ORIGINAL ARTICLE

Long-term effects of a Palaeolithic-type diet in obese

postmenopausal women: a 2-year randomized trial
C Mellberg
1,7
, S Sandberg
2,7
, M Ryberg
1,7
, M Eriksson
3
, S Brage
4
, C Larsson
5,6
, T Olsson
1,8
and B Lindahl
2,8
BACKGROUND/OBJECTIVES: Short-term studies have suggested benecial effects of a Palaeolithic-type diet (PD) on body weight
and metabolic balance. We now report the long-term effects of a PD on anthropometric measurements and metabolic balance in
obese postmenopausal women, in comparison with a diet according to the Nordic Nutrition Recommendations (NNR).
SUBJECTS/METHODS: Seventy obese postmenopausal women (mean age 60 years, body mass index 33 kg/m
2
) were assigned to
an ad libitum PD or NNR diet in a 2-year randomized controlled trial. The primary outcome was change in fat mass as measured by
dual-energy X-ray absorptiometry.
RESULTS: Both groups signicantly decreased total fat mass at 6 months ( 6.5 and2.6 kg) and 24 months ( 4.6 and2.9 kg),
with a more pronounced fat loss in the PD group at 6 months (Po0.001) but not at 24 months (P 0.095). Waist circumference and
sagittal diameter also decreased in both the groups, with a more pronounced decrease in the PD group at 6 months ( 11.1
vs 5.8 cm, P 0.001 and3.7 vs 2.0 cm, Po0.001, respectively). Triglyceride levels decreased signicantly more at 6 and 24
months in the PD group than in the NNR group (Po0.001 and P 0.004). Nitrogen excretion did not differ between the groups.
CONCLUSIONS: A PD has greater benecial effects vs an NNR diet regarding fat mass, abdominal obesity and triglyceride levels in
obese postmenopausal women; effects not sustained for anthropometric measurements at 24 months. Adherence to protein intake
was poor in the PD group. The long-term consequences of these changes remain to be studied.
European Journal of Clinical Nutrition advance online publication, 29 January 2014; doi:10.1038/ejcn.2013.290
Keywords: adipose tissue; diet; insulin resistance; postmenopausal; weight
INTRODUCTION
The incidence of obesity has increased substantially since the early
1980s, resulting in major public health challenges.
1,2
Related to this,
dietary risk factors and physical inactivity collectively accounted for
10% of global deaths and disability-adjusted life years in a recent
comparative risk assessment of the burden of disease and injury.
2
Central fat accumulation, that is, the accumulation of visceral
adipose tissue, is clearly associated with an increased risk of type 2
diabetes mellitus (T2DM) and cardiovascular disease (CVD).
3
In menopause, fat is redistributed from peripheral to central
depots, which is associated with an increased incidence of diabetes
and CVD among postmenopausal women.
4
Diets moderately high in protein (2530% of energy intake) have
been suggested to be benecial for weight loss when used ad
libitum, at least up to 12 months.
5
This includes a more pronounced
reduction in body fat and blood pressure and improved lipid prole
when compared with high-carbohydrate diets. However, it is not
clear if an increased protein intake is sustainable during long-term
diet interventions. An increased intake of monounsaturated fats
(MUFA) and omega-3 fatty acids and moderately decreased intake
of carbohydrates (40% of energy intake) may also be benecial with
regard to weight loss as well as have protective effects on CVD.
6,7
Such dietary properties are possible with a Palaeolithic-type diet
(PD), making it seem worthy of randomized controlled trials.
Short-term studies with this type of diet have suggested benecial
effects on energy intake, weight, waist circumference, body mass
index (BMI), and metabolic balance, including insulin sensitivity, as
well as cardiovascular risk markers, even when administered ad
libitum in comparison with other types of diets.
811
Our hypothesis
was that a PD would be more efcient than a conventional low-
fat/high-bre diet
12
at reducing fat mass during a 2-year
randomized dietary intervention trial in obese postmenopausal
women. Furthermore, we wanted to analyze if a PD would have
benecial effects on cardiovascular risk markers.
SUBJECTS AND METHODS
Subjects
The subjects were recruited in 2007 through advertisements in local
newspapers. In total, 210 women expressed an interest in participating in
the study (Figure 1) out of which 70 postmenopausal non-smoking women
with a BMIX27 kg/m
2
fullled the inclusion criteria. Exclusion criteria
included consumption of a restricted or vegetarian diet, allergy to key
components in the intervention diets, history of heart disease, kidney
1
Department of Public Health and Clinical Medicine, Ume University, Ume, Sweden;
2
Department of Public Health and Clinical Medicine, Occupational and Environmental
Medicine, Ume University, Ume, Sweden;
3
Department of Statistics, USBE, Ume University, Ume, Sweden;
4
MRC Epidemiology Unit, University of Cambridge, Cambridge, UK;
5
Department of Food and Nutrition, and Sport Science, University of Gothenburg, Gothenburg, Sweden and
6
Department of Food and Nutrition, Ume University, Ume,
Sweden. Correspondence: C Mellberg, Department of Public Health and Clinical Medicine, Ume University, Ume SE-90185, Sweden.
E-mail: caroline.mellberg@medicin.umu.se
7
These authors share rst authorship.
8
These authors share senior authorship.
Contributors: Study concept and design; acquisition of data; and drafting the manuscript: CM, SS, MR, CL, TO, and BL. Analysis and interpretation of data and critical revision of the
manuscript for important intellectual content: all the authors. Statistical analysis: CM, SS. ME, and BL. Obtained funding: TO, BL, and MR. Administrative, technical or material
support: SB, CL, TO, and BL. Study supervision: MR, CL, TO, and BL.
Received 13 September 2013; revised 2 December 2013; accepted 3 December 2013
European Journal of Clinical Nutrition (2014), 18
& 2014 Macmillan Publishers Limited All rights reserved 0954-3007/14
www.nature.com/ejcn
disease, hyperthyreosis or hypothyreosis, osteoporosis or diabetes. Other
exclusion criteria were abnormal fasting plasma glucose levels (X7mmol/l),
blood pressure exceeding 150/90 mmHg, hormone replacement therapy,
statins, beta-blockers or any medication for psychiatric disorders.
Three women on monotherapy with an angiotensin-converting enzyme
inhibitor for mild hypertension were included in the study. Each woman
met a physician for clinical assessment, followed by a series of baseline
tests at the Clinical Research Center at Ume University Hospital, Ume,
Sweden.
Diet intervention
After baseline measurements, the women were randomized to a PD or a
Nordic Nutrition Recommendations (NNR) diet for 24 months. All study
personnel (except the dieticians) were blinded to the dietary allocation of
the participants. Both diets were consumed ad libitum. The PD provided
30% of energy intake (E%) from protein, 40 E% fat and 30 E%
carbohydrates and included a recommendation for a high intake of MUFA
and polyunsaturated fatty acids (PUFA). The diet was based on lean meat,
sh, eggs, vegetables, fruits, berries and nuts. Additional fat sources were
avocado and oils (rapeseed and olive oil) used in food preparation and
dressing. Dairy products, cereals, added salt and rened fats and sugar
were excluded. The NNR diet
12
was aiming at a daily intake of 15 E%
protein, 2530 E% fat and 5560 E% carbohydrates, with emphasis on low-
fat dairy products and high-bre products. Each group took part in a total
of 12 group sessions held by a trained study dietician (one dietician per
diet) throughout the 24-month study period. The group sessions consisted
of information on and cooking of the intervention diets, dietary effects on
health, behavioral changes and group discussions. The subjects were given
recipes and written instructions to facilitate the preparation of meals at
home. Eight group sessions (four cooking classes and four follow-up
sessions) were held during the rst 6 months of the intervention.
Additional group meetings were held at 9, 12, 18 and 24 months. The
protocol was in accordance with the Helsinki declaration and approved by
the Regional Ethical Review Board at Ume University, Ume, Sweden. This
trial was registered at clinicaltrials.gov as NCT00692536.
Anthropometry and metabolic parameters
Anthropometric measurements were made at baseline and after 6, 12, 18
and 24 months, as described earlier.
13
The sagittal abdominal diameter
was recorded at the umbilical level as the height of the abdomen
measured when lying down on the examination couch with the legs
straight. Body composition, including an estimation of the total percentage
of body fat, was measured by dual-energy X-ray absorptiometry
(GE Medical Systems, Lunar Prodigy X-ray Tube Housing Assembly, Brand
BX-1L, Model 8743, Madison, WI, USA). Systolic and diastolic blood pressure
was measured twice at 2-min intervals in the sitting position after
5 min rest, using an automatic blood pressure meter (Boso-medicus,
Bosch, Germany). Fasting blood samples were drawn from the antecubital
vein after a resting period. Aliquots were frozen immediately at 20 1C
until analyzed for plasma glucose, plasma insulin, serum cholesterol,
triglycerides, high-density lipoprotein (HDL) cholesterol, tissue
plasminogen activator activity, plasminogen activator inhibitor type 1
(PAI-1) activity and high-sensitivity C-reactive protein (hsCRP). Serum
low-density lipoprotein (LDL) was calculated as (serum cholesterol serum
HDL serum triglycerides)/2.2.
Assessed for eligibility
(n=210)
Excluded (n=140)
Did not meet inclusion criteria (n=140)
Discontinued intervention (n=1)
family reason (n=1)
Analyzed (n=34)
PD group (n=35)
Received allocated intervention (n=35)
Did not receive allocated intervention (n=0)
Discontinued intervention (n=8)
family reason (n=1), lack of time (n=2),
other disease (n=2), disappointed with
allocation (n=1) and problem with
transportation (n=2)
Analyzed (n=27)
NNR diet group (n=35)
Received allocated intervention (n=35)
Did not receive allocated intervention (n=0)
Allocation
Follow-up 24 mos.
Randomized (n=70)
Enrollment
Discontinued intervention (n=7)
Loss of motivation (n=2), other disease (n=3),
lack of time (n=1), unknown reason (n=1)
Analyzed (n=27)
Discontinued intervention (n=5)
Loss of motivation (n=2), other disease
(n=2), disappointed with allocation (n=1)
Analyzed (n=22)
Follow-up 6 mos.
Figure 1. Flow diagram of subjects participation in the trial. Participants were prescribed to eat either a PD or a diet according to the NNR.
Long-term weight loss with a Palaeolithic-type diet
C Mellberg et al
2
European Journal of Clinical Nutrition (2014) 1 8 & 2014 Macmillan Publishers Limited
Dietary assessment
Dietary intake was assessed using the 4-day estimated self-reported food
records conducted at baseline (2 4 days) and monthly until 6 months,
thereafter at 9, 12, 18 and 24 months. Subjects were instructed to keep a
record of all food items consumed over 4 consecutive days (three
weekdays and one weekend day) and to describe and estimate the
amount of food eaten by using coloured food-portion photographs
representing known weights and household measuring utensils (for
example, cup, spoon and standard weight of food items). The reported
food intake was converted to estimates of energy and nutrient intake
using the nutritional analysis package Dietist XP (version 3.0, Kost och
Na ringsdata AB, Bromma, Sweden) based on the food composition
database of the Swedish National Food Administration (2008-03-06).
Adherence to protein intake
Nitrogen excretion in urine (NU) was used as a biomarker for protein
intake, with three 24-h urine samples collected at baseline and after 6 and
24 months. The para-aminobenzoic acid method
14
was used to verify the
completeness of the urine collections. NU was determined using the
Kjeldahl technique with a Kjeltec analyser (model NMKL nr 6, Eurons Food
& Agro AB, Lidko ping, Sweden). Duplicates for 10% of the samples were
included for quality control. The analytical precision was10%.
Measurement of energy expenditure
Resting energy expenditure (REE) was measured at baseline and after 6
and 24 months using indirect calorimetry (Datex-Ohmeda Deltatrac II,
Datex-Ohmeda Inc, Madison, WI, USA) with breath-by-breath sampling.
Free-living physical activity energy expenditure (PAEE) was estimated at
the same time points using data collected over a 7-day period using a
combined accelerometer and heart rate monitor (Actiheart, CamNtech Ltd,
Cambridge, UK) as described previously.
1517
Diet-induced thermogenesis,
that is the production of heat after eating, was xed at 10% of the total
energy expenditure (TEE) for all individuals in the study. TEE was calculated
for each participant as the sum of the PAEE and REE, divided by 0.9 and
expressed as kcal or MJ per 24-h day.
Statistical methods and randomization
The primary outcome of the study was the change in fat mass over a
period of 2 years, which was also used to calculate power. Thirty-ve
subjects were estimated to be needed in each diet arm to achieve a
signicant outcome (Po0.05) with 80% power. Block randomization with a
block size of four and an allocation ratio of 1:1 was done by a statistician
blinded to the study. Most variables were normally distributed, but serum
insulin, serum TG, PAI-1, hsCRP, total lean mass and PUFA (g) required
logarithmic transformation. Differences between groups at baseline were
tested by independent sample t-tests. We used generalized estimating
equations to evaluate repeated measurements over time. A two-sided
P-value o0.05 was considered signicant. Statistical analyses were
performed using the SPSS for Windows (version 20.0, IBM Corporation,
Armonk, NY, USA). The primary analysis in this trial was an intention-
to-treat analysis.
RESULTS
Subject characteristics
No differences were found in the baseline characteristics between
the diet groups, except for higher HDL cholesterol in the PD group
(Tables 1 and 3). A total of 30% (n 21) of the participants were
lost to follow-up (Figure 1). A higher proportion of participants
completed the PD arm than the NNR arm (77% PD, 63% NNR).
Medication use did not change during the study period.
Energy expenditure and dietary intake
REE, PAEE and TEE did not change or differ between the groups
during the study period (Table 2), with the exception of a decrease
in REE at 6 months in the study population as a whole. The change
in reported nutrient intake between baseline, 6 and 24 months
and the difference between the groups at each time point is
presented in Table 2. No difference was found between the diet
groups regarding nutrient intake at baseline. Reported daily
energy intake decreased over time, without signicant differences
between groups (Table 2). The PD group had a 19% and 20%
lower reported energy intake and the NNR group 18% and 12%
lower reported energy intake at 6 and 24 months, respectively.
The PD group reported a signicantly lower intake (E% and
g/day) of carbohydrates, higher intake (E% and g/day) of protein,
MUFA, PUFA, cholesterol and higher total fat (E%), MUFA:SFA
(saturated fatty acid) and PUFA:SFA ratios compared with the NNR
group (Table 2). The PD group reported a more pronounced
change in the ratio (E%) protein:carbohydrates:total fat from
baseline to 6 and 24 months (17:46:33, 23:29:44 and 22:34:40,
respectively) compared with the NNR group (17:45:35, 19:48:32
and 17:43:34, respectively). Target intakes were not fully achieved;
the PD did not reach the target amounts of percentage energy of
protein (30 E%) at 6 and 24 months, and the NNR group did not
reach the target amounts of carbohydrates (5560%).
Adherence to protein intake
A total of 406 urine collections from 65 subjects (34 PD, 31 NNR) at
baseline, 51 subjects (30 PD, 21 NNR) at 6 months and 39 subjects
(21 PD, 18 NNR) at 24 months were available for comparison of
reported protein intake and NU. Mean NU were 13 g/day in both
the groups at baseline (Table 2). There was no difference in NU
within or between groups at 6 or 24 months follow-up, indicating
poor adherence to the target protein intake (30 E%) in the PD
group.
Anthropometry and cardiometabolic risk markers
Both the diet groups decreased their total fat mass: 6.5
and2.6 kg at 6 months; and 4.6 and 2.9 kg at 24 months for
the PD and NNR group, respectively, with a signicant difference
between the groups at 6 months but not at 24 months (Po0.001
and P 0.095 respectively; Figure 2). The PD group also lost more
total lean mass ( 1.3 vs 0.4 kg; P 0.005) during the rst 6
months. Both the diet groups had a signicant weight loss during
the whole study period, with signicantly greater weight loss in
the PD group at all follow-up time points except at 24 months
(Figure 2). The largest weight loss was measured at the 12-month
follow-up; 8.7 kg in the PD group and4.4 kg in the NNR group.
BMI (mean (s.e.m.)) decreased 3.0 (0.30), 3.3 (0.36) and 2.4
(

0.41) kg/m
2
in the PD group and 1.2 (

0.27), 1.7 (

0.38) and
1.4 (

0.34) kg/m
2
in the NNR group at 6, 12 and 24 months,
respectively. The loss in BMI was more pronounced in the PD
group at 6 months (Po0.001) and 12 months (P 0.002) but not
at 24 months (P 0.059). In both the diet groups, waist
circumference decreased signicantly during the whole study
period, with a signicantly more pronounced decrease in the PD
group at 6 months ( 11.1 vs 5.8 cm; P 0.001; Figure 2).
Table 1. Baseline characteristics of the study participants
a
Palaeolithic
diet (n35)
Nordic Nutrition
Recommendations
diet (n35)
P-value
Age (years) 59.55.5 60.35.9 NS
Body weight (kg) 87.0

10.6 86.8

10.0 NS
Body mass index (kg/m
2
) 32.7

3.6 32.6

3.3 NS
Waist circumference (cm) 105.4

10.0 104.7

10.4 NS
Sagittal diameter (cm) 21.7

2.2 21.7

2.2 NS
Total fat mass (kg) 39.87.2 40.98.6 NS
Total lean mass (kg)
b
42.65.0 41.74.0 NS
Abbreviation: NS, not signicant (P40.05).
a
Data are reported as
means.d. Differences between the groups at baseline were tested by
independent sample t-tests.
b
Hypothesis testing after logarithmic trans-
formation.
Long-term weight loss with a Palaeolithic-type diet
C Mellberg et al
3
& 2014 Macmillan Publishers Limited European Journal of Clinical Nutrition (2014) 1 8
Table 2. Changes in energy expenditure, dietary intake and biomarker of protein intake during 2 years of intervention among obese
postmenopausal women
a
Palaeolithic-type
diet
Nordic Nutrition
Recommendations
diet
Model effect
diet time
b
Difference
between
the groups
c
REE (kcal/day) (MJ/day) 0.231
Baseline 1324

21.0 (5.54

0.09) 1292

25.1 (5.41

0.11)
Change 06 months 50.0

22.9 ( 0.21

0.10) 27.8

14.5 ( 0.12

0.06)
Change 024 months 29.9

13.4 ( 0.13

0.06) 3.1

14.8 (0.01

0.06)
PAEE (kcal/day) (MJ/day) 0.559
Baseline 769

33.4 (3.22

0.14) 750

43.0 (3.14

0.18)
Change 06 months 16.733.4 ( 0.070.14) 1.933.4 (0.0080.14)
Change 024 months 33.443.0 (0.140.18) 7.235.8 ( 0.030.15)
TEE (kcal/day) (MJ/day) 0.956
Baseline 2331

47.8 (9.76

0.20) 2288

64.5 (9.58

0.27)
Change 06 months 52.6

45.4 ( 0.22

0.19) 38.2

43.0 ( 0.16

0.18)
Change 024 months 4.78

54.9 (0.02

0.23) 0.72

50.2 (0.003

0.21)
Energy intake (kcal/day) (MJ/day) 0.304
Baseline 2000

59.0 (8.37

0.25) 2019

59.1 (8.45

0.25)
Change 06 months 37561.0 ( 1.570.26) 35961.2 ( 1.500.26)
Change 024 months 40189.5 ( 1.680.38) 25162.2 ( 1.050.26)
Protein (E%) o0.001
Baseline 17.1

0.32 17.2

0.41
Change 06 months 6.30

0.54
z
1.59

0.45
z
o0.001
Change 024 months 4.79

0.72
z
0.16

0.40 o0.001
Protein (g/day) o0.001
Baseline 84.4

2.60 85.2

2.44
Change 06 months 9.343.34
y
8.742.48
z
o0.001
Change 024 months 0.393.28 8.842.99
y
0.037
Carbohydrate (E%) o0.001
Baseline 46.2

0.67 45.3

0.86
Change 06 months 16.9

1.09
z
1.05

0.89 o0.001
Change 024 months 12.7

1.68
z
2.02

1.84 o0.001
Carbohydrate (g/day) o0.001
Baseline 224

7.83 222

8.15
Change 06 months 1047.38
z
40.87.88
z
o0.001
Change 024 months 87.211.9
z
32.59.56
y
o0.001
Total fat (E%) o0.001
Baseline 33.4

0.60 34.6

0.73
Change 06 months 10.1

1.13
z
2.34

0.87
y
o0.001
Change 024 months 7.02

1.40
z
0.26

1.76 0.003
Total fat (g/day) o0.001
Baseline 74.6

2.45 78.6

3.00
Change 06 months 3.433.56 18.03.32
z
o0.001
Change 024 months 3.774.39 9.984.65
z
0.331
SFA (E%) 0.09
Baseline 12.7

0.34 13.5

0.35
Change 06 months 2.87

0.53 1.71

0.43
Change 024 months 2.16

0.54 0.17

1.02
SFA (g/day) 0.227
Baseline 28.2

1.07 30.5

1.32
Change 06 months 10.61.24 8.411.57
Change 024 months 8.911.83 3.682.43
MUFA (E%) o0.001
Baseline 13.0

0.31 13.1

0.38
Change 06 months 7.69

0.69
z
1.07

0.39
y
o0.001
Change 024 months 5.08

0.88
z
0.24

0.65 o0.001
MUFA (g/day) o0.001
Baseline 28.8

1.07 29.6

1.24
Change 06 months 8.201.84
z
7.061.35
z
o0.001
Change 024 months 2.412.02 3.371.60
z
0.025
PUFA (E%) o0.001
Baseline 5.39

0.21 5.54

0.22
Change 06 months 4.32

0.52
z
0.07

0.30 o0.001
Change 024 months 3.19

0.58
z
0.31

0.50 o0.001
Long-term weight loss with a Palaeolithic-type diet
C Mellberg et al
4
European Journal of Clinical Nutrition (2014) 1 8 & 2014 Macmillan Publishers Limited
In addition, the sagittal diameter decreased signicantly over time
in a similar manner in both the groups, with a larger decline in the
PD group at 6 months ( 3.7 vs 2.0 cm; Po0.001; Figure 2).
Concomitantly, hip circumference decreased over time with a
signicant difference between the groups at 6 months ( 6.8 with
PD vs 2.7 cm with NNR at 6 months; Po0.001).
Triglyceride levels decreased signicantly in the PD group over
time with a 0.26 and 0.22 mmol/l difference between the diet
groups at 6 and 24 months (Po0.001 and P 0.004), respectively
(Table 3). Other benecial cardiometabolic changes occurred in
the study population as a whole over time (Table 3). At both 6 and
24 months, diastolic blood pressure, heart rate, CRP, LDL
cholesterol and PAI-1 activity decreased, as well as systolic blood
pressure and total cholesterol at 6 months, and HDL cholesterol
increased at 24 months. No differences were measured over time
or between groups with regard to fasting glucose, fasting insulin
concentrations and tissue plasminogen activator activity.
DISCUSSION
We found that a diet with reduced carbohydrate and SFA intake
and a relative increase in the intake of protein, MUFA and PUFA
has strong and long-lasting effects on fat mass, body weight and
abdominal obesity in postmenopausal women, but there were no
signicant differences in anthropometric measurements at 24
months between the groups. Notably, triglyceride levels
Table 2. (Continued)
Palaeolithic-type
diet
Nordic Nutrition
Recommendations
diet
Model effect
diet time
b
Difference
between
the groups
c
PUFA (g/day) o0.001
Baseline 12.00.59 12.40.59
Change 06 months 5.751.28
z
1.840.59
y
o0.001
Change 024 months 2.821.21
z
1.040.93 0.012
MUFA:SFA (ratio) o0.001
Baseline 1.05

0.04 0.99

0.03
Change 06 months 1.14

0.10
z
0.06

0.03 o0.001
Change 024 months 0.75

0.11
z
0.05

0.05 o0.001
PUFA:SFA (ratio) o0.001
Baseline 0.440.03 0.420.02
Change 06 months 0.600.07
z
0.070.03
z
o0.001
Change 024 months 0.440.07
z
0.040.04 o0.001
Omega 3 fatty acids (g/day) 0.001
Baseline 2.68 (0.18) 2.62 (0.13)
Change 06 months 1.13 (0.29)
z
0.20 (0.23) o0.001
Change 024 months 4.67 (0.52)
z
2.65 (0.44)
z
0.003
Omega-6 fatty acids (g/day) 0.005
Baseline 9.040.45 9.220.52
Change 06 months 3.561.01
z
0.650.81 0.001
Change 024 months 4.790.54
z
4.810.62
z
0.986
Dietary cholesterol (mg/day) o0.001
Baseline 305

14.3 342

17.4
Change 06 months 276

27.8
z
43.6

20.0
z
o0.001
Change 024 months 146

28.2
z
19.0

20.6 o0.001
Dietary bre (E%) 0.869
Baseline 2.450.06 2.180.08
Change 06 months 0.440.12 0.500.15
Change 024 months 0.290.10 0.240.13
Dietary bre (g/day) 0.548
Baseline 24.6

1.03 21.9

0.97
Change 06 months 0.94

1.17 0.35

0.98
Change 024 months 2.70

1.36 0.77

1.15
Sucrose (g/day) 0.547
Baseline 35.32.11 40.93.18
Change 06 months 8.361.77 11.23.24
Change 024 months 7.242.52 11.32.85
Nitrogen in urine (g/day)
d
0.374
Baseline 13.0

0.44 13.0

0.40
Change 06 months 0.04

0.33 0.68

0.43
Change 024 months 0.47

0.54 0.98

0.34
Abbreviations: E %, percentage of energy; MUFA, monounsaturated fat; PAEE, physical activity energy expenditure; PUFA, polyunsaturated fat; REE, resting
energy expenditure; SFA, saturated fat; TEE, total energy expenditure. Generalized estimating equations, with the PD group as a reference group, were used to
estimate the change from baseline to 6 and 24 months within each treatment group and to test the difference between treatment groups.
a
All values are
means.e. obtained from 4 days of estimated food records. Change over time vs baseline and within group;
z
Po0.05,
y
Po0.01 and
z
Po0.001 (only presented
if overall model effect was signicant).
b
The P-value represents the overall model effect of the diet time interaction during the 24-month study.
c
The P-value
represents the effect of the diet time interaction at 6 and 24 months and was only calculated if the overall model effect was signicant (post-hoc analysis).
d
Nitrogen in urine, used as a biomarker for protein intake, was determined by analysis of nitrogen excretion in 24-h urine (Kjeldahl technique).
Long-term weight loss with a Palaeolithic-type diet
C Mellberg et al
5
& 2014 Macmillan Publishers Limited European Journal of Clinical Nutrition (2014) 1 8
decreased signicantly more in the PD group than in the control
group based on the NNR diet. However, adherence to the target
intake of protein was poor in the PD group.
Increased visceral fat mass, associated with liver fat (that is,
ectopic fat accumulation), may contribute to increased risk for the
metabolic and cardiovascular complications after the meno-
pause.
18
Triglyceride elevations are an essential part of the
metabolic dysfunction seen with ectopic fat accumulation. Of
interest is that we recently showed a 5-week PD to be associated
with a 49% decrease in liver fat and increased insulin sensitivity
linked to a 41% decrease in serum triglycerides.
13
A diet with a
similar macronutrient composition as the present study was also
shown previously to reduce visceral fat more than total fat on a
long-term basis, albeit with a high dropout from the study after 2
years.
19
Further studies on the putative effect by a PD on visceral
fat and ectopic fat seem therefore warranted.
The lack of a signicant decrease in fasting glucose and insulin
levels in our study may be explained in part by the fact that
the subjects had normal glucose tolerance at baseline (that is,
obese but healthy), thereby reducing the possibility of improving
the metabolic status and these cardiovascular risk markers. In
contrast, we found signicant effects over time, but not between
the groups, of PAI-1 and CRP levels, with a mean decrease in the
PD group that was approximately twice that of the NNR group.
Analyses of the putative effects of a PD in subjects with a more
pronounced metabolic dysfunction are encouraged by short-term
studies indicating an improvement in the glucose tolerance and
cardiovascular risk markers of patients with T2DM and CVD.
9,10,20
The profound decrease in energy intake (20% vs 12% for the PD
and NNR groups at 24 months) is in line with earlier short-term
studies from us and others.
911,13,20
This may be due to effects on
satiety by increased intake of protein and low energy-dense foods,
as well as PUFAs.
2128
Notably, no signicant differences were
found in urinary nitrogen excretion between the diet groups at
any time point, which may have limited the differences between
the groups. Adherence to the target protein intake was thus poor
in the PD group, in line with earlier studies with the aim of
increasing protein intake.
29,30
There could be several reasons for
this, such as protein-rich food being more expensive, social
inuences on womens food choices or a lower food preference
for protein-rich food among women.
3133
Importantly, this
suggests that other factors than protein content contributes to
the benecial effects of the PD diet. One possibility is the higher
intake of PUFAs in the PD group. In line with this, increased satiety
after intake of long chain omega-3 fatty acids has been
demonstrated.
25
The magnitude of effects on body composition and triglyceride
levels do, however, suggest that larger randomized trials are done
regarding putative differences between a PD and other types of
diet with different macronutrient compositions. The reductions in
fat mass, weight and abdominal obesity were thus profound but
less different between groups than expected from our power
analysis.
Despite the stratication for BMI, the PD group had a more
benecial metabolic prole at baseline, including higher HDL
cholesterol levels. The NNR study group also had higher variability
in some of the study variables, which may have inuenced our
ability to detect differences between the groups. Therefore, we
may have underestimated the effects of the intervention on
various outcome variables. In addition, the dropout rate from the
study was larger than expected (37% and 23%, respectively, for
the NNR and PD group); the latter in line with recent data from
Jonsson et al.
22
We did not include an observational study group. Therefore,
we cannot conclude whether the changes observed in these
intervention groups differ from the natural course regarding
the study parameters in this population. However, our aim was to
investigate the possible effect of a diet regimen in comparison
with the low-fat, high-bre diet in Nordic countries generally
recommended when the study started.
In conclusion, a PD during 2 years with ad libitum intake of
macronutrients, including an increased intake of PUFAs and
Fat mass
0 6 24
Months
30
35
40
45
k
g
PD
NNR
*** NS
Body weight
76
78
80
82
84
86
88
90
0 6 12 18 24
Months
k
g
PD
NNR
* ** ***
NS
Waist circumference
90
95
100
105
110
0 6 24
Months
c
m
PD
NNR
**
NS
Sagittal diameter
16
18
20
22
24
0 6 24
Months
c
m
PD
NNR
***
NS
Figure 2. The effect of diet intervention on different anthropometric measurements. Generalized estimated equations were used, with the
Palaeolithic diet group as a reference group. P-values for the diet time interaction for (a) fat masso0.001, (b) weighto0.001, (c) waist 0.001
and (d) sagittal diametero0.001. Data are presented as mean

s.e., *Po0.05, **Po0.01 and ***Po0.001 for differences between diet groups
at respective time points based on estimated marginal means. NS, not signicant.
Long-term weight loss with a Palaeolithic-type diet
C Mellberg et al
6
European Journal of Clinical Nutrition (2014) 1 8 & 2014 Macmillan Publishers Limited
Table 3. Changes in bioindicators during 2 years of intervention among obese postmenopausal women
a
Palaeolithic-type diet Nordic Nutrition
Recommendations diet
Model effect
diet time
b
Difference
between the
groups
c
Insulin (mIU/l) 0.500
Baseline 8.43

0.69 9.02

0.77
Change 06 months 1.29

0.91 0.10

0.91
Change 024 months 0.18

0.53 0.87

0.77
Glucose (mmol/l) 0.465
Baseline 5.130.15 5.170.20
Change 06 months 0.21

0.15 0.05

0.24
Change 024 months 0.04

0.13 0.002

0.22
Systolic blood pressure (mmHg) 0.293
Baseline 1412.2 1382.2
Change 06 months 12.22.2 8.51.9
Change 024 months 3.73.5 1.72.2
Diastolic blood pressure (mmHg) 0.349
Baseline 83.0

1.3 82.9

1.5
Change 06 months 6.61.1 5.01.5
Change 024 months 4.81.5 1.51.8
Heart rate (beats/min) 0.401
Baseline 72

1.6 71

1.7
Change 06 months 2.2

1.4 3.2

1.1
Change 024 months 3.6

1.8 1.4

1.7
High-sensitivity CRP (mg/l) 0.464
Baseline 2.140.29 2.400.31
Change 06 months 0.42

0.20 0.22

0.17
Change 024 months 0.45

0.23 0.05

0.29
Tissue plasminogen activator (IU/ml) 0.544
Baseline 0.510.05 0.460.06
Change 06 months 0.060.05 0.030.06
Change 024 months 0.05

0.06 0.05

0.06
PAI-1 (U/ml) 0.431
Baseline 21.3

2.7 24.3

2.3
Change 06 months 5.62.3 3.32.3
Change 024 months 4.52.1 2.33.3
Cholesterol (mmol/l) 0.127
Baseline 5.91

0.14 5.52

0.23
Change 06 months 0.67

0.13 0.39

0.15
Change 024 months 0.200.10 0.070.11
HDL (mmol/l) 0.896
Baseline 1.49

0.06* 1.28

0.05
Change 06 months 0.05

0.05 0.04

0.04
Change 024 months 0.16

0.04 0.18

0.04
LDL (mmol/l) 0.291
Baseline 3.870.13 3.640.21
Change 06 months 0.43

0.10 0.29

0.11
Change 024 months 0.26

0.08 0.07

0.09
Triglycerides (mmol/l) 0.001
Baseline 1.220.09 1.270.10
Change 06 months 0.380.07
z
0.120.07 o0.001
Change 024 months 0.230.07
z
0.010.06 0.004
Abbreviations: CRP, C-reactive protein; HDL, high-density lipoprotein; LDL, low-density lipoprotein; PAI-1, plasminogen activator inhibitor type 1. Generalized
estimating equations, with the PD group as a reference group, were used to estimate the change from baseline to 6 and 24 months within each treatment
group and to test differences between the treatment groups.
a
All values are means.e. Change over time vs baseline and within the group;
z
Po0.001 (only
presented if overall model effect was signicant). *P 0.01 for difference between the groups at baseline.
b
The P-value represents the overall model effect of
the diet time interaction during the 24-month study.
c
The P-value represents the effect of the diet time interaction at 6 and 24 months and was only
calculated if the overall model effect was signicant (post-hoc analysis).
Long-term weight loss with a Palaeolithic-type diet
C Mellberg et al
7
& 2014 Macmillan Publishers Limited European Journal of Clinical Nutrition (2014) 1 8
MUFAs, reduces fat mass and abdominal obesity with signicantly
better long-term effect on triglyceride levels vs an NNR diet.
Adherence to the prescribed protein intake was poor in the PD
group, suggesting that other components of the PD diet are of
greater importance. The putative long-term benecial effects of
different components of the PD on obesity-related diseases,
notably T2DM, need to be explored.
CONFLICT OF INTEREST
The authors declare no conict of interest.
ACKNOWLEDGEMENTS
We thank all of the women who participated in this study for their invaluable
patience and cooperation. Erik Ha gg, Jonas Andersson, Lars-Go ran Sjo stro m, Go ran
Ericsson, Inger Arnesjo , Katarina Iselid and Monica Holmgren made important
contributions to screening the health of study subjects and technical assistance.
Johanna Larsson helped process food records. Kate Westgate and Stefanie Mayle
(MRC Epidemiology Unit, University of Cambridge, Cambridge, UK) assisted with
processing physical activity data. Paul Franks contributed important views on
planning the study. This study was supported by grants from The Swedish Council for
Working Life and Social Research (2006-0699 and 2010-0398), the Swedish Research
Council (K2011-12237-15-6), the Swedish Heart and Lung Foundation, the County
Council of Va sterbotten and Ume University, Sweden.
REFERENCES
1 Finucane MM, Stevens GA, Cowan MJ, Danaei G, Lin JK, Paciorek CJ et al. National,
regional, and global trends in body-mass index since 1980: systematic analysis of
health examination surveys and epidemiological studies with 960 country-years
and 9.1 million participants. Lancet 2011; 377: 557567.
2 Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, Adair-Rohani H et al. A com-
parative risk assessment of burden of disease and injury attributable to 67 risk
factors and risk factor clusters in 21 regions, 1990-2010: a systematic analysis for
the Global Burden of Disease Study 2010. Lancet 2012; 380: 22242260.
3 Rexrode KM, Carey VJ, Hennekens CH, Walters EE, Colditz GA, Stampfer MJ et al.
Abdominal adiposity and coronary heart disease in women. JAMA 1998; 280:
18431848.
4 Davis SR, Castelo-Branco C, Chedraui P, Lumsden MA, Nappi RE, Shah D et al.
Understanding weight gain at menopause. Climacteric 2012; 15: 419429.
5 Te Morenga L, Mann J. The role of high-protein diets in body weight management
and health. Br J Nutr 2012; 108(Suppl 2): S130S138.
6 Abete I, Astrup A, Martinez JA, Thorsdottir I, Zulet MA. Obesity and the metabolic
syndrome: role of different dietary macronutrient distribution patterns and
specic nutritional components on weight loss and maintenance. Nutr Rev 2010;
68: 214231.
7 Estruch R, Ros E, Salas-Salvado J, Covas MI, Corella D, Aros F et al. Primary
prevention of cardiovascular disease with a Mediterranean diet. N Engl J Med
2013; 368: 12791290.
8 Frassetto LA, Schloetter M, Mietus-Synder M, Morris RC, Sebastian A. Metabolic
and physiologic improvements from consuming a paleolithic, hunter-gatherer
type diet. Eur J Clin Nutr 2009; 63: 947955.
9 Jonsson T, Granfeldt Y, Ahren B, Branell UC, Palsson G, Hansson A et al. Benecial
effects of a Paleolithic diet on cardiovascular risk factors in type 2 diabetes: a
randomized cross-over pilot study. Cardiovasc Diabetol 2009; 8: 35.
10 Lindeberg S, Jo nsson T, Granfeldt Y, Borgstrand E, Soffman J, Sjo stro m K et al.
A Palaeolithic diet improves glucose tolerance more than a Mediterranean-like diet
in individuals with ischaemic heart disease. Diabetologia 2007; 50: 17951807.
11 Osterdahl M, Kocturk T, Koochek A, Wa ndell PE. Effects of a short-term inter-
vention with a paleolithic diet in healthy volunteers. Eur J Clin Nutr 2008; 62:
682685.
12 Becker W, Lyhne N, Pedersen AN, Aro A, Fogelholm M, Tho rsdottir I et al.
Nordic Nutrition Recommendations 2004 integrating nutrition and physical
activity. Scand J Nutr 2004; 48: 178187.
13 Ryberg M, Sandberg S, Mellberg C, Stegle O, Lindahl B, Larsson C et al.
A Palaeolithic-type diet causes strong tissue-specic effects on ectopic
fat deposition in obese postmenopausal women. J Intern Med 2013; 274:
6776.
14 Bingham S, Cummings JH. The use of 4-aminobenzoic acid as a marker to validate
the completeness of 24 h urine collections in man. Clin Sci (Lond) 1983; 64:
629635.
15 Brage S, Brage N, Franks PW, Ekelund U, Wareham NJ. Reliability and validity of
the combined heart rate and movement sensor Actiheart. Eur J Clin Nutr 2005; 59:
561570.
16 Brage S, Ekelund U, Brage N, Hennings MA, Froberg K, Franks PW et al. Hierarchy
of individual calibration levels for heart rate and accelerometry to measure
physical activity. J Appl Physiol 2007; 103: 682692.
17 Stegle O, Fallert SV, MacKay DJ, Brage S. Gaussian process robust regression for
noisy heart rate data. IEEE Trans Biomed Eng 2008; 55: 21432151.
18 Volzke H, Schwarz S, Baumeister SE, Wallaschofski H, Schwahn C, Grabe HJ et al.
Menopausal status and hepatic steatosis in a general female population. Gut
2007; 56: 594595.
19 Skov AR, Toubro S, Ronn B, Holm L, Astrup A. Randomized trial on protein vs
carbohydrate in ad libitum fat reduced diet for the treatment of obesity. Int J Obes
Relat Metab Disord 1999; 23: 528536.
20 Jonsson T, Granfeldt Y, Erlanson-Albertsson C, Ahren B, Lindeberg S. A paleolithic
diet is more satiating per calorie than a mediterranean-like diet in individuals with
ischemic heart disease. Nutr Metab (Lond) 2010; 7: 85.
21 Halton TL, Hu FB. The effects of high protein diets on thermogenesis, satiety and
weight loss: a critical review. J Am Coll Nutr 2004; 23: 373385.
22 Jonsson T, Granfeldt Y, Lindeberg S, Hallberg AC. Subjective satiety and other
experiences of a Paleolithic diet compared to a diabetes diet in patients with type
2 diabetes. Nutr J 2013; 12: 105.
23 Lejeune MP, Westerterp KR, Adam TC, Luscombe-Marsh ND, Westerterp-Plan-
tenga MS. Ghrelin and glucagon-like peptide 1 concentrations, 24-h satiety, and
energy and substrate metabolism during a high-protein diet and measured in a
respiration chamber. Am J Clin Nutr 2006; 83: 8994.
24 Paddon-Jones D, Westman E, Mattes RD, Wolfe RR, Astrup A, Westerterp-Plan-
tenga M. Protein, weight management, and satiety. Am J Clin Nutr 2008; 87:
1558S1561S.
25 Parra D, Ramel A, Bandarra N, Kiely M, Martinez JA, Thorsdottir I. A diet rich in long
chain omega-3 fatty acids modulates satiety in overweight and obese volunteers
during weight loss. Appetite 2008; 51: 676680.
26 Ruxton CH, Reed SC, Simpson MJ, Millington KJ. The health benets of omega-3
polyunsaturated fatty acids: a review of the evidence. J Hum Nutr Diet 2004; 17:
449459.
27 Sloth B, Due A, Larsen TM, Holst JJ, Heding A, Astrup A. The effect of a high-MUFA,
low-glycaemic index diet and a low-fat diet on appetite and glucose
metabolism during a 6-month weight maintenance period. Br J Nutr 2009; 101:
18461858.
28 Soenen S, Westerterp-Plantenga MS. Proteins and satiety: implications for weight
management. Curr Opin Clin Nutr Metab Care 2008; 11: 747751.
29 de Souza RJ, Bray GA, Carey VJ, Hall KD, LeBoff MS, Loria CM et al. Effects of 4
weight-loss diets differing in fat, protein, and carbohydrate on fat mass, lean
mass, visceral adipose tissue, and hepatic fat: results from the POUNDS LOST trial.
Am J Clin Nutr 2012; 95: 614625.
30 Pagoto SL, Appelhans BM. A call for an end to the diet debates. JAMA 2013; 310:
687688.
31 Beagan BL, Chapman GE. Family inuences on food choice: context of surviving
breast cancer. J Nutr Educ Behav 2004; 36: 320326.
32 Cooke LJ, Wardle J. Age and gender differences in childrens food preferences.
Br J Nutr 2005; 93: 741746.
33 Drewnowski A, Specter SE. Poverty and obesity: the role of energy density and
energy costs. Am J Clin Nutr 2004; 79: 616.
Long-term weight loss with a Palaeolithic-type diet
C Mellberg et al
8
European Journal of Clinical Nutrition (2014) 1 8 & 2014 Macmillan Publishers Limited