Food Security (2019) 11:541–552

https://doi.org/10.1007/s12571-019-00934-x

ORIGINAL PAPER

Heterogeneous factors predict food insecurity among the elderly

in developed countries: insights from a multi-national analysis
of 48 countries
Jae Yeon Park 1,2 & Arlette Saint Ville 1 & Timothy Schwinghamer 2 & Hugo Melgar-Quiñonez 2

Received: 18 October 2018 / Accepted: 1 May 2019 / Published online: 28 May 2019
# International Society for Plant Pathology and Springer Nature B.V. 2019

Abstract
This study aimed to estimate the prevalence of food insecurity and its association with demographic and socioeconomic factors that
characterized elderly survey respondents in 48 developed countries. Cross-sectional data came from the Gallup World Poll for the
years 2014 to 2017 based on respondents aged ≥60 years (n = 53,556). Food insecurity was assessed using the Food Insecurity
Experience Scale, a direct measure of experience-based food insecurity. Descriptive statistics of target populations and the prev-
alence of food insecurity were produced. Demographic variables included age, gender, marital status, household size, education
level, social support, and wellbeing. For socioeconomic variables, income, employment status, urbanicity, and the Community
Basics Index were evaluated. A binary logistic regression model was used and the level of statistical significance was set at 5%.
Respondents aged 60 to 69 years were the most food insecure. Food insecure individuals were more likely to live alone, not have a
partner, and tended to have poorer scores for social support and wellbeing. Also, poor community infrastructure was associated with
food insecurity of elderly people, and there were more food insecure elderly people in urban areas. At the multinational level, results
indicated wide and statistically significant disparities among continents. The study concluded that not only personal factors but also
social conditions could prevent the elderly from achieving full food security status.

Keywords Food insecurity . Food accessibility . Population aging . Elderly . Developed countries

1 Introduction quality of life during these extra years has emerged as one of
the most daunting and costly challenges facing public health
The world is experiencing a situation without precedent, policies (Crimmins and Beltran-Sanchez 2011; Ziliak and
which complicates worsening global food insecurity. As liv- Gundersen 2018). Although several countries have imple-
ing longer does not necessarily connote living healthier, the mented policies and assistance programs to protect health
and food security of citizens, projected trends in population
ageing and a lack of research in the area give rise to concerns
* Jae Yeon Park about the fiscal sustainability of existing health care and pen-
jae.yeon.park@mail.mcgill.ca sion systems to manage the growing number of older adults
(World Health Organization 2015; Canadian Institute for
Arlette Saint Ville
Arlette.saintville@mail.mcgill.ca
Health Information 2017). Additionally, international efforts
have been ongoing to improve conditions faced by the elderly.
Timothy Schwinghamer As food insecurity is inextricably connected with socio-
timothy.schwinghamer@mail.mcgill.ca
economic weaknesses, there has been a common perception
Hugo Melgar-Quiñonez that senior citizens in affluent nations are comfortable with
hugo.melgar-quinonez@mcgill.ca
multiple protections. Such unwarranted assumptions might
1 provide conditions that allow for an unethical acceptability
Department of Human Nutrition, Margaret A. Gilliam Institute for
Global Food Security, McGill University, Sainte-Anne de Bellevue, of the presence of food insecurity among the elderly, especial-
QC H9X 2E5, Canada ly in developed nations. Questions on the food insecurity sta-
2
Department of Plant Science, McGill University, Sainte-Anne de tus of the elderly in developed countries calls for greater ef-
Bellevue, QC H9X 3V9, Canada forts to understand the extent and, if needed, the introduction
542 Park J. et al.

of policies to reduce and prevent this problem (Campbell Table 1 Country economies classification
1991; Hamelin et al. 2008; FRAC 2015). While chronic nu- Threshold GNI/Capita (current US$)
tritional deficiencies are more prevalent in developing coun-
tries, food insecurity has come to be recognized as a rising High-income >12,235
social issue among the elderly in many developed countries Upper-middle income 3956-12,235
(Food and Agriculture Organization 2017). As the need to Lower-middle income 1006-3955
maintain sustainable food security in old age continues to Low-income <1005
grow, many countries may confront crises unless reforms are
Source: World Bank 2017
made and food system targeting improves (Mitchell 1993;
Oliveira and Watanabe 2010).
In recognition of the stated gap, this research aims to ex- 2.1.2 Sociological criteria
amine the food insecurity status of older adults in developed
countries. In doing so, this study is directed towards the de- The Human Development Index (HDI) is a summary measure
tection and generation of empirical data on the factors developed by the United Nations to gauge a country’s level of
influencing food insecurity faced by the elderly. As a result, human development. The higher the HDI the more prosperous
it is hoped that it will contribute to proactive social policy the country is (UNDP 2016). HDI emphasizes that people and
reforms and encourage public sector involvement in their capabilities should be the criteria for assessing the devel-
responding to the problem from the national to local levels. opment of society and it is taken into account under three
dimensions: a long and healthy life, education opportunity,
and a decent standard of living. Countries are grouped into
2 Literature review low-, medium-, high-, and very high-HDI and countries that
have achieved very high HDI are designated as developed.
2.1 Defining the elderly in developed countries
2.2 Current situation of the elderly
Globally there is no official set age that determines when one
is considered an elderly citizen. The criteria commonly chosen Population aging, an inevitable consequence of falling fertility
are chronological, clinical, and socio-demographical which rates and declining mortality rates, is one of the most notable
vary with cultures, eras, and the purposes of the research. global demographic trends of the twenty-first century.
According to the World Health Organization, the chronologi- Regardless of countries’ developmental stages or geographic
cal age of 60 years may be generally used to denote old age, location, the number of persons aged 80 years and over con-
because people at that age tend to start perceiving significant tinues to grow faster than any other age group across the world
life changes with increased major burdens of health care and (United Nations 2015). This global aging process is most ad-
non-communicable diseases (Demirgüç-Kunt et al. 2016; vanced in developed countries. For example, Japan is suffering
World Health Organization 2015). More specific breakdowns from what has been called a Bsuper-aging^ process with 33%
into subgroups, such as ‘young-old’ and ‘middle-old’, may of the population aged 60 years and over in 2017, followed by
help to improve the accuracy of analysis and make more me- Italy (29%), Germany (28%), and Portugal and Finland (27%)
ticulous understanding of the elderly. (United Nations 2017). Adaptation to rapid demographic
The term ‘developed country’, also known as industrialized changes will therefore have to be undertaken much more quick-
country or affluent country, is based on multiple concept clas- ly now than during any other era. This growing proportion of
sifications. It is often used to designate a nation that is more the elderly within national populations brings with it specific
economically advanced with a relatively high level of general concerns because of their unique vulnerabilities.
standard of living and technological infrastructures (UNDP
2016; World Bank 2017). The most commonly used criteria 2.2.1 Vulnerabilities of the elderly in public health
by sector for evaluating a country’s level of development are
described below. The vulnerabilities of the elderly are characterized in public
health policy as: frailty, chronic diseases, mental impairment,
2.1.1 Economic criteria and social challenges. Frailty, in the case of the elderly, can be
defined as the condition of being weak and delicate which
The primary factor used to determine if an economy is consid- often accompanies a number of chronic diseases (Fried et al.
ered either developed or developing is ‘gross national income 2001). Some scientists describe ageing as, Ba slow-motion car
(GNI) per capita’. Using this classification, the World Bank crash – everything just gets wrecked simultaneously^
assigns the world’s economies into four income groups – high, (Kirkwood 2008). When this cumulative decline reaches an
upper-middle, lower-middle, and low (Table 1). aggregate threshold, it can be an indicator of the severity of the
Food insecurity among the elderly in 48 developed countries 543

aging process, physiologic alteration, chronic diseases, im- challenges that they face from difficulties with food available
mune deficiency, or comorbidity over a life span (Clegg to meet their needs, to their gaining access to food, to their
et al. 2013). ability to be involved in preparing, eating and maintaining an
It has been noted that chronic diseases increase with age. adequate diet (Quandt et al. 1998; Quandt and Vitolins 1997).
While there is no evidence to determine inevitable causality Clearly, food insecurity in older adults is a potential risk factor
between aging process and onset of disease, yet age is obvi- for poor nutritional status, unhealthy eating patterns, chronic
ously the biggest risk factor for a broad variety of clinical disease and mental distress (Fernandes et al. 2018).
conditions that often prohibits the elderly from living normal Food insecurity was commonly perceived as a social prob-
lives (Kirkwood 2008). In 2013, WHO estimated the average lem of developing countries. It was widely assumed that the
number of years to be lived in good health as 62 years, with problem related to food insecurity did not take place in devel-
the corresponding value of total life expectancy at birth as oped countries (Campbell 1991; Ishiguro 2014; Strickhouser
71 years. This implies that there is a nine year difference (as et al. 2014). For example, 38% of the elderly population in
of 2013) between good health and life expectancy where the Japan have been struggling to purchase foodstuffs and daily
elderly cannot enjoy a satisfactory life due to disability, frailty, necessities (Ishiguro 2014), and racial and ethnic disparities in
and/or chronic diseases (World Health Organization 2015). health conditions among U.S. older adults has emerged as a
Declines in vision, hearing ability and muscular function are serious social disintegration issue, resulting in wide ranges of
three of the most common symptoms of the ageing process food insecurity (Neff et al. 2009; Institute of Medicine 2011).
(Jaul and Barron 2017). Cardiovascular disease, osteoporosis, While these studies highlight food insecurity issues of concern
dementia, and diabetes are representative types of chronic among the elderly, less is known about food insecurity of the
conditions, commonly found in the elderly (Jaul and Barron elderly in other affluent nations, because previous studies on
2017). The prevalence of multimorbidity increases signifi- food insecurity have focused heavily on children and women
cantly with age and the impact of multimorbidity on quality in developing countries.
of life seems to be much greater than the sum of each effect
(Salive 2013; World Health Organization 2015). 2.3.1 Factors associated with food insecurity of the elderly
The elderly not only suffer from physical decline but also
face mental health issues. Poorer psychological health in older The elderly population presents unique demographic and
populations, particularly in older men, has been identified as a physiological features including functional impairments, eco-
significant problem in terms of prospective negative outcomes nomic insecurity and sensory impairments, which are not nor-
regarding increased utilization of medical care resources, high mally seen in other generations. As a result, this suggests that
rates of suicide risk and the severity of the caregiver burden the risk factors of food insecurity among the elderly should be
(Butcher et al. 2001; Byers and Yaffe 2011; Enache et al. studied separately (Lee and Frongillo 2001) (Fig. 1).
2011). All these issues are interconnected since for a person Although many combined factors affect food insecurity,
to be diagnosed with dementia, he or she can have comorbid financial resources are the most significant risk factor at both
depression (Byers and Yaffe 2011). This suggests that mental individual and community levels. The economic environment
instability can have a negative influence on physical activity such as inflation, unemployment rates, housing price, and tax
and diet quality, and vice versa (Jacka and Berk 2013). rates has considerable impacts on food insecurity (Bickel and
Beyond these functional impairments, the elderly also face Nord 2000; Coleman-Jensen et al. 2016; Cook and Frank
social difficulties. Elderly people frequently experience addi- 2008; Shobe et al. 2018). Food insecurity is most widespread
tional threats to their wellbeing, such as changes in social po- in the lowest-income groups and decreases as income rises
sitions, losing their sense of self-worth, as they lose control of (Strickhouser et al. 2014).
their body and property, or the loss of close relationships due to Health problems contribute to food insecurity in five dis-
death or decline (World Health Organization 2015). These so- tinct ways: (1) limiting access to sufficient food; (2) limiting
cial difficulties may increase their isolation and reduce their the practical use of food; (3) increasing the need for special
ability to find social support during this critical time. nutrients and meal patterns, leading to anxiety about being
able to maintain a constant balance; (4) increasing medical
2.3 Food insecurity of the elderly expenses; and (5) the loss of appetite, resulting in anorexia
of aging (Hall and Brown 2005; Wolfe et al. 1996).
Given the key components of food insecurity: availability, Sociodemographic factors of food security in the elderly
accessibility, utilization and stability, older adults are one of may be broken down roughly into community and individual
the more vulnerable population groups because their uniquely levels. In community-based research, high social capital is
vulnerable position across the four dimensions of food secu- thought of as an indicator of community development and
rity may limit the ability to use food (Hall and Brown 2005). both allow residents to obtain higher quality of food services
Their specific vulnerabilities implies a whole series of and resources (Carter et al. 2012; Cummins et al. 2009; Kay
544 Park J. et al.

Fig. 1 Conceptual framework:

Food insecurity in the elderly.
Source: Wolfe et al. 1996

2005). At the micro-level, having higher education, smaller inclusion criteria for defining developed countries, in this
family size, living with partners, and being retired are all pos- study, was based on the World Bank classification and the
itively associated with food security (Coleman-Jensen et al. Human Development Index by the United Nations
2016; Donald et al. 1998; Hanson et al. 2007; Olayemi 2012). Development Program: 1) high-income economies with a
GNI per capita of $12,476 or more, and 2) graded as ‘very
high Human Development Index’ in the HDI report (UNDP
3 Methodology 2016; World Bank 2017). Chronological age was used to sub-
categorize the elderly into three groups: Byoung-old^ are peo-
This study employed a quantitative cross-sectional research ple in their 60s, Bmiddle-old^ are people in their 70s, and
design, including a multinational-scale measurement, quanti- Boldest-old^ are people more than 80 years old. The number
fication, and comparison of the prevalence of food insecurity. of respondents to the GWP surveys from 2014 to 2017 who
The research was provided as part of a collaboration between were over 60 years old and lived in developed countries was
the McGill University Institute for Global Food Security and 78,100. All individuals with missing data on food insecurity
the Voices of the Hungry Project of the Food and Agriculture and potential predictor variables were not included. Therefore,
Organization of the United Nations. 53,556 individuals were included in the final analysis.

3.1 Objectives 3.3 Food insecurity measurement

The specific objectives of this study were: (1) to examine The prevalence of food insecurity was measured using
macro-level prevalence of food insecurity among the elderly the Food Insecurity Experience Scale Module (FIES), a
in 48 developed countries, and (2) to assess possible risk fac- global standard scale that directly measures the access
tors affecting food insecurity by the elderly in developed dimension of food insecurity, developed by the Food
countries, using cross-nationally comparative methods. and Agriculture Organization (FAO). The FAO carried
out linguistic and cultural adaptations of the FIES before
3.2 Data and study sample surveying nationally representative samples; thus this ap-
proach allows for the identification of national and his-
The data used for this study were collected by Gallup® World torical trends, and comparison of outcomes across coun-
Poll (GWP) from 2014 to 2017, conducted on nationally rep- tries (Food and Agriculture Organization 2016a). The
resentative samples of 1000 adult individuals (aged 15 years FIES is an experience-based metric of severity of food
and older) in over 150 countries. Respondents were selected insecurity, estimated directly from responses to a series
based on probability-based sampling including both rural and of questions regarding food access in their daily lives
urban areas. In the cases of medium- and high-income coun- during the previous 12 months (Ballard et al. 2013). It
tries, interviews were administered via telephone for 30 min is composed of 8 questions which ask people about anx-
and the sampling frame was random-digit-dialing. The data iety and concerns regarding the obtainment of food,
set went through a quality assurance process, including post- followed by compromise in the quantity and the diversity
stratification weights so that projection of results was as close of food intake due to limited money or other resources
as possible to the national population (Gallup Inc. 2018). The (Ballard et al. 2013) (Table 2). The responses are
Food insecurity among the elderly in 48 developed countries 545

Table 2 Items in the food insecurity experience scale survey module

Questions in the Food Insecurity Experience Scale Survey Module for Individuals (FIES SM-1) as fielded in the GWP

During the last 12 months, was there a time when…:

Q1 … you were worried you would not have enough good to eat because of a lack of money or other resources? Worried
Q2 … you were unable to eat healthy and nutritious food because of a lack of money or other resources? Healthy
Q3 … you ate only a few kinds of foods because of a lack of money or other resources? Fewfoods
Q4 … you had to skip a meal because there was not enough money or other resources? Skipped
Q5 … you ate less than you thought you should because of a lack of money or other resources? Ateless
Q6 … your household ran out of food because of a lack of money or other resources? Ranout
Q7 … you were hungry but did not eat because there was not enough money or other resources? Hungry
Q8 … you went without eating for a whole day because of a lack of money or other resources? Whlday

Source: Food and Agriculture Organization 2016a

recorded as simple dichotomous variables (‘yes’ or ‘no’) The Community Basics Index was used to measure
to minimize differences in response styles and to im- overall satisfaction with life regarding environment,
prove the precision of comparability across countries. housing, and community infrastructure. The index is a
Affirmative responses are summed and classified into 4 practical way to evaluate abstract constructs such as the
different food insecurity levels (row score range): food likelihood that one will recommend the neighborhood as
secure (0), mildly food insecure (1–3), moderately food a place to live, or the desirability of infrastructure
insecure (4–6), or severely food insecure (7–8) (Fig. 2) services.
(Food and Agriculture Organization 2016a). The Daily Experience Index was assessed as an indi-
cator of subjective wellbeing on daily health condition,
which estimated individual experience of affective reac-
3.4 Dependent variable
tions against physical and psychosocial health, defined
by Kahneman et al. 1999. It includes a series of ques-
Food insecurity status, measured using the FIES, was
tions related to feelings of pleasure and pain, of joy and
the main dependent variable for this study. An elderly
sorrow, and of satisfaction and dissatisfaction within the
person would be classified as ‘food insecure’ if the re-
whole range of circumstances, from the biological to the
spondent answered at least one of the eight items in the
societal (Kubovy et al. 1999).
affirmative.
The Social Life Index was used to measure perceptions of
respondents on the adequacy of the social support. It is de-
3.5 Independent variables signed to capture a feeling of belonging and an opportunity for
making friends.
Economic variables included: per capita income quin- Every index (Community Basics Index, Daily
tiles. Household income per capita was estimated based Experience Index and Social Life Index) was organized
on monthly income including wages, salaries, remit- into several questions, which had a simple dichotomous
tances from family members, farming, and all other response (Table 3). Each item was scored either 1 for
sources. Respondents were categorized (five categories) yes or 0 for no and all other answers (including nega-
based on per capita income (Gallup Inc. 2018). tive, don’t know and refused). Total scores were com-
Socio-demographic variables included: age groups puted by the mean of valid items multiplied by 100.
(young-old, middle-old, or oldest-old), gender, marital Values for the final analysis ranged from 0 to 100, with
status (living with partner, living alone, or widowed), higher values indicating positive situations. GWP tested
job status (employed, unemployed, or out of the work- Cronbach’s alpha for the reliability of the indices at the
force), highest level of education attained (four-year col- country level. The Community Basics Index, Daily
lege, secondary school, or elementary school), number Experience Index and Social Life Index had scores for
of people living in the household, and urbanicity.

Fig. 2 Definition of food

insecurity levels. Source: Food
and Agriculture Organization
2016b
546 Park J. et al.

Table 3 Description of indices

Indices Index questions

Community basics index • In the city or area where you live, are you satisfied or dissatisfied with the public transportation systems?
• In the city or area where you live, are you satisfied or dissatisfied with the roads and highways?
• In your city or area where you live, are you satisfied or dissatisfied with the quality of air?
• In your city or area where you live, are you satisfied or dissatisfied with the quality of water?
• In your city or area where you live, are you satisfied or dissatisfied with the availability of good affordable housing?
• In the city or area where you live, are you satisfied or dissatisfied with the educational system or the schools?
• In the city or area where you live, are you satisfied or dissatisfied with the availability of quality healthcare?
Daily experience index • Did you feel well-rested yesterday?
• Were you treated with respect all day yesterday?
• Did you smile or laugh a lot yesterday?
• Did you learn or do something interesting yesterday?
• Did you experience the following feelings during a lot of the day yesterday? How about enjoyment?
• Did you experience the following feelings during a lot of the day yesterday? How about physical pain?
• Did you experience the following feelings during a lot of the day yesterday? How about worry?
• Did you experience the following feelings during a lot of the day yesterday? How about sadness?
• Did you experience the following feelings during a lot of the day yesterday? How about stress?
• Did you experience the following feelings during a lot of the day yesterday? How about anger?
Social life index • If you were in trouble, do you have relatives or friends you can count on to help you whenever you need them, or not?
• In the city or area where you live, are you satisfied or dissatisfied with the opportunities to meet people and make friends?

Source: Gallup, Inc. 2017

Cronbach’s alpha of 0.90, 0.72 and 0.65, respectively x1, 2, 3, the independent variables (for this study are: age
(Gallup Inc. 2017). range, region, gender, marital status, urbanicity, household
size, employment, income quintiles, education level,
Community Basics Index, Daily Experience Index, and
3.6 Statistical analysis Social Life Index.)
The statistical analysis was performed using the IBM
Descriptive statistics of the study population and prevalence SPSS® Complex Samples (version 21). Associations
of food insecurity by each variable of interest were analyzed. were considered statistically significant at p < 0.05.
The magnitudes of the associations between food in- All identities of survey participants are strictly confi-
security and independent variables were estimated using dential and approved by the government authorities in
binary logistic regression. This statistical technique is each country. Rigorous research standards and scientifi-
well fitted to examine how a set of predictor variables cally proven methodologies were applied to analyze the
is related to a dichotomous outcome (Osborne 2008). data.
For the data analysis, ‘yes’ and ‘no’ responses to the
FIES items were converted to ones and zeroes, and the
raw FIES score was calculated for each respondent. 4 Results
Then, the logistic regression provides a model of ob-
serving the probability of an individual becoming food 4.1 Descriptive analysis
insecure. Odds ratios were used to quantify the effect of
the covariates which had significant associations with A total of 53,556 elderly individuals were sampled from
food insecurity and the model was adjusted for the co- 48 countries, classified into seven geographic regions.
variate variables listed above. The logistic model equa- Table 4 shows micro levels of food insecurity status
tion is specified explicitly as: by age groups. Oldest-old people were significantly
more likely to be food secure and young-old people
Logit ðpÞ ¼ a þ b1 x1 þ b2 x2 þ b3 x3 þ ⋯ accounted for the highest percentage of both moderate
and severe food insecure categories. The demographic,
Where: socioeconomic and health characteristics of the study
p the probability that an individual will be food insecure population are shown in Table 5. Most of the survey
(food insecure = 1, food secure = 0). respondents were female (56.4%), lived with spouse
a the constant of the equation. (57.7%), lived in urban areas (84.9%) and completed
b1, 2, 3, the coefficient of the independent variables. secondary school (53.0%). About 46.4% of the
Food insecurity among the elderly in 48 developed countries 547

Table 4 Cross-tabulation analysis of food insecurity status and age in Table 5 Demographic and socio-economic characteristics of the elderly
developed countries from 2014 to 2017 in developed countries from 2014 to 2017

Age Food insecurity status (%) Variables n (53,556) % of the sample

n Food Mild food Moderate Severe Age range

secure insecure food food 80 or more (Oldest-old) 6407 12.0
insecure insecure 70 to 79 (Middle-old) 17,918 33.5
60 to 69 years (Young-old) 29,231 54.6
80 or more 6407 5382 726 202 97 Region
(Oldest-old) European Union 32,774 61.2
84.0% 11.3% 3.2% 1.5% EU others 5645 10.5
70 to 79 17,918 14,909 2101 645 263 North America 2352 4.4
(Middle-old) 83.2% 11.7% 3.6% 1.5% Asia 5682 10.6
Australia and New Zealand 3688 6.9
60 to 69 29,231 24,161 3374 1098 598 Latin America and the Caribbean 2079 3.9
(Young-old) 82.7% 11.5% 3.8% 2.0% Middle east 1336 2.5
Linear-by-Linear Association: <0.001 Gender
Male 23,359 43.6
Female 30,197 56.4
Source: Calculated by authors
Marital status
Married/domestic partner 30,887 57.7
elderly lived in a two-person household, followed by Single/never married/separated/divorced 8971 16.7
Widowed 13,698 25.6
35.0% of the elderly living alone, which presented the
Urbanicity
global trends in household size among the elderly Urban 45,488 84.9
population. Rural 8068 15.1
Table 6 shows the percentage distribution of food Household size
1 18,728 35.0
insecurity status among the elderly in developed coun- 2 24,850 46.4
tries by demographic characteristics. Based on this in- 3–4 7590 14.2
formation, in Australia and New Zealand only 7.4% of 5 or more 2388 4.5
Employment
the elderly were food insecure, which was the lowest in Employed 14,200 26.5
the world. The proportion of food insecurity was greater Unemployed 620 1.2
in the female group, the widowed group, and those who Out of the workforce 38,736 72.3
Per Capita Income quintiles
were unemployed. There was slightly higher percentage Richest 20% 13,342 24.9
of food insecurity in rural areas than urban ones but the Fourth 20% 12,374 23.1
difference is meager. Middle 20% 11,734 21.9
Second 20% 9611 17.9
Poorest 20% 1336 2.5
4.2 Adjusted logistic regression Education
College/university 12,011 22.4
Secondary/high school 28,372 53.0
The result of adjusted binary logistic regression allowed Elementary 13,173 24.6
us to observe which contextual variables remained sig- Community basics index
High 33,033 61.7
nificantly associated with food insecurity status among Moderate 14,577 27.2
the elderly (Table 7). Compared to food-secure elderly, Low 5946 11.1
food-insecure elderly were more likely to be poor, com- Daily experience index
plete a lower level of education, and live alone. Also, High 35,657 66.6
Moderate 13,049 24.4
the elderly who were food insecure tended to perceive Low 4850 9.1
their social support as weak, were less satisfied with Social Life Index
their community, and had lower self-rating of wellbeing. High 38,142 71.2
Moderate 12,868 24.0
As age increased, food insecurity rates went down. Low 2546 4.8
For example, people who were young-old were statisti-
cally more likely to be food insecure (OR = 1.7) than Source: Calculated by authors
those in other elderly groups. There was no significant
difference in food insecurity between men and women. indicated that, in large households, the elderly were less
In terms of marital status - single elderly, including likely to be food insecure (OR = 0.6). There was a grad-
never married, separated, and divorced - were more ual increase in odds of food insecurity at the lower
probably experiencing food insecurity (OR = 1.4), as levels of education level and income. Regarding em-
were widowed elderly adults (OR = 1.3). The coefficient ployment status, there was no significant difference in
of household size was estimated positive, which food insecurity between the elderly who were Bout of
548 Park J. et al.

Table 6 Percentages of food insecurity status of the elderly by Table 7 Binary logistic regression analyses of food insecurity on the
demographic characteristics from 2014 to 2017 predictor variables (n = 53,556)

Variables Food security Food Total Variables 95% CI

(%) insecurity (%) (n)
Odds ratio p value
Age range
80 or more (Oldest-old) 84.0 16.0 6407 Age
80 or more (Oldest-old) (Ref.) 1.000
70 to 79 (Middle-old) 83.2 16.8 17,918 70 to 79 (Middle-old) 1.272*** (1.166, 1.388) <0.001
60 to 69 years (Young-old) 82.7 17.3 29,231 60 to 69 years (Young-old) 1.653*** (1.514, 1.805) <0.001
Region Region
European Union (Ref.) 1.000
European Union 82.4 17.6 32,774 EU others 1.089* 0.044
EU others 81.5 18.5 5645 North America 0.849* 0.032
North America 89.7 10.3 2352 Asia 0.848*** <0.001
Australia and New Zealand 0.525*** <0.001
Asia 84.3 15.7 5682 Latin America and the Caribbean 1.679*** <0.001
Australia and New Zealand 92.6 7.4 3688 Middle east 2.191*** <0.001
Latin America and the 72.6 27.4 2079 Gender
Male (Ref.) 1.000
Caribbean
Female 1.033 0.239
Middle east 75.1 24.9 1336
Marital status
Gender Married/domestic partner (Ref.) 1.000
Male 85.6 14.4 23,359 Single/never 1.407*** <0.001
married/separated/divorced
Female 81.0 19.0 30,197
Widowed 1.346*** <0.001
Marital status Urban/rural
Married/domestic partner 86.5 13.5 30,887 Urban (Ref.) 1.000
Rural 1.116** 0.002
Single/never 79.3 20.7 8971 Household size
married/separated/divorced 1 (Ref.) 1.000
Widowed 77.5 22.5 13,698 2 0.719*** (0.662, 0.781) <0.001
Urbanicity 3–4 0.638*** (0.579, 0.703) <0.001
Urban 83.2 16.8 45,488 5 or more 0.619*** (0.542, 0.708) <0.001
Employment
Rural 82.1 17.9 8068 Employed (Ref.) 1.000
Household size Unemployed 2.796*** (2.305, 3.392) <0.001
1 79.3 20.7 18,728 Out of the workforce 1.030 (0.966, 1.100) 0.367
Income quintiles
2 86.4 13.6 24,850 Richest 20% (Ref.) 1.000
3–4 82.5 17.5 7590 Fourth 20% 1.980*** (1.805, 2.172) <0.001
5 or more 77.7 22.3 2388 Middle 20% 2.981*** (2.724, 3.263) <0.001
Second 20% 4.233*** (3.860, 4.643) <0.001
Employment Poorest 20% 5.965*** (5.387, 6.604) <0.001
Employed 87.1 12.9 14,200 Education
Unemployed 56.0 44.0 620 College/university (Ref.) 1.000
Secondary/high school 1.400*** (1.295, 1.515) <0.001
Out of the workforce 81.9 18.1 38,736 Elementary 2.062*** (1.892, 2.248) <0.001
Education Community basics index
High (Ref.) 1.000
College/university 91.5 8.5 12,011
Moderate 1.526*** (1.441, 1.616) <0.001
Secondary/high school 84.3 15.7 28,372 Low 1.922*** (1.784, 2.071) <0.001
Elementary 72.5 27.5 13,173 Daily experience index
High (Ref.) 1.000
Source: Calculated by authors Moderate 2.450*** (2.314, 2.594) <0.001
Low 4.415*** (4.094, 4.761) <0.001
Social life index
the workforce^ and who were employed. Nevertheless, High (Ref.) 1.000
Bunemployed^ seniors were 2.8 times more likely to be Moderate 1.460*** (1.379, 1.546) <0.001
Low 2.318*** (2.102, 2.557) <0.001
food-insecure (OR = 2.8). Community Basics Index
(OR = 1.9), Daily Experience Index (OR = 4.4) and so- The asterisks denote statistical significance at the following levels:
cial support (OR = 2.3) all had positive effect on food *p < 0.05; ** p < 0.01; ***p < 0.001. CI = confidence interval
insecurity among the elderly. Odds ratios were adjusted each of the variables shown
Across the global regions, Oceania (OR = 0.5) and Asia Source: Calculated by authors
(OR = 0.8) were characterized by statistically significantly
Food insecurity among the elderly in 48 developed countries 549

less prevalent food insecurity among the elderly than the negative impact on food accessibility and consumption
European Union members, while food insecurity was more among the elderly in multiple ways. In the same vein,
prevalent in South/Latin America (OR = 1.7) and the Middle social support among the elderly is a particularly key
East (OR = 2.2) than in the European Union. determinant of acquiring appropriate foods for a nutri-
tious diet (Ishikawa et al. 2016). Unfortunately, in con-
trast to this fact, there is a conspicuous increase in the
5 Discussion number of the elderly living alone (United Nations
2017) suggesting that this issue should be of concern
Using data from the GWP, we calculated comparable to welfare policy makers and that additional community
prevalence estimates of food insecurity among the indi- programs are needed to support elderly citizens who
viduals aged 60 years or over from 48 developed coun- live alone. Thus, delivery services by supermarkets,
tries. There were strong differences in individual-level NGOs that provide assistance to go grocery shopping,
food insecurity among global regions. The odds of food and policies that facilitate group trips by the elderly to
insecurity rate among the elderly in Australia and New supermarkets could play important roles in reducing this
Zealand were significantly much lower, even though specific risk factor for the elderly who live alone and
both have been experiencing a large growth in elderly suffer from food insecurity.
population (World Health Organization 2015). Daily experienced wellbeing had a negative associa-
Although health care services and education of the elderly tion with the prevalence of food insecurity among re-
have improved with time, there have been multiple risk factors spondents, which aligned with previous studies. This
that affect food insecurity among the elderly. Our study pro- study was not able to determine whether the relationship
vides insight into the more complex dimensions of food inse- between food insecurity and subjective wellbeing was
curity among the elderly in developed countries, particularly causal, yet, there was a potential for reverse causality
in terms of their ability to access adequate food. between these two variables among elderly persons be-
Consistent with previous studies, we found that respon- cause negative emotions and experience may affect food
dents who were food insecure were more likely to be in the insecurity through high healthcare costs and low moti-
lower income bracket, although there was no clear evidence vation to eat heathily (Lee and Frongillo 2001; Tugade
that employment was associated with food insecurity status et al. 2004).
among the elderly (Lee and Frongillo 2001; Russell et al. Experiences of food insecurity, including disruption
2014; Strickhouser et al. 2014; Fernandes et al. 2018). The of meal quality and worrying about acquiring food, are
strong relationship between income and food security was common among elderly population across countries
confirmed once again. around the world. The significant risk factors of the
Our study found that food insecure elderly respon- problem are likely to occur simultaneously, pushing
dents tended to have lower levels of education. In a the elderly into being much more vulnerable to food
similar context, some studies have found that education insecurity. To assure food security and wellbeing for
level was linked to expenditure for healthy diet the elderly, food security assistance systems should be
(Gucciardi et al. 2009; Kahneman et al. 1999; Public responsive to the heterogeneity and the upward trend of
Library of Science 2013, November 6). These findings the status of elderly adults.
suggest that education influences how the elderly man- In 2002, the Madrid International Plan of Action on
age food quantity as well as dietary quality. Therefore, Ageing, adopted at the Second World Assembly on
policies and social programs in conjunction with food Ageing, drew up a comprehensive action plan for
and nutrition education may reduce the food insecurity Bbuilding a society for all ages^. The Plan highlighted
rate among the elderly. the importance of collaboration between governments,
Overall, 86.5% of elderly persons who lived with a nonprofit organizations, and other actors for success in
spouse were food secure, which is significantly higher ensuring security and dignity of older persons (United
than those in any other marital status. In addition, the Nations 2002). Keeping up with the signs of the times,
study found that respondents in single-person house- this study has produced several important findings to be
holds tended to be the most vulnerable to food insecu- faced by those concernsed about the elderly. In our
rity. These aspects can be explained in more detail by a view, this study has the potential to encourage policy
previous study. Food shopping is an active behavior makers and service providers to think carefully about
which might be difficult to undertake easily by elderly diversity of food security status and risk factors among
people living alone who suffer from a lack of physical the elderly.
functioning and social communication (Thompson et al. Government assistance programs, which can result in in-
2011; Ishikawa et al. 2016). Thus, living alone has a creasing food accessibility of the elderly, would build
550 Park J. et al.

resilience for food and nutrition security, which in turn will Enache, D., Winblad, B., & Aarsland, D. (2011). Depression in dementia:
Epidemiology, mechanisms, and treatment. Current Opinion in
positively impact public health. The prevention of poor health
Psychiatry, 24(6), 461–472.
among the elderly will make a great contribution to individ- Fernandes, S. G., Rodrigues, A. M., Nunes, C., Santos, O., Gregório, M.
uals’ wellbeing as well as social sustainability. If the elderly J., de Sousa, R. D., Dias, S., & Canhão, H. (2018). Food insecurity
live in good health, this unprecedented demographic increase in older adults: Results from the epidemiology of chronic diseases
cohort study 3. Frontiers in Medicine, 5(203). https://doi.org/10.
in the aging population will be regarded not as a burden on
3389/fmed.2018.00203.
society but as growing human resource that might be the key Food and Agriculture Organization (2016a). Methods for estimating com-
to national development. parable rates of food insecurity experienced by adults throughout the
world. Voices of the Hungry. Rome.
Food and Agriculture Organization (2016b). Voices of the hungry:
Compliance with ethical standards Measuring food insecurity through people’s experiences one metric
for the world.
Conflict of interest The authors declare that there is no conflict of Food and Agriculture Organization. (2017). The state of food security and
interest. nutrition in the world 2017. Rome: FAO, IFAD, UNICEF, WFP and
WHO.
FRAC. (2015). A plan of action to end Hinger in America. Washington, DC.
Fried, L. P., Tangen, C. M., Walston, J., Newman, A. B., Hirsch, C.,
Gottdiener, J., et al. (2001). Frailty in older adults: Evidence for a
References phenotype. The Journals of Gerontology. Series A, Biological
Sciences and Medical Sciences, 56(3), M146–M156.
Ballard, T. J., Kepple, A. W., & Cafiero, C. (2013). The food insecurity Gallup Inc (2017). Worldwide research methodology and codebook.
experience scale: Developing a global standard for monitoring hun- Washington, DC.
ger worldwide. Rome: FAO. Gallup Inc (2018). How does the Gallup world poll work? https://www.
Bickel, G., & Nord, M. (2000). Guide to Measuring Hoeusehold Food gallup.com/178667/gallup-world-poll-work.aspx?g_source=link_
Insecurity: Revised 2000. Reports of the Federal Ienteragency Food wwwv9&g_campaign=item_178685&g_medium=copy. Accessed
Security Measurement Project. Alexandria: Food and Nutrition August 18 2018.
Service, USDA. Gucciardi, E., Vogt, J. A., DeMelo, M., & Stewart, D. E. (2009).
Butcher, H. K., Holkup, P. A., & Buckwalter, K. C. (2001). The experi- Exploration of the relationship between household food insecurity
ence of caring for a family member with Alzheimer’s disease. and diabetes in Canada. Diabetes care, 32(12), 2218–2224. https://
Western Journal of Nursing Research, 23(1), 33–55. https://doi. doi.org/10.2337/dc09-0823.
org/10.1177/019394590102300104. Hall, B., & Brown, J. L. (2005). Food security among older adults in the
Byers, A. L., & Yaffe, K. (2011). Depression and risk of developing United States. Topics in Clinical Nutrition, 20(4), 329–338.
dementia. [Review Article. Nature Reviews Neurology, 7(6), 323– Hamelin, A., Mercier, C., & Bédard, A. (2008). Perception of needs and
331. https://doi.org/10.1038/nrneurol.2011.60. responses in food security: Divergence between households and
Campbell, C. C. (1991). Food insecurity: A nutritional outcome or a stakeholders. Public Health Nutrition, 11(12), 1389–1396. https://
predictor variable? Journal of Nutrition, 121(3), 408–415. doi.org/10.1017/S1368980008003406.
Canadian Institute for Health Information. (2017). National health expen- Hanson, K., Sobal, J., & Frongillo, E. (2007). Gender and marital status
diture trends, 1975 to 2017. Ottawa: CIHI. clarify associations between food insecurity and body weight. The
Carter, M. A., Dubois, L., Tremblay, M. S., & aljaard, M. (2012). Local Journal of Nutrition, 137(6), 1460–1465.
social environmental factors are associated with household food Institute of Medicine (2011). Socioeconomic disparities: Food insecurity
insecurity in a longitudinal study of children. BMC Public Health, and obesity. Hunger and Obesity: Understanding a Food Insecurity
12(1038). https://doi.org/10.1186/1471-2458-12-1038. Paradigm: Workshop Summary. Washingeton D.C.
Clegg, A., Young, J., Iliffe, S., Rikkert, M. O., & Rockwood, K. (2013). Ishiguro, K. (2014). Food access among elderly Japanese people. Asian
Frailty in elderly people. The Lancet, 381(9868), 752–762. Social Work and Policy Review, 8(3), 275–279. https://doi.org/10.
1111/aswp.12032.
Coleman-Jensen, A., Matthew, P. R., Christian, A. G., & Anita, S. (2016).
Ishikawa, M., Yokoyama, T., Nakaya, T., Fukuda, Y., Takemi, Y.,
Household food security in the United States in 2015. Economic
Kusama, K., et al. (2016). Food accessibility and perceptions of
Research Service, USDA.
shopping difficulty among elderly living alone in Japan. The
Cook, J. T., & Frank, D. A. (2008). Food security, poverty, and human Journal of Nutrition, Health & Aging, 20(9), 904–911. https://doi.
development in the United States. Annals of the New York Academy org/10.1007/s12603-015-0694-6.
of Sciences, 1136(1), 193–209. Jacka, F. N., & Berk, M. (2013). Depression, diet and exercise. Medical
Crimmins, E. M., & Beltran-Sanchez, H. (2011). Mortality and morbidity Journal of Australia, 199(6 Suppl), S21–S23. https://doi.org/10.
trends: Is there compression of morbidity? Journals of Gerontology: 5694/mja12.10508.
Series B, 66(1), 75–86. https://doi.org/10.1093/geronb/gbq088. Jaul, E., & Barron, J. (2017). Age-related diseases and clinical and public
Cummins, S., Smith, D. M., Taylor, M., Dawson, J., Marshall, D., Sparks, health implications for the 85 years old and over population.
L., & Anderson, A. S. (2009). Variations in fresh fruit and vegetable Frontiers in Public Health, 5(335). https://doi.org/10.3389/fpubh.
quality by store type, urban-rural setting and neighbourhood depri- 2017.00335.
vation in Scotland. Public Health Nutrition, 12(11), 2044–2050. Kahneman, D., Diener, E., & Schwarz, N. (Eds.). (1999). Well-being: the
https://doi.org/10.1017/S1368980009004984. foundations of hedonic psychology. New York, NY, US: Russell
Demirgüç-Kunt, A., Klapper, L., & Panos, G. A. (2016). Saving for Old Sage Foundation.
Age. World Bank. Kay, A. (2005). Social capital, the social economy and community de-
Donald, R., Craig, G., & Victor, O. (1998). Socio-economic determinants velopment. Community Development Journal, 41(2), 160–173.
of food insecurity in the United States: Evidence from the SIPP and Kirkwood, T. B. (2008). A systematic look at an old problem. Nature, 7,
CSFII datasets (F. a. R. E. Division, Trans. Washington, DC: USDA. 451(7179).
Food insecurity among the elderly in 48 developed countries 551

Kubovy, M., Kahneman, D., Diener, E., & Schwarz, N. (1999). Wolfe, W. S., Olson, C. M., Kendall, A., & Frongillo, E. A. (1996).
Wellbeing: The foundations of hedonic psychology. New York: Understanding food insecurity in the elderly: A conceptual frame-
Russell Sage Foundation. work. Journal of Nutrition Education, 28(2), 92–100.
Lee, J. S., & Frongillo, E. (2001). Factors associated with food insecurity World Bank (2017). World development indicators 2017. Washington, DC.
among U.S. elderly persons: Importance of functional impairments. World Health Organization (2015). World report on ageing and health.
The Journals of Gerontology Series B: Psychological Sciences and Geneva.
Social Sciences, 56B(2), S94–S99. Ziliak, J. P., & Gundersen, C. (2018). The state of senior hunger in
Mitchell, O. S. (1993). Retirement Systems in Developed and Developing America 2016: An annual report. Feeding America.
Countries: Institutional features, economic effects, and lessons for econ-
omies in transition. (pp. w4424). National Bureau of Economic
Research.
Neff, R. A., Palmer, A. M., McKenzie, S. E., & Lawrence, R. S. (2009). Jae Yeon Park is an MSc student
Food systems and public health disparities. J Hunger Environ Nutr, in Human Nutrition at McGill
4(3–4), 282–314. https://doi.org/10.1080/19320240903337041. U n i v e r s i t y. S h e h o l d s a
Olayemi, A. O. (2012). Effects of family size on household food security Bachelor ’s degree in Plant
in Osun state. Nigeria. Asian Journal of Agriculture and Rural Bioscience from Kyungpook
Development, 2(2), 136–141. National University in South
Oliveira, L. D. S., & Watanabe, E. A. M. (2010). Assessment of public Korea. Her previous project was
policy food security in countries with different levels development. about functional analysis of mo-
Agroalimentaria, 16(31), 15–29. lecular mechanisms related to en-
vironmental stresses. Currently,
Osborne, J. W. (2008). Best practices in quantitative methods. SAGE
she is studying under the supervi-
Publications, Inc.
sion of Dr. Melgar-Quiñonez fo-
Quandt, S. A., & Vitolins, M. Z. (1997). Meal patterns of older adults in
cusing on the recent trends in so-
rural communities: Life course analysis and implications for under-
cial structure influencing food in-
nutrition. The Journal of Applied Gerontology, 16(2), 152–171.
security in vulnerable popula-
Quandt, S. A., Arcury, T. A., & Bell, R. A. (1998). Self-management of tions. Through this research, she aims to understand the demographic
nutritional risk among older adults: A conceptual model and case and socio-economic factors affecting individual levels of food insecurity.
studies from rural communities. Journal of Aging Studies, 12(4), Furthermore, she is also interested in researching the relationship between
351–368. international agri-food systems and global food security.
Russell, J., Flood, V., Yeatman, H., & Mitchell, P. (2014). Prevalence and
risk factors of food insecurity among a cohort of older australians.
The Journal of Nutrition, Health & Aging, 18(1), 3–8. https://doi.
org/10.1007/s12603-013-0339-6.
Arlette Saint Ville is a Post-
Salive, M. E. (2013). Multimorbidity in older adults. Epidemiologic Doctoral Researcher at McGill
Reviews, 35(1), 75–83. University working with an
Shobe, M. A., Narcisse, M.-R., & Christy, K. (2018). Household financial IDRC-funded project looking at
capital and food security. Journal of Poverty, 22(1), 1–22. sustainable food systems in the
Strickhouser, S., Wright, J. D., & Donley, A. M. (2014). Food insecurity Caribbean. Her research explores
among older adults. Washington, DC: AARP Foundation. how institutions influence food
Thompson, J. L., Bentley, G., Davis, M., Coulson, J., Stathi, A., & Fox, security, the global food and nu-
K. R. (2011). Food shopping habits, physical activity and health- trition security governance archi-
related indicators among adults aged >/=70 years. Public Health tecture, and the interplay between
Nutrition, 14(9), 1640–1649. https://doi.org/10.1017/ food insecurity, governance and
S1368980011000747. conflict. Arlette completed a
Tugade, M. M., Fredrickson, B. L., & Barrett, L. F. (2004). Psychological Bachelor of Environmental
resilience and positive emotional granularity: Examining the bene- Geography degree (Honours) at
fits of positive emotions on coping and health. Journal of Nipissing University, a Master of
Personality, 72(6), 1161–1190. Environmental Studies degree at York University, and PhD at McGill
UNDP. (2016). Human Development Report 2016. New York: UN Plaza. University. Her doctoral research used mixed methods to explain why
United Nations (2002). Political Declaration and Madrid International scientific advances in agriculture, food and environment have not trans-
Plan of Action on Ageing. Paper presented at the second world lated into sustainable food and nutrition security outcomes for the
assembly on ageing, Madrid, 8–12 April 2002. Caribbean as part of the McGill/UWI Caricom Project. Her work has
United Nations (2015). World Population Ageing 2015. Department of appeared in journals such as Food Policy, Food Security, Rural Studies
Economic and Social Affairs, Population Division. and Regional Environmental Change. She worked as an independent
United Nations (2017). World population ageing 2017. Department of international development consultant in the Caribbean for twelve years.
Economic and Social Affairs, Population Division.
552 Park J. et al.

Timothy Schwinghamer is the Dr Melgar-Quiñonez is the

agricultural research biostatisti- Director of the Institute for
cian at the Lethbridge Research Global Food Security. As the
and Development Centre Margaret A. Gilliam Faculty
(Agriculture and Agri-food Scholar in Food Security he holds
Canada). He is the creator of two an appointment in the McGill
packages in R: BsEparaTe^ that School of Dietetics and Human
combines maximum likelihood Nutrition. He obtained a degree
estimation of the parameters of in Medicine (1992) and a doctoral
matrix and third-order tensor nor- degree in Science (1996) from the
mal distributions with unstruc- Friedrich Schiller University,
tured factor variance-covariance Germany. After 9 years’ work as
matrices, and unbiased modified a professor in the Department of
likelihood ratio testing of simple Nutrition at Ohio State University
and double separability for (2003–2012), he moved to
variance-covariance structures; and Brsurface^ that plans rotatable central McGill in 2012. Previously he worked as a researcher in public health
composite design (CCD) experiments and analyzes CCD data. He previ- nutrition at the University of California in Davis (1998–2003) and at the
ously lectured in BQuantitative Methods: Ecology^ at McGill University, Mexican Institute of Public Health (1996–1998). His research program on
a course about methods for ecological data analysis using R. He was a the assessment of household food security includes over 20 countries
research associate (statistician) at the McGill Institute for Global Food around the globe and he maintains a strong collaboration with the
Security (Ste-Anne-de-Bellevue, QC). United Nations Food and Agriculture Organization (FAO).