Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
International Journal of Obesity (2000) 24, 450±459 ß 2000 Macmillan Publishers Ltd All rights reserved 0307±0565/00 $15.00 www.nature.com/ijo Leptin is in¯uenced both by predisposition to obesity and diet composition A Raben1* and A Astrup1 1 Research Department of Human Nutrition, Centre for Food Research, The Royal Veterinary and Agricultural University, Frederiksberg, Denmark OBJECTIVE: (1) To investigate whether plasma leptin concentrations differ between subjects with and without the genetic predisposistion to obesity, and (2) to investigate the effect of dietary manipulations on plasma leptin in these subjects. DESIGN: Fasting and postprandial plasma leptin concentrations were measured before and after 14 days' ad libitum intake of a fat-rich (FAT), starch-rich (STARCH) or sucrose-rich (SUCROSE) diet. On day 15 ad libitum breakfast and lunch were given and blood sampled regularly until 6 p.m. SUBJECTS: Eight normal-weight, post-obese women and 10 matched controls (body mass index, 23.5 0.5 and 22.9 0.3 kg=m2). MEASUREMENTS: Leptin, glucose, insulin, appetite ratings, dietary intake, body weight and composition. RESULTS: Fasting leptin concentration on day 1 or 15 did not differ between post-obese and controls. However, after meal intake leptin increased in post-obese compared with controls on all three diets. In both groups fasting and postprandial leptin concentrations were greater after SUCROSE compared with FAT and STARCH. CONCLUSION: A larger postprandial leptin concentration was observed in post-obese subjects than in controls. This may be related to greater insulin sensitivity in adipose tissue in the post-obese. Furthermore, increased leptin concentrations were found after a sucrose-rich diet in both groups, possibly related to larger postprandial insulin peaks on this diet. Both contentions should, however, be validated by further studies. International Journal of Obesity (2000) 24, 450±459 Keywords: starch; fat; sucrose; post-obese; glucose; insulin; appetite Introduction The OB-protein leptin, secreted from adipose tissue and ®rst described in 1994,1 may be important in the development of obesity by in¯uencing both food intake and energy expenditure. Thus, leptin has been found to decrease food intake and increase energy expenditure both in ob=ob mice2 ± 4 and in non-obese mice.2 In humans, leptin has repeatedly been shown to be positively associated with body weight, body mass index and fat mass.5 ± 8 Also, in humans, homozygous mutation in the genes encoding leptin or the leptin receptor is associated with early-onset obesity, along with no pubertal development and other endocrine de®ciencies.9 Two hypotheses have resulted from these observations. Either a failure to produce leptin in the adipose tissue or a resistance to its action in the central nervous system may result in increased body weight and obesity. The latter hypothesis has, however, recently been challenged by Arch et al.10 They stated that, if leptin resistance is a cause rather than a consequence of obesity, leptin concentrations would *Correspondence: A Raben, Research Department of Human Nutrition, Centre for Food Research, The Royal Veterinary and Agricultural University, 30 Rolighedsvej, DK-1958 Frederiksberg C, Denmark. E-mail: [email protected] Received 29 March 1999; revised 6 September 1999; accepted 4 November 1999 be expected to be raised in subjects who are prone to obesity but who have slimmed. Whether low fasting leptin concentrations predispose to obesity has already been investigated in prospective studies. In one study low leptin concentrations predicted weight gain in 36 very obese, middle-aged Pima Indians.11 However, in another study on 180 moderately obese, non-diabetic Mexicans studied after 3.25 y no such connection was found.8 Likewise, Lindroos et al found that basal, fasting leptin concentrations were not predictive of 4 y weight changes in 49 severely obese women with a familial predisposition to obesity.12 Since the obese state is in itself a confounding factor when trying to disclose the reasons for obesity, studies of leptin concentrations in normal-weight subjects with the genetic predisposition to obesity (ie pre- or postobese) are, however, more interesting. Furthermore, fasting leptin concentrations do not represent the diurnal pattern. Thus differences in leptin concentrations may occur at other times during the day or night. With regard to food intake, leptin in humans has been found to increase during acute and chronic overfeeding,13 and to decrease during fasting and longer-term energy restriction.14,15 Studies on the impact of changes in dietary composition (1 week ± 4 months), when energy balance and body weight is maintained, have in general shown no effect on fasting leptin of major changes in dietary composition Leptin, predisposition to obesity and diet A Raben and A Astrup (fat content 14 ± 60 energy%).16 ± 18 One of these studies found that 8 months' ad libitum low-fat diet decreased leptin only in those who lost body weight, indicating that the diet per se had no effect.18 The decrease in leptin also paralleled a decrease in body fat, resting energy expenditure and fasting insulin.18 Studies on the in¯uence of acute meal intake on postprandial leptin show more discordant results. Thus, some found an effect of meal intake on leptin concentrations,19 ± 22 while others did not.5,17,23 ± 27 These discrepancies may relate to differences in the duration of the postprandial measurement period (1 ± 10 h), different types of test meals or types of subjects used. Especially, a measurement period of at least 8 h seems to be necessary to discover any diet-induced differences in plasma leptin concentrations. The purpose of the present study was two-fold: (1) to compare fasting and postprandial leptin concentrations in normal-weight, post-obese with never-obese subjects, (2) to investigate the impact of a fat-rich, a starch-rich=low-fat (`recommended') or a sucroserich=low-fat diet, on fasting and 8-h postprandial leptin concentrations after 14 days' ad libitum intake in the two study groups. Methods Subjects Eighteen healthy, normal-weight women, eight postobese (PO) and 10 controls (C), matched on a groupbasis for age, weight, height, percentage fat mass, and percentage fat-free mass participated in the study (Table 1). The post-obese women had a family history of obesity (at least one obese parent or sibling), had been more than 10% overweight (average s.e.m. 38 9%), and had been weight-stable for at least 2 months. None had undergone surgical operations to become normal-weight. The study was approved by the Municipal Ethical Committee of Copenhagen and Frederiksberg as in accordance with the Helsinki II declaration. All subjects gave written consent after the experimental procedure had been explained to them. Table 1 Subjects characteristicsa Age (y) Height (m) Weight (kg) Body mass index (kg=m2) Maximal weight (kg) Previous overweightb (%) Fat mass (%) Fat-free mass (%) Post-obese (n 8 ,) Controls (n 10 ,) 40 4 1.67 1.3 65.4 1.2 23.5 0.5 89 4 38 9 28.7 1.1 71.3 1.1 38 3 1.65 1.3 62.1 1.3 22.9 0.3 63 1c 2 6 29.2 0.7 70.8 0.7 Data are means s.e.m. a Measured on day 1 of the ®rst of three ad libitum diets. bSee Ref. 46. cPost-obese vs controls, P < 0.05 by unpaired t-test. 451 Experimental design The experimental design has been described in detail previously.28 In brief, each subject completed three 14 day ad libitum dietary periods, a sucrose-rich (SUCROSE), a starch-rich (STARCH) and a fat-rich (FAT) period. The order of the periods differed, but subjects in the post-obese and control group were following similar diet orders. Each subject was measured at the same time in her menstrual cycle. Before each experimental period subjects were given a standardized, weight-maintenance diet for 3 days (days -2, -1 and 0). After the standard diet, the experimental diets were supplied in ad libitum amounts (see below) to be consumed at home for 14 days. Days 0 and 14 were spent in our respiration chambers. In the morning on day 15 blood samples were taken in the fasted state and postprandially after breakfast and lunch until 6 p.m. Body weight and composition were measured in the fasting state on day 1 and 15. At least 2 weeks and no more than 6 weeks separated the dietary periods. The subjects were instructed not to change their physical activity pattern during or between the three experimental diets. Data for 20 subjects on ad libitum energy intake, body weight and composition, 24 h energy expenditure, substrate oxidations, appetite sensations and habitual food intake, and for 18 subjects (eight PO and 10 C) on plasma catecholamines, blood lipids and coagulation factors, have been published separately.28,29 Diets The standard weight-maintenance diet given before each dietary period contributed 13 energy-percentage (E%) protein, 37 E% fat, 50 E% carbohydrate (9 E% sucrose), 2.9 g=MJ dietary ®ber and had a P=S-ratio of 0.4. The diet was prepared according to each subject individual energy needs adjusted to the nearest 0.5 MJ.30 The average macronutrient compositions of the 14 day ad libitum diets as follows: SUCROSE Ð 59 E% carbohydrate (23 E% sucrose), 28 E% fat and 13 E% protein; FAT Ð 46 E% fat, 41 E% carbohydrate (2 E% sucrose) and 13 E% protein; STARCH Ð 59 E% carbohydrate (2 E% sucrose), 28 E% fat and 13 E% protein. Dietary ®ber amounted to a total of 22, 32 and 20 g=d as on FAT, STARCH and SUCROSE, respectively. The ratio between polyunsaturated and saturated fatty acids (P=S-ratio) was 0.4 on the FAT diet and 0.7 on both the SUCROSE and STARCH diets. The amount of food given for breakfast and lunch on day 15 was similar to the ad libitum amounts consumed on day 14 in the chamber (Table 2). Coffee, tea and water consumption as well as smoking (two PO and one C) were allowed, but the amount and time was copied from the ®rst dietary period. The computer database of foods from the National Food Agency of Denmark (Dankost 2.0) was used in International Journal of Obesity Leptin, predisposition to obesity and diet A Raben and A Astrup 452 Table 2 Ad libitum energy intake on day 15 P-value ANOVA Breakfast Energy (kJ) Lunch Energy (kJ) Breakfast lunch Energy (kJ) Fat Starch Sucrose dg Diet Group PO C 2816 324 2371 192 2141 223A 2307 206 3006 370B 2636 265 0.28 0.03 0.46 PO C 3948 478 3544 310 3022 356 3148 278 4216 400 3188 176 0.07 0.14 0.22 PO C 6764 730A 5915 234 5163 519B 5456 326 7222 658A 5823 364 0.07 0.008 0.24 Data are means s.e.m. PO: post-obese (n 8). C: control (n 10). d g diet grop interaction. Results in the same row with different superscript are signi®cantly different (per diet, n 18; P < 0.05). the calculations of energy and nutrient intake of the diets.31 Anthropometric measurements Body weight was measured Ð blinded for the subjects Ð in the morning of days 1 and 15 after 10 h of fasting and after voiding. The same digital scale was used each time (Seca model 707, Copenhagen). Body composition was subsequently estimated by electrical bioimpedance using an Animeter (HTS-Engineering Inc., Odense, Denmark). Fat mass (FM) and fat-free mass (FFM) were calculated as previously reported.32. Blood sampling On day 15 the subjects left the respiration chamber at 9.00 a.m. After voiding and weighing the subjects lay on a bed in the supine position and a Ven¯on catheter (Viggo, Gothenborg, Sweden) was inserted into an antecubital arm vein. After 10 min rest a fasting blood sample was taken. Breakfast was eaten at 10.00 a.m. and lunch at 2.00 p.m. Blood samples were taken after 15, 30, 60, 120 and 240 min from initiation of breakfast and lunch. The subjects rested in the supine position for 10 min before each blood sample. During the day the subjects could sit, walk quietly or go to the toilet. The type of activity during the day of the ®rst dietary period was noted and repeated thereafter. Appetite sensations Every hour from 9 a.m. (fasting) until 5 p.m. subjects recorded their sensations of hunger, fullness, desire to eat and prospective food consumption on 10 cm visual analog scales (VAS) with words anchored at each end, expressing the most positive (eg good, pleasant) or the most negative rating (eg bad, unpleasant).33 This was done on day 14, while in the respiratory chambers. However, since food and ¯uid intake was repeated on day 15, just postponed by 1 h, appetite data (n 18) could be compared with blood data (n 18). International Journal of Obesity Laboratory analyses Blood was sampled without stasis through an indwelling antecubital cannula. Blood for determination of plasma leptin was sampled in plain tubes, centrifuged at 3000 G and 4 C and stored at ÿ20 C. Leptin was analyzed by Radioimmunoassay using a DRG-Human leptin RIA-kit RIA 1624 from DRG Instruments GmbH, Marburg, Germany. The intraserial coef®cient of variation (CV) was 4.4% and the interserial CV was 4.2%. Blood for glucose analyses was sampled in EDTA=K3=sodium-¯uoride tubes, centrifuged at 1500 g and 20 C and stored at ÿ20 C. Glucose was determined on a Cobas Mira (Roche Diagnostic System), using an end-point analysis with MPR3 Gluco-quant R glucose=HK kinetic 1442457 (Boehringer and Mannheim GmbH Diagnostica) and the hexokinase=G6P-DH method. Blood for insulin was sampled in plain tubes, centrifuged at 3000 g for 10 min at 4 C and stored at ÿ20 C. Serum insulin was determined as a non-competitive sandwich assay with a DAKO insulin kit, code no. K6219 (DAKO A=S, Denmark). Statistical analyses All results are given as means s.e.m. Initial group characteristics were compared by an unpaired t-test (Table 1). The areas under the postprandial curves were calculated separately for each subject as the difference between the integrated area of the response curve and the rectangular area determined by the basal values. Differences in fasting concentrations and areas under the curve (AUC) between the three diets and two groups were tested by parametric analysis of variance (ANOVA) using the GLM procedure in SAS. Factors were group, diet, and group diet using subject (group) as error term for group effects. In case of signi®cance, a t-test on least squares means (for unbalanced designs) was used to test for differences between groups or diets. Differences in the postprandial responses were tested by ANOVA using GLM in SAS and diet time group, diet group, time group, diet time, time, diet Leptin, predisposition to obesity and diet A Raben and A Astrup and group as factors and subject (group) as error term for all group effects. Simple correlation analyses were performed within each diet (n 18). However, in order to explain the observed differences in leptin between diets, two types of analyses were performed. One was simple linear regression analyses using means of groups and diets (n 6). The other was an adjustment of leptin for differences in the relevant parameter (eg body weight). This was done by including the adjustment parameter as a factor (covariate) in the ANOVA. The signi®cance level was set at P < 0.05. Statgraphics Software version 4.2 (Graphic Software Systems Inc., Rockville, MD) and the Statistical Analysis Package version 6.12 for Windows (SAS Institute, Cary, NC) were used in the statistical calculations. Results Energy intake The 14 day ad libitum energy intake (EI) was lowest on STARCH (9.1 0.4 MJ=day) compared with both SUCROSE (10.3 0.5 MJ=day) and FAT (10.3 0.4 MJ=day) (P < 0.05). Compared with the control group, PO consumed signi®cantly more energy on FAT (11.0 0.7 vs 9.7 0.4 MJ=day, P < 0.001) and SUCROSE (11.4 0.7) vs 9.5 0.5 MJ=day, P < 0.0001). On average both groups consumed more energy on SUCROSE than STARCH at breakfast, whereas there were no differences at lunch (Table 2). Total energy intake for both breakfast and lunch was signi®cantly lower for STARCH compared with FAT and SUCROSE (P < 0.01). There were no group differences over the day (Table 2). 453 Leptin Fasting leptin concentrations on days 1 and 15 and changes from day 1 to 15 were similar in PO and C (Table 3). However, in both groups fasting leptin on day 15 and change in fasting leptin (day 15ÿday 1) were higher after SUCROSE compared with the other two diets (diet effect, P < 0.01). After meal intake leptin values initially dropped, but then increased steadily in PO compared with a minor increase in C (group time effect, P < 0.0001; Figure 1). The differences were most pronounced after lunch and were independent on the type of diet. Total areas under the curve (AUC) were higher after SUCROSE than STARCH, but not different between groups (Figure 1). When data were expressed as areas under the curves above baseline (d-AUC's), the d-AUC's were larger in PO on all three diets and only negative values were observed for C (data not shown). There were no diet differences in the d-AUCs, indicating that the higher postprandial concentration and AUC on SUCROSE were due to the higher fasting concentration on this diet compared with FAT and STARCH. Glucose Fasting glucose on days 1 and 15 and the changes from day 1 to day 15 were not different between groups or diets (Table 3). After meal intake on day 15, however, there were signi®cant differences in the glucose response between both diets and groups Table 3 Fasting % plasma concentrations in post-obese and controls before (day 1) and after (day 15) 14 days' ad libitum intake of a diet rich in fat, starch or sucrose P-value ANOVA Fat Before (day1) Leptin (ng=ml) Glucose (mmol=l) Insulin (pmol=l) After (day15) Leptin (ng=ml) Glucose (mmol=l) Insulin (pmol=l) Change (day15ÿday1) Leptin (mg=ml) Glucose (mmol=l) Insulin (pmol=l) Starch Sucrose PO C PO C PO C 8.8 1.6 10.9 1.1 4.73 0.10 4.87 0.13 32 7 34 8 9.9 2.0 10.4 0.9 4.77 0.10 4.82 0.11 32 8 33 7 8.7 1.5 9.8 0.8 4.55 0.08 4.85 0.11 33 9 32 6 PO C PO C PO C 9.7 2.0 10.3 1.5 4.64 0.17 4.76 0.14 39 7 51 6 9.5 1.8 9.3 0.7 4.58 0.12 4.72 0.10 45 6 56 5 11.1 2.2 11.7 1.6 4.46 0.13 4.80 0.10 42 6 64 9 PO C PO C PO C 0.9 0.9a ÿ0.5 1.0 ÿ0.09 0.11 ÿ0.11 0.08 6 6 7 9 ÿ0.4 0.6a ÿ1.1 0.7 ÿ0.19 0.08 0.10 0.06 13 11 23 8 2.4 1.0b 1.9 1.2 ÿ0.09 0.08 ÿ0.05 0.09 9 11 32 10 dg Diet Group 0.35 0.29 0.50 0.07 0.15 0.28 0.89 0.99 0.97 0.73 0.008 0.88 0.18 0.59 0.23 0.61 0.29 0.06 0.78 0.0007 0.43 0.81 0.69 0.61 0.55 0.34 0.22 Data are means s.e.m. PO: eight post-obese women. C: 10 matched control women. d g diet group interaction. Results in the same row with different superscripts are different, P < 0.05 (n 18). International Journal of Obesity Leptin, predisposition to obesity and diet A Raben and A Astrup 454 Figure 1 Mean plasma leptin concentrations in eight postobese (PO) and 10 controls (C) after 14 days' ad libitum intake of a fat-rich (FAT), a starch-rich (STARCH) and a sucrose-rich (SUCROSE) diet. Brkf. breakfast. Signi®cant by ANOVA, P < 0.05: group time, diet (SUCROSE > STARCH) and time. Bottom panel: mean ( s.e.m.) total areas under the curves (AUC). Diet effect, P < 0.01. Different letter: diets differ, P < 0.01. (Figure 2). SUCROSE resulted in the lowest AUC compared with STARCH and FAT and the post-obese subjects had a lower AUC on all three diets compared with controls (Figure 2). International Journal of Obesity Figure 2 Mean plasma glucose concentrations in eight postobese (PO) and 10 controls (C) after 14 days' ad libitum intake of a fat-rich (FAT), a starch-rich (STARCH) and a sucrose-rich (SUCROSE) diet. Brkf. breakfast. Signi®cant by ANOVA, P < 0.01: group time, diet time, group (PO < C, all diets), diet (SUCROSE < STARCH and FAT) and time. Bottom panel: mean ( s.e.m.) total areas under the curves (AUC). Diet effect, P < 0.0001. Group effect, P < 0.01; ***P < 0.0001 for PO vs C. Different letters: diets differ, P < 0.0001. Insulin Fasting insulin concentration were not signi®cantly different between diets or groups, although the group Leptin, predisposition to obesity and diet A Raben and A Astrup 455 effect was near-signi®cant on day 15 (P 0.058; Table 3). There was a signi®cant time diet interaction in postprandial insulin responses with a steeper rise in insulin after SUCROSE, a slower response after STARCH, and a ¯atter response after FAT (Figure 3). A lower AUC was seen in PO compared with C on all three diets (Figure 3). Appetite sensations Comparison of appetite ratings by ANOVA showed a time group interaction for hunger (P < 0.01). Furthermore, a group effect was observed (P < 0.05) with the post-obese subjects feeling more satis®ed (Figure 4) and less hungry, and having less desire to eat (prospective consumption) than the controls on all three diets (P < 0.01). A diet effect was observed for prospective consumption (P < 0.05), with FAT giving higher ratings (more desire to eat) than SUCROSE or STARCH (P < 0.05). The mean ratings (averaged over the day) showed similar signi®cance levels as the ANOVAs concerning group and diet effects. Adjustment of mean ratings for total energy intake during the day did not change these ®ndings. Body weight and composition There were no signi®cant differences between the two groups in 14 day changes of body weight or composition (Table 4). However, a diet effect was observed for changes in body weight and fat-free mass (P < 0.05; Table 4). Thus, body weight was reduced after STARCH compared with SUCROSE, and fatfree mass was increased after SUCROSE compared with FAT or STARCH (Table 4). These ®ndings have been discussed previously.28 Leptin correlations and adjustments Simple correlation analyses within each diet (n 18) showed a positive correlation between change in fasting leptin (day 15ÿday 1) and change in body weight on FAT (r 0.52, P < 0.05) and change in fat mass (kg) on FAT and SUCROSE (r 0.47 ± 0.58, P < 0.05). A positive correlation between fasting leptin and hunger (day 15) was found on STARCH (r 0.53, P < 0.05). Correlation analyses using the six means of diets and groups resulted in no signi®cant correlations between fasting leptin on one hand and glucose, insulin, appetite scores or noradrenaline concentration on the other hand. However, the AUC's for leptin correlated negatively with the AUC's for glucose (r ÿ0.84, P < 0.05), respectively. Furthermore, AUC for leptin correlated positively with AUC for noradrenaline (r 0.87, P < 0.05).28 Adjustment analyses, where all 54 data were included, resulted in the following. The observed diet differences in fasting leptin from day 1 to 15 did not change signi®cantly after adjusting for changes in body weight, fat mass, 14 day energy, Figure 3 Mean plasma insulin concentrations in eight postobese (PO) and 10 controls (C) after 14 days' ad libitum intake of a fat-rich (FAT), a starch-rich (STARCH) and a sucrose-rich (SUCROSE) diet. Brkf. breakfast. Signi®cant by ANOVA, P < 0.05: time diet, group (PO < C, all diets) and time. Bottom panel: mean ( s.e.m.) area under the curves (AUC). Group effect, P < 0.05; *P < 0.05; **P < 0.01 PO vs C. carbohydrate or fat intake. However, after adjusting for 14-day sucrose intake, the diet effect became nonsigni®cant (P 0.36). The diet effect for the areas under the curve (AUC) was still signi®cant after International Journal of Obesity Leptin, predisposition to obesity and diet A Raben and A Astrup 456 d-AUC glucose, d-AUC insulin or d-AUC noradrenaline or the four factors together. Discussion Figure 4 Mean satiety scores for eight post-obese (PO) and 10 controls (C) after 14 days' ad libitum intake of a fat-rich (FAT), a starch-rich (STARCH) and a sucrose-rich (SUCROSE) diet. Brkf. breakfast. Signi®cant by ANOVA, P < 0.05: group (PO < C, all diets) and time. Bottom panels: mean ratings ( s.e.m.) from 9 a.m. to 5 p.m. Group effect, P < 0.05, ***P < 0.001; **P < 0.01 for PO vs C. adjusting for energy intake on day 15, AUC glucose, AUC insulin or AUC noradrenaline or all four factors together. The group effect for d-AUC leptin did not change after adjusting for energy intake on day 15, International Journal of Obesity Two interesting results were obtained from the present study. The ®rst was the postprandial increase in leptin levels observed in post-obese, normal-weight subjects, but not in matched controls. The second was an increased fasting and total area under the curve for leptin after 14 days' ad libitum intake of a sucroserich=low-fat diet compared with a starch-rich=low-fat or a fat-rich diet. The increase in postprandial leptin concentration in our post-obese group was a new ®nding not previously described in the literature. That it was found consistently on all three diets and not abolished by adjusting for differences in energy intake, increases the validity of the observations. It could indicate that these postobese subjects are resistant to leptin and that leptin's normal feedback mechanism does not work optimally in this group. Findings from a recent study on cerebral blood ¯ow in obese support this. Here, a depressed neural activity of hypothalamus in 11 obese women was observed during exposure to food, although leptin concentrations were high at the same time.34 However, it could also be due to an increased insulin sensitivity, especially in the adipose tissue, as has been reported before in similar subjects.35 This is further supported by the previously observed differences in fat metabolism in post-obese subjects. Thus, after a fat-rich meal a smaller increase in plasma triglycerides and a greater suppression of plasma non-esteri®ed fatty acids and fat oxidation was reported, indicating both depressed lipolysis and greater lipid storage after the meal.32 It would also correspond to data from Pima Indians showing that increased insulin sensitivity was a risk factor for weight gain.36 The present ®ndings of lower glucose and insulin concentration in PO than in C on all three diets, despite similar or higher energy intake in PO, also points towards increased insulin sensitivity in PO. Since prolonged hyperinsulinaemia has been found to stimulate leptin release,37 increased insulin sensitivity could have caused the higher postprandial leptin values in the post-obese subjects. The fact that we found no differences between PO and C in fasting leptin concentration could have lead to the assumption that leptin is not a factor predictive of weight gain in these subjects. Two previous studies with post-obese subjects have found similar results.38,39 Thus, no relation between initial fasting leptin and weight change was found in a 4 y prospective study on 14 post-obese, postmenopausal women.39 Similarly, a study in which the post-obese state was obtained by gastroplasty reported no differences between leptin concentration in eight post-obese Leptin, predisposition to obesity and diet A Raben and A Astrup Table 4 Changes % in body weight, fat mass and fat-free mass in post-obese and controls after 14 days' ad libitum intake of a diet rich in fat, starch of sucrose 457 P-value ANOVA Fat Change in body weight (kg) Change in fat mass (kg) Change in fat-free mass PO C PO C PO C ÿ0.6 0.7 ÿ0.2 0.3 ÿ0.2 0.3 ÿ0.2 0.3 ÿ0.3 0.4A ÿ0.1 0.2 Starch Sucrose A ÿ0.7 0.6 ÿ0.7 0.2 ÿ0.5 0.3 ÿ0.3 0.2 ÿ0.3 0.5A ÿ0.4 0.2 B 0.5 0.5 ÿ0.1 0.2 0.1 0.3 ÿ0.3 0.2 0.4 0.3B 0.1 0.2 dg Diet Group 0.31 0.02 0.87 0.54 0.49 0.88 0.50 0.03 0.93 Data are means s.e.m. PO: eight post-obese women. C: 10 matched control women. d g diet group interaction. Results in the same row with different superscripts are different, P < 0.05 (n 18). women (6.4 ng=ml) and eight matched controls (8.9 ng=ml).38 Conversely, eight post-obese subjects (two males, six females) were recently found to have much lower fasting leptin concentration (4.6 ng=ml) than eight controls (11.2 ng=ml, one male, seven females).40 The reason for the discrepancy between these studies is not easy to ®nd, but in general it is dif®cult to rely on a single measurement, especially when the groups are relatively small (n 8 ± 10) and include different sexes. However, with the present results of increased postprandial leptin concentrations in post-obese, considerations on the fasting values may be less relevant. The second interesting ®nding in the present study was the increased leptin concentration after SUCROSE compared with FAT and STARCH. This was found for fasting values and total areas under the curves (AUC's). Since the incremental AUC's (dAUC's) were not different, the larger AUC's were caused by the higher fasting values on SUCROSE. The reason for the increased fasting leptin cannot be found in the fasting glucose and insulin levels, since these were unchanged by the dietary manipulations. Instead it may be that the greater initial postprandial insulin peak on SUCROSE had an overall stimulating effect on leptin secretion on this diet, giving rise to the higher fasting leptin on SUCROSE. Plasma leptin levels do not increase acutely after meal intake, but a slow increase starting 4 h postprandially and reaching a peak 8 h postprandially has been reported.20 Furthermore, the often reported diurnal variation with highest leptin late at night and lowest during the day24 was found to be related not to the day=night factorper se, but to the food intake.20 Thus, meal intake does have an impact on leptin secretion. Several recent studies also support that carbohydrates per se have an effect on leptin secretion. Romon et al reported signi®cant differences in 8 h leptin concentrations between an isoenergetic high-fat and highcarbohydrate meal in 22 non-obese males and females.22 Furthermore, Coppack et al 19 found that in vivo tissue leptin production increased 5 h after a meal with 70 E% carbohydrate, demonstrating both a slow effect of meal intake and a speci®c effect of carbohydrate on leptin secretion. Finally, Astrup et al found that leptin secretion from adipose tissue, mea- sured by the 133Xe-clearance technique and cannulation of the abdominal fat vein, increased steeply 30 min after intake of a mixed meal and returned towards baseline from 60 to 120 min postprandially. There were no observable changes in arterial, venous or arterial-venous leptin concentrations.41 The increase in leptin secretion occurred before the absorption of dietary fat, but simultaneously with the increase in plasma glucose and insulin. This indicates that leptin not only acts as a long-term lipostatic agent, but also as a short-term glucostatic hormone, stimulated acutely by the intake of dietary carbohydrate. Many previous test meal studies using only short measurement periods and plasma samples have therefore not been able to show these changes. In line with the above is also a recent study by Jenkins et al.14 They demonstrated a close correlation between changes in fasting leptin and carbohydrate intake after 4 days of fasting, whereas no such correlations were seen with changes in fat or protein intake. They also concluded that, since these correlations were observed before any marked loss of fat mass, leptin may have a role in defending the body's carbohydrate stores and may therefore be involved in the satiating effects of carbohydrate. In the present study, leptin was measured on ad libitum diets, ie under conditions resembling the subjects' normal life, given these types of diets. Since diets were given ad libitum and subjects lost some body weight on the STARCH diet (0.7 kg), data were adjusted both for differences in energy intake and changes in body weight and fat mass. This did not, however, change the conclusions. Instead the leptin differences could be abolished by adjusting for 14 day average sucrose intake indicating that sucrose per se was the reason for the increased leptin levels. Despite the fact that subjects had a similar ad libitum energy intake on the FAT and SUCROSE diet, leptin did not increase on the FAT diet. The increased leptin on SUCROSE might therefore be one of several mechanisms trying to defend the body against weight gain on such a diet by reducing food intake and increasing energy expenditure. That energy intake was not lower on SUCROSE than on FAT therefore seems puzzling. Maybe the SUCROSE diet was too palatable28 and=or included International Journal of Obesity Leptin, predisposition to obesity and diet A Raben and A Astrup 458 too many ¯uids42 for leptin to have a reducing effect on energy intake compared with the FAT diet. Instead, both in the present and previous studies, sympathetic nervous system activity, measured by energy expenditure and=or catecholamine concentrations, was actually increased on a sucrose-rich diet.28,43,44 The positive relation between AUC leptin and noradrenaline reported here supports this hypothesis. The association of leptin with food intake or appetite sensations in humans has so far only been investigated in a few studies. In a study on 12 normalweight males, insulin and hunger ratings were found to differ signi®cantly, whereas leptin levels were similar 8 h after two different test meals.27 In another study, however, fasting leptin was found to be positively related to fasting satiety in 22 overweight women on caloric restriction.45 In the present study, we saw the opposite on STARCH, ie a positive correlation between fasting hunger and leptin. The latter discrepancies can, however, be related to the use of appetite measurements from single individuals in a correlation analyses. This may give very different and probably often erroneous outcomes due to great interand intraindividual variation.33 In the present study, however, both mean 8 h satiety and leptin concentrations were in general higher in post-obese than in controls. This could support a relation between leptin and satiety in humans, although a bias in the registration of subjective appetite in obese and obesity-susceptible individuals may be suspected. Conclusion Two important ®ndings were obtained from the present study. The ®rst was a larger postprandial increase in leptin levels in post-obese, normal-weight subjects compared with never-obese controls. This might be explained by increased insulin sensitivity in adipose tissue in the post-obese group. The second ®nding was increased leptin levels after 14 days' ad libitum intake of a sucrose-rich=low-fat diet compared with a starchrich=low-fat or a fat-rich diet. This may have been caused by higher postprandial insulin peaks and thereby greater leptin stimulation on the sucrose-rich diet. Further scienti®c support for these contentions is still warranted, however. Acknowledgements The study was supported by the Danish Research and Development Programme for Food Technology 1990 ± 1994, Danisco Sugar, and the Danish Medical Research Council, grant no. 12-9537-3. MasterFoods a.s. and Toms Chokolade A=S generously sponsored foods for the study. The authors gratefully thank International Journal of Obesity Bente Knap, Inge Timmermann, Jannie Mùller Larsen, Charlotte Kostecki, Karina Graff Larsen, Lone Kistrup Larsen, Lis Kristoffersen and Trine Jessen for expert technical assistance. References 1 Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Positional cloning of the mouse obese gene and its human homologue. Nature 1994; 372: 425 ± 432. 2 Pelleymounter MA, Cullen MJ, Baker MB, Hecht R, Winters D, Boone T, Collins F. Effect of the obese gene product on body weight regulation in ob=ob mice. Science 1995; 269: 540 ± 543. 3 Camp®eld LA, Smith FJ, Guisez Y, Devos R, Burn P. Recombinant mouse OB protein: Evidence for a peripheral signal linking adiposity and central neural networks. Science 1995; 269: 546 ± 549. 4 Halaas JL, Gajiwala KS, Maffei M, Cohen SL, Chait BT, Rabinowitz D, Lallone, RL, Burley SK, Friedman JM. Weightreducing effects of the plasma protein encoded by the obese gene. Science 1995; 269: 543 ± 546. 5 Considine RV, Sinha MK, Heiman ML, Kriauciunas A, Stephens TW, Nyce MR, Ohannesian JP, Marco CC, McKee LJ, Bauer TL, Caro JF. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. New Engl J Med 1996; 334(5): 292 ± 295. 6 Ostlund RE, Yang JW, Klein S, Gingerich R. Relation between plasma leptin concentration and body fat, gender, diet, age, and metabolic covariates. J Clin Endocrinol Metab 1996; 81(11): 3909 ± 3913. 7 Ma Z, Gingerich RL, Santiago JV, Klein S, Smith CH, Landt M. Radioimmunoassay of leptin in human plasma. Clin Chem 1996; 42(6): 942 ± 946. 8 Haffner SM, MykkaÈinen LA, Gonzalez CC, Stern MP. Leptin concentrations do not predict weight gain: the Mexico City Diabetes Study. Int J Obes 1998; 22: 695 ± 699. 9 Clement K, Vaisse C, Lahlou N, Cabrol S, Pelloux V, Cassuto D, Gourmelen M, Dina C, Chambaz J, Lacorte JM, Basdevant A, Bougneres P, Lebouc Y, Froguel P, Guy-Grand B. A mutation in the human leptin receptor gene causes obesity and pituitary dysfunction. Nature 1998; 392: 398 ± 401. 10 Arch JRS, Stock MJ, Trayhurn P. Leptin resistance in obese humans: does it exist and what does it mean? (Commentary) Int J Obes 1998; 22: 1159 ± 1163. 11 Ravussin E,PratleyRE,MaffeiM,WangH,FriedmanJM,Bennett PK, Bogardens C. Relatively low plasma leptin levels predict weight gain in Pima Indians. Nature Med 1997; 3: 238 ± 240. 12 Lindroos AK, Lissner L, Carlsson B, Carlsson L, Torgerson J, Karlsson C, StenloÈf K, SjoÈstroÈm L. Familial predisposition for obesity may modify the predictive value of serum leptin concentrations for long-term weight change in obese women Am J Clin Nutr 1998; 67: 1119 ± 1123. 13 Kolaczynski JW, Ohannesian JP, Considine RV, Marco CC, Caro JF. Response of leptin to short-term and prolonged overfeeding in humans. J Clin Endocrinol Metab 1996; 81(11): 4162 ± 4165. 14 Jenkins AB, Markovic TP, Fleury A, Campbell LV. Carbohydrate intake and short-term regulation of leptin in humans. Diabetologia 1997; 40: 348 ± 351. 15 Dubuc GR, Phinney SD, Stern JS, Havel PJ. Changes of serum leptin and endocrine and metabolic parameters after 7 days of energy restriction in men and women. Metabolism 1998; 47(4): 429 ± 434. 16 Schrauwen P, van Marken Lichtenbelt WD, Westerterp KR, Saris WHM. Effect of diet composition on leptin concentration in lean subjects. Metabolism 1997; 46(4): 420 ± 424. 17 Weigle DS, Duell PB, Connor WE, Steiner RA, Soules MR, Kuijper JL. Effect of fasting, refeeding, and dietary fat restriction on plasma leptin levels. J Clin Endocrinol Metab 1997; 82(2): 561 ± 565. Leptin, predisposition to obesity and diet A Raben and A Astrup 18 Havel PJ, Kasim-Karakas S, Mueller W, Johnson PR, Gingerich RL, Stern JS. Relationship of plasma insulin and adiposity in normal weight and over-weight women: effects of dietary fat content and sustained weight loss. J Clin Endocrinol Metab 1996 81(12): 4406 ± 4413. 19 Coppack SW, Stanner S, Rawesh, A, Goodrich S, MohamedAli V. In vivo adipose tissue leptin production before and after a high-carbohydrate meal. Int J Obes 1997; 21(Suppl 2): 77. 20 Dallongeville J, Hecquet B, Lebel P, Edme J-L, le Fur C, Fruchart J-C, Auwerx J, Romon M. Short term response of circulating leptin to feeding and fasting in man: in¯uence of circadian cycle. Int J Obes 1998; 22: 728 ± 733. 21 Pratley RE, Nicolson M, Bogardus C, Ravussin E. Plasma leptin responses to fasting in Pima Indians. Am J Physiol 1997; 273: E644 ± E649. 22 Romon M, Lebel P, Velly C, Fruchart JC, Dallongeville J. In¯uence of a high fat or a high carbohydrate meal on leptin postprandial response in humans. Int J Obes 1998; 22(Suppl 3): S173. 23 Maffei M, Halaas J, Ravussin E, Pratley RE, Lee GH, Zhang Y, Fei H, Kim S, Lallone R, Ranganathan S, Kern PA, Friedman JM. Leptin levels in human and rodent: measurement of plasma leptin and ob RNA in obese and weightreduced subjects. Nature Med 1995; 1: 1155 ± 1161. 24 Sinha MK, Ohannesian JP, Heiman ML, Kriauciunas A, Stephens TW, Magosin S, Marco C, Caro JF. Nocturnal rise of leptin in lean, obese, and non-insulin-dependent diabetes mellitus subjects. J Clin Invest 1996; 97: 1344 ± 1347. 25 Dagogo-Jack S, Fanelli C, Paramore D, Brothers J, Landt M. Plasma leptin and insulin relationships in obese and non-obese humans. Diabetes 1996; 45: 695 ± 698. 26 Clapham JC, Smith SA, Moore GBT, Hughes MG, Azam H, Scott A, Jung RT. Plasma leptin concentrations and OB gene expression in subcutaneous adipose tissue are not regulated acutely by physiological hyperinsulinaemia in lean and obese humans. Int J Obes 1997; 21: 179 ± 183. 27 Joannic J-L, Oppert J-M, Lahlou N, Basdevant A, Auboiron S, Raison J, Bornet F, Guy-Grand B. Plasma leptin and hunger ratings in healthy humans. Appetite 1998; 30: 129 ± 138. 28 Raben A, Macdonald I, Astrup A. Replacement of dietary fat by sucrose or starch: Effects on 14 d ad libitum energy intake, energy expenditure and body weight in formerly obese and never-obese subjects. Int J Obes 1997; 21: 846 ± 859. 29 Marckmann P, Raben A, Astrup A. Ad libitum intake of lowfat diets rich in either starch foods or sucrose: effects on blood lipids, factor VII coagulant activity and ®brinogen. Metabolism 2000 (in press). 30 FAO=WHO=UNU. Energy and protein requirements. Report of a joint FAO= WHO=UNU Expert Consultation, Technical Report Series 1985; No. 74. World Health Organization: Geneva. 31 Mùller A. Danish Food Tables (Levnedsmiddeltabeller). Storkùkkencenteret, levnedsmiddelstyrelsen, 1989. 32 Raben A, Andersen HB, Christensen NJ, Madsen J, Holst JJ, Astrup A. Evidence for an abnormal postprandial response in women predisposed to obesity. Am J Physiol 1994; 267: E549 ± E559. 33 Raben A, Tagliabue A, Astrup A. The reproducibility of subjective appetite scores. Br J Nutr 1995; 73: 517 ± 530. 34 Karhunen LJ, Lappalainen RI, Vanninen EJ, Kuikka JT, Uusitupa MIJ. Serum leptin and regional cerebral blood ¯ow during exposure to food in obese and normal-weight women. Neuroendocrinol 1999; 69: 154 ± 159. 35 Toubro S, Western P, BuÈlow J, MacDonald I, Raben A, Christensen NJ, Madsen J, Astrup A. Insulin sensitivity in postobese women. Clin Sci 1994; 87: 407 ± 413. 36 Swinburn BA, Nyomba BL, Saad MF, Zurlo F, Raz I, Knowler WC, Lillioja S, Bogardus C, Ravussin E. Insulin resistance associated with lower rates of weight gain in Pima Indians. J Clin Invest 1991; 88(1): 168 ± 173. 37 Boden G, Chen X, Kolaczynski JW, Polansky M. Effects of prolonged hyperinsulinemia on serum leptin in normal human subjects. J Clin Invest 1997; 100: 1107 ± 1113. 38 Letiexhe MR, Scheen AJ, Sodoyez-Goffaux F, Caro JF, LefeÁbvre PJ. Relationships between basal plasma leptin, fasting total or true insulin levels and insulin sensitivity index S in lean, obese and post-obese females. Int J Obes 1997; 21(Suppl 2): S101. 39 Nagy TR, Davies SL, Hunter GR, Darnell B, Weinsier L. Serum leptin concentrations and weight gain in post-obese, postmenopausal women. Obes Res 1998; 6(4): 257 ± 261. 40 Filozof C, MuruÂa C, Sanchez P, Brailovsky C, Gonzalez C, Perman M, Ravussin E. Low rates of fat oxidation and low plasma leptin concentration in weight-stable post-obese subjects. Int J Obes 1998; 22(Suppl 3): S165. 41 Astrup A, Ranneries C, Simonsen L, BuÈlow J, Friedman J. Leptin: a short-term acting glucostatic satiety hormone? Int J Obes 1997; 21(Suppl 2):S13. 42 De Castro JM. The effects of the spontaneous ingestion of particular foods or beverages on the meal pattern and overall nutrient intake of humans. Physiol Behav 1993; 53: 1133 ± 1144. 43 Blaak EE, Saris WHM. Postprandial thermogenesis and substrate utilization after ingestion of different dietary carbohydrates. Metabolism 1996; 45: 1235 ± 1242. 44 Raben A, Kiens B, Richter EA. Differences in glycemia, hormonal response and energy expenditure after a meal rich in mono- and disaccharides compared to a meal rich in polysaccharides in physically ®t and sedentary subjects. Clin Phys 1994; 14: 267 ± 280. 45 Heini AF, Lara-Castro C, Kirk KA, Considine RV, Caro JF, Weinsier RL. Association of leptin and hunger-satiety ratings in obese women. Int J Obes 1998; 22: 1084 ± 1087. 46 Metropolitan Life Insurance Company. Metropolitan height and weight tables for men and women, according to frame, ages 25 ± 29. Stat Bull 1983; 64: 2 ± 9. 459 International Journal of Obesity