one-year weight maintenance after significant weight loss in healthy

7/31/2019 One-Year Weight Maintenance After Significant Weight Loss in Healthy

1/12

One-year weight maintenance after significant weight loss in healthy

overweight and obese subjects: does diet composition matter?13

Elizabeth A Delbridge, Luke A Prendergast, Janet E Pritchard, and Joseph Proietto

ABSTRACT

Background: For many people, maintenance of weight loss is elu-

sive. Whereas high-protein (HP) diets have been found to be supe-

rior to high-carbohydrate (HC) diets for weight loss in the short

term, their benefits long term are unclear, particularly for weight

maintenance. Furthermore, the literature lacks consensus on the

long-term effects of an HP diet on cardiovascular disease risk

factors.

Objective: The objective was to investigate whether macronutrientdietary composition plays a role in weight maintenance and in

improvement of cardiovascular disease risk factors.

Design: The study comprised 2 phases. Phase 1 featured a very-

low-energy diet for 3 mo. In phase 2, the subjects were randomly

assigned to an HP or an HC diet for 12 mo. The diets were iso-

caloric, tightly controlled, and individually prescribed for weight

maintenance. The subjects were overweight or obese but otherwise

healthy men and women.

Results: The subjects lost an average of 16.5 kg during phase 1 and

maintained a mean (6SEM) weight loss of 14.5 6 1.2 kg (P ,

0.001) during phase 2; no significant differences between groups

were observed. By the end of the study, reductions in systolic blood

pressure were 14.3 6 2.4 mm Hg for the HP group and 7.7 6

2.2 mm Hg for the HC group (P , 0.045). Forty-seven percent of

the 180 subjects who began the study completed both phases.

Conclusions: The results indicate that the protein or carbohydrate

content of the diet has no effect on successful weight-loss mainte-

nance. A general linear model analysis indicated that dietary

treatment (HP or HC) was a significant factor in systolic blood pres-

sure change and in favor of the HP diet. This trial was registered at

www.clinicaltrials.gov as NCT 00625236. Am J Clin Nutr

2009;90:120314.

INTRODUCTION

In recent years there has been renewed interest in high-protein

diets for weight loss. An increase in the protein content of the diet

is usually accompanied by a reduction in carbohydrate content;

thus, the terms high protein and low carbohydrate are used

interchangeably in the literature; despite the protein content of

dietary treatment ranging from 25% (1) to 45% (2). Many studies

with a duration of up to 6 mo have found that diets high in protein

achieve greater weight loss than do high-carbohydrate diets

(24).

Very-low-energy diets (VLEDs) have been used effectively for

at least the past 20 y to achieve large and rapid weight losses.

Whereas one analysis of the literature found that long-term

weight regain experienced by individuals who followed a VLED

is no worse than weight regain experienced by individuals who

followed low-energy diets (5), another review found that indi-

viduals who followed VLEDs maintained a significantly greater

weight losses than did subjects who followed a low-energy

balanced diet (6). The type of diet and intensity and duration of

follow-up are likely to have an effect on the long-term success of

weight maintenance.

Little research has been done on the role of a high-protein (HP)diet on the maintenance of weight loss. Lejeune et al (7) showed

that, in overweight and obese subjects who followed a VLED for

4 wk (mean weight loss: 6.36 2.0 kg), a higher-protein, weight-

maintenance diet resulted in significantly less weight regain over

the subsequent 6 mo (0.8 compared with 3.0 kg; P, 0.05)). The

higher-protein diet consisted of 18% protein and was compared

with a 15% protein control diet. The higher-protein group

consumed the additional protein in the form of protein supple-

ments, whereas the control group consumed their usual diet

without additional supplements. Weight maintenance after

weight loss was also investigated by Claessens et al (8). In their

study, overweight and obese subjects underwent weight loss by

following a VLED for 5 to 6 wk. The subjects were then ran-domly assigned to an HP or high-carbohydrate (HC) weight-

maintenance diet for 12 wk. Subjects in the HP diet group

consumed 25% of their energy intake as protein and protein

supplements twice daily. The HC group consumed 55% of their

energy as carbohydrate, along with twice daily maltodextrin

supplements. The HP group maintained greater weight loss than

did the HC group.

The purpose of our study was to achieve a large and rapid

weight loss using a VLED and to then investigate the effects of 12

mo of weight maintenance with a low-fat, HP diet or a low-fat,

HC diet on body weight in overweight and obese men and

women. The subjects were free-living and were not provided with

any food or supplements during the weight-maintenance period.

1From the Department of Medicine, Heidelberg Repatriation Hospital,

Heidelberg, Australia (EAD and JP); the Department of Mathematics and

Statistics, La Trobe University, Bundoora, Australia (LAP); and the Depart-

ment of Physiology, University of Melbourne, Parkville, Australia (JEP).2 Supported in part by Meat and Livestock Australia.3 Address correspondence to EA Delbridge, Department of Medicine, Build-

ing 24, HeidelbergRepatriation Hospital, 300 Waterdale Road,Heidelberg3081,

Victoria, Australia. E-mail: [email protected].

Received November 9, 2008. Accepted for publication August 28, 2009.

First published online September 30, 2009; doi: 10.3945/ajcn.2008.27209.

Am J Clin Nutr 2009;90:120314. Printed in USA. 2009 American Society for Nutrition 1203

byguest

onJune20,2012

www.ajcn.org

Downloadedfrom
http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/http://www.ajcn.org/


2/12

SUBJECTS AND METHODS

The study was approved by the Human Research Ethics

Committees of the Royal Melbourne and Austin Hospitals. The

subjects were recruited by newspaper advertisement and word of

mouth. They were men and women aged 18 to 75 y whosebody mass indexes (BMI; in kg/m2) were 30 or 27 and hadcomorbidities. In providing their informed consent to partici-

pate, the subjects indicated their willingness to be randomly

assigned to study groups and to adhere to the study protocol.

Individuals with a history or presence of significant disease,

endocrine disorder, psychiatric illness, and alcohol or drug abuse

were excluded. In addition, women were excluded if they were

lactating, pregnant, or planned to become pregnant during the

study.

Study design

The study had a parallel, randomized design that consisted of

2 time phases. Phase 1 was a 3-mo intensive weight-loss phase

that all subjects began at the start of the study. Phase 2 refers to

the 12-mo weight-maintenance phase during which eligible

subjects were placed on either an HC or an HP diet after the

completion of phase 1. Phase 1+2 refers to the entire study

duration, ie, both phase 1 and phase 2.

Dietary instructions

During phase 1, the subjects were provided with the com-

mercially available VLED, (Optifast; Nestle Nutrition, Frankfurt,

Germany). As per the manufacturers recommendations, the

VLED was consumed 3 times/d as a meal replacement for a 12-

wk period and was taken as a milkshake, soup, or dessert. The

subjects were also permitted to consume up to 2 cups of low

starch vegetables daily with a small amount of oil (5 mL) and

a minimum of 2 L water or low-energy drinks. This providedsubjects with 500550 kcal/d. The subjects attended the clinic

individually each fortnight to be weighed, counseled on the use

of the VLED, and be provided with a 2-wk supply of the VLED.

The subjects were required to lose 10% of their body weightduring phase 1 to progress to the weight-maintenance phase

(phase 2). In phase 2, the subjects were randomly assigned to the

HC or HP dietary group (see Statistics). Harris-Benedict equa-

tions were used to estimate individual basal metabolic rate

(BMR) (9). Each individual subjects total energy expenditure

was calculated as BMR multiplied by an activity factor of 1.3,

which was considered appropriate for weight maintenance in

mildly active adults. The subjects were counseled to consume an

energy intake consistent with weight maintenance, and this in-formation was reinforced at each of the monthly visits attended

by subjects during phase 2, particularly if subjects began to

regain weight. The subjects received meal plans and recipes

appropriate to the dietary group to which they had been allo-

cated. They bought and prepared their own food and were in-

vited to attend group cooking classes that were specific to the

HP or HC dietary treatments.

Subjects in the HP group were instructed to consume 30% of

their intake as protein, whereas subjects in the HC group were

instructed to consume 15% of their intake as protein. Both

dietary groups were recommended to consume lean red meat

34 times/wk to meet the recommended intakes of iron and zinc

in line with Australian Dietary Guidelines. Both groups were

advised to reduce their fat intake to ,30% of their intake, with

a particular emphasis on reducing saturated fat. Carbohydrates

with a low glycemic index (GI) were recommended to both

groups. Issues such as comfort eating, snacking, reading food

labels, and healthy carryout and eating-out options were ad-

dressed during monthly counseling sessions. The subjects were

also encouraged to practice healthy behaviors such as aerobic

exercise 3 times/wk.

Data collection

The screening visit included a medical history, physical ex-

amination, and blood test. Body weight was measured at each

visit at approximately the same time of day by using the same

digital calibrated scales. The subjects wore street clothing, re-

moved their shoes, and voided their bladder before being

weighed. At each visit, bioelectrical impedance analysis was used

to measure body composition (Tanita TBF-300; WW Wedder-

burn Pty, Ltd, Sydney, Australia) by using the standard adult

mode of measurement. At each visit, waist and hip circum-

ferences were measured to the nearest 0.5 cm, and the mean of 2readings taken with a spring-loaded tape measure was recorded.

Waist circumference was measured at the level of the umbilicus

and hip circumference at the level of the greater trochanters.

Dietary compliance was estimated by using 3-d food diaries. The

subjects were instructed by a dietitian to provide as much in-

formation on food and drinks consumed (eg, brand name, vol-

ume, weight, and ingredients) over 3 consecutive days (2

weekdays and 1 weekend day). The food diaries and instructions

for their use were provided at months 8 and 14 of the study. The

subjects were asked to complete the diaries immediately before

their scheduled visits at months 9 and 15. The diaries provided

data on energy intake midway through and at the end of phase 2

the weight-maintenance phase. Food diaries were analyzed byusing Foodworks Professional Edition version 3.02.581 (Xyris

Software, Highgate Hill, Australia). Twenty-four-hour urine

samples were collected to coincide with food diary collection

midway through (month 9) and at the end of phase 2 (month 15).

The subjects were thoroughly educated on the procedure for

collecting an accurate 24-h urine sample and were also provided

with written instructions. The subjects were instructed to collect

all urine excreted after the first volume voided in the morning

until and including the first morning urine voided the following

day. Blood samples were collected via venipuncture after an

overnight fast. The analyses were performed by the Pathology

Department at the Royal Melbourne Hospital.

Statistics

The subjects were randomly assigned to the HP or HC group

by using randomly chosen block sizes of 2, 4, and 6 within 8 strata

that accounted for sex, BMI (,35 or 35), and age (,40 or40 y). After successfully completing phase 1, eligible subjectswere allocated to the next available treatment in the computer-

generated randomized list of treatments for their corresponding

strata.

All P values were 2-sided, and a P value 0.05 was con-sidered to indicate statistical significance. Values are provided as

mean 6 SEM unless otherwise indicated. The analysis was

1204 DELBRIDGE ET AL

byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


3/12

performed by using Minitab statistical software (version 14.0;

Minitab, Inc, State College, PA).

The statistical analysis of study completion measurements

consisted of 2 distinct analyses. The first was a primary intention to

treat (ITT) analysis that consisted of 2 different approaches to

missing data. The first approach used was the last observation

carried forward (LOCF). The second approach was return to

baseline (RTB) for study dropouts. The second analysis considered

study completers only (CO). For each measurement analysis, 2general linear models (GLMs) were fit: the first dealing with

measurement changes over phase 1+2 from baseline and the

second dealing with measurement changes during phase 2. Each

GLM had treatment group (HC or HP), sex (male or female), and

the interaction between treatment and sex as fixed effects.

The response variable for the first GLM (phase 1+2) was the

change in the measurement of interest at study completion

compared with baseline and included the measurement at baseline,

baseline weight, and age as covariates. The response variable for

the second GLM (phase 2) was the change in the measurement of

interest at study completion compared with the measurement at

completion of phase 1. Change of measurement at completion of

phase 1 compared with baseline and age were included as covari-ates. To assess the strength of linear relations between paired data,

we used Pearson correlation coefficients. The associated P value is

reported to test for nonzero correlation.

RESULTS

Baseline characteristics

The subjects were well matched according to baseline char-

acteristics (Table 1). There were no significant differences in

baseline demographic, anthropometric, or metabolic variables

between the subjects who went on to be randomly assigned to

either of the 2 dietary groups in phase 2.

Attrition

As can be seen in Figure 1, 180 persons were screened to take

part in the study. One man was excluded at the screening stage

because of a diagnosis of hemochromatosis. Of the 179 subjects

that began phase 1 of the study, 20 voluntarily withdrew because

of intolerance to the meal replacement. Fourteen subjects did not

succeed in losing the required 10% of their body weight. Four of

these 14 subjects were not included in the analysis because they

had not adequately complied with the study protocol because

they attended less than one-half of the required visits during

phase 1. Four subjects who successfully completed phase 1

decided not to continue in the study. One hundred and forty-one

(70 men and 71 women) subjects (78% of the initial number

recruited) completed phase 1 and progressed to phase 2, in

which they were randomly assigned to either an HP or HC diet.

There were 35 men in both the HC and HP diet groups and 35women in the HC group and 36 in the HP group.

After randomization, approximately equal numbers of subjects

withdrew from each dietary treatment group. Three subjects who

withdrew from the HP group were not included in the analysis

because they did not attend a sufficient number of scheduled

appointments to be considered compliant with the protocol. One

subject withdrew from the study after randomization and before

commencing any of the phase 2 procedures. No subject withdrew

because of symptomatic adverse effects of the diet. Forty-two

subjects completed the HP diet, and 42 subjects completed the HC

diet, which indicated that 59.6% of the 141 subjects who com-

menced phase 2 completed this phase. Except for age and blood

pressure measurements, there were no significant differences inbaseline characteristics between the subjects who completed the

study and those who dropped out. On average, the subjects who

completed the study were older (46.4 6 1.1 y) than those who

dropped out (40.0 6 1.0 y) (P , 0.001). Compared with the

subjects who dropped out, the subjects who completed the study

also had, on average, a higher baseline systolic blood pressure

(136.3 6 1.8 mm Hg compared with 129.1 6 1.6 mm Hg; P =

0.004) and a higher baseline diastolic blood pressure (85.3 6 1.2

mm Hg compared with 82.0 6 1.0 mm Hg; P = 0.032).

Anthropometric data

At the completion of phase 1, the mean weight loss experi-

enced by all subjects was 16.5 6 0.5 kg (P , 0.001), as seen in

Table 2. This equated to a mean weight loss of 14.7% (99% CI:

215.47,213.92; P, 0.001). All anthropometric measurements

decreased significantly during phase 1 for all subjects. The

subjects who completed both phases of the trial experienced

a small but significant increase in weight and BMI during phase

2 (weight: 3.6 6 0.9kg, P , 0.001; BMI: 1.260.3, P , 0.001),

as evident in Table 3. However, weight at completion of the

TABLE 1

Baseline (before phase 1) characteristics of the subjects who began phase 21

Characteristic HC (n = 35 M, 35 F) HP (n = 35M, 36 F)

Age (y) 44 6 1.1 43.7 6 1.4

Weight (kg) 109.4 6 2.6 114.0 6 3.0

BMI (kg/m2

) 38.6 6 0.8 39.3 6 0.8

Waist circumference (cm) 113.5 6 1.8 116.5 6 2.0

Fat (%) 42.4 6 1.0 (63) 41.7 6 1.0 (68)

Systolic blood pressure (mm Hg) 131 6 2.1 135 6 1.8

Diastolic blood pressure (mm Hg) 83 6 1.3 85 6 1.3

Total cholesterol (mmol/L) 5.4 6 0.1 (66) 5.5 6 0.2 (60)

HDL (mmol/L) 1.2 6 0.04 (65) 1.2 6 0.03 (59)

Triglycerides (mmol/L) 1.9 6 0.2 (65) 2.1 6 0.4 (61)

1All values are means 6 SEMs; n in parentheses. HC, high carbohydrate; HP, high protein. There were no significant

differences between groups.

HIGH-PROTEIN DIET AND WEIGHT MAINTENANCE 1205

byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


4/12

study was on average 14.5 6 1.2 kg (P , 0.001) less than the

baseline weight for all completing subjects, which equates to

a decrease of 12.98% (99% CI:214.96,211.00) in body weight

(Table 4). There was no difference between the HP and HC

groups in weight or BMI changes during phase 1 or phase 2:

15% of the HC group compared with 26% of the HP group lostadditional weight during phase 2 (P = 0.205). During phase 2,

42.5% of the HC group and 47.6% of the HP group regained

,20% of the initial weight lost during phase 1.

Whereas all subjects experienced a significant decrease in both

fat mass and fat-free mass during phase 1, all subjects experi-

enced a significant increase in fat mass and no significant change

in fat-free mass in phase 2. When the results for phase 1 and 2 are

taken together, from the beginning to the end of the study, all

subjects experienced significant decreases in weight, fat mass,

and fat-free mass, with no difference observed between the

groups. Significant reductions in waist and hip circumferences

were achieved during phase 1, with no further changes in these

measurements during phase 2. The reductions, however, were

maintained until the end of the study by those completing the

study (Table 4).

The ITT analysis of those who withdrew from the study (Table

4), in which baseline weights (RTB) or the last recorded (LOCF)

weight were carried forward and analyzed, still showed that there

was no difference in weight loss between the HP and HC dietary

groups after 12 mo of dietary counseling.

Lipids

As is evident in Table 1, mean lipid concentrations were in the

normal range at baseline for all subjects. For all subjects who

completed phase 1, total cholesterol and triglycerides decreased

significantly after the VLED (Table 2). At the end of the study,

total and LDL-cholesterol and triglyceride concentrations in

those subjects who had gone on to complete phase 2 (CO)

remained significantly reduced compared with baseline con-

centrations (Table 4). Although the changes in LDL during both

phases I and II were not significant, LDL cholesterol decreased

significantly from baseline by the end of phase 2 (Table 4). There

was no change in HDL during phase 1 (Table 2); however, duringthe 12 mo of the dietary treatment in phase 2, a significant in-

crease in mean HDL (P , 0.001) was observed (Table 3) and

was sustained until the end of the study by the completing

subjects (Table 4). There was no difference in lipid concen-

trations between the 2 dietary groups at baseline or at the end of

phase 1 or phase 2. The significant changes in total, LDL, and

HDL cholesterol and in triglycerides experienced by the subjects

who completed phases I and II were not observed in the ITT

analyses (Table 4).

Blood pressure

Blood pressure measured in the 2 dietary groups was within thenormal range at the commencement of the study (Table 1) and

remained so for the duration of the study. Pairwise comparisons

across all subjects showed a significant mean decrease in SBP of

13.2 6 1.4 mm Hg (P , 0.001) from baseline to the end of

phase 1 (Table 2). After the subjects were randomly assigned to

the treatment groups at month 3, no significant difference in

mean SBP decrease from baseline was detected between the 2

groups (CO) at this time (P = 0.375). The mean decrease in SBP

during phase 1 was 12.3 6 2.1 mm Hg for the HC group and

14.9 6 2.1 mm Hg for the HP group (Table 2).

Although subjects in both treatment groups experienced a sig-

nificant increase in SBP during phase 2 (4.7 6 1.9 mm Hg, P ,

0.014; Table 3), all subjects who completed the study experiencedan overall decrease in SBP from baseline of 11.1 6 1.7 mm Hg

(P , 0.001) (Table 4). For those subjects who completed the

study (CO), the mean decrease in SBP from baseline to the end of

the study was 14.3 6 2.4 mm Hg for the HP group and 7.7 6 2.2

mm Hg for the HC group (Table 4), and the difference between

the 2 groups was statistically significant (P , 0.045). When this

finding was investigated further by using a GLM, the dietary

treatment (HP) was found to be a significant fixed effect in

influencing SBP in subjects who completed the study and in the

ITT analyses (Table 5). The estimated model coefficient for

treatment indicated that a larger decreases in SBP can be expected

with an HP diet after weight loss. The subjects baseline SBP and

FIGURE 1. Status of subjects throughout the trial. 1Subjects with

a diagnosis of hemochromatosis that precluded participation.2

Poorcompliance with protocol; did not attend a sufficient number of studyvisits.

3Decided not to continue in study after randomization.


byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


5/12

baseline weight, along with age, were found to be significant

covariates.

Further statistical analyses were then conducted to investigate

the effect of the dietary treatments on SBP during phase 2. A

GLM was used that considered sex, treatment, sex treatment

interaction, age, and a reduction in SBP during phase 1 ( Table

6). It was found that the dietary treatment was a significant fixedeffect in the CO and LOCF analyses and was nearly significant

in the RTB analysis; the estimated model coefficient indicated

larger expected increases in SBP for the HC diet. The SBP re-

duction at month 3 was a significant covariate in the CO and ITT

analyses.

A significant mean decrease in DBP of 8.56 1.1 mm Hg (P,

0.001) was detected across all subjects at the end of month 3

(Table 2), with no significant difference between treatment

group reductions in DBP after randomization (P = 0.283). As

was observed for SBP, all CO subjects experienced a further

small but significant increase in DBP during phase 2 (3.7 6 1.4

mm Hg, P , 0.012; Table 3), but there was no difference in the

increase between the 2 dietary groups. Again a GLM was usedthat considered sex, treatment, sex treatment interaction, age,

and a reduction in DBP during phase 1. Only the reduction in

DBP reduction at month 3 was a significant covariate for pre-

dicting reduction in DBP during phase 2 for the CO and in the

ITT analysis (Table 6).

At the studys completion, the mean decrease in DBP from

baseline for the CO was 7.8 6 1.8 mm Hg for the HP group and

2.4 6 1.7 mm Hg for the HC group (P , 0.034) (Table 4).

However, pairwise comparisons to account for individual sub-

jects showed that both groups experienced comparable mean

increases in DBP of 3.7 mm Hg from month 3 to study

completion (Table 3). A GLM using dietary treatment as a fixed

effect did not show it to be significant. The covariate baseline

DBP was significant for the CO and in the ITT, whereas weight

was a significant covariate for the CO (Table 5).

Dietary complianceCompliance with the dietary recommendations was assessed

with food diaries and by the objective measures of urea ex-

cretion and weight maintenance. Food diaries in conjunction

with 24-h urine collections were satisfactorily completed by 6

menand 5 women in the HP group at month9 and by 10 men and

6 women in the HP group at month 15. In the HC group,

complete food diaries and 24-h urine collections were ade-

quately completed by 5 men and 5 women at month 9 and by 9

men and 4 women at month 15. There was no difference in

weight regain between those subjects who did and did not

supply food diaries in conjunction with urine collections; the

subjects who did comply regained 2.9 6 0.7 kg and those who

did not comply regained 3.1 6 0.5 kg by the end of the study(P = 0.375)

Estimated energy requirements for weight maintenance were

calculated for the mean body weight of the men and women.

When these estimates were compared with the energy intakes

calculated from the subjects food records, it was apparent that,

on average, women in both study groups had underestimated

their food intake and thus their energy intakes (mean calculated

intake = 2044 kcal; mean reported intake = 1370 kcal; P ,

0.001). The men, however, did not show such disparity in their

reported and calculated energy intakes (mean calculated in-

take = 1771 kcal; mean reported intake = 1765 kcal; P =

0.961).

TABLE 2

Measurement changes from baseline to the end of phase 1 for all individuals who completed phase 1 or who were

randomly assigned to the high-carbohydrate (HC) or high-protein (HP) group at the end of phase 11

Measurement change Al l (n = 159) P HC (n = 70) HP (n = 71) P

Weight (kg) 216.5 6 0.5 ,0.001 217.6 6 0.8 217.4 6 0.7 0.837

BMI (kg/m2

) 25.7 6 0.2 ,0.001 26.2 6 0.3 25.9 6 0.2 0.379

Waist (cm)2

214.2 6 0.5 ,0.001 214.3 6 0.8 215.2 6 0.7 0.394

Hips (cm)3

211.0 6 0.3 ,0.001 210.8 6 0.5 211.7 6 0.4 0.172

Fat mass (kg) 212.8 6 0.6 ,0.001 213.8 6 0.84 213.5 6 0.9 0.725

Fat-free mass (kg) 23.4 6 0.4 ,0.001 23.7 6 0.4 23.6 6 0.7 0.908

Cholesterol change4

20.65 6 0.08 ,0.001 20.65 6 0.11 20.59 6 0.09 0.641

HDL change5

20.003 6 0.02 0.878 20.015 6 0.02 0.003 6 0.03 0.605

LDL change6

0.14 6 0.5 0.763 0.59 6 0.92 20.33 6 0.09 0.323

Triglyceride change7 20.90 6 0.19 ,0.001 20.87 6 0.16 20.62 6 0.13 0.210

SBP change8

213.2 6 1.4 ,0.001 212.3 6 2.1 214.9 6 2.1 0.375

DBP change8

28.5 6 1.1 ,0.001 27.4 6 1.4 29.8 6 1.8 0.283

Pulse change9

24.7 6 0.9 ,0.001 24.5 6 1.2 25.3 6 1.6 0.668

1All values are means 6 SEMs; n = 159 subjects who completed phase 1, regardless of whether or not they continued

on to phase 2. DBP, diastolic blood pressure; SBP, systolic blood pressure.2 n = 149 (all), n = 66 (HC), n = 68 (HP).3 n = 148 (all), n = 65 (HC), n = 68 (HP).4 n = 125 (all), n = 63 (HC), n = 54 (HP).5 n = 124 (all), n = 62 (HC), n = 54 (HP).6

n = 117 (all), n = 58 (HC), n = 52 (HP).7 n = 125 (all), n = 62 (HC), n = 55 (HP).8 n = 130 (all), n = 62 (HC), n = 59 (HP).9

n = 98 (all), n = 45 (HC), n = 46 (HP).


byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


6/12

TABLE

3

Measurementchangesfromtheendofphas

e1totheendofphase2forallindividualswhocompletedphase1orwhowererandomlyassignedtothehigh-carbohydrate(HC)or

high-protein(HP)group

1

CO

CO

LOC

F

RTB

Measurementchange

All(n=82)

P

HC(n=40)

HP(n=4

2)

P

HC(n=70)

HP(n

=68)

P

HC(n=70)

HP(n=68)

P

Weight(kg)

3.6

6

0.9

,0.0

01

4.3

6

1.4

3.0

61

.1

0.4

76

3.8

6

0.9

3.1

6

0.8

0.5

44

9.5

6

1.2

8.3

6

1.2

0.4

82

BMI(kg/m2)

1.2

6

0.3

,0.0

01

1.4

6

0.5

1.0

60

.4

0.5

88

1.3

6

0.3

1.1

6

0.3

0.6

24

3.3

6

0.4

2.8

6

0.4

0.3

67

Waist(cm)2

0.0

16

0.8

8

0.9

94

0.9

26

1.5

20.8

161

.0

0.3

38

0.0

56

0.9

4

0.68

6

0.7

2

0.5

99

6.1

6

1.2

5.5

6

1.2

0.7

35

Hips(cm)2

0.4

36

0.7

7

0.5

73

0.8

96

1.2

0

0.0

260

.94

0.5

82

0.4

36

0.7

2

1.02

6

0.6

3

0.5

34

5.2

6

0.9

4.4

6

0.9

0.5

59

Waist-hipratio

2

20.0

06

0.0

0

0.2

44

0.0

06

0.0

1

20.0

160

.00

0.1

83

0.0

06

0.0

0

0.00

6

0.0

0

0.4

82

0.0

26

0.0

05

0.0

16

0.0

1

0.4

43

Fatmass(kg)

3.7

6

1.2

0.0

04

3.2

6

1.4

4.2

62

.2

0.6

85

1.8

6

1.0

3.0

6

1.2

0.4

30

6.6

6

1.3

7.0

6

1.3

0.8

27

Fat-freemass(kg)

0.6

16

0.3

6

0.0

94

0.8

96

0.4

3

0.3

460

.58

0.4

48

0.8

56

0.5

6

20.17

6

0.3

8

0.1

49

2.1

6

0.6

0.7

86

0.5

5

0.1

26

Totalcholesterol3

0.4

56

0.1

1

,0.0

01

0.3

66

0.1

7

0.5

460

.13

0.3

94

0.4

16

0.1

1

0.40

6

0.0

9

0.9

45

0.4

86

0.1

0

0.4

16

0.0

9

0.6

44

HDLcholesterol

0.1

76

0.0

2

,0.0

01

0.1

66

0.0

3

0.1

860

.04

0.7

60

0.1

36

0.0

2

0.15

6

0.0

3

0.6

11

0.1

26

0.0

2

0.1

26

0.0

3

0.9

49

LDLcholesterol3

20.5

16

0.7

1

0.4

79

21.4

16

1.5

0

0.3

360

.09

0.2

49

20.6

06

0.7

6

0.24

6

0.0

6

0.2

73

20.5

76

0.7

6

0.2

16

0.0

7

0.3

15

Triglycerides

3

0.2

16

0.0

9

0.0

24

0.3

06

0.1

1

0.1

360

.15

0.3

35

0.2

16

0.0

6

0.09

6

0.0

9

0.2

41

0.3

86

0.0

8

0.2

76

0.1

2

0.4

22

SBP4

4.7

6

1.9

0.0

14

7.5

6

2.3

1.9

62

.8

0.1

29

5.8

6

1.5

1.2

6

1.7

0.0

42

6.9

6

1.5

4.7

6

1.9

0.3

64

DBP4

3.7

6

1.4

0.0

12

3.7

6

1.9

3.7

62

.2

0.9

86

3.4

6

1.5

2.3

6

1.4

0.5

80

5.3

6

1.3

4.1

6

1.5

0.5

43

Pulse5

21.4

6

1.3

0.2

93

20.6

6

2.1

22.3

61

.5

0.5

28

21.1

6

0.9

23.4

6

1.2

0.1

36

20.2

46

1.1

21.1

6

1.5

0.6

69

1

Allvaluesaremeans6

SEMs.CO,

completersonly;LOCF,

lastobservationcarriedforward;RTB,returntobaseline;SBP,s

ystolicbloodpressure;DBP,diastolicbloodpressure.

2

n

=76(all),n

=36(HC/CO),n

=40(HP/CO).

3

n

=75(all),n

=36(HC/CO),n

=39(HP/CO).

4

n

=67(all),n

=33(HC/CO),n

=34(HP/CO).

5

n

=51(all),n

=27(HC/CO),n

=24(HP/CO).


byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


7/12

TABLE

4

Measurementchangesfrombaselinetothee

ndofphase2forallindividualswhocomplete

dphases1and2orwhowererandomlyassignedtothehigh-carbohydrate(HC)orhigh-pro

tein(HP)groupattheend

ofphase11

Measurementchange

CO

CO

LOCF

RTB

All(n=82)

P

HC(n=40)

HP(n=

42)

P

HC(n=70)

HP(n=68)

P

HC(n=70)

HP(n=68)

P

Weight(kg)

214.5

6

1.2

,0.0

01

214.3

6

2.0

214.8

6

1.5

0.8

41

213.8

6

1.3

21

4.3

6

1.1

0.7

70

28.1

6

1.4

29.1

6

1.3

0.6

10

BMI(kg/m2)

25.1

6

0.5

,0.0

01

25.1

6

0.8

25.1

6

0.5

0.9

79

24.9

6

0.5

2

4.8

6

0.4

0.8

82

22.9

6

0.5

23.2

6

0.4

0.7

47

Waist(cm)2

216.0

6

1.1

,0.0

01

215.4

6

1.7

216.5

6

1.4

0.5

91

214.1

6

1.1

21

4.5

6

1.1

0.7

93

28.1

6

1.3

29.7

6

1.3

0.3

76

Hips(cm)2

211.5

6

1.0

,0.0

01

210.5

6

1.7

212.4

6

1.2

0.3

65

210.3

6

1.0

21

0.7

6

0.9

0.7

64

25.5

6

1.1

27.3

6

1.0

0.2

47

Waist-hipratio

2

20.0

56

0.0

0

,0.0

01

20.0

56

0.0

1

20.0

56

0.0

1

0.6

02

20.0

46

0.0

0

20

.046

0.0

0

0.8

06

20.0

36

0.0

0

20.0

36

0.0

0

0.9

99

Fatmass(kg)

29.9

6

1.4

,0.0

01

210.5

6

1.7

29.3

6

2.1

0.6

59

210.9

6

1.2

2

9.9

6

1.5

0.5

85

25.5

6

1.1

25.7

6

1.4

0.9

28

Fat-freemass(kg)

23.6

6

0.4

,0.0

01

22.9

6

0.5

24.2

6

0.6

0.1

14

22.9

6

0.5

2

3.6

6

0.6

0.3

97

21.5

6

0.3

22.6

6

0.5

0.0

66

Totalcholesterol3

20.3

96

0.0

9

,0.0

01

20.3

06

0.1

4

20.4

86

0.1

1

0.3

44

20.2

26

0.1

0

20

.286

0.0

9

0.6

18

20.1

56

0.0

7

20.2

76

0.0

7

0.2

29

HDLcholesterol3

0.2

06

0.0

2

,0.0

01

0.2

16

0.0

3

0.1

96

0.0

4

0.7

39

0.1

16

0.0

3

0

.146

0.0

3

0.4

75

0.1

06

0.0

2

0.1

16

0.0

3

0.7

47

LDLcholesterol4

20.3

06

0.0

9

0.0

01

20.2

56

0.1

3

20.3

56

0.1

2

0.5

69

20.1

66

0.0

9

20

.176

0.0

9

0.9

40

20.1

26

0.0

6

20.2

16

0.0

7

0.3

31

Triglycerides

3

20.7

46

0.1

3

,0.0

01

20.9

06

0.2

3

20.6

06

0.1

5

0.2

70

20.6

26

0.1

3

20

.566

0.1

2

0.7

38

20.4

46

0.1

2

20.2

56

0.1

0

0.5

55

SBP5

211.1

6

1.7

,0.0

01

27.7

6

2.2

214.3

6

2.4

0.0

45

25.0

6

1.6

21

1.7

6

1.8

0.0

06

23.9

6

1.2

28.2

6

1.6

0.0

37

DBP5

25.2

6

1.3

,0.0

01

22.4

6

1.7

27.8

6

1.8

0.0

34

23.1

6

1.4

2

6.3

6

1.5

0.1

34

21.2

6

0.9

24.5

6

1.1

0.0

27

Pulse5

27.4

6

1.4

,0.0

01

26.2

6

1.7

28.6

6

2.1

0.3

98

23.9

46

1.1

2

7.1

6

1.4

0.0

84

23.1

6

0.9

24.7

6

1.3

0.3

11

1

Allvaluesaremeans6

SEMs.CO,

completersonly;LOCF,

lastobservationcarriedforward;RTB,returntobaseline;SBP,s

ystolicbloodpressure;DBP,diastolicblood

pressure.

2

n

=77(all),n

=37(HC/CO),n

=40(HP/CO).

3

n

=66(all),n

=33(HC/CO),n

=33(HP/CO).

4

n

=60(all),n

=28(HC/CO),n

=32(HP/CO).

5

n

=75(all),n

=36(HC/CO),n

=39(HP/CO).


byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


8/12

Actual protein intake was calculated from urinary urea mea-

sured in 24-h urine volume by using the following formulas (10):

Urinary urea in mg=dL urinary urea

3242h urine volume=100 1

Urinary urea nitrogen Equation 130:47 2

Nonurea nitrogen mg=d weight of the subject kg331

3

Protein intake g=d Equation 2 Equation 3=100036:25

4

The subjects were assumed to be in nitrogen balance and from

this we could calculate nitrogen and thus protein intake. From

measured urinary urea, the average protein intake at month 9 (half

way through phase 2) was calculated to be 95.9 g/d for the HC

group and 120.8 g/d for the HP group (Table 7). At month 15(end of phase 2), the mean protein intakes for the HC and HP

groups had increased significantly from month 9 to 110.2 and

136.1 g/d respectively (P , 0.006).

The mean (6SD) total urine collected was 2370 6 808 mL

for the whole study population, with volumes ranging from 1092

to 3959 mL for the 3 time points at which it was measured. This

is in line with expected urine output over 24 h. There were no

significant differences in urine volume between the HP and

HC groups at any time point. The correlation between protein

intake estimated from food diaries and calculated from urine

urea output was investigated by using Pearsons correlation

coefficients. A significant positive correlation of 0.329 (P =

0.036) was seen at month 9 and 0.353 (P = 0.010) at month 15.A paired-sample t test showed no difference in protein intake

estimated by food diaries or calculated by urinary urea for the

HC group at month 9 (Table 7). However, the protein intake

calculated from urine urea was significantly higher than the

estimated intake from food diaries for HC group at month 15

and for the HP group at months 9 and 15. Two-sample t tests

showed no significant difference in protein intake between the

TABLE 5

P values for fixed effects and covariates analysis in which the response is the measurement change from baseline to study completion for the general linear

models (phases 1 + 2)1

Measurement

SBP DBP

LOCF RTB CO LOCF RTB CO

Fixed effects

Treatment 0.006 0.050 0.032 0.179 0.028 0.064

Sex 0.753 0.776 0.927 0.059 0.591 0.287

Treatment sex 0.841 0.325 0.602 0.988 0.821 0.656

Covariates

Baseline ,0.001 (20.57) ,0.001 (20.36) ,0.001 (20.48) ,0.001 (20.78) ,0.001 (20.37) ,0.001 (20.63)

Weight ,0.001 (0.20) 0.002 (0.14) 0.008 (0.25) 0.058 (0.07) 0.074 (0.05) 0.043 (0.12)

Age 0.004 (0.30) 0.217 (0.12) 0.015 (0.40) 0.298 (0.08) 0.520 (20.04) 0.781 (0.03)

1Coefficients for covariates are in parentheses. SBP, systolic blood pressure; DBP, diastolic blood pressure; CO, completers only; LOCF, last observation

carried forward; RTB, return to baseline.

TABLE 6

P values for fixed effects and covariates analysis in which the response is the measurement change from month 3 to study completion for the general linear

model (phase 2)1

Measurement

SBP DBP

CO LOCF RTB CO LOCF RTB

Fixed effects

Treatment 0.003 0.004 0.051 0.195 0.254 0.045

Sex 0.317 0.299 0.829 0.290 0.104 0.477

Treatment sex 0.970 0.131 0.837 0.807 0.654 0.933

Covariates

Month 3 change ,0.001 (20.57) ,0.001 (20.42) ,0.001 (20.61) ,0.001 (20.54) ,0.001 (20.47) ,0.001 (20.73)

Age 0.305 (0.14) 0.163 (0.13) 0.901 (0.01) 0.550 (20.07) 0.815 (0.02) 0.255 (20.07)

1Coefficients for covariates are in parentheses. SBP, systolic blood pressure; DBP, diastolic blood pressure; CO, completers only; LOCF, last observation

carried forward; RTB, return to baseline.


byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


9/12

HC and HP groups estimated from food diaries at month 9, but

there were significant differences between the protein intakes

estimated from food diaries at month 15 and calculated from

urinary urea between the 2 groups at months 9 and 15 (Table 7).

The food diaries of the HC group indicated that the groupconsumed, on average, significantly greater than the recom-

mended 15% of their energy from protein at both months 9 and 15

and significantly less than the recommended 55% of their energy

from carbohydrate (Figure 2). The mean macronutrient con-

sumption of the HP group reported in food diaries was not sig-

nificantly different from the recommended intakes. Because of

the small sample sizes, the P values reported for this analysis

were obtained by using Wilcoxons signed rank test. Both the HP

and HC groups consumed minimal amounts of alcohol (,

3% ofenergy intake), and there were no significant differences between

the 2 groups in their consumption at either time point.

The mean fiber intake reported in the food diaries was 21.7 g/d,

with no significant difference between the 2 groups or over the 2

TABLE 7

Protein intake reported in food diaries and calculated from urinary urea in the high-carbohydrate (HC) and high-protein (HP) groups

Month 0 Month 9 Month 15

n

Calculated from

urine urea n Food diary

Calculated from

urine urea P1 n2 Food diary

Calculated from

urine urea P1

g/d g/d g/d g/d g/d

HC group 65 97.2 6 4.83

10 90.8 6 6.2 95.9 6 3.6 0.553 13 87.4 6 8.1 110.2 6 4.8 0.004

HP group 66 100.36 3.6 11 103.9 6 10.1 120.8 6 5.2 0.011 16 112.6 6 7.6 136.1 6 5.7 0.035

P4

0.145 0.139 0.007 0.001 0.040

1Paired-sample t test (intragroup comparison).

2Approximately one-third of the participants who provided food diaries and 24-h urine samples at month 9 also provided them at

month 15.3

Mean 6 SEM (all such values).4

Independent sample t test (intergroup comparison).

FIGURE 2. Recommended compared with reported macronutrient intakes in the high-protein (HP) and high-carbohydrate (HC) diet groups at months 9 and

15. Month 9 represents the half-way point of phase 2, and month 15 represents the end of phase 2. HP group: n = 11 at month 9, and n = 16 at month 15; HCgroup: n = 10 at month 9, and n = 13 at month 15. The dashed lines represent recommended intakes (HC diet: 15% protein, 55% carbohydrate, and 30% fat;HP: 30% protein, 40% carbohydrate, and 30% fat). The bold lines represent the median data for the group, and P values represent the differences betweenreported and recommended intakes.


byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


10/12

time periods. There were no significant differences in any other

reported nutrient intakes between the 2 groups.

DISCUSSION

Weight maintenance

Dietary studies investigating HP diets are usually conducted in

the context of weight loss. Although randomized trials have

indicated that carbohydrate-restricted diets are advantageous

compared with higher-carbohydrate diets in achieving weight

loss over 6 mo, longer trials have shown that the advantage is not

sustained for 1 y (4, 11, 12). In this study, significant weight loss

after 3 mo of intensive dietary treatment was maintained for

a year by following either an HC or an HP weight-maintenance

diet. Others have found that diet composition does not influence

weight regain when protein intake is kept constant and energy

intake is ad libitum (13).

Body composition

Some studies have found the loss of fat mass to be significantly

greater in subjects following an HP diet as opposed to a higher-

carbohydrate diet (3, 14), even with equivalent energy intakes and

weight loss between the groups (15). However, the findings are

inconsistent. One study observed that subjects following a low-

carbohydrate diet lost significantly more lean tissue than did

subjects following a low-fat, energy-restricted diet (16). Clifton

et al pooled data from 3 studies of 12-wk intakes of isocaloric HP

(2740% energy) and standard protein (1520% of energy)

intakes (17). No difference was found between the diets for

weight, fat mass, and lean mass loss.

Cardiovascular disease risk factors

Numerous studies have provided convincing evidence for the

benefits of HP diets on cardiovascular disease risk factors (18).

Energy-restricted, HP diets have frequently been found to de-

crease triglycerides (4, 19, 20) and increase HDL cholesterol (4,

20) compared with energy-restricted, higher-carbohydrate diets.

These improvements occurred even when there was no difference

in weight loss between the HP and HC or low-carbohydrate

treatment groups (15, 21). When protein intake was constant but

carbohydrate intake was reduced from 60% to 40%, the same

reduced triglyceride and increased HDL pattern followed (22).

It is plausible that the decreased carbohydrate-load charac-

teristic of an HP diet can induce a reduction in blood pressurethrough a reduced insulin response. Insulin has been found to

increase blood pressure (23, 24); subsequently, a reduction in

insulin secretion may decrease blood pressure. Although there is

evidence to suggest that HP diets induce a reduction in systolic

and diastolic blood pressure in the short term (25, 26), a sys-

tematic review of low-carbohydrate diets found no change is

systolic blood pressure in participants following such diets (27).

In this study, all subjects experienced a reduction in blood

pressure during phase 1 while following the VLED. However,

during phase 2, only subjects in the HP diet group were able to

sustain the reduction. The blood pressure of subjects in the HC

group significantly increased during this time.

Diet

It has been suggested that, compared with high-GI carbohy-

drates, low-GI carbohydrates may delay the return of hunger and

reduce subsequent food intake (28). In the present study, both

groups were prescribed low-GI carbohydrates, which possibly

contributed to the overall satiety of both diets.

The increased satiety arising from an HP diet has been well

established. In an ad libitum setting, this results in a spontaneousdecrease in food intake (3, 14, 29). However, in this study, the

subjects in the HC and HP groups were prescribed isocaloric meal

plans, and the data suggest that the subjects were compliant with

macronutrient recommendations. Had the dietary prescriptions

been more along the lines of ad libitum HP or HC diets,

a spontaneous reduction in food intake may have been observed.

Any potential satiating effect of protein on appetite may have

been obscured in this study by the macronutrient and energy

restrictions imposed on subjects through the meal plans provided.

Dietary compliance

A strength of this study was its free-living setting, in which nofood was provided to subjects. However, this strength was also

a weakness because it is difficult to measure dietary compliance.

Reported energy intakes in both the HP and HC groups remained

isocaloric, which supported the lack of difference in body weight.

The difference in reported and calculated energy intakes for

women suggested that the women had underestimated their food

and thus energy intakes. Such a discrepancy was not found for the

men. These findings are not unprecedented. A comparison of

energy intakes determined by food diaries and energy expen-

diture measured by doubly labeled water found that middle-aged

women taking part in a long-term diet-intervention trial under-

estimated their energy intake (30). In the current study, neither

group reported a significant increase in energy intake betweenmonths 9 and 15, despite a significant increase in weight ex-

perienced by the HP group (month 9, 1520 kcal; month 15, 1589

kcal; P = 0.620) and the HC group (month 9, 1639 kcal; month

15, 1678 calories; P = 0.864).

In this study, urinary nitrogen was not measured, but rather was

calculated from urinary urea. Although this technique is not

commonly used, its use in dietary studies such as this is not

unprecedented (31, 32), and it has been found to be a valuable

approach to assessing nutritional therapy (33).

According to the analysis of food diaries, the HP group

reported excellent compliance with their macronutrient recom-

mendations at months 9 and 15 (Figure 2). The protein intake of

this group, calculated from urinary urea, supported the food diaryreports. The HC group, however, struggled to consume the

recommended amount of carbohydrate (55%) and to limit their

protein intake to 15%. The most recent data on Australian food

intake reports that the average intake of protein for Australian

males and females is 17% of energy intake (34). Our results

suggest that subjects in the HC group were unable to sustain the

prescribed lowered intake of protein. Although the HC groups

mean protein intake during the study (21.7% of energy intake;

87.4 g/d indicated by the food diary or 110.2 g/d calculated) was

significantly higher than the 15% prescribed, it was still sig-

nificantly lower than the HP groups intake of 30.2% at both time

points. (Table 7 and Figure 2). Although satiety was not formally


byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


11/12

measured in this study, it is possible that the HC group was not

sufficiently satiated with the 15% protein recommendation. It is

also tenable that the HP group may have been more truthful in

recording their food intake. The authors are cognizant that the

numbers of subjects completing food diaries and urine collections

was disappointingly low, and, as such, are conservative in their

declarations of dietary compliance.

Attrition

Dropout rates in year-long studies of persons following HP

diets have ranged from 31% to 48% and from 37% to 50% for

persons following low-fat diets (35). The percentage of subjects

who dropped out of the weight-maintenance dietary phase (phase

2) of our study was in keeping with these findings (41%). The

percentage of subjects who began and completed both phases 1

and II of the dietary treatment was 47%. Although the number of

subjects who withdrew from the study was high, the rate of

attrition in both dietary treatment groups was equal, which

suggests that the reason for withdrawal was not diet related.

To preserve the validity of comparisons between treatment

groups, the ideal outcome would have been to follow all subjectswho underwent randomization to completion of the study (36).

However, although this was not possible, we do not view this as

a limitation of the study because the available data were handled

by using the most appropriate statistical methods (CO, LOCF,

and RTB). Furthermore, the sample size was large and the

analyses using the RTB and LOCF methods supported the

findings of those that completed the study, which thus

strengthened the integrity of our results and provided reassurance

that the results were not unduly biased by attrition.

Conclusions

The present study investigated whether dietary compositionplays a role in improving weight maintenance after weight loss.

Both the HP and HC groups sustained weight losses that were

statistically and clinically significant. This indicated that free-

living overweight and obese people were able to comply with

dietary recommendations and keep weight off over 12 mo. No

statistically significant difference in maintained weight loss was

observed between an HP and an HC diet. The effect of the in-

creased protein intake on appetite may have been masked by the

energy prescription imposed on subjects.

The authors responsibilities were as followsEAD: had full access to all

of thedata in the study andtakes responsibility forthe integrity of the data and

the accuracy of the data analysis; LAP: performed the statistical analyses;

JEP: contributed to the writing and editing of the text; and JP: oversaw

the study design, was involved in the data collection, and contributed

to the writing and editing of the text. JP sits on the Advisory Board for Opti-

fast, the VLED product used in phase 1 of the trial. None of the other authors

had a conflict of interest to disclose.

REFERENCES1. McMillan-Price J, Petocz P, Atkinson F, et al. Comparison of 4 diets of

varying glycemic load on weight loss and cardiovascular risk reduction

in overweight and obese young adults: a randomized controlled trial.Arch Intern Med 2006;166:146675.

2. Baba NH, Sawaya S, Torbay N, et al. High protein vs high carbohydratehypoenergetic diet for the treatment of obese hyperinsulinemic subjects.Int J Obes Relat Metab Disord 1999;23:12026.

3. Skov AR, Toubro S, Ronn B, et al. Randomized trial on protein vs

carbohydrate in ad libitum fat reduced diet for the treatment of obesity.

Int J Obes Relat Metab Disord 1999;23:52836.4. Foster GD, Wyatt HR, Hill JO, et al. A randomized trial of a low-

carbohydrate diet for obesity. N Engl J Med 2003;348:208290.5. Mustajoki P, Pekkarinen T. Very low energy diets in the treatment of

obesity. Obes Rev 2001;2:6172.6. Anderson JW, Konz EC, Frederich RC, Wood CL. Long-term weight-

loss maintenance: a meta-analysis of US studies. Am J Clin Nutr

2001;74:57984.7. Lejeune MP, Kovacs EM, Westerterp-Plantenga MS. Additional protein

intake limits weight regain after weight loss in humans. Br J Nutr 2005;

93:2819.8. Claessens M, van Baak MA, Monsheimer S, Saris WH. The effect of

a low-fat, high-protein or high-carbohydrate ad libitum diet on weight

loss maintenance and metabolic risk factors. Int J Obes (Lond) 2009;33:

296304.9. Harris JA, Benedict FG. A biometric study of basal metabolism in man.

Washington, DC: Carnegie Institute of Washington, 1919 (publication

279).

10. Viswanathan VSC, Varadharani MP, Nair BM, Jayaraman M,

Ramachandran A. Prevalence of albuminuria among vegetarian and non-

vegetarian south Indian diabetic patients. Ind J Nephrol 2002;12:736.11. Dansinger ML, Gleason JA, Griffith JL, et al. Comparison of the Atkins,

Ornish, Weight Watchers, and Zone diets for weight loss and heart

disease risk reduction: a randomized trial. JAMA 2005;293:4353.

12. Stern L, Iqbal N, Seshadri P, et al. The effects of low-carbohydrateversus conventional weight loss diets in severely obese adults: one-year

follow-up of a randomized trial. Ann Intern Med 2004;140:77885.13. Due A, Larsen TM, Mu H, et al. Comparison of 3 ad libitum diets for

weight-loss maintenance, risk of cardiovascular disease, and diabetes:

a 6 mo randomized, controlled trial. Am J Clin Nutr 2008;88:123241.14. Brehm BJ, Seeley RJ, Daniels SR, DAlessio DA. A randomized trial

comparing a very low carbohydrate diet and a calorie-restricted low fat

diet on body weight and cardiovascular risk factors in healthy women.

J Clin Endocrinol Metab 2003;88:161723.

15. Lasker DA, Evans EM, Layman DK. Moderate carbohydrate, moderate

protein weight loss diet reduces cardiovascular disease risk compared to

high carbohydrate, low protein diet in obese adults: a randomized

clinical trial. Nutr Metab 2008;5:30.

16. Meckling KA, OSullivan C, Saari D. Comparison of a low-fat diet to

a low-carbohydrate diet on weight loss, body composition, and risk

factors for diabetes and cardiovascular disease in free-living, overweightmen and women. J Clin Endocrinol Metab 2004;89:271723.

17. Clifton PM, Bastiaans K, Keogh JB. High protein diets decrease total

and abdominal fat and improve CVD risk profile in overweight and

obese men and women with elevated triacylglycerol. Nutr Metab

Cardiovasc Dis 2009;19:54854.

18. Hession M, Rolland C, Kulkarni U, et al. Systematic overview of ran-

domized controlled trials of low-carbohydrate vs. low-fat/low-calorie

diets in the management of obesity and its comorbidities. Obes Rev

2009;10:3650.19. Samaha FF, Iqbal N, Seshadri P, et al. A low-carbohydrate as compared

with a low-fat diet in severe obesity. N Engl J Med 2003;348:207481.20. Yancy WS Jr, Olsen MK, Guyton JR, et al. A low-carbohydrate, keto-

genic diet versus a low-fat diet to treat obesity and hyperlipidemia:

a randomized, controlled trial. Ann Intern Med 2004;140:76977.21. Tay J, Brinkworth GD, Noakes M, et al. Metabolic effects of weight loss

on a very-low-carbohydrate diet compared with an isocaloric high-carbohydrate diet in abdominally obese subjects. J Am Coll Cardiol

2008;51:5967.22. McLaughlin T, Carter S, Lamendola C, et al. Effects of moderate var-

iations in macronutrients composition on weight loss and reduction in

cardiovascular disease risk in obese, insulin-resistant adults. Am J Clin

Nutr 2006;84:81321.23. Genev NM, Lau IT, Willey KA, et al. Does insulin therapy have a hy-

pertensive effect in type 2 diabetes? J Cardiovasc Pharmacol 1998;32:

3941.24. Kanoun F, Ben Amor Z, Zouari B, Ben Khalifa F. Insulin therapy may

increase blood pressure levels in type 2 diabetes mellitus. Diabetes

Metab 2001;27:695700.25. Nobels F, van Gaal L, de Leeuw I. Weight reduction with a high protein,

low carbohydrate, calorie-restriced diet: effects on blood pressure, glu-

cose and insulin levels. Neth J Med 1989;35:295302.


byguest

onJune20,2012

www.ajcn.org

Downloadedfrom


12/12

26. Sargrad KR, Homko C, Mozzoli M, Boden G. Effect of high protein vshigh carbohydrate intake on insulin sensitivity, body weight, hemoglo-bin A1c, and blood pressure in patients with type 2 diabetes mellitus.J Am Diet Assoc 2005;105:57380.

27. Bravata DM, Sanders L, Huang J, et al. Efficacy and safety of low-carbohydrate diets: a systematic review. JAMA 2003;289:183750.

28. Roberts SB. Glycemic index and satiety. Nutr Clin Care 2003;6:206.29. Weigle DS, Breen PA, Matthys CC, et al. A high-protein diet induces

sustained reductions in appetite, ad libitum caloric intake, and bodyweight despite compensatory changes in diurnal plasma leptin and

ghrelin concentrations. Am J Clin Nutr 2005;82:418.30. Martin LJ, Su W, Jones PJ, et al. Comparison of energy intakes de-

termined by food records and doubly labeled water in women par-

ticipating in a dietary-intervention trial. Am J Clin Nutr 1996;63:48390.

31. Kroke A, Klipstein-Grobusch K, Voss et al. Validation of a self-administered food-frequency questionnaire administered in the Euro-pean Prospective Investigation into cancer and nutrition (EPIC) study:

comparison of energy, protein, and macronutrient intakes estimated withthe doubly labeled water, urinary nitrogen, and repeated 24-h dietaryrecall methods. Am J Clin Nutr 1999;70:43947.

32. Johnston CS, Tjonn SL, Swan PD. High-protein, low-fat diets are ef-fective for weight loss and favorably alter biomarkers in healthy adults.J Nutr 2004;134:58691.

33. Flatt JP, Blackburn GL, Bistrian BR. Urinary urea excretion and eval-uation of the nitrogen balance. JPEN J Parenter Enteral Nutr 1992;16:1912.

34. Australian Bureau of Statistics. National Nutrition Survey: nutrient in-

takes and physical measurements, Australia. Canberra, Australia:Commonwealth Department of Health and Family Services, 1995.

35. Nordmann AJ, Nordmann A, Briel M, et al. Effects of low-carbohydrate

vs low-fat diets on weight loss and cardiovascular risk factors: a meta-analysis of randomized controlled trials. Arch Intern Med 2006;166:28593.

36. Ware JH. Interpreting incomplete data in studies of diet and weight loss.N Engl J Med 2003;348:21367.


byguest

onJune20,2012

www.ajcn.org

Downloadedfrom

one-year weight maintenance after significant weight loss in healthy

Documents