PAPER. Effect of a high-protein, energy-restricted diet on weight loss and energy expenditure after weight stabilization in hyperinsulinemic subjects

Transcription

1 (2003) 27, & 2003 Nature Publishing Group All rights reserved /03 $ PAPER Effect of a high-protein, energy-restricted diet on weight loss and energy expenditure after weight stabilization in hyperinsulinemic subjects ND Luscombe 1, PM Clifton 2, M Noakes 2, E Farnsworth 2 and G Wittert 3 1 Department of Physiology, University of Adelaide, South Australia; 2 CSIRO Health Science and Nutrition, Adelaide, South Australia; and 3 Department of Medicine, University of Adelaide, South Australia Objective: To determine the effect of replacing some dietary carbohydrate with protein, during energy restriction, on weight loss, total energy expenditure (TEE), resting energy expenditure (REE), respiratory quotient (RQ), and the thermic effect of feeding (TEF) in subjects with hyperinsulinemia. Design: Parallel, clinical intervention study of 12 weeks energy restriction (6.5 MJ/day) and 4 weeks energy balance (8.2 MJ/day) in two groups of subjects randomly assigned to either a high-protein (HP) diet (27% of energy (%E) as protein, 45%E as carbohydrate) or a lower-protein (LP) diet (16%E as protein, 57%E as carbohydrate). Subjects: A total of 36 obese nondiabetic volunteers with hyperinsulinemia (10 males/26 females, aged y, BMI kg/ m 2, fasting insulin mu/l). Measurements: Body weight and composition, TEE, REE, and RQ were measured at baseline and at week 16. In addition, the TEF to an HP or LP meal was determined for 3 h, at baseline and at week 16. Results: After 16 weeks, weight loss was similar in response to each diet; the overall decrease was kg (Po0.001), of which kg was fat (Po0.001). REE fell similarly with each diet; the overall decrease was kj/day (Po0.001). The TEF was 2% greater after the HP than after the LP meal at baseline (Po0.01) and 0.8% greater at week 16 (P ¼ 0.35). After 16 weeks, the TEF was not reduced in either dietary group. There was no change in TEE after 16 weeks. Conclusion: In subjects with hyperinsulinemia an energy-restrictive diet containing an increased protein-to-carbohydrate ratio does not enhance weight loss or significantly affect energy expenditure. Caloric restriction, rather than the macronutrient composition of the diet, is the most important determinant of weight loss. (2003) 27, doi: /sj.ijo Keywords: diet; protein; weight loss; energy expenditure; insulin resistance Introduction Insulin resistance affects approximately 25 35% of Westernized populations 1 and is associated with obesity, 2,3 type II diabetes, 4 and cardiovascular disease (CVD). 5 In overweight patients, moderate weight loss (B kg), 6 8 particularly reductions in visceral adipose tissue, 9,10 improves insulin resistance. Since the 1960s, high-protein (HP) diets with emphasis on some degree of carbohydrate restriction have been popular with the dieting public. There is at present, however, limited scientific evidence to support the concepts *Correspondence: Dr G Wittert, Department of Medicine, University of Adelaide, Royal Adelaide Hospital, Adelaide, SA 5000, Australia. gwittert@medicine.adelaide.edu.au Received 21 February 2002; revised 22 July 2002; accepted 27 September 2002 that HP diets facilitate weight loss, ameliorate insulin resistance, and improve health. 11,12 Low-fat (r30%) diets with an increased protein content have been shown to enhance weight loss 13,14 and improve insulin sensitivity. 13,15,16 In 13 hyperinsulinemic men, Baba et al 13 showed that with two isocaloric diets, greater weight loss occurred in the group consuming the HP diet. It has been proposed that weight loss on a HP diet may be facilitated, in part, by an increase in thermogenesis that may subsequently blunt the fall in resting and total energy expenditure (TEE) that is often observed during weight loss. 13 The reduced fall in resting energy expenditure (REE) may also be related to a preservation of lean mass that has been observed in obese nondiabetic women after energy restriction on low-fat diets containing an increased protein (436% of energy) content Acute feeding studies in lean

2 and obese nondiabetic subjects have shown that protein can exert up to three times the thermic response as compared to isocaloric preloads of either carbohydrate or fat. 19,20 Two studies, one in 13 hyperinsulinemic men 13 and a second in eight overweight subjects, 21 showed that resting and TEE were reduced less after an energy-restrictive HP diet (45 and 36% protein) than after isocaloric standard protein diets (15 and 12% protein). Besides these two small studies, no others have examined the effect of energy restriction and increased dietary protein on the changes in body weight, energy expenditure, and thermogenesis in obese subjects with or without insulin resistance. In addition, neither of the aforementioned studies 13 simultaneously examined changes in total and REE, and the thermic effect of feeding (TEF) in response to HP diets. The aim of this study was to compare the effect of two isocaloric diets either high or lower in dietary protein (30 vs 15% of energy as protein) on weight loss, TEE, REE, the TEF, and respiratory quotient (RQ) in obese hyperinsulinemic subjects after energy restriction and subsequent weight stabilization. We hypothesized that an increase in dietary protein may enhance weight loss independent of dietary calories as a result of increased thermogenesis that may consequently preserve REE and TEE. 55%E from carbohydrate). The fatty acid profiles for each diet were matched (8%E as saturated, 12%E monounsaturated, and 5%E polyunsaturated fatty acids). The diets were prescriptive fixed menu plans and subjects were supplied with key foods that made up 60% of their energy intake to assist with compliance. The key foods supplied were preweighed meat and poultry, shortbread biscuits, Canola Lite TM margarine, and Sunola TM oil (MeadowLea Foods Ltd, Mascot, NSW, Australia), plus Kraft Free TM (3% fat) cheese (Kraft Foods Ltd, Melbourne, Vic, Australia), skim milk powder, and diet yoghurt for the HP diet, and sultanas and rice for the LP diet. Further differences between the two diets are described elsewhere. 23 Each fortnight, subjects visited the same research dietitian who provided detailed dietary instruction and assessment. Weighed daily food records were completed on 3 consecutive days (2 weeks and 1 weekend day) at each 2-week period, and energy and macronutrient intakes were determined using Diet 1 Nutritional software (Xyris Software, Highgate Hill, Queensland, Australia). 24 This program is based on Australian food composition tables and food manufacturers data. The database has been extensively modified by our research dietitians to add new foods and recipes, and there were no missing values for the nutrients of interest. 583 Research design and methods Subjects A total of 36 volunteers with hyperinsulinemia (10 male and 26 female) were recruited by public advertisement. Subjects were included if they were aged between 20 and 65 y, had a fasting plasma insulin greater than 12 mu/l, and a body mass index (BMI) between 27 and 43 kg/m 2. Data from the American National Health and Nutrition Examination Survey showed that BMI values of 27.8 kg/m 2 for men and 27.3 kg/m 2 for women were the cutoff points associated with a substantially increased risk of morbidity and mortality. 22 Subjects with a BMI greater than 43 kg/m 2 were excluded because of technical difficulties in performing DEXA. Exclusion criteria included women who were not postmenopausal or infertile, type I or II diabetes, proteinuria, or a history of liver, unstable cardiovascular, respiratory, gastrointestinal disease or malignancy. All subjects were asked to maintain exercise routines at levels established prior to the study. Subjects on antihypertensive or lipid lowering medication were asked to maintain the same dose throughout the study. All subjects gave informed written consent to participate in the study, which was approved by the Human Ethics Committees of the Commonwealth Scientific Industrial Research Organisation, and the Royal Adelaide Hospital. Diets The prescribed diets were (a) HP diet (30% of energy (%E) from protein (B110 g/day), 40%E from carbohydrate) and (b) lower-protein (LP) diet (15%E from protein (B60 g/day), Experimental design The study was conducted on an outpatient basis. Subjects were matched on the basis of fasting plasma insulin, BMI, age, and gender. The two matched groups were randomly assigned to either the HP or LP diet. Both the HP and LP groups underwent 12 weeks of energy restriction (B30% restriction of total energy), plus 4 weeks at energy balance on the same macronutrient composition. At weeks 0 and 16, body weight and composition, TEE, REE, RQ, the TEF, and energy expenditure attributable to physical activity (PAEE) were assessed. All measurements were performed by the same investigator. Subjects were instructed to maintain similar activity and food intake throughout the assessment of energy expenditure. The 3- day food and activity diaries were used to assess compliance beginning the day prior to measurements being made at baseline and at week 16. In addition, 24-h urine samples were assessed for urea and creatinine at weeks 0 and 16. Methods Body weight and body composition. Body weight (Mercury digital scales, model AMZ14) was recorded in light clothing. Total fat mass (TFM) and total lean mass (TLM) were assessed by whole-body dual-energy X-ray absorptiometry (DEXA) (Norland densitometer XR36; Norland Medical Systems, Fort Atkinson, Wisconsin, USA; CV of % for TFM and % for TLM).

3 584 Total energy expenditure. TEE was measured using the [ 14 C]- bicarbonate urea method, 25 an isotopic dilution technique that indirectly calculates TEE from the measured specific activity of CO 2 incorporated into urinary urea. The single measurement of TEE takes 48 h. An area of skin on the abdomen was anesthetized with 2% lignocaine hydrochloride (Delta West Pty Ltd; Bentley, Western Australia), and a small flexible infusion cannula (MMT-316 Sof-set Infusion Set (60 cm tube); MiniMed Technologies; West Chatswood, NSW, Australia) was inserted subcutaneously while subjects were supine. A bolus priming dose of 14 C-urea (B0.25 mci) (Amersham Pharmacia Biotech; Castle Hill NSW, Australia) was administered, and thereafter the infusion cannula was connected to a 20 ml syringe (Terumo Syringe with Leur Lock tip; Terumo Corporation, Springfield, SA, Australia) containing ml of a sterile and pyrogen-free [ 14 C]-bicarbonate solution ( mci/ml). The labeled solution was administered over 48 h (commencing immediately after the measurement of REE on day 1) by constant infusion using a mini-pump syringe driver (SIMS Graseby MS16A Syringe driver; SIMS Australasia PTY.Ltd; Bundall QLD, Australia). Subjects completed a 24 h urine save (commenced after the first morning specimen of day 2, until, and including, the first morning specimen of day 3) while doing their normal activities at home or at work. Aliquots of urine were stored at 201C until analyzed for the specific activity urea using a method based on that of Elia et al. 25 After removal of atmospheric CO 2 dissolved in the urine, urease was added and the CO 2 released from urea was sequestered by a CO 2 trapping solution. The specific activity of the trapped CO 2 was determined by scintillation counting, and thereafter total CO 2 production per day was indirectly calculated using the equation 25 CO 2 production ðmol=dayþ ¼ 0:950:85infused bicarbonate ðdpm=dayþ Specific activity of urea ðdpm=molþ TEE was calculated from net CO 2 production assuming an energy equivalent of 535 kj/mol CO 2 (which approximates the value obtained in subjects close to nutrient balance, who have an RQ of B0.85). 26 Preliminary studies in our group have established that the [ 14 C]-bicarbonate urea method has an intra-assay CV of % in lean subjects (eight men) and % in obese subjects (six men/nine women). Within subjects the variation between study days was 4.871% in lean subjects and % in obese subjects. REE and RQ Fasting REE and RQ were measured over 30 min by indirect calorimetry using a ventilated canopy and Deltatract metabolic monitor (Datex Division Instrumentarium Corp., Helsinki, Finland). The instrument was calibrated before each set of measurements. After an 8 h fast, subjects lay supine on a bed in a thermoneutral environment with a clear plastic hood over their head and shoulders, and the REE and RQ were recorded for 30 min. The first 10 min of data were discarded to ensure all subjects had reached equilibrium, and the remaining 20 min of data were averaged and represented the values for fasting REE and RQ. The intraindividual CV of the Deltatrac system was established to be % for fasting REE, % for fasting RQ, and % for TEF. 27 Postprandial RQ and the thermic effect of food (TEF). After the measurement of REE, each subject consumed a 2715 kj HP (32%E as protein, 54%E as carbohydrate) or a 2747 kj LP (10%E as protein, 77%E as carbohydrate) test meal that was representative of their study diet (Appendix), within 20 min. The compositions of the two test meals were different to the macronutrient composition of the overall study diet because the aim was to match the caloric content while maximizing the protein-to-carbohydrate exchange between the two single meals. Thereafter, subjects returned to the hood for 180 min during which RQ and REE values were recorded every 20 min. Fasting RQ was subtracted from every 20-min postprandial RQ value, and the 20-min values over the 180- min period were averaged to determine the mean change in RQ for the subsequent 3 h. Similarly, TEF was calculated from fasting and postprandial REE values. Physical activity energy expenditure. Physical activity was assessed using a 3-day physical activity diary, and an estimation equation, in order to determine whether or not physical activity remained stable over the 3 days of energy expenditure measurements. Individuals were instructed on how to complete diaries and provided with written guidelines along with an example. The investigator reviewed this record with the subject at the completion of the TEE measurement and subjectively graded each day s overall activity from a table that specified physical activity indexes (multiples of resting metabolic rate) for eight different levels of activity. 28 In addition, on the day that TEE was measured (ie day 2), energy expenditure because of physical activity was calculated using the semiqualitative estimation equation: [ 14 C]-bicarbonate derived PAEE (kj/day) ¼ 0.9([ 14 C]-bicarbonate TEE) REE The constant of 0.9 was based on the assumption that 10% of TEE is because of the TEF. 19 Statistical analyses All data are presented as means7s.e.m. Statistical analysis was performed using SPSS 10.0 for windows (SPSS Inc., Chicago, USA). The effect of 12 weeks of energy restriction and 4 weeks of weight maintenance was assessed using repeated-measures ANOVA with time as the within-subject factor, and diet and gender as the between-subject factors.

4 The study had 80% power (a 0.05) to detect a 3.6 kg difference in body weight, a 2.7 kg difference in TFM, a 1.8 kg difference in TLM, a 7% difference in REE, and an 8% difference in TEE between the dietary groups. The study was not sufficiently powered to examine gender specific changes. Significance was set at Po0.05. Results Subject compliance to the study the diets All 36 subjects completed the study. The physical characteristics of subjects in the HP and LP groups at week 0 are shown in Table 1. There was no difference in any of the variables between diet groups. The energy and macronutrient contents derived from the subjects daily weighed food checklists are shown in Table 2, and the energy content and macronutrient composition were similar to those of the prescribed diets. During energy balance, the macronutrient composition of both diets remained the same as during the energy-restriction phase, but the energy content of each diet increased (Po0.001). From weeks 0 to 16, the urea/creatinine ratio increased by % ( ) on the HP diet (P ¼ 0.026), but remained unchanged (from to ) on the LP diet. Body weight and body composition Figure 1 depicts the change in body weight over the duration of the study for the HP and LP dietary groups. After 12 weeks of energy restriction and 4 weeks of energy balance, the overall mean decrease in body weight was kg. There was no effect of diet composition on the reduction in body weight ( vs kg, on the HP and LP diets, respectively). Men were 18.4% heavier than women at both weeks 0 and 16 (Po0.001 for overall effect of gender), and they lost 41.9% more weight than women after 16 weeks of dietary intervention (Po0.001). There was, however, no dietby-gender interaction on the decrease in body weight. After the 16 weeks of dietary intervention, total body fat and abdominal fat mass both decreased ( and kg, respectively) from weeks 0 to 16 (Po0.001). TLM was reduced by kg (Po0.001). Diet composition had no effect on the reductions in total and abdominal fat mass, or on TLM. There was no effect of gender, nor any diet-by-gender interaction on the reductions in total and abdominal fat mass, or on TLM. Total and REE Overall, there was a nonsignificant decrease (4.1%) in TEE from kj/day at week 0 to kj/day at week 16 (Table 3). Diet composition had no affect on TEE. At both weeks 0 and 16, men had a greater TEE than women ( vs 9955 kj/day, Po0.001 for overall effect of gender). After 12 weeks of energy restriction and 4 weeks of energy balance, the overall mean reduction in REE was 8.8% from at baseline to kj/day by week 16 (Po0.001) (Table 3). There was no effect of diet composition on REE (Table 3), regardless of whether REE was expressed as an absolute value, or as REE per kg of body weight or TLM. At both weeks 0 and 16, men had a greater REE than women (9117 vs 6842 kj/day, Po0.001 for overall effect of gender), but there was no effect of gender on the reduction observed for REE. RQ, the TEF, and physical activity energy expenditure Overall, the mean fasting RQ increased from at baseline to at week 16 (Po0.001). Diet composition had a significant effect on fasting RQ such that the increase in the LP diet group (from at baseline to at week 16) was 62% greater than in the increase in the HP group 585 Table 1 Subject characteristics at baseline a LP diet HP diet Combined n (6M/13F) n (4M/13F) n (10M/26F) Age (y)* Body weight (kg)* BMI (kg/m 2 ) TFM (kg) TLM (kg)* Insulin (mu/l)* Glucose (mmol/l)* TFM, total body fat mass; TLM, total body lean mass; BMI, body mass index (weight in kg/height in m 2 ). Baseline measurements at week 0 were assessed using a two-anova with diet and gender as the fixed factors to determine whether there were differences between the LP and HP dietary groups before the intervention commenced. a Data are expressed as means7s.e.m. Subject characteristics for the LP and HP groups were not different. * Significant differences between men and women, with the men having a greater body weight (Po0.001), lean mass (Po0.001), and insulin (P=0.027) and glucose concentration (P=0.029) than women. Women were older than men (P=0.027).

5 586 Table 2 Dietary composition of the study diets derived from subjects daily weighed food records a LP diet (n=19) HP diet (n=17) ER EB ER EB Energy (kj/day) w z w z Protein (%E) w w z z Carbohydrate (%E) w w z z Total fat (%E) w w w w SFA (%E) w w w w MUFA (%E) w w w w PUFA (%E) w z w w Fiber (g/day) w w z z Cholesterol (mg/mj) w w z z Three days (2 week and 1 weekend day) of dietary data were analyzed at weeks 2, 4, 6, 8, 10, and 12 during the energy restrictive (ER) phase of the study, and at weeks 14 and 16 during energy balance (EB). No differences were found between the six diet records in the ER phase or between the two records for the EB phase, so data for recordings in each phase were averaged. %E, % of total energy; SFA, saturated fatty acid; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid. ER and EB data were compared using repeated-measures ANOVA with diet and gender as between subject factors. a Data are expressed as means7s.e.m. Common superscripts denote that the composition of the allocated diets were not different (P>0.05). Figure 1 Body weight during 12 weeks of energy restriction (denoted by the solid arrow 2) and 4 weeks of energy balance (denoted by the hatched arrow ) on a normal protein (LP, n ¼ 19) or high-protein diet (HP, n ¼ 17). Body weights were measured at weeks 0, 4, 8, 10, 12 (energy restriction), and 16 (end of energy balance) of the dietary intervention. Data were compared using repeated-measures ANOVA with diet and gender as between subject factors. Values are means7s.e.m. *Significant decrease in weight from weeks 0 to 12 (Po0.001), but there was no effect of diet, or any time-by-diet interactions. From weeks 12 to 16, there was no decrease in weight. (from at baseline to at week 16) (time-by-diet effect, P ¼ 0.02) (Table 3). There was no effect of gender on fasting RQ. At week 0, postprandial RQ after the LP meal was approximately 54.5% greater than after the HP meal ( vs , Po0.001) (Table 3). At week 16, postprandial RQ remained 33.3% greater following the LP meal as compared to the HP meal, but the effect of diet was no longer significant (0.009 vs 0.006, P ¼ 0.06). There was a significant time-by-diet effect on postprandial RQ (P ¼ 0.003); in the LP group postprandial RQ decreased from at baseline to at week 16, and increased from to in the HP group. There was no effect of gender on postprandial RQ. Postprandial RQ was not related to either the 5.3 mu/l decrease in fasting plasma insulin (r ¼ 0.02, P ¼ 0.93) or the 11.3% decrease in insulin area under the curve (r ¼ 0.05, P ¼ 0.78). At week 0, the TEF after the HP meal (expressed either as an absolute value or as a percentage of energy intake at the test meal) was 2% greater than for the LP meal (9.1 vs 7.1%, P ¼ 0.009) (Table 3). At week 16, TEF remained 0.8% greater following the HP as compared to the LP meal, but the effect of diet composition was nonsignificant (8.6 vs 7.8%, P ¼ 0.3) (Table 3). After 12 weeks of energy restriction and 4 weeks of energy balance, TEF was not significantly reduced from baseline in either the LP or HP group (mean change of and 0.56%, respectively) (Table 3). There was no effect of gender on TEF. There was no correlation between TEF at weeks 0 and 16, and the change in body weight. At baseline as compared to week 16, energy expenditure because of physical activity was not significantly different ( vs kj/day). Diet composition had no effect on energy expenditure because of physical activity (Table 3). In accordance with this result, we also observed that the physical activity index (a multiple of REE) derived from 3-day physical activity diaries was not different at week 16 as compared to week 0 ( vs ). There was no significant effect of gender on physical activity.

6 Table 3 Energy expenditure variables and respiratory quotient at weeks 0 and 16 of dietary intervention a 587 LP diet HP diet Combined n (6M/13F) n (4M/13F) n (10M/26F) TEE (kj/day) w Wk Wk Change REE (kj/day) wz Wk Wk Change TEF (%EI) y Wk Wk Change PAEE (kj/day) Wk Wk Change Fasting RQ zz Wk Wk Change Postprandial RQ zyz Wk Wk Change TEF, mean thermic response to a 2747 kj LP (10% protein, 77% carbohydrate) test meal, or a 2715 kj HP (32% protein, 54% carbohydrate) test meal over 3 h. The TEF is expressed as a % of energy intake (EI) at the test meal; fasting RQ, ratio of VCO 2 /VO 2 ; postprandial RQ, mean increase above fasting RQ over 3 h after test meal (increase/3 h); change, the difference between week 16 values and week 0. Wk 0 and Wk 16 data were compared using repeated-measures ANOVA with diet and gender as between subject factors. a Data are means7s.e.m. w Significant effect of gender at both weeks 0 and 16, Po z Significant effect of time from week 0 to 16, Po y Significant effect of diet at weeks 0, P= z Significant time-by-diet effect, Po0.02. Discussion The findings from this study showed that, over 16 weeks, total energy intake and not the protein-to-carbohydrate ratio of the diet is the determinant for weight loss. This observation is in accordance with previous findings from our group 23 in obese subjects with type II diabetes. It also agrees with several other short-term studies (7 days to 10 weeks) in obese nondiabetic subjects that showed equivalent weight loss (mean change 3 8%) on energy-restricted diets ( MJ/day) with increased protein (36 49% of energy as protein) as compared to diets with LP contents (10 37%). 16,21 In contrast, a study in 13 hyperinsulinemic men showed 28% more weight loss (8.3 vs 6.0 kg) over 4 weeks on a diet containing 45% protein as compared to an isocaloric (B7.4 MJ/day) diet with 12% protein. 13 A significant difference in total body water loss ( 1.0 vs 0.3 l) between the HP and LP diets, rather than fat loss ( 7.1 vs 6.3 kg), may explain some of the differential weight loss in Baba et al s 13 study. In the present study, total body water was not measured, and the 0.6 kg (8.4%) greater loss of fat mass and the 0.5 kg (29.4%) smaller loss of lean mass on the HP as compared to the LP diet were not statistically significant and possibly explain why the decrease in body weight was not greater on the HP diet. In Baba et al s 13 study there was an 11.5% smaller decrease in REE (equates to a difference of 1053 kj/day) on the HP as compared to their LP diet. A blunting in REE of this magnitude would be sufficient to explain the differential weight loss between Baba et al s 13 two diets over the 4 weeks of energy restriction. In the present study, the fall in REE was not blunted by the increased protein content of the HP diet (even when adjusted for body weight or lean mass) and we observed a similar result in our previous study in subjects with type II diabetes. 27 Both of our studies had sufficient power (80%, a 0.05) to detect a 3.6 kg

7 588 difference in the mean body weight, a 2.7 kg difference in fat mass, a 1.8 kg difference in lean mass and a 7% difference in REE, between the HP and LP dietary groups. In response to energy restriction, 29 a 13% decrease of body weight in obese subjects resulted in a 9% decrease in REE, which was not significant once normalized for lean mass. There was no difference in lean mass between the groups in our study. Disparity in outcomes for weight loss and REE between the study of Baba et al 13 and our study may have been a consequence of two-fold greater dietary protein content in the HP diet that they used as compared to protein content of the HP diet that we used. We found a small (4.2%) but not statistically significant decrease in TEE after weight loss. Our study had enough power (80%, a 0.05) to detect an overall fall in TEE of 8%. In all, 15 of the 36 subjects had a decrease in TEE of 8% or more, but in the remaining 21 subjects there was a large degree of variance in the change in TEE; seven individuals experienced a decrease of less than 8% and 14 experienced an increase in TEE. This is the first time we have examined the effect of HP diets on TEE. Our findings are in contrast to those of Whitehead et al, 21 the only study to have shown that an increased protein intake during energy restriction blunts both the fall in TEE as well as the sleeping metabolic rate. However, since the fall in sleeping metabolic was significantly less in Whitehead et al s 21 HP group than in their LP group, it is not surprising that the fall in TEE was also significantly blunted. Discrepant findings between Whitehead et al 21 and the present study may also be the result of a larger difference in the protein content between the HP and LP diets (21% difference in their study vs 12% difference in this study), or they may be because of the larger variability observed in the change in TEE in our subjects as compared to theirs. We used the [ 14 C]-bicarbonate urea method to measure TEE whereas Whitehead et al 21 used whole-body indirect calorimetry. The [ 14 C]-bicarbonate urea method, which does not restrict subjects to an artificial living environment, has been shown to measure TEE within 2 5% of the value derived from whole-body indirect calorimetry. 26,30,31 Our observation that TEE was not significantly decreased after weight loss is consistent with the findings reported by others. 32,33 Both Rumpler et al 32 and Amatruda et al 33 measured TEE after their subjects body weights had been stabilized at the reduced level for a period of time. Therefore, the timing of the energy expenditure measurements (ie during dynamic weight loss vs after stabilization at reduced weight) may have also impacted on the disparity in outcomes between the study of Whitehead et al 21 and our study. In accordance with previous findings from our group, 27 as well as others, 19,34 we observed that the TEF was greater after the HP meal than after the LP meal, in subjects with hyperinsulinemia. However, the increased TEF in the HP group only wasted 0.8 2% more energy over the 3 h than the TEF in the LP group, which was not sufficient to influence weight loss in this study. Over 6 months, during energy balance, and assuming that our subjects eat three 2.7 MJ meals per day (or a total of 8.1 MJ/day), the extra kj/day being burnt on the HP as compared to the LP diet may equate to a difference of only kg between the two diets. Accordingly, given the large with-subject day-today variation in the measurement of TEF and the dominant effects of energy intake and physical activity on energy balance, the impact of TEF on either weight loss or weight maintenance is likely to be minimal, at least in the intermediate term. Since the objective of this study was to determine the effect of an HP energy-restrictive diet on weight loss and the changes in energy expenditure, our subjects were asked to maintain physical activity constant throughout the study. Our findings showed that the average daily level of physical activity, assessed from the 3-day activity diaries and a published table of activity indexes, was similar during the 3 days of energy expenditure measurements at both weeks 0 and 16. We also found that the average energy expended because of physical activity as derived from the labeled bicarbonate (PAEE) studies was similar at weeks 0 and 16. There was, however, substantial variation in the change in PAEE between our individuals, and since the mean error in the measurement of PAEE is 41% (B950 kj/day), it is likely that small changes in PAEE were not detected. The greater increase in postprandial RQ after the LP meal, as well as the greater increase in fasting RQ after 16 weeks on the LP diet, presumably reflects the greater carbohydrate content of the LP diet used in this study, and is consistent with our previous findings in subjects with type II diabetes. 27 The magnitude of the increase in postprandial RQ was, however, smaller than expected. This suggests that, in our subjects with insulin resistance, the ability to switch from predominantly lipid oxidation, during fasting, to increased glucose uptake, oxidation and storage after feeding, was reduced. A study by Zurlo et al 35 in 152 nondiabetic Pima Indians showed that high fasting insulin concentrations were correlated with low 24-h RQ, which is consistent with the notion that insulin resistance is associated with lower rates of carbohydrate oxidation. After weight loss, we speculated that an improvement in insulin sensitivity may enhance carbohydrate oxidation as a consequent of improvements in glucose uptake and the suppression of both hepatic glucose release and the release of free fatty acids from the adipose tissue. 35 After 16 weeks, however, significant decreases in fasting insulin and postprandial insulin area under the curve were not related to the increase in postprandial RQ. The implication of having a reduced capacity to switch from fat to carbohydrate oxidation is that it is more difficult to maintain weight loss as compared to someone who has a larger capacity to switch between dietary substrates. We conclude that total energy intake and not the proteinto-carbohydrate ratio of the diet is the major determinant of weight loss in insulin-resistant subjects. The HP diet did not blunt the decrease in energy expenditure that may be

8 associated with weight loss. It remains to be determined whether increased protein diets are of benefit in the maintenance of weight loss, or associated with better compliance with energy restriction in the long term. Acknowledgements This study was supported by a National Health and Medical Research Grant # and the Diary Research and Development Corporation of Australia. We also gratefully acknowledge Rosemary McArthur, Anne McGuffin, Jodie Avery, and Eleni Argyriou for assistance in performing these studies References 1 Kelly GS. Insulin resistance: lifestyle and nutritional interventions. Altern Med Rev 2000; 5: Kopelman P. Obesity as a medical problem. Nature 2000; 404: Belfiore F, Iannello S. Insulin resistance in obesity: metabolic mechanisms and measurement methods. Mol Genet Metab 1998; 65: Boden G. Role of fatty acids in the pathogenesis of insulin resistance and NIDDM. Diabetes 1997; 46: Despres JP, Lamarche B, Mauriege P, Cantin B, Lupien PJ, Dagenais GR. Risk factors for ischaemic heart disease: is it time to measure insulin? Eur Heart J 1996; 17: Reaven G. Syndrome X. J Clin Endocrinol Metab 2001; 3: McLaughlin T, Abbasi F, Kim HS, Lamendola C, Schaaf P, Reaven G. Relationship between insulin resistance, weight loss, and coronary heart disease risk in healthy, obese women. Metabolism 2001; 50: Golay A, Eigenheer C, Morel Y, Kujawski P, Lehmann T, de Tonnac N. Weight-loss with low or high carbohydrate diet? Int J Obes Relat Metab Disord 1996; 20: Goodpaster BH, Kelley DE, Wing RR, Meier A, Thaete FL. Effects of weight loss on regional fat distribution and insulin sensitivity in obesity. Diabetes 1999; 48: Pascale RW, Wing RR, Blair EH, Harvey JR, Guare JC. The effect of weight loss on change in waist-to-hip ratio in patients with type II diabetes. Int J Obes Relat Metab Disord 1992; 16: St. Jeor ST, Howard BV, Prewitt ET, Bovee V, Bazzarre T, Eckel RH. Dietary protein and weight reduction: a statement for healthcare professionals from the Nutrition Committee of the Council on nutrition, physical activity, and metabolism of the American Heart Association. Circulation 2001; 104: Wylie-Rosett J. Fat substitutes and health: an advisory from the nutrition committee of the American Heart Association. Circulation 2002; 105: Baba NH, Sawaya S, Torbay N, Habbal Z, Azar S, Hashim SA. High protein vs high carbohydrate hypoenergetic diet for the treatment of obese hyperinsulinemic subjects. Int J Obes Relat Metab Disord 1999; 23: Skov AR, Toubro S, Ronn B, Holm L, Astrup A. Randomized trial on protein vs carbohydrate in ad libitum fat reduced diet for the treatment of obesity. Int J Obes Relat Metab Disord 1999; 23: Seino Y, Seino S, Ikeda M, Matsukura S, Imura H. Beneficial effects of high protein diet in treatment of mild diabetes. Hum Nutr Appl Nutr 1983; 37A: Piatti PM, Monti F, Fermo I, Baruffaldi L, Nasser R, Santambrogio G, Librenti MC, Galli, Kienle M, Pontiroli AE, Pozza G. Hypocaloric high-protein diet improves glucose oxidation and spares lean body mass: comparison to hypocaloric high-carbohydrate diet. Metabolism 1994; 43: Vazquez JA, Kazi U, Madani N. Protein metabolism during weight reduction with very-low-energy diets: evaluation of the independent effects of protein and carbohydrate on protein sparing. Am J Clin Nutr 1995; 62: Hoffer LJ, Bistrian BR, Young VR, Blackburn GL, Matthews DE. Metabolic effects of very low calorie weight reduction diets. J Clin Invest 1984; 73: Westerterp KR, Wilson SA, Rolland V. Diet induced thermogenesis measured over 24h in a respiration chamber: effect of diet composition. Int J Obes Relat Metab Disord 1999; 23: Tappy L, Jequier E, Acheson K. Thermic effect of infused amino acids in healthy humans and in subjects with insulin resistance. Am J Clin Nutr 1993; 57: Whitehead JM, McNeill G, Smith JS. The effect of protein intake on 24-h energy expenditure during energy restriction. Int J Obes Relat Metab Disord 1996; 20: Kuczmarski RJ, Flegal KM, Campbell SM, Johnson CL. Increasing prevalence of overweight among US adults. The National Health and Nutrition Examination Surveys, 1960 to JAMA 1994; 272: Parker B, Noakes M, Luscombe N, Clifton P. Effect of a high protein, high-monounsaturated fat weight loss diet on glycemic control and lipid levels in Type 2 diabetes. Diabetes Care 2002; 25: Cashell K, English R, Lewis J. Composition of foods. AGPS: Canberra, Australia; Elia M, Jones MG, Jennings G, Poppitt SD, Fuller NJ, Murgatroyd PR, Jebb SA. Estimating energy expenditure from specific activity of urine urea during lengthy subcutaneous NaH14CO3 infusion. Am J Physiol 1995; 269: E172 E Elia M. Energy equivalents of CO2 and their importance in assessing energy expenditure when using tracer techniques. Am J Physiol 1991; 260: E75 E Luscombe ND, Clifton PM, Noakes M, Parker B, Wittert G. Effects of energy-restricted diets containing increased protein on weight loss, resting energy expenditiure, and the thermic effect of feeding in Type 2 diabetes. Diabetes Care 2002; 24: NHMRC. Report of National Health and Medical Research Council. In: Recommended dietary intakes for use in Australia. Australian Government Publishing Services: Canberra; Ravussin E, Burnand B, Schutz Y, Jequier E. Energy expenditure before and during energy restriction in obese patients. Am J Clin Nutr 1985; 41: Gibney E, Elia M, Jebb SA, Murgatroyd P, Jennings G. Total energy expenditure in patients with small-cell lung cancerfresults of a validated study using the bicarbonate-urea method. Metab: Clin Exp 1997; 46: Paton NJ, Elia M, Jebb SA, Jennings G, Macallan DC, Griffin GE. Total energy expenditure and physical activity measured with the bicarbonate-urea method in patients with human immunodeficiency virus infection. Clinical Science 1996; 91: Rumpler WV, Seale JL, Miles CW, Bodwell CE. Energy-intake restriction and diet-composition effects on energy expenditure in men. Am J Clin Nutr 1991; 53: Amatruda JM, Statt MC, Welle SL. Total and resting energy expenditure in obese women reduced to ideal body weight. J Clin Invest 1993; 92: Westerterp PM, Rolland V, Wilson SA, Westerterp KR. Satiety related to 24 h diet-induced thermogenesis during high protein/ carbohydrate vs high fat diets measured in a respiration chamber. Eur J Clin Nutr 1999; 53: Zurlo F, Lillioja S, Puente A, Nyomba B, Raz I, Saad M, Swinburn B, Knowler W, Bogardus C, Ravussin E. Low ratio of fat to carbohydrate oxidation as predictor of weight gain: study of 24-h RQ. Am J Physiol 1990; 259: E650 E

9 590 Appendix This contains an itemized composition of the LP and HP test meals used in the assessment of the meal tolerance and TEF tests (see Table 4). Table 4 Itemized composition of the LP and HP test meals Food Measure Energy (kj) Protein (g) Fat (g) CHO (g) LP test meal White bread Four slices Cream cheese 20 g Strawberry jam 20 g Dried Apricots 40 g Orange juice 500 g Total % Energy HP test meal White bread Two slices Lean leg ham 50 g Kraft freet cheese 60 g Pura classic litet Flavored milk 525 g Total % Energy a Kraft freet cheese (Kraft Foods Ltd; Melbourne, Australia); Pura classic litet flavored milk (National Foods Ltd; Salisbury, South Australia).