Many of these
diseases manifest with complications. For example, T2DM is often caused by, or
associated with obesity, a proinflammatory condition that will introduce additional
variation in physiology and responsiveness to diet and drugs. Symptoms associated
with the metabolic syndrome dyslipidemia, hypertension, insulin resistance and
hyperinsulinemia (5659) further complicate even sophisticated classification
schemes for diabetes (60). Many of the primary causes of T2DM, such as glycemic
control and insulin regulation, change gradually over many years in genetically
susceptible individuals exposed to improper diets and lack of physical activity. The
resulting physiologies and responses to increased glucose in turn create conditions
for the development of complications. These physiological abnormalities may have
overlapping molecular and genetic causes to further confuse diagnosis and the
development of targeted treatment options (4).
As an example of the genetic and molecular complexities of chronic diseases,
genetic analyses suggest that at least 24 chromosomal regions are linked to T2DM.
Seven are statistically significant (log of the odds = LOD > 3.6), and 17 others
suggestive [2.0 < LOD < 3.6; (61)]. One or more genes at each of these loci contribute
to the development of the T2DM but to different amounts in different individuals
depending on function, genegene interactions, and geneenvironment interactions.
Association studies have linked 51 genes to T2DM or metabolic conditions associ-
ated with this disease, some but not all of which map to the T2DM loci (reviewed
in Reference 4). It is not surprising, then, that chronic diseases and their complications
are difficult to study accurately, diagnose precisely, and treat effectively. Similarly,
there may be multiple healthy states, each produced by different combinations of
genes at different chromosomal loci. Developing optimal diets to maintain health in
different healthy individuals is the flip side of nutrigenomic efforts to understand
disease processes caused by imbalanced nutrient intakes in genetically susceptible
individuals.
17.2.4 COMPLEXITY OF DATA SETS

Although these challenges seem intractable, high-throughput technologies of genom-
ics, transcriptomics, proteomics, and metabolomics, will produce large data sets
(also called high-dimensional) of nutrients, gene variants, genetic ancestry data,
physiological measurements of proteins and metabolites, as well as physical activity
levels and other personal habits. Results will than have to be extracted from these
interacting data sets to produce knowledge of geneenvironment interactions for
each individual. The existing paradigm of analyzing biological data often assumes
linear relationships, such as dose-dependent effects. However, genes, nutrients, and
their interactions may not necessarily be related linearly.
Analyzing complex data sets is not limited to biological data, and other disciplines
have developed algorithms that can discover nonlinear patterns in high-dimensional
data. Dimensionality reduction is a mathematical method of mapping multidimen-
sional data into a space of fewer dimensions (62,63), generating a deeper under-
standing of complex biological processes (e.g., Reference 64 and review in Reference
65). Among these methods are isomap (46,62,63), factor analysis (http://www2.
chass.ncsu.edu/garson/pa765/factor.htm), and classification tree analysis
(http://www.statsoft.com/textbook/stclatre.html), all of which can be used to analyze
complex patterns of gene variants and their influence on subphenotypes (insulin
levels or fasting glucose levels, etc.) in response to different dietary intakes. The
behavior of interacting genetic and environmental systems cannot yet be predicted
accurately and, hence, multiple approaches may be needed to extract knowledge
from experimental data. Nevertheless, data generated from different high throughput
technologies may be analyzed with more confidence when other paradigms (non-
linearity) are considered possible.
17.3 THE CHANGING SHAPE

OF BIOMEDICAL RESEARCH
The cultural and scientific view that science is conducted by a lone researcher
developed during the early part of the scientific revolution. The practice of conduct-
ing science began to change in the early 20th century with the development of teams
consisting of a senior mentor, technicians, graduate students, and postdoctoral fel-
lows, usually at one academic institution or company. Although there are lab-to-lab
collaborations within a nation and internationally, these typically result from friend-
ships and contacts and are ad hoc. The U.S. governmentsponsored Manhattan
project and subsequent development of national laboratories focusing on high-energy
physics and its applications accelerated an era of collaboration consisting of
specialized teams within a country and sometimes across national boundaries.
Biological research has been slower to adopt this large-scale collaboration model.
However, the complexities of understanding health and disease processes in light of
the nutritional, genetic, and cultural variation are changing the structure of science
teams. Many current scientific experimental designs and projects are reductionist in
nature, which still allows single-investigator-led teams. Focusing on one pathway of
a complex process is unlikely to determine the full spectrum of genenutrient inter-
actions that produce health or lead to disease (14). A systems biology approach (66)
using technologies that analyze DNA variation, transcription regulation, protein
amounts or activities, or metabolite concentrations (6769) will be required for
progress to be made. No single research team has the expertise or resources for
completely analyzing how the many environmental factors, including thousands of
food chemicals, alter the expression of genetic information in an individual.
Hence, the new paradigm emerging in the 21st century for biological research
is based on the model of the large-scale genome projects (13). Spurred by these
successes, scientists in many disciplines are organizing national and international
efforts for their individual fields of research for providing resources and best
practices (examples in Table 17.1).
TABLE 17.1
Examples of Investigator and Center Networks
Topic or Field Title Goal Reference
Mouse genetics Collaborative Develop inbred and congenic 108

Cross Project strains as a community resource
for complex trait analyses
Pharmacogenetics Pharmacogenetics Database for pharmacogenomics http://www.nigms.n
Research results ih.gov/pharmacog
Network enetics/goals.html
Heart, lung, blood, PROGENI Administrative and data http://www.biostat
and sleep Program for coordinating center .wustl.edu/proge
disorders study Genetic ni/
Interaction
Human HuGENet Analyses of human genome 109,110
epidemiology variation on population health
Human P3G the Public International network promoting http://www.p3gco
epidemiology Population transparency and collaboration nsortium.org
Project in for human epidemiology studies
Genomics
Inflammatory IBD Network Consortium for understanding the 111
bowel disease genetic contributions to this
disease
Nutrigenomics researchers have also been successful in developing collab-

orative centers supported by national governments (Table 17.2). Centers and
multi-institutional programs supplement the extensive national and international
collaborations that several individual laboratories have developed (reviewed in
Reference 70).
The new model emerging from these collaborative projects is one in which
individually funded laboratories in separate institutions contribute expertise to a
large study that analyzes multiple biological parameters. For example, an epide-
miologic study of nutrientgene interactions affecting some phenotype may require
biostatisticians for designing the experiments with science specialists and ethicists
(and perhaps nonscience community members), clinicians who see patients, genet-
icists for analyzing SNPs and chromosomal inheritance, nutritionists monitoring
and assessing dietary intakes, metabolomics researchers determining serum and
urine metabolite concentrations, proteomic specialists studying multiple enzymatic
pathways in various tissues, neuroscientists examining NMR brain spectra, phys-
iologists interpreting hormone levels, bioinformaticists capturing and managing
the data sets, and biocomputationists using high-dimensionality reduction and
other algorithms for analyzing the data. A type of specialist currently in short
supply in our scientific communities is the individual who can organize the
team and interpret the data and results from multiple disciplines. Funding such an
enterprise would add to the development of this integrative science and, more
importantly, to the type of data and results that could be applied to complex
TABLE 17.2
Examples of Multi-Institutional Centers for Studying
GeneNutrient Interactions
Partners/Host Institutiona Agency Focus
AFMNet Advanced Foods and Materials Network

http://www.afmnet.ca/
Fourteen universities across Networks of Centres of Structure dynamics function of foods
Canada Excellence and biomaterials
Canadian National Functional foods and nutraceuticals
Government
Genetics, ethics, economics,
environment, law, and society
(GE3LS)
ARS Childrens Nutrition Research Center

http://www.ars.usda.gov/main/site_main.htm?modecode=62-50-00-00
Baylor College of Medicine USDA Natural products and nuclear receptors
Texas Childrens Hospital
United States Department of
Agriculture Agricultural
Research Service (USDA
ARS)
DIoGenes Diet, Obesity, and Genes

http://www.diogenes-eu.org/
Thirty research partners European Union Obesity and
Sixth Framework projects: Macronutrient composition of diet
Four major industries
Food Quality and Safety Genenutrient interactions
Priority Genes and diet at the population level
Consumer attitudes and behavior
Food technology
Ensure that science yields applications
Data hub
EARNEST Early Nutrition Programming Project

http://www.earlynutrition.org
Thirty-eight research European Union Effects of early programming on later
partners Sixth Framework projects: chronic diseases
Sixteen European countries Food Quality and Safety Definition of critical periods in fetal and
Priority early life on later disease
(continued)

Genetic determinants on early

programming effects and on
subsequent outcome
Role of specific nutrients and their
interactions in the maternal and infant
diet on programming
Mechanisms for early programming of
later disease and their risk factors
Development strategies for treating and
preventing adverse programming effects
of early nutrition
How knowledge about early
programming affects consumer
behavior
Impact of early nutrition on the economic
burden of adult ill-health
New technologies for potential
economic advantage.
EuroFIR European Food Information Resource Network

http://www.eurofir.net/
Forty academic and small- to European Union Authoritative source of food
medium-sized enterprises Sixth Framework projects: composition data in Europe
Food Quality and Safety
Twenty-one European
Priority
countries
HealthGrain Exploiting Bioactivity of European Cereal Grains for Improved Nutrition

and Health Benefits
http://www.healthgrain.org/pub/
Forty-three organizations European Union Consumer research
Sixth Framework projects:
Fifteen European Countries Plant breeding and biotechnology
Priority Technology and processing
Nutrition
Dissemination and training

HEPADIP Hepatic and Adipose Tissue in the Metabolic Syndrome

http://www.hepadip.org/
Twenty-seven partner European Union Biology of adipose and liver tissue
institutions Sixth Framework projects: Integrative studies of adipose and liver in
Thirteen countries rats
Priority
Integrative studies of adipose and liver in
humans
Clinical and genetic epidemiology studies
Development and integration of omics and
bioinformatics
Target validation and development
LipGene Diet, genomics, and the metabolic syndrome: an integrated nutrition, agrofood,
social, and economic analysis
http://www.lipgene.org
Twenty-five partners European Union Role of dietary fats and genetic variation
Sixth Framework projects: Create agrofoods of LC n-3 PUFA
Ten European countries
Priority Alter milk fat from cows and fats in poultry
Consumer attitudes: MS and agrofood
technologies
Economic barriers to new agrofood
technologies
Diet vs. drug costs in management of the MS
Dissemination and education program
NCMHD Center of Excellence in Nutritional Genomics

http://nutrigenomics.ucdavis.edu
University of NIH National Center for Biological basis of health disparities
CaliforniaDavis Minority Education and outreach
Health and Health
Childrens Hospital of
Oakland Research Institute Disparities
USDA Western Human
Nutrition Research Center
Ethnic Health Institute
University of Illinois
Chicago Laboratory of
Nutrigenomic
(continued)

NuGO European Nutrigenomics Organization

http://www.nugo.org
Twenty-three partner European Union Develop and integrate genomic
institutions in ten European Sixth Framework projects: technologies for European
countries Food Quality and Safety Priority nutritional sciences
One collaborative center Facilitate the application of
technologies in nutritional
research worldwide
Create the leading international virtual
center of excellence in nutrigenomics
Train a new generation of European
scientists to use postgenomic
technologies
Nutrigenomics New Zealand

http://www.nutrigenomics.org.nz/
University of Auckland New Zealand Foundation for Inflammatory bowel diseases
Research and Science
Crown Research Institutes:
Technology (FRST)
AgResearch
HortResearch
Crop and Food Research
The UAB Center for Nutrient Gene Interactions in Cancer Prevention

http://207.203.38.82:8888/ssg/cngi2/index.htm
University of Alabama NIH National Cancer Institute Dietary polyphenols and
Birmingham Division of Cancer Prevention chemoprevention
The Northern California Soybean
Cancer Center Statistical analyses of high-
The University of dimensional data
Missouri-Rolla
a For a full list of partner institutions and countries, see URL for each center.
Source: Adapted from Kaput, J. 2007. Nutrigenomics 2006 Update, Clinical Chemistry and Lab-
oratory Medicine, in press.
biological phenotypes such as chronic diseases. A functioning alliance producing

evidence-based recommendations to address personal and public health issues
must be accomplished first, given the current nation (or region)-based funding
mechanisms of the early 21st century.
17.4 THE ROAD MAP FOR HARNESSING

NUTRIGENOMICS FOR PERSONAL
AND PUBLIC HEALTH
The development of regional and national nutrigenomic centers and programs (70),
although encouraging and essential, is nonetheless insufficient to address the com-
plexities of nutrientgene interactions. The reasons were outlined previously in the
diversity challenges of nutrient complexity of foods, genetic heterogeneity, intricacies
of metabolic processes, and the need for large data sets. This realization led to a
call for international strategic alliances to harness nutritional genomics for public
and personal health (39), a position statement coauthored by 89 scientists from
22 countries. Table 17.3 provides an outline of key steps in harnessing nutritional
genomics data for developing the knowledge for personalized nutrition.
17.4.1 THE STATUS OF THE NUTRIGENOMICS SOCIETY 2006

The shift from small- to large-scale projects will be challenging because of the
sociology of modern science, the source of funding, and how they are managed, and
how they are executed. These challenges can be met through the creation of a central
organizing center focused on promoting, organizing and communicating interest,
developing community-based resources, developing best practices, and developing
global databases for data sharing.
TABLE 17.3
Road Map for Nutritional Genomics Research
Topic Rationale Outcome
Data federation Scalable databases share genetic, Allow for greater statistical
phenotypic, dietary, nutritional status, and power across populations
other environmental and cultural information
Larger study Statistical power limitations Improve reliability of results
populations
Phenotypes More reliable analyses and consistency Improve reliability of results
Nutrient intakes Quantitative measurements Improve reliability of results
Genomic controls Ancestry background analyses Control for epistasis
Study Inclusion of diverse cultures and genotypes More complete understanding
participants of nutrientgene interactions
Assess other Affects expression of genetic info Reduce noise in data sets
environmental
variables
Partnerships Resource building for capturing nontraditional More complete understanding
among data and interactions among data sets of nutrientgene interactions
academia,
industry, society
Source: Adapted from Kaput, J., Ordovas, J.M., Ferguson, L. et al. 2005. Br J Nutr 94: 62332.
17.4.1.1 Promoting, Organizing, and Communicating Interest
The only nutrigenomics organization whose mission is multinational and that is not
explicitly linked to a specific research program is The European Nutrigenomics
Organisation (NuGO). The mission of NuGO is to develop, integrate, and facilitate
genomic technologies, infrastructure, and research for nutritional science, to train a
new generation of nutrigenomics scientists, in order to improve the impact of nutri-
tion and genetics in health promotion and disease prevention. NuGO has recently
established the Nutrigenomics Society to provide its leadership and resources to all
in the international nutrigenomic community.
The first step of this effort was the creation of a central Web site (http://www.
nugo.org) that provides information to the researcher in pages called the Nutrigenomics
Information Portal (NIP) (71). Divided into Research, Technology, Outreach, and
Partnerships, the NIP provides a meeting and resource site for those interested in
nutrigenomics topics. A second section of the Web site, labeled Facts about Nutrige-
nomics (FAN), is designed for the educated public. FAN describes the field, key terms
and concepts and information about the collaborative effort. The final section of the
public pages are entitled Nutrigenomics Stakeholders (NS), which is a section focusing
on industry, health professional, regulators, politicians, patient organizations, and gov-
ernment departments. Many who use the Web recognize the need for reliable infor-
mation from experts, and the content of the Nutrigenomics Society intends to provide
balanced and accurate descriptions of nutrigenomics science and its applications. The
Web site is an ongoing effort that was initiated in mid-2006 (71). The ultimate goal
of the Web site registration of individuals and projects, resource development, and
best practices is to create data that can be shared across studies.
In addition to the education and outreach efforts of NuGO, the NCMHD Center of
Excellence in Nutritional Genomics at the University of CaliforniaDavis has a nutrige-
nomics LISTSERV to alert subscribers to various scientific and public news articles,
nutrigenomic activities, and issues facing the field. Started in July of 2002, the LIST-
SERV currently has over 1600 subscribers (January 2007) in over 50 countries.
17.4.1.2 Best Practices
Many scientists doing research at the turn of the millennium were taught reductionist
approaches to science. This philosophy and the associated methods have contributed
immensely to scientific progress and will always be needed to understand individual
pathways and reactions of complex networks. However, reductionism cannot fully
explain complex, interacting systems because expression of genetic information relies
on environmental factors that change during life. Developmental and aging processes
of biological systems also vary over time, possibly altering geneenvironment inter-
actions. Such complexity is rarely accounted for in reductionist experimental designs.
Nutrigenomics projects therefore will require best practices developed by experts
in different disciplines. NuGO began this process in early 2004 with ongoing meet-
ings and discussions among European scientists at different institutions. This effort
has generated three best practices papers in transcriptomics (72), metabolomics (68),
and proteomics (73). These recommendations focus on the technologies of each
discipline rather than experimental designs for human, laboratory animal, cell culture
experiments, or for pre- and postexperiment biocomputation methods. Because tech-
nologies and experience change, these best practices will require ongoing discus-
sions, which will, in the near future, include non-European scientists. Online forums
will be created to foster interactions across national and physical divides.
17.4.1.3 Global Biological Databases
The challenges inherent in nutrigenomic studies, specifically genetic variation, pop-

ulation architecture, individual and cultural food preferences, among others, requires
large study populations and consistent phenotypes, common data elements, and
database interoperability. None of these can be achieved without forethought and
planning. Hence, the development of global biological databases focused on data
from the omics technologies of genomics, transcriptomics, proteomics, and metab-
olomics will be necessary. Developing robust nutrient intake databases that account
for the many cultural practices associated with food systems throughout the world
will be one of the more significant challenges for the nutrigenomic community. The
first database to be developed through this effort is the genenutrient database, a
collaboration between the Genetics and Nutrition Laboratory at Jean Mayer USDA
Human Nutrition Research Center on Aging at Tufts University and NuGO. This
database will consist of published reports of SNP (gene variant)nutrientphenotype
associations.
17.5 THE EXTRAORDINARY CHALLENGES

OF INVOLVING SCIENTISTS AND THE PUBLIC
IN DEVELOPING COUNTRIES
Much, if not all of, the research described in this and other review articles about
nutrigenomics are derived from studies conducted in developed countries. Caucasians,
African Americans (whose ancestors for the most part are West Africans), Asian
Americans of various ancestries, and in a few instances, with other ethnicities [e.g.,
Indians, Malays, and Chinese (74) or Koreans (75)] have been study participants.
Representatives from other ancestral and ethnic groups are rare or missing from
most nutritional genomic research projects. This is not unique to nutrigenomics, but
to nutrition (76), genetics (77) and other scientific disciplines. The rationale for this
focus was that funding was obtained from national agencies and also because
including other genetic makeups confounded the analyses because ancestral back-
grounds could not be quantitatively measured. Based on data from the Human
Genome and HapMap projects, no population can serve as the reference group. A
full scientific description of nutrientgene interactions requires inclusion of as many
ancestral groups and cultures as possible because epistatic, and perhaps epigenetic
interactions (78), may differ in genetic background and may alter statistical
genenutrientphenotype associations. The different SNPs responsible for lactase
persistence in African and European descendents (see previous text) exemplifies the
need to include other genetic makeups in research and application processes.
17.5.1 A HUMANITARIAN NEED ALSO EXISTS

The World Health Organizations (WHO) Child and Adolescent Health Development
Web site (http://www.who.int/child-adolescent-health/dev.htm) states, The future
of human societies depends on children being able to achieve their optimal physical
growth and psychological development. Data from numerous scientific studies
have demonstrated that improper diets block both physical and mental growth and
development (reviewed in Reference 59, 79, and 80). Because nutrient intake
requirements are usually determined from studies of individuals in developed coun-
tries, who represent only a subset of the human genetic diversity (see earlier text),
controversy exists as to whether international nutrient intake standards are valid
(81,82). Some experimental data pertinent to this controversy exist. Additional
dietary folate in individuals with variants of methylene tetrahydrofolate reductase
[MTHRF; (83)] individuals with polymorphisms in alcohol-metabolizing genes
experience skin flushing (84), and individuals carrying certain variants of angio-
tensinogen are sensitive to increased salt intakes (85). The likelihood of inheriting
these gene variants differs because allele frequencies differ among ethnic popula-
tions. If nutrient intakes are to be optimized for health of individuals in developed
countries, it follows that this process should also be applied to those in developing
countries. One could make a strong argument that nutrigenomics research should
first focus on populations in most need of food because delivering the appropriate
nutrients would do the most good in terms of personal and public health (86). For
example, Neel proposed the thrifty gene hypothesis in 1962 (87), which posited
that certain genetic makeups are more susceptible to energy-dense foods because
their metabolism utilizes energy from foods more efficiently (8890). If this is true
for calorie intake, it is likely to be true for concentrations of at least some macro-
and micronutrients (81,91,92).
Imbalances of nutrient intake may be most devastating during key nutritional
developmental windows (59,79,80) Developmental windows are periods during
which perturbations of the intrauterine, and probably early postnatal, growth envi-
ronment have major effects on biological processes. These critical periods may allow
for permanent programming of cardiovascular and metabolic functions, and are
predicted to determine final health outcomes in adults.
17.5.2 FUNDING NUTRIGENOMIC RESEARCH

IN DEVELOPING COUNTRIES
Obtaining funds for nutrigenomic research efforts in developing countries will

not be easy. The 10/90 gap describes the current biomedical funding scheme in
the world: only 10% of the U.S. research budget is spent on conditions and
diseases that affect 90% of the worlds population (93). This gap is narrowing
only because of the rapid increase in obesity and T2DM in developing countries
and a significant percentage of the Wests research funds are spent on those
diseases, although, in populations where the genetic makeups are not reflective
of the worlds diversity.
Because analyzing genenutrient interactions responsible for chronic diseases
requires inclusion of many different genetic makeups, the increased funding within
developed countries may soon enable inclusion of scientists and citizens of devel-
oping countries to participate in the research and its applications. Genetic testing
costs are continually falling and economies of scale may allow the analyses of
individuals in different populations at only incrementally higher costs. Collaborative
efforts will likely allow the transfer of technologies and information which may aid
these disadvantaged nations economically (9496).
Several different groups are exploring the moral issues of doing research in
resource-poor countries (97) and in developing ethical statements and procedures
for such research. The Nuffield Council on Bioethics (http://www.nuffieldbioeth-
ics.org/) has thoroughly reviewed this topic in 1999 (98), with follow-up discussions
published in 2005 (99). The Joint Centre for Bioethics of the University of Toronto
has provided guidelines for ethics and reasons for using technology and genomics
for science and economic development in disadvantaged countries (95,96,100).
The need for addressing the imbalances in nutrient availability, genetic diversity,
and economic development in developing countries has been more eloquently
expressed by nutritionists (76), geneticists (77), economists (100,101), and ethicists
(9599). Disadvantaged individuals and populations can, and should be, included
in new and ongoing studies through appropriate collaborations with scientists, eth-
icists, public health officials, and institutions in developing countries as can those
most vulnerable subpopulations in developed countries. Science and morality justify
these efforts.
These efforts are similar to the recent description of the new nutrition project
(102104), an elucidation of the precepts in the Giessen Declaration (105). That
declaration calls for the expansion of the definition, concepts, and dimensions of
nutritional research and its applications to include biological, social, and environ-
mental principles and methods. The new nutrition would address personal, popu-
lation, and planetary health through an integrative approach of studying and
understanding food systems. The authors of the Giessen Declaration believe that
nutrition sciences should be, and are rightfully, involved with addressing global food
and nutrition insecurity and inadequacy and the new epidemics of obesity, diabetes,
and other chronic diseases whose incidence is rising in developed and developing
countries. The sentiments of the Giessen Declaration and its accompanying expla-
nations are similar to those expressed by an informal international network of
nutrigenomic and related scientists (39).
17.5.3 ISSUES FACING COLLABORATIVE RESEARCH IN DEVELOPING

COUNTRIES: THE ETHIOPIAN EXAMPLE
Although the case for inclusion and benefit sharing is compelling, the challenges
are many. Conditions in Ethiopia may serve as an example of some of these chal-
lenges. Ethiopias Human Development Index (HDI, http://hdr.undp.org/statistics/
data/country_fact_sheets/cty_fs_ETH.html) is 0.361 (out of 1), which is a measure
of living a long life (life expectancy 47 years in 2003, and falling), being educated
(literacy rates are gender specific but average about 45%), and income (GDP
US$711). Although these statistics alone would skew statistical analyses of most
sociological and biological experiments, the physical and health-care infrastructure
issues are even more significant. Approximately 70% of urban dwellers have access
to safe drinking water with per-liter consumption of about 7 l. Only 15 to 20% of
country or rural inhabitants have access, with some consumption estimates of
between 3 and 12 l per day. The recommended amounts for basic needs of drinking,
culinary, personal hygiene, and other domestic uses is 20 to 50 l per day for rural
and urban residents, respectively (106). Accessibility is defined as the geographic
proximity of safe source of drinking water points within 15 minutes walk or a radius
of 1 km from users homes (106) [Emphasis added]. In Ethiopia, 21% of the rural
population, 84% of the urban dwellers, and 30% of the countrys inhabitants have
access to safe drinking water (called drinking water coverage). These numbers
translate to significant biological consequences:
Children under age 5 have 4 to 7 episodes of diarrhea per year (106)

As many as 92.7% of school-children under the age of 15 in the Lake
Awassa area of Southern Ethiopia had at least one helminthic infection.
The most prevalent parasites were A. lumbricoides, hookworm species, T.
trichura, and S. mansoni.
Of the estimated 5000 street children (no permanent home or sleeping
quarters) in Awassa (Southern Nations, Nationalities, and People Region,
Ethiopia), 42.6% had malaria-like febrile illnesses, 33.1% had respiratory
illness, and 4.5% had diarrheal illnesses. Almost 50% of the children ate
only when meals were available (107); that is, half of the street kids in
Awassa did not know where or when their next meal was going to be
available.
The federal Ethiopian government promises full health care, but only ~52% of the
population has access (Ethiopian Ministry of Health, Annex 1). The number of
physicians per 1000 people is 0.029 in Ethiopia, and 2.3 in the U.S. The point of
this listing is that examining nutrientgene interactions, or simply nutrient require-
ments (76), will require significant planning and adjustments of protocols many
researchers take for granted there are few studies that monitor water intake or
energy expenditure of walking to water sources. Nutrigenomics research cannot
address all of these issues but can, and should, focus on extending its research and
humanitarian efforts to those in developing and transitional countries.

Understanding genenutrient interactions and applying that knowledge for personal
and public health will require strategic collaborations, resource and data sharing,
and joint funding strategies that cross discipline and national boundaries. It is not
likely that nutrigenomics will fulfill its promise of personalized nutrition without
comparing responses of genenutrient interactions across many different genetic
makeups. Genedisease studies have demonstrated the need to study multiple ances-
tral groups (51). The effort to begin those collaborations and international networking
was initiated with the publication of a position statement authored by 89 individuals
from 22 countries (39). From this effort, an international Nutrigenomics Society
(71) is emerging that will encourage scientists from any related discipline to join
and participate. The organization of this democratic, ground-up approach will be
done initially through an international Web site hosted by NuGO (71). The interna-
tional team is committed to including people of all nations and ethnicities in the
research and benefits of that research, in large part to reduce the disparities in health
care and living standards across the world. Significant challenges for this research
effort must be overcome in developed economies the effort requires mainly
political will. For developing countries facing infectious diseases, lack of access to
safe water, food insecurity, poor sanitation and health care, the issues are significantly
greater. Genetic variation is likely to produce differences in requirements of micro-
nutrients needed to maintain health and allow for full attainment of physical and
mental health. Hence, the perspective for nutrigenomics is not only improved
nutrition for individuals in developed countries, but also the potential for evidence-
based improvements of public health for citizens of the world.
ACKNOWLEDGMENTS
The preparation of this manuscript was supported by the National Center for Minority
Health and Health Disparities Center of Excellence in Nutritional Genomics, Grant MD
00222 and from the European Union, EU FP6 NoE Grant, Contract No. CT2004-505944.
KEY READINGS
1. NCMHD Center of Excellence in Nutritional Genomics Web site: http://nutrigenomics.
ucdavis.edu.
2. NuGO Web site: http://www.nugo.org.
3. The Case for Strategic International Alliances to HarnessNutritional Genomics for
Public and Personal Health (39).
4. A variant of the gene encoding leukotriene A4 hydrolase confers ethnicity-specific
risk of myocardial infarction (51).
5. Developmental origins of the metabolic syndrome: prediction, plasticity, and
programming (59).
REFERENCES
1. Lander, ES, Linton, LM, Birren, B et al. 2001. Initial sequencing and analysis of the
human genome. Nature 409: 860921.
2. Venter, JC, Adams, MD, Myers, EW et al. 2001. The sequence of the human genome.
Science 291: 130451.
3. The International HapMap, C. 2005. A haplotype map of the human genome. Nature
437: 12991320.
4. Kaput, J, Noble, J, Hatipoglu, B, Kohrs, K, Dawson, K, and Bartholomew, A. 2007.
Application of Nutrigenomic Concepts to Type 2 Diabetes Mellitus. Nutrition, Metab-
olism, and Cardiovascular Disease, 17: 89103.
5. Kaput, J, Perlina, A, Hatipoglu, B, Bartholomew, A, and Nikolsky, Y. 2007. Nutrigenom-
ics: Concepts and Applications to Clinical Medicine. Pharmacogenomics, 8: 369390.
6. Willett, W, The Pursuit of Optimal Diets: A Progress Report, in Nutritional Genomics.

Discovering the Path to Personalized Nutrition, J Kaput, Rodriguez, RL, Eds., 2006,
John Wiley and Sons: Hoboken, NJ, pp. 3756.
7. Kaput, J and Rodriguez, RL. 2004. Nutritional genomics: the next frontier in the
postgenomic era. Physiol Genomics 16: 16677.
8. Ordovas, JM. 2004. The quest for cardiovascular health in the genomic era: nutrige-
netics and plasma lipoproteins. Proc Nutr Soc 63: 14552.
9. DeBusk, RM, Fogarty, CP, Ordovas, JM et al. 2005. Nutritional genomics in practice:
where do we begin? J Am Diet Assoc 105: 58998.
10. Fogg-Johnson, N and Kaput, J. 2003. Nutrigenomics: An Emerging Scientific Disci-
pline. Food Technology 57: 6167.
11. Ferguson, LR and Kaput, J. 2004. Nutrigenomics and the New Zealand Food Industry.
Food New Zealand (the journal of the New Zealand Institute of Food Science and
Technology): 2936.
12. Hirschhorn, JN, Lohmueller, K, Byrne, E et al. 2002. A comprehensive review of
genetic association studies. Genet Med 4: 4561.
13. Newton-Cheh, C and Hirschhorn, JN. 2005. Genetic association studies of complex
traits: design and analysis issues. Mutat Res 573: 5469.
14. Ioannidis, JP. 2005. Why most published research findings are false. PLoS Med 2:
e124.
15. Cardon, LR and Bell, JI. 2001. Association study designs for complex diseases. Nat
Rev Genet 2: 919.
16. Lander, E and Kruglyak, L. 1995. Genetic dissection of complex traits: guidelines
for interpreting and reporting linkage results [see comments]. Nat Genet 11: 2417.
17. Tabor, HK, Risch, NJ, and Myers, RM. 2002. OPINION: Candidate-gene approaches
for studying complex genetic traits: practical considerations. Nat Rev Genet 3: 3917.
18. Risch, N. 1997. Evolving methods in genetic epidemiology. II. Genetic linkage from
an epidemiologic perspective. Epidemiol Rev 19: 2432.
19. Ordovas, JM and Corella, D. 2004. Nutritional genomics. Annu Rev Genomics Hum
Genet 5: 71118.
20. Kaput, J. 2004. Diet-disease gene interactions. Nutrition 20: 2631.
21. Ordovas, JMaC, D., Gene-Environment Interactions: Defining the Playfield, in Nutri-
tional Genomics. Discovering the Path to Personalized Nutrition, J Kaput, Rodriguez,
RL, Eds. 2006, John Wiley and Sons: Hoboken, NJ. pp. 5776.
22. Kolonel, LN, Altshuler, D, and Henderson, BE. 2004. The multiethnic cohort study:
exploring genes, lifestyle and cancer risk. Nat Rev Cancer 4: 51927.
23. Schuster, GU, Nutrients and Gene Expression, in Nutritional Genomics. Discovering
the Path to Personalized Nutrition, J Kaput, Rodriguez, RL, Eds. 2006, John Wiley
and Sons: Hoboken, NJ. pp. 153176.
24. Guo, SaS, G., Green Tea Polyphenols and Cancer Prevention, in Nutritional Genomics.
Discovering the Path to Personalized Nutrition, J Kaput, Rodriguez, RL, Eds. 2006,
John Wiley and Sons: Hoboken, NJ. pp. 177206.
25. Goldberg, JP, Belury, MA, Elam, P et al. 2004. The obesity crisis: don't blame it on
the pyramid. J Am Diet Assoc 104: 11417.
26. Jequier, E. 2002. Pathways to obesity. Int J Obes Relat Metab Disord 26(Suppl. 2): S127.
27. Rutishauser, IH. 2005. Dietary intake measurements. Public Health Nutr 8: 11007.
28. Corella, D and Ordovas, JM. 2005. Single nucleotide polymorphisms that Influence
lipid metabolism: interaction with dietary factors. Annu Rev Nutr 25: 341390.
29. Corella, D and Ordovas, JM. 2004. The metabolic syndrome: a crossroad for genotype-
phenotype associations in atherosclerosis. Curr Atheroscler Rep 6: 18696.
30. Reynolds, TL, Nemeth, MA, Glenn, KC et al. 2005. Natural variability of metabolites
in maize grain: differences due to genetic background. J Agric Food Chem 53: 100617.
31. Haytowitz, D, Pehrsson, PR, and Holden, JM. 2002. The Identification of Key Foods
for Food Composition Research. J Good Composition Anal 15: 183194.
32. Kim, KH and Sobal, J. 2004. Religion, social support, fat intake and physical activity.
Public Health Nutr 7: 77381.
33. Irwin, M. 2002. Effects of sleep and sleep loss on immunity and cytokines. Brain
Behav Immun 16: 50312.
34. Redwine, L, Hauger, RL, Gillin, JC et al. 2000. Effects of sleep and sleep deprivation
on interleukin-6, growth hormone, cortisol, and melatonin levels in humans. J Clin
Endocrinol Metab 85: 3597603.
35. Gleeson, M, Nieman, DC, and Pedersen, BK. 2004. Exercise, nutrition and immune
function. J Sports Sci 22: 11525.
36. Nieman, DC, Davis, JM, Henson, DA et al. 2003. Carbohydrate ingestion influences
skeletal muscle cytokine mRNA and plasma cytokine levels after a 3-h run. J Appl
Physiol 94: 191725.
37. Nieman, DC, Dumke, CI, Henson, DA et al. 2003. Immune and oxidative changes
during and following the Western States Endurance Run. Int J Sports Med 24: 5417.
38. Nieman, DC, Davis, JM, Brown, VA et al. 2004. Influence of carbohydrate ingestion
on immune changes after 2 h of intensive resistance training. J Appl Physiol 96:
12928.
39. Kaput, J, Ordovas, JM, Ferguson, L et al. 2005. The case for strategic international
alliances to harness nutritional genomics for public and personal health. Br J Nutr
94: 62332.
40. Patterson, N, Hattangadi, N, Lane, B et al. 2004. Methods for high-density admixture
mapping of disease genes. Am J Hum Genet 74: 9791000.
41. Tishkoff, SA and Verrelli, BC. 2003. Patterns of human genetic diversity: implications
for human evolutionary history and disease. Annu Rev Genomics Hum Genet 4:
293340.
42. Jorde, LB and Wooding, SP. 2004. Genetic variation, classification and race. Nat
Genet 36(Suppl. 1): S2833.
43. Enattah, NS, Sahi, T, Savilahti, E et al. 2002. Identification of a variant associated
with adult-type hypolactasia. Nat Genet 30: 2337.
44. Tishkoff, SA, Reed, FA, Ranciaro, A et al. 2007. Convergent adaptation of human
lactase persistence in Africa and Europe. Nat Genet 39: 3140.
45. Chiu, S, Diament, A.L., Fisler, J.S., and Warden, C.H., Gene-Gene Epistasis and Gene
Environment Interactions Influence Diabetes and Obesity, in Nutritional Genomics.
Discovering the Path to Personalized Nutrition, J Kaput, Rodriguez, RL, Eds. 2006,
John Wiley and Sons: Hoboken, NJ. pp. 135152.
46. Moore, JH. 2004. Computational analysis of gene-gene interactions using multifactor
dimensionality reduction. Expert Rev Mol Diagn 4: 795803.
47. Carlborg, O and Haley, CS. 2004. Epistasis: too often neglected in complex trait
studies? Nat Rev Genet 5: 61825.
48. Nakahara, K and Carthew, RW. 2004. Expanding roles for miRNAs and siRNAs in
cell regulation. Curr Opin Cell Biol 16: 12733.
49. Hartman, JL, Garvik, B, and Hartwell, L. 2001. Principles for the buffering of genetic
variation. Science 291: 10014.
50. Hartman, JL, IV, Genetic and Molecular Buffering of Phenotypes, in Nutritional
Genomics. Discovering the Path to Personalized Nutrition, J Kaput, Rodriguez, RL,
Eds. 2006, John Wiley and Sons: Hoboken, NJ. pp. 105134.
51. Helgadottir, A, Manolescu, A, Helgason, A et al. 2006. A variant of the gene encoding
leukotriene A4 hydrolase confers ethnicity-specific risk of myocardial infarction. Nat
Genet 38: 6874.
52. Klos, KLE, Kardia, SLR, Hixson, JE et al., Linkage Analysis of Plasma ApoE in Three
Ethnic Groups: Multiple Genes with Context-Dependent Effects. 2005. p. 157167.
53. Kittles, RA and Weiss, KM. 2003. Race, ancestry, and genes: implications for defining
disease risk. Annu Rev Genomics Hum Genet 4: 3367.
54. Shriver, MD, Mei, R, Parra, EJ et al. 2005. Large-scale SNP analysis reveals clustered
and continuous patterns of human genetic variation. Hum Genomics 2: 8189.
55. Kaput, J. An Introduction and Overview of Nutritional Genomics: Application to
Type 2 Diabetes and International Nutrigenomics, in Nutritional Genomics. Discov-
ering the Path to Personalized Nutrition, J Kaput, Rodriguez, RL, Eds. 2006, John
Wiley and Sons: Hoboken, NJ. pp. 136.
56. Steinmetz, A. 2003. Treatment of diabetic dyslipoproteinemia. Exp Clin Endocrinol
Diabetes 111: 23945.
57. Cordain, L, Eades, MR, and Eades, MD. 2003. Hyperinsulinemic diseases of civilization:
more than just Syndrome X. Comp Biochem Physiol A Mol Integr Physiol 136: 95112.
58. Curtis, J and Wilson, C. 2005. Preventing type 2 diabetes mellitus. J Am Board Fam
Pract 18: 3743.
59. McMillen, IC and Robinson, JS. 2005. Developmental origins of the metabolic
syndrome: prediction, plasticity, and programming. Physiol Rev 85: 571633.
60. Rosenzweig, JL, Weinger, K, Poirier-Solomon, L et al. 2002. Use of a disease severity
index for evaluation of healthcare costs and management of comorbidities of patients
with diabetes mellitus. Am J Manage Care 8: 9508.
61. Florez, JC, Hirschhorn, J, and Altshuler, D. 2003. The inherited basis of diabetes
mellitus: implications for the genetic analysis of complex traits. Annu Rev Genomics
Hum Genet 4: 25791.
62. Roweis, ST and Saul, LK. 2000. Nonlinear dimensionality reduction by locally linear
embedding. Science 290: 23236.
63. Tenenbaum, JB, de Silva, V, and Langford, JC. 2000. A global geometric framework
for nonlinear dimensionality reduction. Science 290: 231923.
64. Dawson, K, Rodriguez, RL, and Malyj, W. 2005. Sample phenotype clusters in high-
density oligonucleotide microarray data sets are revealed using Isomap, a nonlinear
algorithm. BMC Bioinformatics 6: 195.
65. Kaput, J and Dawson, K. 2007. Complexity of type 2 diabetes mellitus data sets
emerging from nutrigenomic research: a case for dimensionality reduction? Mutation
Research. In press.
66. Hood, L, Heath, JR, Phelps, ME et al. 2004. Systems biology and new technologies
enable predictive and preventative medicine. Science 306: 6403.
67. Bijlsma, S, Bobeldijk, I, Verheij, ER et al. 2006. Large-scale human metabolomics
studies: a strategy for data (pre-) processing and validation. Anal Chem 78: 56774.
68. Gibney, MJ, Walsh, M, Brennan, L et al. 2005. Metabolomics in human nutrition:
opportunities and challenges. Am J Clin Nutr 82: 497503.
69. van Ommen, B. 2004. Nutrigenomics: exploiting systems biology in the nutrition and
health arenas. Nutrition 20: 48.
70. Kaput, J. 2007. Nutrigenomics 2006 Update. Clinical Chemistry and Laboratory
Medicine, 45: 279287.
71. Kaput, J, Astley, S, Renkema, M, Ordovas, J, and van Ommen, B. 2006. Harnessing
nutrigenomics: development of web-based communication, databases, resources, and
tools. Genes Nutr. 1: 512.
72. Garosi, P, De Filippo, C, van Erk, M et al. 2005. Defining best practice for microarray
analyses in nutrigenomic studies. Br J Nutr 93: 42532.
73. Fuchs, D, Winkelmann, I, Johnson, IT et al. 2005. Proteomics in nutrition research:
principles, technologies and applications. Br J Nutr 94: 30214.
74. Qi, L, Tai, ES, Tan, CE et al. 2005. Intragenic linkage disequilibrium structure of the
human perilipin gene (PLIN) and haplotype association with increased obesity risk
in a multiethnic Asian population. J Mol Med.
75. Shin, HD, Kim, KS, Cha, MH et al. 2005. The effects of UCP-1 polymorphisms on
obesity phenotypes among Korean female subjects. Biochem Biophys Res Commun
335: 62430.
76. Uauy, R. 2005. Defining and addressing the nutritional needs of populations. Public
Health Nutr 8: 77380.
77. Consortium, IH. 2004. Integrating ethics and science in the International HapMap
Project. Nat Rev Genet 5: 46775.
78. Jiang, YH, Bressler, J, and Beaudet, AL. 2004. Epigenetics and human disease. Annu
Rev Genomics Hum Genet 5: 479510.
79. Lucas, A. 1998. Programming by early nutrition: an experimental approach. J Nutr
128: 401S406S.
80. Gallou-Kabani, C and Junien, C. 2005. Nutritional epigenomics of metabolic syn-
drome: new perspective against the epidemic. Diabetes 54: 1899906.
81. Victora, CG, Morris, SS, Barros, FC et al. 1998. The NCHS reference and the growth
of breast- and bottle-fed infants. J Nutr 128: 11348.
82. Hughes, RG and Lawrence, MA. 2005. Globalization, food and health in Pacific
Island countries. Asia Pac J Clin Nutr 14: 298306.
83. Wilcken, B, Bamforth, F, Li, Z et al. 2003. Geographical and ethnic variation of the
677C>T allele of 5,10 methylenetetrahydrofolate reductase (MTHFR): findings from
over 7000 newborns from 16 areas world wide. J Med Genet 40: 61925.
84. Chen, WJ, Chen, CC, Yu, JM et al. 1998. Self-reported flushing and genotypes of
ALDH2, ADH2, and ADH3 among Taiwanese Han. Alcohol Clin Exp Res 22:
104852.
85. Nakajima, T, Wooding, S, Sakagami, T et al. 2004. Natural selection and population
history in the human angiotensinogen gene (AGT): 736 complete AGT sequences in
chromosomes from around the world. Am J Hum Genet 74: 898916.
86. Darnton-Hill, I, Margetts, B, and Deckelbaum, R. 2004. Public health nutrition and
genetics: implications for nutrition policy and promotion. Proc Nutr Soc 63: 17385.
87. Neel, JV. 1962. Diabetes mellitus: a thrifty genotype rendered detrimental by
progress? Am J Hum Genet 14: 35362.
88. Lieberman, LS. 2003. Dietary, Evolutionary, and Modernizing Influences on the
Prevalence of Type 2 Diabetes. Annu Rev Nutr:
89. Epping-Jordan, JE, Galea, G, Tukuitonga, C et al. 2005. Preventing chronic diseases:
taking stepwise action. Lancet 366: 166771.
90. James, PT, Leach, R, Kalamara, E et al. 2001. The worldwide obesity epidemic. Obes
Res 9(Suppl. 4): 228S233S.
91. Iannotti, LL, Tielsch, JM, Black, MM et al. 2006. Iron supplementation in early
childhood: health benefits and risks. Am J Clin Nutr 84: 126176.
92. McCann, JC and Ames, BN, Is docosahexaenoic acid, an n-3 long-chain polyunsaturated
fatty acid, required for development of normal brain function? An overview of evidence
from cognitive and behavioral tests in humans and animals. 2005. pp. 281295.
93. Research, GFfH, The 10/90 Report on Health Research 2001 2002. 2002, Global
Forum for Health Research, World Health Organization: Geneva, Switzerland.
94. Acharya, T, Daar, AS, Thorsteinsdottir, H et al. 2004. Strengthening the role of
genomics in global health. PLoS Med 1: e40.
95. Daar, AS, Thorsteinsdottir, H, Martin, DK et al. 2002. Top ten biotechnologies for
improving health in developing countries. Nat Genet 32: 22932.
96. Singer, PA and Daar, AS. 2001. Harnessing genomics and biotechnology to improve
global health equity. Science 294: 879.
97. Marshall, PA. 2005. Human rights, cultural pluralism, and international health
research. Theor Med Bioeth 26: 52957.
98. Bioethics, NCo. 2002. The ethics of research related to healthcare in developing
countries: 190.
99. Bioethics, NCo. 2005. The ethics of research related to healthcare in developing
countries: a follow-up Discussion Paper:
100. Sachs, JD and Reid, WV. 2006. Environment. Investments toward sustainable devel-
opment. Science 312: 1002.
101. Sachs, JD, The End of Poverty. Economic Possibilities for Our Time. 2005, New York:
The Penguin Press.
102. Leitzmann, C and Cannon, G. 2005. Dimensions, domains and principles of the new
nutrition science. Public Health Nutr 8: 78794.
103. Beauman, C, Cannon, G, Elmadfa, I et al. 2005. The principles, definition and
dimensions of the new nutrition science. Public Health Nutr 8: 6958.
104. Cannon, G and Leitzmann, C. 2005. The new nutrition science project. Public Health
Nutr 8: 67394.
105. 2005. The Giessen Declaration. Public Health Nutr 8: 7836.
106. Kunie, A and Ahmed, A. 2005. An overview of environmental health status in Ethiopia
with particular emphasis to its organization, drinking water and sanitation: a literature
survey. Ethiop J Health Dev 19: 89103.
107. Sorsa, S, Kidanemariam, T, and Erosie, L. 2002. Health Problems of street children
and women in Awassa, Southern Ethiopia. Ethiop J Health Dev 16: 129137.
108. Churchill, GA, Airey, DC, Allayee, H et al. 2004. The Collaborative Cross, a com-
munity resource for the genetic analysis of complex traits. Nat Genet 36: 11337.
109. Khoury, MJ. 2004. The case for a global human genome epidemiology initiative. Nat
Genet 36: 10271028.
110. Ioannidis, JPA, Gwinn, M, Little, J et al. 2006. A road map for efficient and reliable
human genome epidemiology. Nat Genet 38: 35.
111. Cavanaugh, JA. 2003. IBD International Genetics Consortium: international cooper-
ation making sense of complex disease. Inflamm Bowel Dis 9: 1903.
18 The Future of Foods

Heribert J. Watzke and J. Bruce German
CONTENTS
18.1 Introduction ..................................................................................................261

18.2 The Values of Industrial Foods....................................................................262
18.2.1 Integrating Food Values ...................................................................263
18.2.2 Food Values Meeting Consumer Desires.........................................263
18.3 The Tools of the Future ...............................................................................265
18.3.1 Life Sciences and Genomics............................................................265
18.3.2 Computing and Scientific Knowledge Management.......................266
18.3.3 Consumer Demands in Future .........................................................267
18.3.4 The Technologies for the Future .....................................................267
18.3.5 Analytic Technologies for Health Assessment ................................269
18.3.6 Technologies for Food Production...................................................271
18.3.7 Self-Assembly Processes in Food Biomaterials..............................271
18.4 Applying Consumer Health, Biological Knowledge,
and Advanced Technologies Practice to Personalized Foods .....................272
18.4.1 Individual Health..............................................................................272
18.4.2 Personalizing Foods and Diets ........................................................273
18.5 Personalizing Food: Future Perspectives.....................................................273
18.6 Conclusions ..................................................................................................274
Key Readings.........................................................................................................275
References..............................................................................................................276
18.1 INTRODUCTION
Although not typically appreciated, processed foods are perhaps the most critical
elements to the success of modern civilization. In fact, it is a tribute to the success of
the processed food industry that urbanized societies are largely blissfully unaware of
the multiple values inherent in their foods that make life as we know it possible. The
core values of processing of foods are primarily to convert unstable, unsafe, and
sparingly nutritious agricultural commodities into safe, stable, and completely nourish-
ing foods. Industrialized societies continue to add and combine more values within
the foods that they consume. This chapter will first examine the core values of indus-
trialized foods and then suggest where the most obvious opportunities lie to increase
or add value in the future. Finally, the chapter will examine the advances in science
and technologies that are emerging, or would be necessary, to enable such a future.
261
18.2 THE VALUES OF INDUSTRIAL FOODS

Over the past century, as income levels have risen and societies have been able to
invest in the science and technologies of food production, additional values have
been added to processed foods. Foods are increasingly more convenient, freeing a
larger and larger segment of the population from the need to prepare foods for
immediate consumption. Foods are becoming progressively more delicious. Eating
is a physiologically pleasurable experience, and safe, nourishing foods also enhance
the quality of daily life through enhanced taste, flavor, texture, and even color. With
greater income, society is able to afford increasingly more diverse foods in which
the various elements of food delight (flavor, texture, color, etc.) are simultaneously
enhanced. Foods are becoming more and more nutritious. In the latter half of the
20th century, a growing awareness of the ability of different food components to
alter physiology, metabolism, and immunologic functions has led to the develop-
ment of functional foods in which these health values are explicitly added to
foods targeted at subsets of consumers (e.g., athletes, the infirm, the elderly)
(Clydesdale, 1997).
The processed food industry of today is based on a remarkable depth and breadth
of scientific research and development. Modern processed foods are possible only
because of the detailed knowledge that this scientific activity has produced. This is
certainly in part a detailed compositional description of the agricultural commodities
that underpin foods as well as the compositions of the assembled final products.
Information detailing the compositions of commodities include the amount and types
of different proteins (from plant, animal, and microbial sources), carbohydrates
(glucose, simple sugars, complex polymers of different sugar backbones, and
polysaccharides), lipids (triacylglycerides as fats and oils from plants and animals,
complex membrane lipids, and sterols), nucleic and organic acids, alcohols, and a
wide array of plant secondary metabolites (carotenoids, flavonoids, anthocyanins,
etc.). These are the basic ingredients of foods.
The knowledge of food components also extends to their macroscopic struc-
tures. The processing of basic commodities and their formulation into food products
can alter both the relative composition of the basic molecules and also their cellular
and biological structures. Food processing is designed to reformulate the compo-
sitions of foods, mixing ingredients from diverse commodity resources, and to
remodel and synthesize new structures. Gels, fibers, emulsions, and foams are
highly desirable food structures that are explicitly formed during food processing
and preparation. These structures contribute to the stability, shape, texture, and
organoleptic quality of the final foods. Considerable research has been directed
toward understanding these basic structures, their formation, stability, and organo-
leptic properties (Tolstogustov, 1987).
The basic compositional and descriptive knowledge of commodities now also
extends to detailed genomic sequences of an increasingly complete subset of agri-
cultural commodities (European Plant Science Organization, 2005; Womack, 2005;
McCarthy et al., 2006) and the molecular composition of these plants, animals, and
microorganisms that we convert into the foods that we eat (Catchpole et al., 2005).
Equally important, scientific research is building the functional annotation of these
The Future of Foods 263
components both in terms of their structures, stabilities, and chemical interactions

as raw materials and final foods (Sagalowicz et al., 2006; Tolstogustov, 1986) and
in terms of their nutritional, safety, and toxicological implications for the humans
that finally consume them (German et al., 2006; McCarthy et al., 2006). This
remarkably diverse and predictive scientific knowledge is translated into practice by
the myriad technologies that are used to process commodities into foods. Technologies
that process foods range from careful control of heating and biological fermentations
to highly complex engineering systems that simultaneously deliver pressure, temperature,
mechanical, and biological inputs to high-throughput, multicomponent, sterile
process assembly lines (Bruin and Jongen, 2003).
18.2.1 INTEGRATING FOOD VALUES

Perhaps the most remarkable requirement of industrialized foods is the need to
integrate multiple values within each food. It is not sufficient that a food be just safe
or convenient. Foods must be simultaneously safe, nourishing, stable, convenient,
affordable, and delightful. Furthermore, this requirement extends to any updation of
food products. Any subsequent enhancement in an existing foods value must do so
without compromising the other value assets. This places a great deal of pressure
on all aspects of the food development process. Improving stability and convenience
cannot lead to deterioration of flavor. Nor can improvements in flavor compromise
safety. The first generation of functional foods failed because, although they
contained added health values, the ingredients and technologies necessary to add
these values compromised the perceived values of the foods (taste and convenience).
Hence, within the very competitive food marketplace, they simply failed to attract
consumers away from their existing choices. The next generation of functional foods
recognized this need, and functional health values have been incorporated into foods
that also deliver all of the values that consumers have come to expect.
18.2.2 FOOD VALUES MEETING CONSUMER DESIRES

Innovations in processed foods are thus an intersection of (1) desired consumer
values, (2) molecular scientific knowledge, and (3) efficient industrial technologies
(Figure 18.1). The future will certainly see a continuous series of developments in
Consumer
Demands
Food
Scientic Industrial
Knowledge Technology
FIGURE 18.1 Key inputs to food design, formulation, and process innovations.
Product-centered/Commodity-driven
Product matches expectations
Sensory Nutrition
Functionality Optimization Shelf-Life
STOP
Food Safety
Cost Cost Cost Cost
Packaging Home
Raw Ingredients Body
Processing PRODUCT Storage Preparation
Material Preparation Eects
Distribution Eating
FIGURE 18.2a Current approach to food processing, designed to assemble food products
according to the consumers anticipated view of particular products or product categories.
all three of these areas, with progressive enhancements in the values provided by
foods. Although many trends will be discernible, one unifying theme is the increasing
personalization of foods, and this singular trend toward more personalized food
product propositions will transform all of the values of processed foods.
The basic driving forces of the current industrial food sector are illustrated in
Figure 18.2a. Food production today is a product-centric, commodity-driven indus-
try. With food safety as the critical, nonnegotiable core, the food product itself is
the defining element. Once established in the consumers mind as a viable product,
whether a fast food hamburger or a sweet carbonated beverage, the consumer is
buying a familiar product with known organoleptic properties. The profit drivers all
lie within the ability of competitors to drive the costs of delivering that same product
lower. Nutrition is only considered as an ingredient additive at the end (vitamin-
enriched), without compromising the basic product perception. This is a model that,
while providing food products at spectacular volumes and low prices, is providing
very little net value either to the consumer or to the production chain itself. Agri-
culture is essentially eating itself alive trying to provide cheaper and cheaper food
products, and the consumers are literally eating themselves to death because of the
poor nutritional contents. It is now possible to foresee an entirely different model
(Figure 18.2b) in which the consumer becomes the defining element. In this model
Consumer-centered/Benet-driven
Food delivers benets
Raw Packaging Home Delight

Ingredients Processing PRODUCT Storage Preparation Health
Material
Food Safety
Preparation Distribution Eating Performance
Value Value Value Value Quality & Safety Value
Soft Rening Bio-inspired Processing p.o.s. Production Individualization

Indiv
p.o.s. = point of sales
FIGURE 18.2b Proposed approach to foods for the future, designed around the consumer
and delivering foods that address the specific needs, desires, and aspirations of each consumer
for their individual health and well-being.
the goal is to provide values to the consumer. With the consumer as the goal, many
values such as health, performance, and even enjoyment and delight become
attributes that enhance the overall value proposition to the consumer. In this model,
any stage in the agricultural process that is capable of adding, or enhancing, benefits
to the consumer creates value at that level, from seed producer through farmer
through processor to delivery system. It is critical to such a model that the variation
among consumers be recognized, and thus foods must become more personalized.
Consumers themselves will be the major drivers of the trend to personalization.
The average consumer is increasingly gaining awareness of his or her wants and
needs, and growing scientific knowledge and technological advances will race to
meet them. Examples of this ongoing progressive trend are numerous, yet the
diversity of foods and consumers makes it difficult to predict the future. Nonetheless,
it is possible to categorize innovations within each of the previously stated areas
(desired consumer values, molecular scientific knowledge, and efficient industrial
technologies) to gain a sense of predictability. From the consumer perspective, the
major trend is toward individualization of health. The last decade has seen an
explosion in the personalization of consumer products of virtually all kinds from
computers to entertainment, from transportation to cosmetics (German et al., 2005b).
This trend has been evident in foods, extending to all of the values of foods, but
will increasingly include health as a personal value proposition. Consumers know
more about their own health issues and the potential of foods to alter them. Although
it has been a core value of processed foods to ensure safety, the future safety of
foods will take on a more personal dimension. Foods will include means of making
consumers themselves safer. Initially this will mean more specific control of food
processing and specific labeling (no nuts, gluten-free, etc.). Eventually foods will
incorporate various means of ensuring that however consumers choose to assemble
their diet, risks to safety will be minimized.
18.3 THE TOOLS OF THE FUTURE

18.3.1 LIFE SCIENCES AND GENOMICS
The major growth dimension in science is the developing biological understanding
of virtually all life-forms, because of the genomics revolution and the sequencing
of the entire genome of humans, of hundreds of viruses, bacteria, plants, and animals,
and even entire ecosystems (Miller et al., 2004; McCarthy et al., 2006; Morris,
2006). This biological knowledge of genomes and the rapidly assembling annotation
of their functions are dramatically accelerating the rate at which we are acquiring
a truly mechanistic understanding of human health and disease, and how diet impacts
both. This knowledge is also much more accessible to other fields of research, and
thus this basic genomics knowledge will also empower biotechnologies to interact
with our lives in multiple ways. The scientific understanding of microbial, plant,
and animal physiology, and metabolism will lead to a greater diversity of agricultural
commodities that can serve as input to the food chain (Delmer, 2005). But even greater
innovations will come as new biological applications are brought forward to partic-
ipate in the actual processing of foods, i.e., bioguided processing (Ward et al., 2004).
It is certainly not revolutionary for a wine maker who has used yeast to ferment
grape juice into wine to imagine a life-form participating in the processing of an
agricultural commodity. Nonetheless, the improvements in molecular dexterity, cost,
and control that biological strategies contribute to food processing will lead to a
dramatic increase in biologically based innovations in foods. Many of the compo-
nents of human milk are active catalysts whose benefits to the infant are provided
by the catalytic activity itself (e.g., bile-salt-stimulated lipase makes lipids more
available for absorption). At present such activities are destroyed during processing,
whereas in future they will be retained. Other molecules present in milk are highly
dependent on the subtleties of molecular structure lost during processing, for exam-
ple, the milk fat globule in which surface carbohydrates capable of binding microbial
toxins are lost because of homogenization. Again, processes designed to maintain
these structures will be valuable but likely only to a subset of consumers; hence,
the need to personalize the value chain from processing to product.
18.3.2 COMPUTING AND SCIENTIFIC KNOWLEDGE MANAGEMENT

Technology continues to be driven by computerization, and the proliferation of all
aspects of information technologies is moving rapidly through the food process
industry as the cost of acquiring and managing information plummets. Nonetheless,
the key to the applications of the massive computing power now available to food-
related biotechnologies is to coordinate if not unify the terminologies that are
used for cross-disciplinary studies in food, agriculture, and nutrition (Lange et al.,
2007). Equally important for the successful application of computational mathemat-
ics now bringing remarkable, and genuinely revolutionary, understanding of basic
biological processes (Grigorov, 2005) to improving health through foods is to
standardize the database structures of these same fields so that advances in the larger
fields of biology and medicine can be readily translated into food and nutrition
(Lemay et al., 2007). It is particularly discouraging that health itself remains a poorly
defined concept in scientific terms. Computing power is most effective when auto-
mated across large and dissimilar data sets. However, if data sets are not comparable,
they must be hand-curated, reducing all of the computing power back to the snails
pace of individual comparisons. If we examine fields in which database consolidation
has occurred (communication and satellite imaging, for example), it is possible to
see that applications can not only be made commercially viable (mobile phone
technologies), but astonishing cost-efficiencies are possible (mobile phone technol-
ogies have supplanted hard line technologies in the developing world). The example
of global financial transactions is not just illustrative of what is possible today in
terms of personalized technologies, it is most illuminating by virtue of how little
value is necessary to make it commercially viable. The routine use of electronic
transactions (credit card purchasing) that has spread throughout the world in order
to carry out large financial transactions is now used for on-the-spot payment of a
single cup of coffee. If such a trivial value is sufficient to drive globally accessible
credit card technology into coffee shops, then it is only the imagination of the
scientist that limits what would be possible for foods and health if we could move
to a more standardized and computer-friendly information system.
18.3.3 CONSUMER DEMANDS IN FUTURE

The last 50 years have seen a continuous trend toward providing products to con-
sumers that maximize their ability to make immediate choices about the foods they
wish to eat (Moskowitz et al., 2005). This drive to provide immediate food choices
has literally transformed the competitive food marketplace. Not just convenience,
but safety, stability, packaging, distribution channels, and now nutrition, all have
been driven relentlessly forward by a marketplace catering to individual consumers
wishing to customize their personal eating opportunities. The emerging demand for
nutritious foods is now moving not simply away from foods that are perceived as
unhealthy but the trends are instead toward foods that can improve the personal
health of consumers (Sloane, 2006). This desire is in part being driven by the reality
of the worsening health of the average consumer.
The aging, sedentary population in the Western world is increasingly becoming
overweight, hypertensive, and diabetic, and all of these conditions are accelerated
by poor diets (Alberti, 2001). Importantly, these poor diets are not deficient in
essential nutrients (vitamins, minerals) but are instead unbalanced in terms of cal-
ories and macronutrients (Popkin, 2006). These macronutrient imbalances lead to
chronic dysregulation of normal metabolism within susceptible individuals.
Although such metabolic disorders are eventually devastating to health and lead to
diseases different from deficiencies of essential nutrients, they develop gradually as
endogenous, noncommunicable diseases (Quam et al., 2006). Furthermore, whereas
the first wave of metabolic disease atherosclerosis was caused by an invisible,
progressive accumulation of lipoproteins in artery walls, and thus was not evident
to consumers, obesity, hypertension, and diabetes are readily sensed or measured,
and consumers are vividly aware of their own deteriorating health although they do
not know how to change their health trajectories (Petrovici and Ritson, 2006). This
crisis in health-care problems owing to chronic degenerative diseases is becoming
rapidly evident as the population ages. Even if the formal policies and infrastructures
of public health are unable to mount an effective change that resolves these health
issues, consumers are increasingly demanding that their food provide health protec-
tion. It has not escaped the attention of the food industry that the only significant
growth in food marketing value is occurring in the health-promoting food sector
(Clydesdale, 2004). Consumers are increasingly recognizing the importance of diets
to their health, and will be demanding products and services to meet their desires.
18.3.4 THE TECHNOLOGIES FOR THE FUTURE

The genomics revolution has enhanced not only the knowledge of biology, but has
dramatically altered the way research in the life sciences is conducted. Various aspects
of biology are being discovered at unprecedented rates from the development of
diseases in humans to the production of biofuels by plants and microorganisms (Collins
et al., 2003; Pharkya et al., 2004). Although the initial stages of the biological revolution
in agriculture has been to narrow genetic diversity of crops due largely to the agronomic
success of genetically modified core commodities such as soy and corn, this is not
sustainable in the long term. What our rapidly growing knowledge of plant and animal
genomics portends for the future is the increased domestication of a much larger and
more diverse agriculture. Human agriculture is based on a very limited subset of all
living organisms. The reason for this is that most organisms evolved under the constant
Darwinian selective pressure to avoid being eaten. As a result of this pressure, most
plants elaborated secondary metabolism leading to the production of compounds to
discourage animal predation. Antinutritive, and overtly toxic alkaloids, flavonoids,
protease inhibitors, phytates, etc., render most existing wild plants inedible. However,
as we gain mastery of the genomics of plants and the history of their evolution becomes
clear, it will be possible to selectively eliminate the antinutritional and toxic elements.
With such tools, plant geneticists will be able to dramatically expand the available
genetic stocks for agriculture and increase the diversity of inputs to the food supply.
What the organic food movement of today is struggling to provide as value proposi-
tions, with increased sustainability, environmental protection, and species diversity
will be dramatically enhanced by the ability to understand the genetics of all of the
agricultural players from plants to animals, and even microorganisms.
In addition to making plant diversity an asset to food production, genomics will
increase our biotechnology dexterity with a wide array of microorganisms. Bacteria,
yeasts, and molds represent a very successful traditional means of food processing
providing increased safety, stability, and nutritional quality while modifying many
of the taste, flavor, and textural properties of biomaterials from cultured vegetables,
meats, cheeses, and cereals. The combination of knowledge and biological tools that
microbial genomics is providing will revolutionize the bioprocessing of foods.
Not only bacteria, yeasts, and molds but viruses too will be recruited to the task of
improving food values.
Although agricultural species diversity will increase dramatically, one of the
interesting outcomes of this knowledge explosion involves understanding the diver-
sity within species. Research is revealing all of the ways that humans differ from
each other, one of them certainly being the way different humans respond to diet.
The fields of nutrigenomics and nutrigenetics are rapidly compiling examples of
genetic variation in the normal population that predicts variation in health outcome.
From dietary fat and lipoprotein atherogenicity (Krauss, 2001; Ordovas, 2003) to
dietary carbohydrate and weight loss (Martinez et al., 2003), to dietary vitamins and
risk of birth defects (Whetstine et al., 2002), genetic variation is being demonstrated
to be a key component of the variation in human health that is ascribable to diet.
The most widely used index of metabolic health is the level of circulating lipids in
blood, primarily cholesterol and triglycerides. Just by looking at the mechanisms
by which lipids are cleared from blood, we see a vivid story of the genetic variation
within the human population. Clearance of triglyceride-rich lipoproteins is affected
by polymorphisms at several gene loci, including apoB (Rantala et al., 2000), apoAV
(Lai et al., 2003), apo E (Ostos et al., 1998), and lipoprotein lipase (Mero et al.,
1999). The consequences of these genotypic differences are that individuals can be
assigned to varying responses to, for example, high- and low-fat diets. The size and
atherogenicity of cholesterol-rich, low-density lipoproteins, and their response to
diet are mitigated by heritable genes that can be demonstrated to persist in the
offspring of phenotypic carriers (Krauss, 2001). Again, as a result of defining these
genotypes, individuals who vary in response to diet will eventually be
assigned/guided to lower or higher saturated fat-containing diets.
The success of these emerging studies implies that parts of our health will be
predictable from simple genotypic analyses. However, this is not the only basis on
which individuals in the population differ with respect to responses to diet. We are
all exposed to different environments, (exercise, diets, etc.), and these are now
recognized as increasingly important to the net effects of dietary choices on our
health (Bateson et al., 2004). If multiple variables relate to the response of health
to diet beyond genotype of individuals, then health itself will have to be explicitly
measured using technologies akin to disease diagnostics. However, these assessors
will need to be more accurate and capture the fluctuations in metabolism and
physiology that correspond to variations in aspects of health, not simply the diag-
nostics of disease (German et al., 2005a).
18.3.5 ANALYTIC TECHNOLOGIES FOR HEALTH ASSESSMENT

Currently, diagnostic tests use the measurement of surrogate endpoints, or biomar-
kers of a disease process, to assess the progress of an individual toward a disease
state or to monitor the effectiveness of a treatment. This approach is successful and
becoming increasingly automated for diagnosing many of the diseases of humans
including infection and overt toxicities where a specific molecule or group of
molecules present in blood or tissues are uniquely characteristic of the disease state,
and are otherwise foreign to the matrix being sampled in healthy individuals (Sajda,
2006). Diseases that are caused by an imbalance in endogenous metabolism are
more difficult to diagnose and predict because they lack unambiguous markers and
are, instead, characterized by changes only in the absolute concentration of meta-
bolites that are otherwise always present in the tissue or fluid (Schoenhagen and
Nissen, 2006). In these cases, the absolute, measured concentration of the metabo-
lites must be compared to a database of metabolite concentrations for the entire
population. Even when these databases are assembled and the utility of a single
biomarker for monitoring disease risk is well established (e.g., plasma cholesterol
in atherosclerosis and plasma glucose in diabetes), these single measurements cannot
by themselves identify the biochemical pathway responsible for the problem (e.g.,
in this individual, why is glucose high?). Ideal assessment technologies are more
comprehensive and reflect the spectrum of metabolic states that dictate overall
immediate health as well as predict future health risk. Assessment should be capable
of identifying both the presence of a dysregulation and the responsible molecular
mechanism, and ideally predict a logical strategy for intervening in the process in
each individual (German et al., 2005a). For example, cholesterol levels are now
routinely measured in a substantial fraction of the population as a means of identi-
fying those at risk for developing heart disease. However, the basic technologies
that measure cholesterol, even when ascribing the cholesterol to specific lipoprotein
classes, are unable to distinguish the underlying mechanisms in each individual that
are responsible for the elevated cholesterol. They simply note that the individual is
potentially at risk because of this reading. If the measures of cholesterol were
broadened to include a relatively few additional metabolites, much more information
could be gained as to the underlying cause. For example, it has been well estab-
lished that the basic causes of elevated cholesterol in humans are not all the same.
Some individuals have high levels of circulating cholesterol because their hepatic
and peripheral biosynthesis rates are high (Schmitz and Langmann, 2006). Some
have high cholesterol because they absorb inordinately high levels from the intestine,
and still others have high cholesterol because they secrete less sterol both as cho-
lesterol and bile acids from the liver as bile (Plosch et al., 2005). By measuring
sterol biosynthesis intermediates quantitatively in plasma, relative sterol synthesis
rates can be estimated, thus identifying those who synthesize high levels of choles-
terol; by measuring the immediate products of bile acid synthesis in plasma, those
who produce and excrete less bile can be identified, and by measuring phytosterols
in plasma, those individuals who hyperabsorb sterols can be identified (reviewed in
German et al., 2005a). Now, with each of the different causes of high cholesterol
identified within individuals, appropriate strategies for effective intervention can also
be designed. Cholesterol synthesis drugs are highly effective in those who oversyn-
thesize cholesterol, but are relatively ineffective in those who hyperabsorb. Similarly,
sterol absorption inhibitors are highly effective in those who hyperabsorb sterols.
The new sciences of omics genomics, transcriptomics, proteomics and
metabolomics have emerged as a result of the developments of complete databases
of genomes and their products, and the technologies to simultaneously analyze all,
or significant subsets of all members of a biological class of molecules (genes,
mRNAs, proteins, and metabolites) (Raamsdonk et al., 2001). These tools have
dramatic implications for the routine assessment of health because of their potential
to literally measure a complete picture of functional biology in a tissue or fluid
compartment at once (Nicholson and Wilson, 2003). Omic assessment technologies,
and in particular metabolomics, thus offer a unique opportunity for clinical chemistry
not only to monitor the progress of disease states but also to characterize and define
how to maintain health. With such information on health status, individuals will be
able to gain knowledge of their unique metabolic needs, and most importantly, will
be able to act on these needs with diet as the major intervention. Assessment conducted
routinely (e.g., monthly, yearly) will also provide feedback regarding the success of
all food, drug, and lifestyle changes within a time frame that reinforces them or
redirects them as appropriate. Again, this type of assessment will bring a much more
empowered consumer population to the marketplace for agriculture and its products
foods. Even more importantly, with assessed health status of individuals in place as
a consumer asset, it will be possible for diets to be much more effective in preventing
diseases long before they occur, at which points foods become a competitive alter-
native to drugs in maintaining health and preventing disease.
There is an important economic argument for disease prevention that undermines
most attempts to bring preventive medicine into practice. How much will a consumer
pay for a drug/therapeutic/bioactive substance to prevent a disease that they will
never have? The answer is discouragingly simple, very little. Drugs are perceived
to be a valuable product category precisely because they are designed to cure existing
disease. In prevention, you are currently healthy and by following a particular
regimen of food consumption, behavior, or lifestyle, you are unlikely to develop a
future disease. Hence, the value of any single therapeutic preventive is very low. In
order to be successful, preventive strategies have to be very cost-effective. In fact,
it is precisely because foods are already routinely consumed, carry an existing value
proposition, and can easily deliver multiple biological activities that they are so
appropriate as the engines of prevention.
18.3.6 TECHNOLOGIES FOR FOOD PRODUCTION

The technological advances of the first half of the 20th century saw the burgeoning
science of chemistry yielding a wide range of applications, for example, the petro-
leum chemistry revolutionizing fuel and transportation, and the fertilizer chemistry
driving the green revolution. This was also true for the food processing industry.
From the separation and use of chemicals as ingredients and processing aids that
improved the safety, quality, and stability of agricultural commodities and food
products (acids, gases, emulsifiers, colors flavors, etc.), to the synthesis and enrich-
ment of vitamins for nutrient adequacy, the food processing industry was able to
apply the breakthroughs of chemistry to the improvement of food process operations
and the routine manufacturing of foods. Todays research in food engineering is
actively pursuing various means of expanding the ways in which foods are processed
within the existing unit operations, and with the same goals in mind. There are many
examples of novel technologies that could potentially enhance the cost and effec-
tiveness of existing unit operations in food processing. For example, it is possible
to render pathogenic bacteria inactive by high pressure, and the use of high-pressure
sterilization of processed foods is being studied intensively (Gould, 2000).
Nonetheless, these cannot be considered to be the major focus of todays techno-
logical innovation in foods. The true technological innovation that is emerging is
information itself. The technological developments of computing in the latter half
of the 20th century are driving incredible progress in the acquisition, storage, and
mathematical manipulation of information. Such information is providing levels of
control of food production, processing, distribution, and marketing not previously
possible. Interestingly, information management of health is the largest growth
industry of the worlds largest information highway, the Internet. It is certainly
inevitable that the future will move toward information transfer so effectively that
personal health knowledge becomes part of each individuals daily decision making,
of which diet will be a part. The only questions are how fast and how the food
industry will accelerate this trend.
18.3.7 SELF- ASSEMBLY PROCESSES IN FOOD BIOMATERIALS

The gradual migration of food processing from large centralized factories producing
large numbers of identical products to processing that is capable of assembling
more customized formulations appropriate for a more individualized health prop-
osition will require that foods be assembled in smaller unit operations. Although
it is currently assumed that this will require factory-like operations, this is not
necessarily true. The goals of current processing of ensuring microbial safety while
maintaining basic nutrient composition will broaden to include the added value of
retaining many of the inherent biological properties of the commodities from which
the biomaterials originally derive (Ward et al., 2004). Furthermore, although con-
siderable external energy is input during processing to manipulate food materials
in order to produce multiple phases (foams, emulsions, etc.) as the means to
formulate, stabilize, and texturize food products, many of these desired goals could
be achieved without the macroscopic, high-energy unit operations currently used
today. In fact, to the extent possible, food processing will take advantage of the
properties of the molecules within biomaterials to spontaneously self-assemble into
desired forms (Leser et al., 2006). Self-assembly of biomolecules is not a new
concept invented for the food sciences life itself is based upon it. The most
remarkable property of life is that every organism, from the smallest to the largest,
is ostensibly self-assembled from simple biopolymers. As the sciences of genomics
begin to understand the more structure-specific properties of molecules as gene
products, the self-assembly properties of biomolecules even complex colloidal
properties will similarly be understood with sufficient accuracy to control them (de
Campo et al., 2004). There are already many examples of adapting the natural
properties of biopolymers for the net benefit of food products. Cheese is an obvious
illustration of this principle: the properties of casein micelles are captured for
aggregation into three-dimensional gel networks.
18.4 APPLYING CONSUMER HEALTH, BIOLOGICAL

KNOWLEDGE, AND ADVANCED TECHNOLOGIES
PRACTICE TO PERSONALIZED FOODS
18.4.1 INDIVIDUAL HEALTH
The personalization of health will require three elements: assessment technologies,
health trajectory prediction, and intervention strategies. Assessment was discussed
earlier, and it is clear from the rapid rise of personalized medicine that the technologies
to assess health are emerging for therapeutic medicine and will be immediately
translated into other aspects of health care. However, managing an existing disease
is not the same as projecting the health trajectory of healthy individuals according to
an assessment even of a large number of metabolites. Decades of research were
needed to recognize that high levels of cholesterol in blood were predictive of future
heart disease, that altered hepatic cholesterol metabolism was responsible for high
cholesterol levels in blood, and that intervening in this metabolism through diet or
drugs led to a lower risk of heart disease. Nonetheless, this success provides confi-
dence that it is possible to act on systemic health without having a full mathematical,
systems biology solution for the entire organisms physiology. In essence, health
solutions can be pursued empirically to a large extent, both in populations and in
individuals. This will be possible if (1) assessment technologies are developed and
put in place at high throughput to measure large numbers of individuals, (2) databases
chronicling the outcomes of assessment with health outcomes are constructed and
annotated appropriately, and (3) individual assessment results are able to be compared
with outcomes from the larger databases, and assessment converted to projections
and interventions as appropriate. The advantages of a combined empiricalmechanistic
approach to health assessment and improvement are that we can start now; the
discovery process will find the most common health issues first, thus providing
value to the largest number of individuals immediately, and value can be gained as
both health improvement and commercial development proceed and as results emerge.
18.4.2 PERSONALIZING FOODS AND DIETS

The food processing industry today has achieved unprecedented success in providing
safe and affordable foods based on manufacturing large quantities of relatively stan-
dardized products in factory operations engineered to deliver well-defined and vali-
dated unit operations. The advantages of the economies of scale that are possible
with this process model have led to not only cost of production efficiencies
extending from agricultural commodity production to food product assembly to
market distribution but to dramatic improvements in other aspects of manufactur-
ing, including waste management, environmental protection, and worker safety. This
very visible success would seem to be incompatible with a personalized approach to
food production. In fact, an important unanswered question is whether personalization
of diets according to individual needs for health is even possible both technologically
and within the economic framework of modern society. Although we cannot be certain
how the development of personalizing foods will proceed, it is appropriate to examine
models of existing successes in personalizing consumer products based on science,
technology, and consumer values. Perhaps the most extreme example of employing
very sophisticated technology to deliver a frankly minor consumer value is global
positioning satellite (GPS) localization of golf course distances. At various golf
courses throughout the U.S. today, the carts that conduct individual golfers around
the course are equipped with a GPS-locating system whose only consumer value is
to indicate the distance from the cart (or the device when held in the golfers hand)
to the hole. The technology necessary to this product is truly astonishing. An orbiting
satellite network, using Einsteinian mathematics to triangulate distance coordinates
accurately against the massive database of the earths geographical surface that both
the individual (personal) device on the fairway and the hole in the green can be
measured to within a meters distance. With this personalized golf location system
already in place as an example, it is difficult to argue that a more personalized
approach to dietary guidance for health is either scientifically or technologically
impossible. The question is simply, how will it be done?
18.5 PERSONALIZING FOOD: FUTURE PERSPECTIVES

There are unquestionably important hurdles to the widespread conversion of indus-
trial food processing into a more personalized food delivery system. Traditional food
preparation within families was obviously very personal as each individual within
the family unit ate according to the expertise of the family representative (typically,
the mother) who acted as the diet designer for each family member and processed
raw commodities into foods within family meals. Anecdote and cultural history
combined with close individual inspection were the basis of diet decisions. The
question is thus not whether diets and foods themselves can be personalized, but
how can the modern, highly efficient industrial processing of food, which gained its
value largely by centralizing food processing and achieving economies of scale, be
converted to an industrial system of personal food production. The values safety,
affordability, convenience, and delight that brought industrialized foods such
success will continue to operate. To succeed, food production will invariably need
to maintain all of these values and yet become more personal with respect to health.
First and foremost for any personalization is flexibility in product composition.

No matter what the differences in individual human health problems and require-
ments are, they will be resolved in large part by consuming diets of different
molecular compositions. How then can one imagine that different compositions of
foods, decided now at the stage of factory formulation, can be delivered to different
consumers at the point of their purchase, or even the point of consumption?
One model would be to increase the number and diversity of food products that
are produced by the existing factory manufacturing processes. In such a model, the
customization would be performed at the point of distribution using a system of
guided choices. A customer would employ a personal knowledge device (the hand-
held computer is often cited), and a computer would indicate from the wide range
of specific food choices those that are appropriate for each individual to select.
Although such a model seems attractive in that it would create minimal disruption
in existing manufacturing and distribution, it fails in many respects to deliver on the
potential values that personalized foods could, and should, provide. It is also unlikely
that the added inventory control, dilution of market space, and factory flexibility
necessary to such a system would maintain cost-effectiveness.
An alternative model would be to alter the processes of food formulation such
that much of the final compositions would be dictated at the point of purchase,
preparation, or consumption. Such a model would require that intelligent systems
assemble foods much nearer the point at which they are consumed. Although sound-
ing somewhat outrageous for application to existing food products, there are exam-
ples where such a conceptual change has already occurred. Nespresso is a product
category in which a substantial part of the processing of espresso coffee (highly
controlled temperature and pressure extraction of ground beans) is accomplished
within the kitchen or even at the table of the consumer. Different flavor options are
available to personalize the coffee-drinking experience, and the processing is in
large part performed immediately before consumption. In fact, moving from a
one-size-fits-all model to a more plastic and dynamic system of food formulation
would not be limited to compositions for health. All of the consumer food values
that could be discerned as being desirable for each consumer could also be potentially
available. For example, if the composition of a meal/food were being customized
for a single individual, its composition for health (macronutrients, micronutrients,
and bioactive molecules) would be obvious, but so too would its composition for a
more personal preference for flavor, texture, and even convenience. So too would
safety be a customized value, assuring individuals with allergies or intolerances to
particular food components that these have been eliminated.
18.6 CONCLUSIONS
The growing emphasis in the food industry in the 21st century is, and will continue
to be, to deliver safe, convenient, affordable, and delicious food products that provide
the means to assemble diets that maintain and improve the health of consumers. The
power of nutrition research and food processing to solve health problems is illustrated
by the success of identifying the essential nutrients, recognizing their absolute
requirements, and retaining, enriching, and fortifying the food supply to ensure that
deficiency diseases could be prevented. The success is vividly illustrated by the

observation that the population is now unaware of, i.e., has never seen the phenotypes
of diseases caused by classic deficiency diseases: iodinegoiter, vitamin Cscurvy,
and vitamin Ablindness. Through the past three decades, diseases caused by unbal-
anced diets have become literally epidemic across the developed world (Alberti,
2001). These diseases, attributed at least in part to poorly balanced diets, include
atherosclerosis, obesity, diabetes, atherosclerosis, hypertension, and osteoporosis.
Food and nutrition research must now adapt to this new challenge and consider
strategies that can reverse this trend. A combination of research knowledge and food
applications must provide consumers with diets that maintain the value systems of
safety, quality, stability, convenience, and cost. These values are not sufficient; diets
must also ensure that each consumer as an individual maintains optimal health within
the lifestyle that he or she chooses to pursue. The growing scientific knowledge of
the diversity of human genetics, lifestyles, and environments is leading to the con-
clusion that humans are sufficiently different that the same diet will not be optimal
for all, and that some form of personal diet is necessary. This realization would
appear to be devastating to the current system of industrial food production based
on large factories producing homogeneous products with high tolerances of safety,
quality, and cost. Indeed, the components associated with personalizing diet and
health are substantial. First and foremost, personal human health assessment is
necessary. The obvious first step, measuring cholesterol, proved to be the key to
identifying individuals with high cholesterol and at risk of future heart disease. Now
it will be necessary to assess various other aspects of health and identify those
individuals who are at risk of metabolic disorders (obesity, diabetes, hypertension,
and osteoporosis) and other health implications, including, for example, immuno-
logical issues such as immune senescence, allergy, and inflammation. Second, once
an individual is assessed and the scientific knowledge of what the assessment results
imply in terms of dietary composition is in place, it will be necessary to provide
that individual with foods customized to that composition. Such a customization of
industrial food processing seems at first impractical. Nevertheless, the food industry
is already highly plastic in many aspects. Dispensing technologies are already in
place to apportion sweetener and whiteners to coffees, and hence the expansion of
such capabilities into a first-generation personal dispensing system could be done
immediately. What will drive the rate of arrival and acceptance of these technologies
will be the ingenuity and creativity of scientists, technologists, and engineers
throughout the food industry, and the values that they provide to consumers.
KEY READINGS
Collins FS, Green ED, Guttmacher AE, Guyer MS; U.S. National Human Genome Research Institute.
2003 A vision for the future of genomics research. Nature. 24; 422(6934): 83547.
Morris CE, Sands DC. The breeders dilemma yield or nutrition? Nat Biotechnol. Sep-
tember 2006; 24(9): 107880.
Sturino JM, Klaenhammer TR. Engineered bacteriophage-defence systems in bioprocessing.
Nat Rev Microbiol. May 2006; 4(5): 395404.
REFERENCES
Alberti G., 2001, Noncommunicable diseases: tomorrows pandemics, Bull. WHO 79, 907.
Bateson P., D. Barker, T. Clutton-Brock, D. Deb, B. DUdine, R. A, Foley, P. Gluckman, K.
Godfrey, T. Kirkwood. M. M. Lahr, J. McNamara, N. B. Metcalfe, P. Monaghan, H.
G. Spencer and S. E. Sultan, 2004, Developmental plasticity and human health, Nature
430(6998), 419421.
Bruin S. and ThRG Jongen, 2003, Food process engineering: the last 25 years and the
challenges ahead, Crit. Rev. Food Sci. Food Safety 2, 4281.
Catchpole G. S., M. Beckmann, D. P. Enot, M. Mondhe, B. Zywicki, J. Taylor, N. Hardy, A.
Smith, R. D. King, D. B. Kell D.B., O. Fiehn and J. Draper, 2005, Hierarchical
metabolomics demonstrates substantial compositional similarity between genetically
modified and conventional potato crops, Proc. Natl. Acad. Sci. USA 102, 1445814462.
Clydesdale F. M., 1997, A proposal for the establishment of scientific criteria for health claims
for functional foods, Nutr. Rev. 55, 413422.
Clydesdale F., 2004, IFT Expert Panel Report on Functional Foods: Opportunities and Chal-
lenges, Food Technol. 58, 3540.
Collins F.S., E.D. Green, A.E. Guttmacher and M.S. Guyer; US National Human Genome Research
Institute. 2003 A vision for the future of genomics research. Nature. 24; 422(6934): 83547.
de Campo L., A.Yaghmur, L. Sagalowicz, M.E. Leser, H. Watzke and O. Glatter, Reversible
phase transitions in emulsified nanostructured lipid systems. Langmuir. June 22, 2004;
20(13): 525461.
Delmer D. P., 2005, Agriculture in the developing world: Connecting innovations in plant
research to downstream applications, Proc. Natl. Acad. Sci. USA 102, 1573915746.
European Plant Science Organization, 2005, European plant science: a field of opportunities,
J. Exp. Bot. 56, 1699709.
German J. B, F. L. Schanbacher, B. Lnnerdal, J. Medrano, M. a. McGuire, J. L. McManaman,
D. M. Rocke, T. P. Smith, M. C. Neville, P. Donnelly, M. C. Lange and R. E. Ward,
2006, International milk genomics consortium, Trends Food Sci.Technol. 17, 656661.
German J. B., B. D. Hammock and S. M. Watkins, 2005a, Metabolomics: building on a
century of biochemistry to guide human health, Metabolomics 1, 38.
German J. B., C. Yeretzian and H. J. Watzke, 2005b, Personalizing foods for health and
preference, Food Technol. 58, 2631.
Gould G. W., 2000, Preservation: past, present and future, Br. Med. Bull. 56, 8496.
Grigorov M. G., 2005, Global properties of biological networks, Drug Discov. Today 10,
365372.
Krauss R. M., 2001, Dietary and genetic effects on low-density lipoprotein heterogeneity,
Annu. Rev. Nutr. 21, 283295.
Lai C. Q., E. S. Tai, C. E. Tan, J. Cutter, S. K. Chew, Y. P. Zhu, X. Adiconis and J. M. Ordovas,
2003, The apolipoprotein A5 locus is a strong determinant of plasma triglyceride
concentrations across ethnic groups in Singapore, J. Lipid Res. 44, 23652373.
Lange M. C., D. G. Lemay and J. B. German, 2007, A multi-ontology framework to guide
agriculture and food towards diet and health, J. Sci. Food Agric. (in press).
Lemay D. G., A. M. Zivkovic and J. B. German, 2007, Building the bridges to bioinformatics
in nutrition research, Am. J. Clin. Nutr. 13 Apr 2007, online, 10.1002/jsfa. 2832.
Leser M. E., L. Sagalowicz, M. Michel, and H. J. Watzke, Self-assembly of polar food lipids.
Adv. Colloid Interface Sci. 2006 Nov 16; 123126: 12536.
Martinez J. A., M. S. Corbalan, A. Sanchez-Villegas, L. Forga, A. Marti and M. A. Martinez-
Gonzalez, 2003, Obesity risk is associated with carbohydrate intake in women
carrying the Gln27Glu beta2-adrenoceptor polymorphism, J. Nutr. 133, 25492554.
McCarthy F. M., N. Wang, G. B. Magee, B. Nanduri, M. L. Lawrence, E. B. Camon, D. G.

Barrell, D. P. Hill, M. E. Dolan, W. P. Williams, D. S. Luthe, S. M. Bridges and
S. C. Burgess, 2006, AgBase: a functional genomics resource for agriculture, BMC
Genomics 7, 229.
Mero N., L. Suurinkeroinen, M. Syvanne, P. Knudsen, H. Yki-Jarvinen and M. R. Taskinen,
1999, Delayed clearance of postprandial large TG-rich particles in normolipidemic
carriers of LPL Asn291Ser gene variant, J. Lipid Res. 40, 16631670.
Miller W., K. D. Makova, A. Nekrutenko and R. C. Hardison, 2004, Comparative genomics,
Annu. Rev. Genomics Hum. Genet. 5, 1556.
Morris R. M., 2006, Environmental genomics: exploring ecological sequence space, Curr.
Biol. 6, R499501.
Moskowitz, H. R., J. B. German and I. S. Saguy, 2005, Unveiling health attitudes and creating
good-for-you foods: the genomics metaphor, consumer innovative web-based tech-
nologies, Crit. Rev. Food Sci. Nutr. 45,165191.
Nicholson J. K. and I. D. Wilson, 2003, Opinion: understanding global systems biology:
metabonomics and the continuum of metabolism, Nat. Rev. Drug Discov. 2, 668676.
Ordovas J. M., 2003, Cardiovascular disease genetics: a long and winding road, Curr. Opin.
Lipidol. 14, 4754.
Ostos M. A., J. Lopez-Miranda, J. M. Ordovas, C. Marin, A. Blanco, P. Castro, F. Lopez-
Segura, J. Jimenez-Pereperez and F. J, Perez-Jimenez, 1998, Dietary fat clearance is
modulated by genetic variation in apolipoprotein A-IV gene locus, J. Lipid Res. 39,
24932500.
Petrovici D. A. and C. Ritson, 2006, Factors influencing consumer dietary health preventative
behaviours, BMC Public Health 6, 222.
Pharkya P., A.P. Burgard, and C.D. Maranas 2004 OptStrain: a computational framework for
redesign of microbial production systems. Genome Res. 14(11): 236776.
Plosch T., A. Kosters, A. K. Groen and F. Kuipers, 2005, The ABC of hepatic and intestinal
cholesterol transport, Handb. Exp. Pharmacol. 170, 465482.
Popkin B. M., 2006, Global nutrition dynamics: the world is shifting rapidly toward a diet
linked with noncommunicable diseases, Am. J. Clin. Nutr. 84, 289298.
Quam L., R. Smith and D. Yach, 2006, Rising to the global challenge of the chronic disease
epidemic, Lancet, 368, 12211223.
Raamsdonk L. M., B.Teusink, D. Broadhurst, N. Zhang, A. Hayes, M. C. Walsh, J. A. Berden,
K. M. Brindle, D. B. Kell, J. J. Rowland, H. V. Westerhoff, K. van Dam and S. G.
Oliver, 2001, A functional genomics strategy that uses metabolome data to reveal the
phenotype of silent mutations, Nat. Biotechnol. 19, 4550.
Rantala M., T. T. Rantala, M. J. Savolainen, Y. Friedlander and Y. A. Kesaniemi, 2000,
Apolipoprotein B gene polymorphisms and serum lipids: meta-analysis of the role
of genetic variation in responsiveness to diet, Am. J. Clin. Nutr. 71, 713724.
Sagalowicz L., M. E. Leser, H. J. Watzke and M. Michel, 2006, Monoglyceride self-assembly
structures as delivery vehicles, Trends Food Sci. Technol. 17, 204214.
Sajda P., 2006, Machine learning for detection and diagnosis of disease, Annu. Rev. Biomed.
Eng. 8, 537565.
Schmitz G. and T. Langmann, 2006, Pharmacogenomics of cholesterol-lowering therapy,
Vascul. Pharmacol. 44, 7589.
Schoenhagen P. and S. E. Nissen, 2006, Identification of the metabolic syndrome and imaging
of subclinical coronary artery disease: early markers of cardiovascular risk, J. Car-
diovasc. Nurs. 21, 291297.
Sloane E. A., 2006, Top ten functional food trends, Food Technol. 60, 33.
Tolstoguzov V., 1986, Functional properties of protein-polysaccharide mixtures, in Functional

Properties of Food Marcomolecules, Eds. J. R. Mitchell and D. A. Ledward, Elsevier
Applied Science, London, pp. 385415.
Ward R. E, H. J. Watzke, R. Jimnez-Flores and J. B. German 2004, Bioguided processing:
a paradigm change in food production, Food Technol. 58, 4448.
Whetstine J. R., T. L. Witt and L. H. Matherly, 2002, The human reduced folate carrier gene
is regulated by the AP2 and sp1 transcription factor families and a functional 61-base
pair polymorphism, J. Biol. Chem. 277, 4387343880.
Womack J. E., 2005, Advances in livestock genomics: opening the barn door. Genome Res.
15, 1699705.
Index
A B
Adipocyte fatty-acid-binding protein (-FABP), 20 Beta-carotene, 9697
Adiponectin, 92 Biofluids, 1620
Adipose tissue Biomarker(s), 269
nutrients and, 9394 blood glucose, 67
plasticity, 9293 candidates, 4
remodeling, 97 inflammation, 67
vascularity, 90 insulin resistance, 67
Advanced Foods and Materials Network, 245 profiling, 10
Alcohol, 118, 137 protein, 19
bowel cancer risk and, 80 proteome analysis and discovery of, 1820
dependence, 118, 122, 124125, 135, 222 rheumatoid arthritis, 19
diabetes and, 64, 66 universal, 30
metabolizing genes, 252 Bladder cancer, 96
predictors of consumption, 124 Bowel cancer, 78. See also Colorectal cancer
Alcoholism, 118, 122, 124125, 135, 222 Breast cancer, 96, 157
Alzheimers disease, 19, 49, 51 angiogenesis and, 95
Angiogenesis, 89, 97 antioxidants and, 80
adipose tissue plasticity, 9293 carotenoids and, 82
cancer and, 90, 95 folate intake and, 80
gene expression and, 94 genes and, 53, 78, 79, 137
inhibition, 92, 94, 96 hereditary, 137
main principles, 9095 risk, 230
nitric oxide and, 92 genetic testing for, 216
personalized nutrition and, 9597 wine and, 230
polymorphism and, 97
polyphenols and, 9596
regulation, 92, 94, 97 C
Antioxidant(s), 49, 52
breast cancer and, 80 Calcium, 19, 65, 104, 241
dietary and, 55 Cancer, 4, 8, 49, 53
mitochondrial defense, 80 angiogenesis in, 90, 95
phytochemicals as, 117 biology, 7677
polyphenol, 55 bladder, 96
prostate cancer and, 81 breast, 53, 78, 80, 95, 96, 137, 157, 216, 230,
Antioxidant response element, 52 231 (See also Breast cancer)
ARS Childrens Nutrition Research Center, 245 colon, 51, 75, 95
Arthritis, 27, 110 colorectal, 52, 78, 79, 82, 95, 136, 137
rheumatoid, 19, 50, 51, 95 diagnostics, 1819
Aspirin, colorectal cancer and, 82 diet and, 76, 8082, 170, 222
Assessment technology, 269271 gastric, 52, 78, 83
Atherosclerosis, 20 genes and, 53, 75, 76, 78, 79, 137
angiogenesis and, 95 immune response and, 54
development, 267 incidence, 137
inflammation and, 51 initiation, 51
measure, 37 liver, 140, 141
Athletes, 160163 lung, 53, 78, 95, 96
279
mortality, 76, 116 Collaborative Cross Project, 244

ovarian, 78 Colon cancer, 51, 75, 95
pancreatic, 78 Colonic microflora, 29
polyphenols and, 55 Colorectal cancer, 52, 78, 79
prevention, 7584 angiogenesis and, 95
prostate, 53, 56, 78, 81, 95, 96 aspirin and, 82
proteomics, 19 hereditary non-polyposis, 136
PTC/PROP sensitivity and, 112 lifestyle factors and, 137
risk, 117, 223, 230 prevention, 82
diet and, 170, 222 Comparative universality, 225
diet-gene reactions and, 76, 8082 Consumers, 157160, 213, 267
genetic influences on, 7879 COPD. See Chronic obstructive pulmonary
polymorphisms and, 78, 79, 81 disease (COPD)
selenium and, 56 Copper, 54
survival, 76 Coronary artery disease, 34, 52, 102, 103
nutritional factors in, 8283 Cost-benefit evaluation, 175
treatment, 7584, 96 Crohns disease, 50
Carcinogenesis, 137 Cystic fibrosis, 154
Cardioprotective foods, 154
Cardiovascular disease, 8, 49
age-related, 135 D
chronic inflammation and, 49, 51
diet-gene interaction and, 83 Diabetes, 39, 201, 267, 275
dietary changes and, 34 diagnosis, 20
glucose and, 34 heritability, 138
lipid metabolism and, 35 inflammation and, 49
metabolic profiling in, 27 obesity and, 6, 154, 239
mortality, 116 polymorphism and, 65, 101, 108
programming, 104 PTC/PROP sensitivity and, 122
risk, 3343, 108, 170 risk, 41, 6467, 170, 178
traits, 34 assessing, 269
wine and, 135 type I, 19, 52
Catechins, 55, 95 gene, 53
Chamomile tea, 29 polypeptides in, 19
Chronic inflammation, 4957 type II, 4, 8, 138139, 222
Alzheimers disease and, 49, 51 birth weight and, 102
asthma and, 49 diet and, 9, 138139
cancer and, 49, 51 genotype programming interactions in,
cardiovascular disease and, 49, 51 108110
cytokine release and, 51 inflammation and, 8
diabetes and, 49, 51 oxidative phosphorylation and, 9
dietary modulation, 5356 pathogenesis, 6267
disease and, 51 personalized prevention of, 6168
gene-diet interactions in, 56 prevention, 9, 6169, 93, 95
genetic origins, 5153 risk, 41, 6467
nature of, 4951 individual assessment of, 68
nutrient requirements in, 53 treatment, 93
rheumatoid arthritis and, 51 Diabetic retinopathy, 90, 95
Chronic obstructive pulmonary disease (COPD), Diet-gene interactions, 35, 136, 238
51, 55, 103 cancer risk and, 76, 8082
Cigarette smoking, 28 cardiovascular disease and, 83
Coffee, 28 gene expression and, 238
Cogent Syndicated Genomics Attitudes Trends obesity and, 141
(CGAT) survey, 205206 phenotypical intermediaries of, 134
Cognitive learning theory, 174 in type II diabetes, 138
Index 281
Dietary patterns, characterization of, 2930 cystic fibrosis, 154

DioGenes, 245 diabetes, 39, 201, 267, 275
Disease(s) diagnosis, 20
Alzheimers, 19, 49, 51 heritability, 138
biomarkers for, 10 inflammation and, 49
cancer, 4, 8, 49, 53 obesity and, 6, 154, 239
angiogenesis in, 90, 95 polymorphism and, 65, 101, 108
biology, 7677 PTC/PROP sensitivity and, 122
bladder, 96 risk, 41, 6467, 170, 178
breast, 53, 78, 80, 95, 96, 137, 157, 216, 230, assessing, 269
231 type I, 19, 52
colon, 51, 75, 95 gene, 53
colorectal, 52, 78, 79, 82, 95, 136, 137 polypeptides in, 19
diagnostics, 1819 type II, 4, 8, 138139, 222
diet and, 76, 8082, 170, 222 birth weight and, 102
gastric, 52, 78, 83 diet and, 9, 138139
genes and, 53, 75, 76, 78, 79, 137 genotype programming interactions in,
immune response and, 54 108110
incidence, 137 inflammation and, 8
initiation, 51 oxidative phosphorylation and, 9
liver, 140, 141 pathogenesis, 6267
lung, 53, 78, 95, 96 personalized prevention of, 6168
mortality, 76, 116 prevention, 9, 6169, 93, 95
ovarian, 78 risk, 41, 6467
pancreatic, 78 treatment, 93
polyphenols and, 55 diabetic retinopathy, 90, 95
prevention, 7584 heart, 21
prostate, 53, 56, 78, 81, 95, 96 coronary artery, 34, 52, 102, 103
proteomics, 19 diet and, 117
PTC/PROP sensitivity and, 112 inflammation and, 55
risk, 117, 223, 230 risk, 173, 178, 269
diet and, 170, 222 Huntingtons, 158
diet-gene reactions and, 76, 8082 PTC/PROP sensitivity and, 122
genetic influences on, 7879 ulcerative colitis, 50
polymorphisms and, 78, 79, 81 Doping, 161
selenium and, 56 Drug(s), 142, 165, 221, 270
survival, 76 antidiabetic, 66
nutritional factors in, 8283 compliance, 143
treatment, 7584, 96 costs, 160, 166, 247, 271
cardiovascular, 8, 49 foods vs., 225, 228
age-related, 135 genes as targets for, 138
chronic inflammation and, 49, 51 responsiveness to, 30, 31
diet-gene interaction and, 83 selection, 238
dietary changes and, 34 treatment, 10, 62, 68
glucose and, 34
lipid metabolism and, 35
metabolic profiling in, 27 E
mortality, 116
programming, 104 Early Nutrition Programming Project,
risk, 3343, 108, 170 245246
traits, 34 Eating behavior, genetics of, 127129
wine and, 135 Elaboration likelihood model, 171
chronic, 135136 Electrospray ionization, 15
coronary artery, 34, 52, 102, 103 Endothelial hyperpolarizing factor (EDHF), 95
Crohns, 50 Epigenetic silencing, 76, 78, 137
Ethics, 223232 assay, 35

areas of concern, 225232 blood cell, 9
personalized nutrition for consumer, cancer and, 75, 76
230232 diet-gene interactions and, 238
relationship between food and drugs, differential, 8
225228 disease-specific patterns, 7, 9
relationship between personalized and public fetal, 106
health food policies, 228230 Forkhead box C2, 94
perspectives, 223225 genome changes and altered, 77
European Food Information Resource Network, glucocorticoids as regulators, 107
246 high-fat feeding and, 9
European Nutrigenomics Organisation, 31, 135, inflammation, 8
180, 248 interindividual variation, 7
long-term changes, 107
modification, 80
F nutrition and, 4, 9, 28, 35, 154, 216, 238
patterns, 83
Fatty acid(s)
plasminogen, 93
binding protein, 20, 36, 93
profiling, 4, 8, 10
chaperone, 20
regulation, 77, 107, 239
free, 41
SREBP-1, 95
lipid metabolism and, 93
VEGF, 91
long-chain, 92
vitamins and, 239
metabolism, 35
Gene-nutrient interactions, 35, 136, 238, 245248
obesity and, 6
cancer risk and, 76, 8082
regulated nuclear receptors, 36
cardiovascular disease and, 83
sensor, 6
gene expression and, 238
transcription factors, 35
obesity and, 141
Fingerprint analysis, 14
phenotypical intermediaries of, 134
Finnish Diabetes Risk Score, 68
in type II diabetes, 138
Flavonoids, 95, 96, 262
Gene silencing, 107, 154, 161
Food(s)
Genes-as-destiny concept, 157
chemicals in, 239240
Genetic diversity, 240241
consumer desires and values of, 263265
Genetic marketing, 155
databases, 240
Genetic polymorphism(s), 8, 49
drugs vs., 225, 228
cancer risk and, 78
integrating values, 263
immune processes and, 51
personalization, 273, 274
selenium and, 53, 56
processed, 262265
Genetic testing, 159160
Food biomaterials, 271272
for breast cancer risk, 216
Food composition databases, 240
cost, 144, 208, 253
Food production, 273
issues associated with, 207209
technologies for, 271
emotional, 209
Forkhead box C2, 94
employer discrimination, 208
government discrimination, 208
G insurance discrimination, 207208
moral, 208209
Galactosemia, 154 privacy, 207208
Gastric cancer, 52, 78, 83 location, 211212
Gene doping, 161 misuse of information, 210
Gene expression, 10, 28 moral reservations about, 208
aberrant, 75, 76 nutrigenomics and, 170, 178
in angiogenesis, 9596 roots, 215
angiogenesis regulating factor, 94 Genetically tailored products, 211
animal models, 8 Genistein, 96
Index 283
Genomics, 206209 genes and, 5152

applications role of various nutrients in, 54
barriers to adoption of, 212214 selenium and, 55
awareness and interest, 206209 Individual prevention, 222
interest vs. concerns, 214215 Inflammation
life sciences and, 265266 acute, 50
Giessen Declaration, 253 chronic, 4957
Global positioning satellite locating system, 273 Alzheimers disease and, 49, 51
Glucocorticoids, 107 asthma and, 49
Glucose, 28, 40, 138, 242 cancer and, 49, 51
biomarker of blood, 67 cardiovascular disease and, 49, 51
blood profiles, 20 cytokine release and, 51
cardiovascular disease and, 34 diabetes and, 49, 51
free fatty acids and, 41 dietary modulation, 5356
impaired tolerance (See Diabetes) disease and, 51
intolerance, birthweight and, 102 gene-diet interactions in, 56
metabolism, 36 genetic origins, 5153
pancreatic beta-cell toxicity and, 63 nature of, 4951
synthesis, 6 nutrient requirements in, 53
transporters, 138 rheumatoid arthritis and, 51
uptake, inhibition of, 96 Interleukin(s), 5152, 90
Green tea, 95, 96, 120 Interleukin I, gene variations, 52
Growth hormone, 111 International Food Information Council (IFIC)
survey, 206
Iron, 54, 105, 135
H Isoelectric focusing, 15
HapMap, 238, 251
Health assessment, analytic technologies for,
L
269271
Health belief model, 172 Leptin, 92
Health-care providers, 163165 Li-Fraumeni syndrome, 136
Health communication, threat appraisal in, Liddles syndrome, 136
172174 LipGene Diet, 247
HealthGrain, 246 Liver cancer, 140, 141
Heart disease, 21. See also Cardiovascular disease Lung cancer, 53, 78, 95, 96, 230
coronary artery, 34, 52, 102, 103
diet and, 117
inflammation and, 55 M
risk, 173, 178, 269
Macronutrient imbalance, 267
HEPADIP, 247
MALDI-TOF-MS, 15, 19, 20
Herceptin, 83, 157
Manganese superoxide dismutase (MnSOD),
Heuristic-systematic model, 172
8081
HuGENet, 244
Mass spectrometry, 24
Human Genome Project, 238, 251
Metabolic programming, 101111
Human leukocyte antigen (HLA), 52
fetal nutrition-genotype interaction, 108111
Human Plasma Proteome Project (HPPP), 17
mechanistic basis of, 106107
Human Proteome Project (HPP), 17
Metabolic signatures, 2328
Huntingtons disease, 158
Metabolic syndrome, 8, 101
Hypercholesterolemia, familial, 154
biomarker, 20
birthweight and, 102, 103
prevalence, 116
I
prevention and treatment, 93
IBD Network, 244 symptoms, 242
Immune response, 17, 49 website, 247
Metabolism, 7, 153 Nutrigenetics, 35. See also Nutrigenomics

bone, mapping genetics of, 110 Nutrigenomics, 29, 34, 35, 8997. See also
byproducts, 90 Personalized nutrition
capture of fluctuations in, 269 applications, 151166, 200
cholesterol, 108, 272 athletes and, 160163
description of human, 13 best practices, 250251
dysregulation, 267 blood cell profiling and, 10
endogenous, 2728 chronic inflammation and, 4957
imbalance in, 269 consumer interest, 157160
energy, 17 current controversies, 153
fatty acid, 35 defined, 178
folate, 107 desired benefits, 210
genetic influences, 153 in developing countries, 251254
glucose, 36 existing trends and, 159
interorgan, 16 finding research in developing countries,
lipid, 33, 35, 222 252253
cardiovascular disease and, 35 global biological databases, 251
disordered, maternal undernutrition and, health-care providers and, 163165
105 humanitarian need, 252
fatty acids and, 93 issues facing collaborative research,
genetic diet interactions and, 42, 44 253254
plant extracts and, 41 molecular nutrition and, 5
PPARs and, 36 objectives, 154, 170
postprandial, 43 personalized nutrition and, 187 (See also
protein, 17 Personalized nutrition)
proteomics and, 21 promise, 154155, 157
transcription factor PPAR and, 6 promoting, organizing and communication
Metabolomic(s) interest, 250
applications, 2930 public health promise, 156, 232
defined, 24 research, 249
in personalized nutrition, 31 road map for harnessing, 249251
phenotype and, 28 society, 249251
profiles, 23, 27, 31 study design, 7, 8
diet and, 24 tests, 153
plasma, 28 Nutritional compliance, 165
targeted, 24 Nutritional genomics, 35. See also Nutrigenomics
universal, 24 Nutritional programming, 111
urinary, 28
research, bottlenecks in, 3031
Metabonomics, 24 O
Minerals, 54, 267. See also specific elements
Molecular nutrition, 5 Obesity, 6
Multiple sclerosis, 19 biomarker, 20, 61
childhood, 158
diabetes and, 6, 63, 154, 239
diet-gene interactions and, 141
N genetics of, 3641
NCMHD Center of Excellence in Nutritional incidence, 76, 239
Genomics, 247 inflammatory conditions associated with,
Neurodegenerative disorders, 19, 27 8
Nitric oxide, 92, 96 insulin resistance and, 63
Nuclear magnetic resonance, 24 polymorphism and, 40
Nutrient-activated transcription factors, 5 prevalence, 116
Nutrient-metabolism-genotype dependence, 94 surgical therapy of, 68
Nutrient sensors, 94 visceral, 6, 41, 61
Index 285
Obstructive pulmonary disease, chronic, 51, 55, opportunity for, 209212

103 polymorphism measurement and, 111
Onions, 28 programming and, 111
Osteoarthritis, 27, 110 public health and, 133145
Osteoporosis, 110111, 222, 275 resource credibility, 212
polymorphism and, 110 role of physician in, 211
Ovarian cancer, 78 as self care, 164
structure, 222
taste as gatekeeper of, 115129
P terminology, 215
transcriptomics in, 10
Pancreatic cancer, 78 vitamins in, 210, 211
Partial least squares discriminant analysis Pharmacogenetics Research Network, 244
(PLS-DA), 26 Pharmacogenomics, 215
Personalization Phenotype, metabolomic profiles and, 28
consumer evaluation, 195, 197 Phenylalanine, 136
cost and benefits involved, 192195 Phenylalanine hydroxylase, 136
at interactive process level, 193194 Phenylketonuria, 136
at outcome level, 193 Plant diversity, 268
value to company, 194195 Plant polyphenols, 54, 5556, 57
definitions, 188 Plasma, proteins identified in human, 17
key elements of, 187, 189190 Polymorphism(s), 38, 56, 64
process models, 190192 aflatoxin-metabolizing gene, 140
product, 189 alcohol-metabolizing gene, 252
Personalized marketing, 155 angiogenesis and, 97
Personalized nutrition, 10. See also Nutrigenomics cancer and, 78, 79, 81
angiogenesis and, 89, 97 defined, 14
applications, 210212, 215 detection, 14
benefits, 210 diabetes and, 65, 101, 108
cancer prevention and treatment, 7584 glucose levels and, 40
communication related to, 216217 growth hormone gene, 111
consumer attitudes toward, 197, 205217 HLA, 52
critical success factors from marketing view, immune response and, 49, 5153
198201 lactase phlorizin hydrolase gene, 240241
accessibility, 199 measurement, personalized nutrition and,
actionability, 199 111
measurability, 199 nutrient sensors, 94
responsiveness, 199 obesity and, 40
substantive of segments, 199 osteoporosis and, 110
defined, 187 PAH, 136
description, 223 at perilipin locus, 3941
discussion framework, 142144 PTC gene, 123
ethical perspectives, 223225 (See also Ethics) single-nucleotide, 30, 36
ethics, 221232 taste receptor gene, 126
factor and factor levels applied in within transcriptions fractions, 36
scenario-rating tasks, 196 triglyceride levels and, 268
future of, 11, 129, 145, 201202 triglycerides and, 40
individual power and, 211 vitamin D receptor gene, 110
interest in and favorability toward, 209210 Polyphenols, 54, 5556, 57, 9597, 248
metabolic programming, 101111 PPAR transcription factors, 6
metabolomics in, 31 Precaution adoption model, 171
motivations for, 8384, 222 Preemptive regulation, 158
objectives, 165, 222 Pregnancy, metabolic programming in,
opening black box, 223 101111
operational dimensions, 196198 Principle component analysis (PCA), 24, 26
Product personalization, 189 Social judgment theory, 171

Program for Genetic Interaction, 244 Social learning theory, 174
Prostate cancer, 53, 56, 78, 95, 96
antioxidants and, 81
Protection motivation theory, 172 T
Protein(s), 136, 170, 241, 243
Taste, 115129
biomarker, 19
biology of, 116122
C-reactive, 19
genetics of, 115
calcium-binding, 19
modalities, 125127
chip-based sample arrays, 1516, 20
fat, 126127
clusters, 18
salt, 126
diets restricted in, cardiovascular effects of, 104
sour, 126
disabled, 77
sweet and umani, 125126
fatty acid binding, 20, 36, 93
phenylthiocarbamide sensitivity, 117
in human plasma, 17
PTC/PROP sensitivity, 118122
intake, 103
disease and, 122
metabolism, 17
receptors, 124125
origin and nature, 17
Tea, 29, 55, 95, 120
turnover, 28
bitter compounds in, 117
Protein kinase C, 56
tumorigenesis and, 96
Proteome analysis, 1820
Theory of planned behavior, 172
future perspectives, 2021
Thrifty gene hypotheses, 252
urine, 19
Transcription factors, 56
Proteomics
Transcriptomics, 69
cancer, 19
adipose tissue, 8
techniques employed, 1416
blood cell, 9
2D-PAGE, 14
defined, 6
Psoriasis, 51, 95
muscle, 89
PTC gene, 122125
in personalized nutrition, 10
PTC/PROP sensitivity, 118122
studies, 79
disease and, 122
design of, 67
receptors, 124125
requirements for human, 78
Public health, 156, 232, 249251
technology, 6
personalized nutrition and, 133145
Trastuzumab, 83, 157
Public Population Project in Genomics, 244
Trauma, 19, 135
Trimethylamine oxide, 29
Tumor(s), 51, 84
Q
malignant, 76
QT syndrome, 136 molecular portraits, 83
suppressor genes, 76, 77
Tumorigenesis, 51, 90, 96. See also Cancer
R
Regulatory impact analysis, 158
U
Retinoic acid, 5, 97
Retinopathy, diabetic, 90, 95 UAB Center for Nutrient Gene Interactions in
Rheumatoid arthritis, 19, 50, 51, 95 Cancer Prevention, 248
Ulcerative colitis, 50
Universal biomarker, 30
S University of Alabama, 248
SELDI-TOF, 16
Selection perception, 171
V
Selenium, 53, 55, 81
antibody production and, 54 Vascular endothelial growth factor, 90, 91
genetic polymorphisms and, 56 Vitamin(s), 137, 209, 210, 222, 267
Index 287
cystic fibrosis and, 154 W

deficiency, 156, 222, 275
gene expression and, 239 Wine, 55, 95, 117, 135, 227, 230
immune system and, 54 cancer and, 230
in personalized nutrition, 210, 211
Vitamin A, 5, 135, 275
Vitamin C, 54, 55, 156, 275 Z
Vitamin D receptor, 101, 110 Zinc, 54, 105, 135
Vitamin E, 54, 55

Personalized Nutrition

Uploaded by

Copyright:

Available Formats

Personalized Nutrition

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Personalized Nutrition

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What information is provided on the title pages?

What information is provided on the title pages?

9281_C000.

fm Page i Thursday, July 26, 2007 7:30 PM

Boca Raton London New York

CRC Press is an imprint of the

No claim to original U.S. Government works

International Standard Book Number-10: 0-8493-9281-0 (Hardcover)

Library of Congress Cataloging-in-Publication Data

Personalized nutrition : principles and applications / editors, Frans Kok, Laura

Visit the Taylor & Francis Web site at

Chapter 1 Nutrigenomics and Transcriptomics: Applications

Chapter 2 Exploring the Proteome for Markers of Health.................................13

Chapter 3 Metabolomics and the Personalized Metabolic Signature ................23

Chapter 4 Nutrition, Genomics, and Cardiovascular Disease Risk ...................33

Chapter 5 Nutrigenomics and Chronic Inflammation ........................................49

Chapter 6 Personalized Prevention of Type 2 Diabetes.....................................61

Chapter 7 Toward Personalized Nutrition for the Prevention

Chapter 8 Nutrigenomics and Angiogenesis in Obesity ....................................89

Chapter 9 Metabolic Programming during Pregnancy: Implications for

Chapter 10 Taste as the Gatekeeper of Personalized Nutrition .........................115

Chapter 11 Personalized Nutrition and Public Health .......................................133

Chapter 12 Imminent Applications of Nutrigenomics:

Chapter 13 The Personal Factor in Nutrition Communication ..........................169

Chapter 14 A Marketing and Consumer Behavior Perspective

Chapter 15 U.S. Consumer Attitudes toward Personalized Nutrition................205

Chapter 16 Ethics of Personalized Nutrition......................................................221

Chapter 17 International Efforts on Nutrigenomic Health

Chapter 18 The Future of Foods ........................................................................261

1. To what extent is this personal diet-and-health relationship practically

Ben van Ommen

Next, the consequences of metabolic programming during pregnancy and the

Laura Bouwman is a PhD researcher in the department of communication science at

Elizabeth Baily Aldona Dembinska-Kiec

Gerrit J. Hiddink Simon C. Langley-Evans

Jim Kaput Martin Philpott

David B. Schmidt Marianne Walsh

Pieter vant Veer

1.1 Introduction ......................................................................................................3

4 Personalized Nutrition: Principles and Applications

a biomarker profiling approach. The mechanistic approach is a commonly applied

Nutrigenomics and Transcriptomics 5

1.2 NUTRIENTS AND TRANSCRIPTION

mRNA Transcriptomics Disease

FIGURE 1.1 The smart combination of molecular nutrition and nutrigenomics.

6 Personalized Nutrition: Principles and Applications

1.2.2 ROLE OF THE TRANSCRIPTION FACTOR PPAR

1.3.2 TRANSCRIPTOMICS AND STUDY DESIGN

Nutrigenomics and Transcriptomics 7

1.4 APPLICATION OF TRANCRIPTOMICS

8 Personalized Nutrition: Principles and Applications

1.4.2 APPLICATION OF ADIPOSE TISSUE TRANSCRIPTOMICS

1.4.3 APPLICATION OF MUSCLE TRANSCRIPTOMICS

Nutrigenomics and Transcriptomics 9

1.4.4 APPLICATION OF BLOOD CELL TRANSCRIPTOMICS

10 Personalized Nutrition: Principles and Applications

1.4.5 BIOMARKER PROFILING

1.5 HOW CAN TRANSCRIPTOMICS BE USED