Additional protein intake limits weight regain after weight loss in humans

British Journal of Nutrition (2005), 93, 281–289 Additional protein intake limits weight regain after weight loss in humans Manuela P. G. M. Lejeune*, Eva M. R. Kovacs and Margriet S. Westerterp-Plantenga Department of Human Biology, Maastricht University, P.O. Box 616, 6200 MD Maastricht, The Netherlands (Received 28 January 2004 – Revised 7 July 2004 – Accepted 21 September 2004) Since long-term weight maintenance (WM) is a major problem, interventions to improve WM are needed. The aim of the study was toinvestigate whether the addition of protein to the diet might limit weight regain after a weight loss of 5 – 10 % in overweight subjects. In arandomised parallel study design, 113 overweight subjects (BMI 29·3 (SD 2·5) kg/m2); age 45·1 (SD 10·4) years) followed a very-low-energy diet for 4 weeks, after which there was a 6-month period of WM. During WM, subjects were randomised into either a proteingroup or a control group. The protein group received 30 g/d protein in addition to their own usual diet. During the very-low-energydiet, no differences were observed between the groups. During WM, the protein group showed a higher protein intake (18 % v. 15 %;P, 0·05), a lower weight regain (0·8 v. 3·0 kg; P, 0·05), a decreased waist circumference (2 1·2 (SD 0·7) v. 0·5 (SD 0·5 ) cm; P, 0·05)and a smaller increase in respiratory quotient (0·03 (SD 0·01) v. 0·07 0·01; (SD/)P , 0·05) compared with the control group. Weightregain in the protein group consisted of only fat-free mass, whereas the control group gained fat mass as well. Satiety in the fastedstate before breakfast increased significantly more in the protein group than in the control group. After 6 months follow-up, bodyweight showed a significant group £ time interaction. A protein intake of 18 % compared with 15 % resulted in improved WM in over-weight subjects after a weight loss of 7·5 %. This improved WM implied several factors, i.e. improved body composition, fat distribution,substrate oxidation and satiety.
Substrate oxidation: Weight maintenance: Body composition: High protein A common treatment for obesity is weight reduction.
than is expected from its energy content alone (Anderson Although short-term weight loss programmes have proved 1994). Several studies have shown that a high-protein to be successful, long-term weight maintenance (WM) is a lunch decreases later food intake more than a lunch low major problem (Kramer et al. 1989; Wadden et al. 1994; in protein (Booth et al. 1970; Barkeling et al. 1990; Pasman et al. 1997a,b; Westerterp-Plantenga et al. 1998).
Latner & Schwartz, 1999). Westerterp-Plantenga et al.
Successful weight maintenance is of importance for lowering (1999) showed an increased satiety effect of a high-protein risk factors for cardiovascular and other diseases (Goldstein, diet despite similar energy intake. Finally, protein has also 1992; Wing et al. 1992; Van Gaal et al. 1997). To improve the been shown to have low energy efficiency during overfeed- metabolic profile, it is not necessary to achieve the ideal body ing (Dulloo & Jacquet, 1999; Stock, 1999), a situation that weight: a weight reduction of 5 – 10 % is often sufficient to is comparable to a weight regain situation. Although the induce a clinically relevant effect (Goldstein, 1992). To pre- effect of dietary protein has been examined in weight- serve these beneficial effects of weight reduction, an loss studies (Skov et al. 1999; Dumesnil et al. 2001; improvement in long-term WM is necessary.
Layman et al. 2003), the effect of additional protein Pasman et al. (1999) showed that weight regain was intake on WM has not yet been studied. The aim of this slower when the body composition of the weight regained study was to investigate whether the addition of protein consisted of a greater fat-free mass (FFM) due to physical to the diet might limit weight regain after a weight loss activity. We hypothesise that weight regain may be limited of 5 – 10 % in moderately overweight subjects.
if the inevitable increase in body weight consists of a largerFFM, for instance achieved by the consumption of anappropriate substrate. For this, we suggest an increased protein intake, because of its potential to increase FFM (Jean et al. 2001). Furthermore, it is known that, of allthe macronutrients, protein seems to be the most satiating.
One hundred and forty moderately overweight male and Protein consumption suppresses short-term food intake female subjects (BMI 25 – 35 kg/m2, age 18 – 60 years) more than that of fats or carbohydrates, and even more were recruited for this study by advertisements in local Abbreviations: FFM, fat-free mass; FM, fat mass; PAL, physical activity level; REE, resting energy expenditure; RQ, respiratory quotient; TBW, total body water; TEE, total energy expenditure; TFEQ, Three Factor Eating Questionnaire; VLED, very-low-energy diet; WM, weight maintenance.
* Corresponding author: Ms M. P. G. M. Lejeune, fax þ 31 43 367 09 76, email [email protected] BMI was calculated as body weight divided by height2 All the subjects were in good health, were non-smokers, were not using medication and were at most moderate alco- Waist:hip ratio. The distribution of fat was investigated hol users. A written informed consent was obtained from by measuring the waist and hip circumferences, and then all the participants. The Medical Ethics Committee of the calculating the waist:hip ratio. The waist circumference Academic Hospital in Maastricht approved the study.
was measured at the site of the smallest circumference Twenty subjects did not start the study due to relocation, between the rib cage and the iliac crest, with the subjects a change of job or an inability to fulfil the schedule with standing. The hip circumference was measured at the site visits to the university. Intention-to-treat applied for 120 of the largest circumference between the waist and the subjects. During the first week, seven subjects dropped thighs. The waist:hip ratio was calculated by dividing the out because of difficulty maintaining the very-low-energy waist circumference by the hip circumference.
diet (VLED). They had been meant to participate in the Body composition. Total body water (TBW) was additional-protein group. Their baseline characteristics measured using the 2H2O dilution technique (Schoeller did not affect the averages of the baseline measurements et al. 1980; Van Marken Lichtenbelt et al. 1994). In the of the treatment group. Thus, 113 subjects completed the evening, the subjects collected a background urine study. Subject characteristics are shown in sample and then ingested a dose of 2H-enriched water(2H2O), after which they refrained from consuming fluidand food. The following morning, a urine sample from the second voiding was collected between 08.00 h and After the subjects’ baseline characteristics had been deter- 10.00 h. The concentration of 2H in the urine samples mined, they were divided into two similar groups, stratified was measured using an isotope ratio mass spectrometer for gender, BMI, age, eating behaviour (Three Factor (Micromass Optima, Manchester, UK). TBW was obtained Eating Questionnaire (TFEQ); Stunkard & Messick, by dividing the measured 2H dilution space by 1·04 1985; Westerterp-Plantenga et al. 1999) and resting (Schoeller et al. 1980). FFM was calculated by dividing energy expenditure (REE). Both groups followed a the TBW by the constant hydration factor 0·73, which VLED intervention for 4 weeks in order to initiate can be used for adult subjects (Schoeller, 1996). The weight loss. After this weight loss period, a WM period weight loss and weight maintenance periods were long of 6 months followed. Subjects in both groups visited the enough to establish a stable hydration of FFM. Fat mass university six times. During these visits, the measurements (FM) was calculated by subtracting FFM from body were carried out as described in the Measurements section weight. FM expressed as a percentage of total body mass below. Moreover, the subjects were asked by a dietitian how they felt, how they considered taking part in the Attitude towards eating. To determine whether attitude study and whether they had any specific questions. Further- towards food intake changed during the experiment, a more, dietary counselling was provided upon request for all Dutch translation of the TFEQ was used (Stunkard & Mes- sick, 1985; Westerterp-Plantenga et al. 1999). The first One of the groups was provided with additional protein factor of the TFEQ (F1) measures cognitive restrained (protein group, n 53), while the other group did not receive eating: control of food intake by thought and will-power.
this additional protein (control group, n 60). Since, similar The second factor (F2) represents disinhibition: an inciden- to previous meal-replacement studies (Ditschuneit et al.
tal inability to resist eating cues, or inhibition of dietary 1999; Flechtner-Mors et al. 2000; Rothacker 2000; restraint, and emotional eating. The third factor (F3) exam- Ashley et al. 2001; Allison et al. 2003), no placebo was ines the subjective feeling of general hunger.
used for this control group, we included repeated measure- In addition, the Herman – Polivy questionnaire was used ments of dietary restraint to check whether additional diet- to determine the frequency of dieting (Herman & Polivy, differently. After 6 months of WM, the additional protein Post-absorptive appetite profile. To determine the post- intake was stopped. Six months after WM, all the subjects absorptive appetite profile, hunger and satiety were rated were asked to return to the university for a follow-up on anchored 100 mm visual analogue scales in the morning measurement of body weight. For this measurement, thirty-one subjects in the protein group and thirty-nine sub- Blood parameters. A fasting blood sample of 10 ml was jects in the control group returned.
taken and mixed with EDTA to prevent clotting. Plasmawas obtained by centrifugation (at 3000 U/min for 10 minat 48C), frozen in liquid nitrogen and stored at 2 808C until further analysis. Plasma glucose concentrations were The following measurements were executed to determine determined using the hexokinase method (Glucose HK 125 kit; ABX diagnostics, Montpellier, France). The Body weight and BMI. Body weight was measured on a Wako NEFA C-kit (Wako Chemicals, Neuss, Germany) digital balance (model 707, Seca, Hamburg, Germany; was used to determine free fatty acid concentrations. Insu- weighing accuracy of 0·1 kg) with subjects in their under- lin concentrations were measured using the RIA-kit (Insu- wear, in a fasted state and after voiding their bladder.
lin RIA-100; Kabi-Pharmacia, Uppsala, Sweden). The Height was measured using a wall-mounted stadiometer glycerol kinase method was used to determine glycerol (model 220, Seca, Hamburg, Germany).
concentrations (Boehringer Mannheim GmbH, Mannheim, Additional protein intake and weight maintenance Germany). Triacylglycerol was measured using the GPO- & Webster, 1977) was taken for the usual energy storage trinder kit (Sigma Diagnostics Inc., St Louis, MO, USA).
of FM and FFM (A). If body weight gain consisted of The b-hydroxybutyrate dehydrogenase method (Sigma only FFM while FM decreased, 52 MJ/kg FFM gain Diagnostics Inc., St Louis, MO, USA) was used to deter- (Pullar & Webster, 1977) and 30 MJ/kg FM loss were mine b-hydroxybutyrate concentration. Leptin concen- trations were measured using the human leptin RIA-kit ES (MJ/d) ¼ (D body weight (kg) £ 30)/number (Linco Research Inc., St Charles, MO, USA).
Adverse events. Adverse events during treatment were ES (MJ/d)¼ ((DFFM (kg) £ 52) 2 (DFM (kg) £ 30))/ recorded and the severity and outcome specified.
REE and substrate oxidation. In a random subset of To see whether a higher percentage of energy ingested as subjects (n 75), REE and substrate oxidation were protein in the diet could lower the energy efficiency, measured by means of an open-circuit ventilated hood which is known already as the ‘Stock’ hypothesis system. REE was measured in the morning with subjects (Dulloo & Jacquet, 1999; Stock, 1999), we used the fol- in a fasted state while lying supine for 30 min. Gas ana- lowing equation to calculate energy efficiency (Eeff): lyses were performed by a paramagnetic oxygen analyser(type 500A, Servomex, Crowborough, Sussex, UK) and Eeff ðkg=MJÞ ¼ body weight regain ðkgÞ=ðEI ðMJ=dÞ an IR carbon dioxide analyser (type 500A, Servomex, Crowborough, Sussex, UK), similar to the analysissystem described by Schoffelen et al. (1997). Calculation of REE was based upon the Weir’s formulas (Weir,1949). Respiratory quotient (RQ) was calculated as CO2 The VLED provided 2·1 MJ/d (carbohydrate – protein – fat 42:44:14 energy percentage) (Modifast, Novartis Nutrition, Physical activity. The same subset of subjects who Breda, The Netherlands) and was supplied in three sachets underwent metabolic testing was used to measure physical daily that were dissolved in water to obtain a milk shake, activity level (PAL). PAL was determined using a uniaxial pudding, soup or muesli. Vegetables and fruit were allowed accelerometer (CSA, CSAB, Inc. Stamford, CT, USA.) in addition to Modifast. The aim was a body weight loss of (Ekelund et al. 2000), or a tri-axial accelerometer for movement registration (Tracmor Maastricht University,Maastricht, The Netherlands.) (Goris et al. 2001), during 1 week. Subjects wore the CSA or Tracmor duringwaking hours on a belt at waist level on their back. The After the VLED period, the WM period started, in which different accelerometers were randomised over the two all subjects were allowed to eat their habitual diet again.
groups. Half the subjects in the protein group and the con- During the WM phase, the subjects in the protein group trol group received the CSA, the other half received the received 30 g additional protein per day. This was provided Tracmor. Subjects received the same accelerometer every as a sachet of pure protein (protein source calcium casei- nate, 1·4 % calcium) per day to be dissolved in water, PAL was calculated using the following equations: giving rise to a vanilla-flavoured drink. The protein drinkcontained no carbohydrate or fat. Subjects were required to consume the protein drink at lunch or in the afternoon.
¼ ð0·000001379 £ ðcounts=day £ 5ÞÞ þ 1·113 In this way, we aimed at an energy intake comprising18 – 20 % protein/d, depending on the subject’s usual ¼ 21·259 þ ð1·552 £ REEÞ þ ð0·076 £ counts=minÞ Data are presented as mean and standard deviation (SD).
in which TEE is total energy expenditure (MJ/d) and REE A two-factor repeated measures ANOVA was carried out is resting energy expenditure (MJ/d).
to determine possible differences between the protein and Protein intake. Compliance with additional protein control groups in all measured parameters over time.
intake was checked by taking 24 h urine samples after 3 When appropriate, a factorial ANOVA was used for analys- months WM and analysing these for nitrogen. Energy ing differences between the two groups. Post hoc analyses intake from protein was calculated from the 24 h nitrogen were made with the Scheffe F-test. A P value , 0·05 was output according to the formula of Isaksson (1980): regarded as statistically significant. Statistical procedures were performed using Statview SE þ Graphics (AbacusConcepts, Berkeley, CA, USA).
¼ ðnitrogen output in 24 h urine ðg=dÞ þ 2 gÞ £ 6·25 Energy intake and energy efficiency. Energy intake wascalculated as TEE plus energy storage (ES). Energy storage The effects did not differ between men and women so data was calculated from the composition of the energy stored.
from both genders were analysed together. No adverse A figure of 30 MJ per kilogram body weight gain (Pullar events occurred. Subjects continuously reported that they Additional protein intake and weight maintenance felt positive about taking part in the study. They confirmed that ample attention was given to their questions.
During the VLED period, the changes described belowoccurred, which did not differ between the subsequent pro- tein and control groups Subjects lost a significant of their original body weight (P, 0·0001). The mean change in body weight for both groups over time is shown in This consisted of 4·0 (SD 1·7) kg FM and 2·3 (SD 1·7) kg FFM. Waist circumference was also significantly reducedover time. Attitude towards eating showed some significant changes over time Cognitive restraint (F 1, TFEQ)increased significantly; disinhibition (F 2, TFEQ) and gen- eral hunger scores (F 3, TFEQ) decreased during weight loss REE and RQ decreased during weight loss TEE decreased in both groups, but this only reached significance in the control group. Fasting blood glu- cose, insulin, triacylglycerol and leptin levels showed a sig- nificant decrease with weight loss, and b-hydroxybutyrate, glycerol and free fatty acids showed a significant increase Compliance with the additional protein was shown by a higher amount of nitrogen in 24 h urine in the protein group compared with the control group (14·3 ( v. 11·2 (SD 3·5) g/d; P, 0·05). The protein intake in the protein group was 101·7 (SD 22·2) g/d, which was signifi- cantly higher than the 82·7 (SD 22·0) g/d in the control group (P, 0·05). When expressed as percentage of energy intake from protein, significant differences per- sisted, with 18 % in the protein group and 15 % in the During WM, percentage body weight regain, as well as rate of regain, was significantly lower in the protein group than the control group The netpercentage body mass lost after the WM period compared Fig. 1. The change in body weight over time for the protein group (X) (n 53) and the control group (W) (n 60). Values are means with standard deviations of the mean shown by vertical bars.
Table 3. Fasting blood-parameters of the protein (n 32) and the control (n 43) group at baseline, after 4 weeks on a very-low-energy diet(VLED; Modifast) and after 3 months weight maintenance * P, 0·05 compared with baseline (repeated measures ANOVA).
† P, 0·05 compared with control group (factorial ANOVA).
‡ P, 0·05 interaction treatment £ time (two-factor repeated measures ANOVA).
with baseline was significantly higher in the protein group REE returned to almost baseline values in both groups than the control group (6·7 (SD 7·2) % v. 3·8 (SD 4·8) %; during WM To assess possible differences in P, 0·05). The protein group showed a decrease in waist REE adjusted for FFM between groups, we analysed the circumference during WM, whereas the control group residuals of the regression of REE v. FFM. The residual showed an increase (2 1·2 (SD 0·7) cm v. 0·5 (SD 0·5) analysis showed no significant group £ time interaction.
cm; P, 0·05). FFM stayed significantly lower during The increase in RQ during WM was significantly smaller WM compared with baseline in both groups. FM in theprotein group continued to decrease during WM, whileFM increased in the control group (2 0·9 (SD 0·7) kg v.
1·7 (SD 0·4) kg; P, 0·005) Cognitive restraintscores stayed significantly higher during WM comparedwith baseline in both groups. No differences were seen inthe change in cognitive restraint during the study betweenthe groups. Disinhibition stayed significantly lower inboth groups, and general hunger decreased in bothgroups. Satiety in the fasted state before breakfastincreased significantly during WM in the protein groupbut not in the control group The increasein satiety was significantly higher in the protein groupthan the control group. The hunger scores in the fastedstate before breakfast did not change over time ineither group.
Fig. 3. The changes in body weight (B), fat-free mass (B) and fat Fig. 2. The weight regain during weight maintenance (expressed as mass (A) for (a) the protein group (n 53) and (b) the control group percentage of the weight loss) for the protein (B) (n 53) and the (n 60) during the study compared with baseline. Values are means control (A) (n 60) groups. Values are means with standard deviations with standard deviation of the mean shown by vertical bars. VLED, of the mean shown by vertical bars. *P, 0·05 compared with control change after the very-low-energy diet period; 3 months WM, group, †P, 0·05 compared with 1 month weight maintenance, change after 3 months weight maintenance; 6 months WM, change ‡P, 0·05 compared with 2 months weight maintenance.
Additional protein intake and weight maintenance in the protein group than the control group (0·03 (SD 0·01) The composition of the body mass regained was v. 0·07 (SD 0·01); P, 0·05), indicating a smaller decrease in different for the protein and the control groups. The body fat oxidation in the protein group than in the control group.
mass regained in the protein group consisted only The increase in RQ was not related to weight regain of FFM, whereas the FM still decreased during WM, (P. 0·05). No changes in PAL were seen over time and which resulted in a lower percentage of body fat. In the between groups A trend for a lower energy effi- control group, the composition of the body mass regain ciency of the protein group than the control group was was FM as well as FFM. The observation that body mass found after 6 months WM (3·0 £ 1024 (SD 0·003) kg/MJ regain consisting only of FFM results in a slower weight v. 0·001 (SD 0·002) kg/MJ; P¼ 0·12).
regain is in line with our hypothesis that when the Leptin concentrations stayed significantly lower after composition of the weight regained consists of a greater treatment compared with baseline in the protein group FFM, the inevitable increase in body mass will be Glucose concentrations stayed significantly slower. Here, however, we have shown that this was due lower after treatment compared with baseline in the control to increased protein intake without a change in physical group. Insulin concentrations stayed significantly lower activity, whereas in the study of Pasman et al. (1999), after treatment compared with baseline in both groups.
the cause appeared to be increased physical activity.
b-Hydroxybutyrate, glycerol and free fatty acid concen- In addition to a beneficial effect of dietary protein during trations returned almost to baseline values during weight weight loss (Skov et al. 1999; Astrup, 2001; Dumesnil et al.
regain Triacylglycerol concentrations during 2001; Layman et al. 2003), protein also appears to support WM were significantly lower in the protein group than Although subjects were asked to consume 30 g/d protein The change in body weight of the subset of subjects in addition to their usual diet, the analysis of the results (protein n 31, control n 39) who were also measured showed a mean difference of 19 g/d between the protein after 6 months follow-up is shown in The analysis and the control group. This could imply that subjects of the follow-up measurement of body weight showed a in the protein group replaced part of their usual protein significant group £ time interaction.
intake by the protein that was provided to them. Sincethis was a study under free-living conditions, the carbo-hydrate and fat content of the weight maintenance diet With respect to satiety, we found a greater increase in post- In the present study, we investigated whether the addition absorptive satiety in the protein group during WM, although of protein to the diet might limit weight regain after a their energy intake did not differ significantly from that of the weight loss of 5 – 10 % in moderately overweight men control group. Short-term differences in digestion, absorp- and women. The results show that subjects who consumed tion and energy expenditure between different protein 18 % of their energy intake as protein regained less weight sources have been shown (Boirie et al. 1997; Mikkelsen during 6 months WM (0·8 kg), compared with subjects who et al. 2000). Boirie et al. (1997) introduced the concept of consumed 15 % of their energy intake as protein (3·0 kg).
‘fast’ and ‘slow’ proteins, based upon the differences in This was independent of changes in cognitive restraint, digestion and absorption of these proteins. These short- physical activity, REE, TEE and hunger scores since term effects may be related to increases in the concentrations these parameters did not differ between groups.
of gut hormones, such as glucagon-like peptide-1 (Flint et al.
1998; Gutzwiller et al. 1999; Naslund et al. 1999) and chole-cystokinin (Schick et al. 1991; Burton-Freeman et al. 2002;Kissileff et al. 2003). However, this probably does not applyto post-absorptive satiety. Animal protein has been shown tointroduce a higher energy expenditure (Mikkelsen et al.
2000), and, in the longer term, the higher post-absorptivesatiety and thermogenesis were sustained with high-proteindiets consisting of a variety of proteins from different sources(Dulloo & Jacquet, 1999; Westerterp-Plantenga et al. 1999;Dumesnil et al. 2001; Westerterp-Plantenga et al. 2004).
Moreover, a relationship between satiety and thermogenesishas been shown (Crovetti et al. 1998; Westerterp-Plantengaet al. 1999) with using high-protein diets. We thereforesuggest that the higher post-absorptive satiety is due toincreased thermogenesis.
The observation that the increase in REE as a function of Fig. 4. The change in body weight for the subjects with follow-up in FFM during WM did not differ significantly might be due to the protein group (B) (n 31) and the control group (A) (n 39) during the lack of a significant difference in the increase of FFM.
the study compared with baseline. Values are means with standard The slower increase in RQ in the protein group reflects a deviations of the mean shown by vertical bars. *P, 0·05 compared with the control group. VLED, change after the very-low-energy Although the treatment with respect to number of visits, diet period; WM 1, 2, 3, 6 change after 1, 2, 3 and 6 months weightmaintenance.
measurements and attention was identical in both groups, there was no placebo used for the additional protein, simi- lar to previous meal-replacement studies (Ditschuneit et al.
1999; Flechtner-Mors et al. 2000; Rothacker, 2000; Ashley Allison DB, Gadbury G, Schwartz LG, Murugesan R, Kraker JL, et al. 2001; Allison et al. 2003). The issue of the placebo is Heshka S, Fontaine KR & Heymsfield SB (2003) A novel soy- a difficult one. It would be possible to exchange the protein based meal replacement formula for weight loss among obese drink for a carbohydrate or fat drink, but then one needs a individuals: a randomized controlled clinical trial. Eur J ClinNutr 57, 514 – 522.
more complete study that also addresses the specific carbo- Anderson GH (1994) Regulation of food intake. In Modern Nutri- hydrate and fat effect. This was beyond the scope of the tion in Health and Disease, 8th ed. pp. 524 – 536 [M Shils, present study. However, we checked whether the results J Olson and M Shike, editors]. Malvern: Lea & Febiger.
in both groups could be differently related to changes in Ashley JM, St. Jeor ST, Perumean-Chaney S, Schrage J & Bovee eating behaviour. The increase in cognitive restraint V (2001) Meal replacements in weight intervention. Obes Res during the study did not differ between the protein and con- trol group so this cannot explain the limited body weight Astrup A (2001) Healthy lifestyles in Europe: prevention of obes- regain during WM in the protein group compared with ity and type II diabetes by diet and physical activity. Public the control group. It shows that, without a difference in increase in cognitive restraint, the protein group main- Barkeling B, Rossner S & Bjorvell H (1990) Effects of a high- protein meal (meat) and a high-carbohydrate meal (vegetarian) on satiety measured by automated computerized monitoring of Since reporting of food intake in humans cannot be subsequent food intake, motivation to eat and food preferences.
expected to be completely reliable because of reasons of Int J Obes Relat Metab Disord 14, 743 – 751.
privacy (Blundell, 2000), and thus often leads to incorrect Blundell JE (2000) What foods do people habitually eat? A conclusions (Goris & Westerterp, 1999, 2000; Goris et al.
dilemma for nutrition, an enigma for psychology. Am J Clin 2000), we chose to calculate energy intake from energy expenditure and energy storage. The calculated energy Boirie Y, Dangin M, Gachon P,Vasson MP, Maubois JL & intake was not significantly different between both Beaufrere B (1997) Slow and fast dietary proteins differently groups and therefore cannot explain the difference in modulate postprandial protein accretion. Proc Nat Acad SciUSA 94, 14930 – 14935.
Booth DA, Chase A & Campbell AT (1970) Relative effective- Zemel (2003) showed that a low calcium intake (400 mg/d) ness of protein in the late stages of appetite suppression in leads to a smaller body weight loss compared with a high cal- man. Physiol Behav 5, 1299 – 1302.
cium intake. Since the average calcium intake in the Dutch Burton-Freeman B, Davis PA & Schneeman BO (2002) Plasma population is already rather high (800 – 1600 mg/d; Hulshof cholecystokinin is associated with subjective measures of sati- et al. 2003), the additional calcium intake in the protein ety in women. Am J Clin Nutr 76, 659 – 667.
group (420 mg/d) could have contributed to the difference Crovetti R, Porrini M, Santangelo A & Testolin G (1998) The in body weight regain in the present study, yet is unlikely influence of thermic effect of food on satiety. Eur J Clin The analysis of the follow-up measurement of body Ditschuneit HH, Flechtner-Mors M, Johnson TD & Adler G (1999) Metabolic and weight-loss effects of a long-term diet- weight showed a significant difference between both ary intervention in obese patients. Am J Clin Nutr 69, groups after 3 and 6 months WM. The change in body weight over time in the control group appeared to continue Dulloo AG & Jacquet J (1999) Low-protein overfeeding: a tool to as during the first 6 months, whereas the body weight over unmask susceptibility to obesity in humans. Int J Obes Relat time in the protein group appeared to be maintained.
Unfortunately, body composition was not determined Dumesnil JG, Turgeon J, Tremblay A, et al. (2001) Effect of a low- during follow-up. This needs to be repeated in order to glycaemic index-low-fat-high protein diet on the atherogenic be able to indicate whether body composition is still a metabolic risk profile of abdominally obese men. Br J Nutr main factor in determining the difference in WM between the groups. With respect to metabolic syndrome risk Ekelund U, Yngve A, Sjo¨stro¨m M & Westerterp KR (2000) Field evaluation of the Computer Science and Application’s inc.
factors, WM as well as waist circumference was more activity monitor during running and skating training in adoles- favourable in the protein group than in the control group.
cent athletes. Int J Sports Med 21, 586 – 592.
In conclusion, a protein intake of 18 % v. 15 % resulted Flechtner-Mors M, Ditschuneit HH, Johnson TD, Suchard MA & in improved WM over 6 – 12 months in moderately over- Adler G (2000) Metabolic and weight loss effects of long-term weight subjects after a weight loss of 7·5 %. This improved dietary intervention in obese patients: four-year results. Obes WM implied several factors, i.e. improved body compo- sition, fat distribution, substrate oxidation and satiety.
Flint A, Raben A, Astrup A & Holst J (1998) Glucagon-like pep- The result suggests that improved WM is possible when tide 1 promotes satiety and suppresses energy intake in humans.
Goldstein DJ (1992) Beneficial effects of modest weight loss. Int J Obes Relat Metab Disord 16, 397 – 415.
Goris AHC, Meijer EP, Kester A & Westerterp KR (2001) Use of a triaxial accelerometer to validate reported food intakes. Am J We thank Ilse Nijs, Joan Senden, Wendy Sluijsmans and Loek Wouters for their contributions to the study. This Goris AH & Westerterp KR (1999) Underreporting of habitual study was supported by Novartis CH, Consumer Health food intake is explained by undereating in highly motivated lean women. J Nutr 129, 878 – 882.
Additional protein intake and weight maintenance Goris AH & Westerterp KR (2000) Improved reporting of habit- Pullar JD & Webster AJF (1977) The energy cost of fat and ual food intake after confrontation with earlier results on food protein deposition in the rat. Br J Nutr 37, 355 – 363.
reporting. Br J Nutr 83, 363 – 369.
Rothacker DQ (2000) Five-year self-management of weight using Goris AH, Westerterp-Plantenga MS & Westerterp KR (2000) meal replacements: comparison with matched controls in rural Undereating and underreporting of habitual food intake in Wisconsin. Nutrition 16, 344 – 348.
obese men: selective underreporting of fat intake. Am J Clin Schick RR, Schusdziarra V, Mossner J, Neuberger J, Schroder B, Segmuller R, Maier V & Classen M (1991) Effect of CCK on Gutzwiller JP, Drewe J, Goke B, Schmidt H, Rohrer B, Lareid V food intake in man: physiological or pharmacological effect? Z & Beglinger C (1999) Glucagon-like peptide-1 promotes sati- ety and reduces food intake in patients with diabetes mellitus Schoeller DA (1996) Hydrometry. In Human Body Composition, type 2. Am J Physiol 276, R1541 – R1544.
pp. 25 – 43 [AF Roche, SB Heymsfield and TG Lohmans, Herman CP & Polivy J (1980) Restrained eating. In Obesity, pp.
editors]. Champaign, IL: Human Kinetics.
208 – 224 [AJ Stunkard, editor]. Philadelphia: WB Saunders.
Schoeller DA, Van Santen E, Peterson DW, Diez W, Jaspan J Hulshof KFAM, Brussaard JH, Kruizinga AG, Telman J & Lo¨wik & Klein PD (1980) Total body water measurement in humans MRH (2003) Socio-economic status, dietary intake and 10y with 18O and 2H labeled water. Am J Clin Nutr 33, 2686 – 2693.
trends: the Dutch National Food Consumption Survey. Eur J Schoffelen PFM, Westerterp KR, Saris WHM & Ten Hoor F (1997) A dual respiration chamber with automated calibration.
Isaksson B (1980) Urinary nitrogen output as a validity test in dietary surveys. Am J Clin Nutr 33, 4 – 5.
Skov AR, Toubro S, Ronn B, Holm L & Astrup A (1999) Ran- Jean C, Rome S, Mathe´ V, Huneau JF, Aattouri N, Fromentin G, domized trial on protein vs carbohydrate in ad libitum fat Fromentin G, Achagiotis CL & Tome D (2001) Metabolic evi- reduced diet for the treatment of obesity. Int J Obes Relat dence for adaptation to a high protein diet in rats. J Nutr 131, Stock MJ (1999) Gluttony and thermogenesis revisited. Int J Obes Kissileff HR, Carretta JC, Geliebter A & Pi-Sunyer FX (2003) Relat Metab Disord 23, 1105 – 1117.
Cholecystokinin and stomach distension combine to reduce Stunkard AJ & Messick S (1985) The three factor eating question- food intake in humans. Am J Physiol Regul Integr Comp naire to measure dietary restraint, disinhibition and hunger.
Kramer FM, Jeffery RW, Foster JL & Snell MK (1989) Long- Van Gaal LF, Wauters MA & De Leeuw IH (1997) The beneficial term follow-up of behavioral treatment for obesity: patterns effects of modest weight loss on cardiovascular risk factors.
of weight regain among men and women. Int J Obes Relat Int J Obes Relat Metab Disord 21, S5 – S9.
Van Marken Lichtenbelt WD, Westerterp KR & Wouters L Latner JD & Schwartz M (1999) The effects of a high-carbo- (1994) Deuterium dilution as a method for determining total hydrate, high-protein or balanced lunch upon later food body water: effect of test protocol and sampling time. Br intake and hunger ratings. Appetite 33, 119 – 128.
Layman DK, Boileau RA, Erickson DJ, Painter JE, Shine H, Wadden TA, Foster GD & Letizia KA (1994) One-year beha- Santher C & Christon DD (2003) A reduced ratio of dietary vioral treatment of obesity: comparison of moderate and carbohydrate to protein improves body composition and severe caloric restriction and the effects of weight maintenance blood lipid profiles during weight loss in adult women. J therapy. J Consult Clin Psychol 62, 165 – 171.
Weir JBDV (1949) New methods for calculating metabolic rate Mikkelsen PB, Toubro S & Astrup A (2000) Effect of fat-reduced with special references to protein metabolism. J Physiol 109, diets on 24-h energy expenditure: comparisons between animal protein, vegetable protein, and carbohydrate. Am J Clin Nutr Westerterp-Plantenga MS, Kempen KPG & Saris WHM (1998) Determinants of weight maintenance in women after diet- Naslund E, Barkeling B, King N, Gutniak M, Blundell JE, induced weight reduction. Int J Obes Relat Metab Disord 22, Holst JJ, Rossner S & Hellstrom PM (1999) Energy intake and appetite are suppressed by glucagon-like peptide-1 Westerterp-Plantenga MS, Lejeune MPGM, Nijs I, Van Ooijen M (GLP-1) in obese men. Int J Obes Relat Metab Disord 23, & Kovacs EMR (2004) High protein intake sustains weight maintenance after body weight loss in humans. Int J Obes Pasman WJ, Saris WHM, Muls E, Vansant G & Westerterp-Plan- tenga MS (1999) The effect of exercise training on long-term weight maintenance in weight-reduced men. Metabolism 48, Westerterp KR (1999) Satiety related to 24 h diet-induced thermogenesis during high protein/carbohydrate vs. high fat Pasman WJ, Westerterp-Plantenga MS, Muls E, Vansant G, Van diets measured in a respiration chamber. Eur J Clin Nutr 53, Ree J & Saris WHM (1997a) The effectiveness of long-term fiber supplementation on weight maintenance in weight reduced Wing RR, Jeffery RW, Burton LR, Thorson C, Kuller LH & women. Int J Obes Relat Metab Disord 21, 548 – 555.
Folsom AR (1992) Change in waist-hip ratio with weight Pasman WJ, Westerterp-Plantenga MS & Saris WHM (1997b) loss and its association with change in cardiovascular risk fac- The effectiveness of long-term supplementation of carbo- tors. Am J Clin Nutr 55, 1086 – 1092.
hydrate, chromium, fiber and caffeine on weight maintenance.
Zemel MB (2003) Mechanisms of dairy modulation of adiposity.
Int J Obes Relat Metab Disord 21, 1143 – 1151.

Source: http://modicur.de/assets/protein-intake-limits-weight-regain.pdf