| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
University of Pennsylvania School of Medicine, Department of Psychiatry (T.A.W., G.D.F., D.A.A., D.B.S.); Thomas Jefferson University, Department of Medicine (R.V.C., J.S.C.), Philadelphia, Pennsylvania 19104
Address all correspondence and requests for reprints to: Thomas A. Wadden, Ph.D., University of Pennsylvania, 3600 Market Street, Suite 738, Philadelphia, Pennsylvania 19104. E-mail: Wadden{at}mail.med.upenn.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Leptin falls not only with long-term reductions in body weight (and fat) but also in response to short-term decreases in energy intake. Two human studies of fasting, one of 36 hours and the other of 52 hours, found that leptin declined by 35% and 72%, respectively (4, 5). Such reductions are likely to be associated with a series of neurohumoral events that ultimately increase appetitive behavior and decrease energy expenditure in an effort to restore energy balance (6). Leptin levels increase rapidly when caloric restriction is terminated (4).
In previous studies of obese dieters, the effects on leptin of weight reduction have been confounded with those of the caloric restriction required to induce weight loss (1, 2). Differences in caloric restriction may well explain why subjects treated by Considine et al. (1) displayed substantially greater reductions in leptin than those treated by Wing et al. (2), despite the two groups having lost the same amount of weight (i.e. about 10% of initial weight). Thus, the present study used stepwise multiple regression analysis to examine the separate (i.e. independent) effects on leptin of degree of caloric restriction and weight loss. Leptin was measured 5 times during a 40-week period in women randomly assigned to initially consume: 1) a 1000 kcal/day diet comprised of portion-controlled foods; or 2) a 1200 kcal/day balanced deficit diet composed of self-selected foods. We predicted that short-term reductions in leptin (during the first 10 weeks) would be significantly greater in subjects who were prescribed the 1000 kcal/day diet because of their greater caloric restriction. By contrast, we anticipated that long-term reductions in leptin, following the period of marked caloric restriction, would be related primarily to changes in weight and fat.
We selected these two diets based on earlier findings that persons who consumed a portion-controlled diet of 1200 kcal/day lost significantly more weight than individuals asked to consume the same number of calories but from a self-selected diet of conventional foods (7). Several studies have shown that obese individuals typically underestimate their food intake by 20–50% when eating a diet of conventional foods (8, 9). Thus, we anticipated that the actual difference in caloric intake between subjects in our two dietary conditions would be at least 400 kcal/day, which would result in significantly larger weight losses in women who consumed the portion-controlled diet.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Subjects were 49 women with a mean age of 45.0 ± 9.6 yr,
weight of 97.1 ± 13.8 kg, and body mass index (BMI) of 36.4
± 4.5 kg/m2 (see Table 1
) who
were free of contraindications to treatment, described previously (10).
Subjects gave their written informed consent to particiapte in the
study, which was approved by the University of Pennsylvania’s
Committee on Studies Involving Human Beings.
|
All subjects attended 90-min group treatment sessions (of 7–10 participants each) that met weekly for the first 20 weeks and every other week from weeks 21–40. They were instructed in behavioral methods of weight control (11) and prescribed the diets described below.
Dietary conditions
Low-calorie diet (LCD) (N = 25). These subjects were asked to maintain their usual food intake during the first week. During weeks 2–13, they were prescribed a diet of approximately 1000 kcal/day that consisted of four servings daily of a liquid-meal replacement, which was combined with a frozen-food dinner entree and one fruit and two vegetable servings. Each serving of the liquid diet provided 160 kcal, 14 g of protein, 20 g of carbohydrate, and 3 g of fat (OPTIFAST 800; Novartis Nutrition Co., Minneapolis, MN). For dinner, subjects chose from among 10 frozen entrees, each of which provided approximately 250 kcal, 18 g protein, 30 g carbohydrate, and 6 g fat. Beginning at week 14, subjects decreased their consumption of the liquid diet while increasing their consumption of conventional foods so that by week 17 they were instructed to consume approximately 1200 kcal/day, with 12–15% of calories from protein, 55–60% from carbohydrate, and 25–30% from fat. From weeks 21–40, subjects were instructed to consume between 1200–1800 kcal/day, depending on their desire to lose more weight.
Balanced-deficit diet (BDD) (N = 24).These subjects were instructed to maintain their usual food intake the first week. From weeks 2–20 they were instructed to consume a self-selected diet of conventional foods of approximately 1200 kcal/day, with 12–15% of calories from protein, 55–60% from carbohydrate, and 25–30% from fat. During weeks 21–40, subjects were instructed to consume a diet of 1200–1800 kcal/day, depending on their desired weight change.
Dependent measures
Body composition and fat distribution.Body composition was measured by densitometry at baseline and weeks 20 and 40 using methods described previously (10, 12). Body fat distribution was assessed at baseline by waist and hip circumferences, using World Health Organization (13) criteria.
Serum leptin.Blood samples were obtained, following a 12-h overnight fast, at baseline and weeks 6, 10, 20, and 40. Serum leptin was analyzed by radioimmunoassay in the laboratory of Drs. Caro and Considine, using a commercially available kit (Linco Research, St. Charles MO), as described previously (1). This method has a within-assay variation of 3.4% and a between-assay variation of 3.6%. Serum insulin was measured by radioimmunoassay (Linco Research, St. Charles MO) and glucose by the glucose oxidase method with a glucose analyzer 2 (Beckman, Brea, CA). The latter measures were obtained at baseline and weeks 6, 10, and 20.
Resting energy expenditure (REE).REE was measured at baseline and weeks 20 and 40 by indirect calorimetry (DeltaTrac, SensorMedics, Yorba Linda, CA), following methods described previously (10, 12). All assessments were conducted between 0700 and 1000 hr.
Statistical Analyses
The relationships between baseline leptin levels and subjects’ baseline characteristics including weight, BMI, and body fat were examined using Pearson product moment correlations. Differences among conditions in changes in weight, fat, leptin, and the other measures were examined using analysis of variance. The separate effects of dietary condition and weight loss (entered in that order) on changes in leptin were examined by stepwise multiple regression (14). Twenty-one of 25 subjects in the LCD condition completed the 40-week study, as compared with 17 of 24 in the BDD condition, a difference that was not statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Table 1
presents subjects’
pretreatment characteristics and their correlation with baseline leptin
levels. Body fat, hip circumference, and BMI all correlated positively
with leptin (all rs > 0.61). Stepwise multiple
regression revealed that, of the three variables, body fat was the
strongest predictor, accounting for 43% of the variance in baseline
leptin; the two other variables did not contribute significantly. Thus,
the greater the subjects’ body fat, the higher their serum leptin.
Short-term changes in weight and leptin
Women prescribed the portion-controlled LCD lost significantly
more weight at weeks 6 and 10 than those who consumed the self-selected
1200 kcal/day BDD (both P values < 0.05; see Table 2). At week 10, for example, weight had
fallen by 7.4 ± 2.2% in the LCD subjects, as compared with only
4.4 ± 4.0% in the BDD subjects (see Fig. 1A). Differences in weight loss were
associated with significant differences between conditions in
self-reported caloric intake at week 10 (P < 0.004),
as shown in Table 3.
|
|
|
). Stepwise multiple regression showed that dietary condition (i.e. degree of caloric restriction) accounted for 14% of the variance in the change in leptin at week 6 and that weight loss did not add significantly to the variance. (The same findings were obtained when weight loss was entered first in the regression.) At week 10, the two variables together accounted for 36% of the variance in the change in leptin, with dietary condition accounting for 21%. This finding indicates that severity of caloric restriction and weight loss contributed independently to the reduction in leptin.
Long-term changes in weight, body fat, and leptin
At week 20, the LCD subjects had lost 13.2 ± 5.7% of
initial weight as compared with 9.2 ± 7.0% for the BDD subjects.
There were no significant differences between conditions in weight loss
or self-reported caloric intake at this time. Weight losses of the two
conditions also did not differ significantly at week 40. As shown in
Table 2, subjects in both conditions lost
large amounts of fat at weeks 20 and 40, but there were no significant
differences between groups.
Leptin. There were no significant differences between conditions in reductions in leptin (as measured from baseline) at weeks 20 or 40. A paired t test showed that the 37.4 ± 25.9% decline in leptin in the LCD subjects at week 40 was significantly smaller than the 55.8 ± 13.6% reduction observed at week 6 (P < 0.001), despite these subjects having lost three times as much weight at week 40 as at week 6. The results of the same comparison in the BDD subjects approached but did not reach statistical significance (P < 0.10).
Weight loss at weeks 20 and 40 accounted for 28% and 47% of the variance in the change in leptin at these two times, respectively. (Reductions in body fat accounted for 25% and 44% of the variance in changes in leptin at weeks 20 and 40, respectively.) At neither time did dietary condition account for a significant amount of the variance in the change in leptin, presumably because subjects in the two groups consumed similar diets at these times.
Predictors of weight loss. Baseline leptin levels did not predict weight loss at any time during treatment (rs ranged from -0.03 to 0.10), and the fall in leptin at week 20 did not correlate with subsequent weight loss at week 40. By contrast, initial body weight correlated with weight loss at all periods (rs ranged from 0.36 to 0.41, all P values < 0.04).
Changes in REE, insulin, and glucose
An ANOVA with repeated measures showed that REE fell significantly
at weeks 20 (P < 0.001) and 40 (P <
0.001), but there were no significant differences between conditions in
reductions on this measure (see Table 2). Analysis of covariance showed
that the decline in REE was no greater than that anticipated with
reductions in weight and/or fat-free mass. Reductions in REE at week 40
correlated significantly with those in leptin at this time
(r = 0.41, P < 0.05). The correlation
(r = 0.25) at week 20 was not statistically
significant.
Insulin and glucose. Fasting insulin levels fell significantly from baseline at week 6 (P < 0.01) and week 10 (P < 0.01) but not at week 20. At week 6, the decline in the LCD subjects (-4.7 ± 5.0 mg/dL) was significantly greater than that in the BDD subjects (-1.7 ± 3.7 mg/dL) (P < 0.05). Reductions in insulin and leptin were positively correlated at week 6 (r = 0.29, P < 0.07), week 10 (r = 0.38, P < 0.01), and week 20 (r = 0.57, P < 0.01).
Glucose fell significantly from baseline at week 6 (mean = -9.7 ± 9.1) (P < 0.01), but not at weeks 10 or 20. There were no significant differences between conditions in changes in glucose, and reductions in this variable did not correlate significantly with reductions in leptin.
End-of-treatment correlates of serum leptin
Body fat, BMI, and body weight remained highly correlated with
serum leptin at the end of treatment, after a mean loss of 12.5 ±
9.7 kg (see Table 4). Correlations tended to be higher
after weight loss than before. Despite the consistency over time in the
relationship between body fat and serum leptin, a repeated measures
ANOVA showed that leptin concentrations per kilogram of fat (nanogram
per deciliter/fat in kilogram) declined significantly from baseline to
week 40 (i.e. values of 1.07 ± 0.36 and 0.85 ±
0.34, respectively; P < 0.001). The same finding was
obtained when leptin was compared at the two periods, covarying body
fat. To ensure that this apparent reduction in leptin concentration was
not attributable to caloric restriction, we analyzed only the data of
23 subjects who, from week 36 to week 40, either lost no weight or
gained a small amount (mean weight change of 23 subjects was +0.99 ±
1.01 kg). Baseline and week 40 values of leptin per kilogram body fat
remained significantly different (values of 1.12 ± 0.41 and
0.92 ± 0.34, respectively; P < 0.002).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Leptin fell precipitously during the first 6 weeks in response to caloric restriction, with significantly greater reductions in women who consumed the 1000 kcal/day portion-controlled diet than the 1200 kcal/day self-selected diet. In the LCD subjects, the 55% reduction in leptin at week 6 was 10 times greater than the relative reduction in weight. Results of the regression analysis showed that dietary condition (i.e. caloric restriction), not weight loss, contributed significantly to the variance in the reduction in leptin at week 6. At week 10, when subjects had lost more weight, both the degree of caloric restriction and weight loss contributed significantly (and independently) to changes in leptin. Further studies are needed to determine if there is a monotonic relationship between the degree of caloric restriction and the reduction in leptin, as we suspect there is.
Leptin levels increased in the LCD subjects when they discontinued the 1000 kcal/day portion-controlled diet (at week 14) and began consuming a higher calorie diet of conventional foods. By week 40, leptin levels had increased significantly above their nadir at week 6, despite the LCD subjects’ continued weight loss during this time.
While short-term reductions in leptin were affected primarily by caloric restriction, long-term reductions were related most strongly to decreases in body weight and fat. At weeks 20 and 40, 25% and 44% of the variance in the reduction in leptin, respectively, was accounted for by decreases in body fat. This finding extends that reported by Wing et al. (2). As anticipated, dietary condition did not account for any of the variance in the change in leptin at weeks 20 or 40, because subjects in the two groups were prescribed the same diet at these times.
Changes in insulin in the LCD subjects paralleled those in leptin; the largest reductions were observed during the first 10 weeks, when subjects consumed the 1000 kcal/day portion-controlled diet. Thereafter, insulin levels increased as subjects increased their caloric intake. We have previously reported a similar pattern of change in REE, which fell sharply in the first few weeks that subjects consumed a very-low-calorie diet but then rose when calorie intake was increased (10, 12). Further studies are needed to illuminate the nature of the relationships between leptin, REE, and insulin, all of which are highly responsive to short-term energy restriction.
We anticipated that long-term reductions in body fat would be associated with reduced serum leptin levels, given the strong positive relationship observed in this study and others between these two variables (6, 7, 8, 9). This expectation was confirmed. What was not fully anticipated was that leptin concentrations per kilogram of body fat would be significantly lower after weight loss than before. Analysis of covariance, controlling for body fat, showed that the reduced leptin concentrations following treatment exceeded those anticipated with the reduction in body fat. This greater-than-anticipated reduction did not appear to be a consequence of subjects’ being in negative energy balance, as the analysis was limited to subjects who did not lose weight between weeks 36 and 40. The significance of this finding is not clear. It is possible that the reduced leptin concentrations represent a normalization of leptin function, given Klein et al.’s (18) finding that leptin concentrations increase disproportionately with increasing body fat. This abnormality may decrease as subjects achieve a more normal body composition. The greater-than-expected reduction in leptin concentrations also could have occurred if subjects lost proportionally more subcutaneous than omental fat. Leptin messenger RNA (mRNA) levels are substantially greater in subcutaneous than in omental adipocyctes, particularly in women (19). Thus, a greater reduction in subcutaneous fat could result in greater-than-expected reductions in leptin concentrations.
This study would have been improved by our having had greater control over subjects’ food intake, particularly of persons in the self-selected diet (BDD) condition. These women were asked initially to consume a 1200 kcal/day, but we believe that their calorie intake was substantially higher, given their significantly smaller weight loss during the first 10 weeks than women in the LCD condition, in addition to the well-documented tendency for obese individuals to under-report their food intake (8, 9).
Neither baseline leptin values nor reductions in leptin at week 20 predicted reductions in weight or fat at the end of treatment (i.e. week 40). Wing et al. (2) similarly found that end-of-treatment reductions in leptin did not predict weight change during follow-up. Further studies, however, are needed in view of Ravussin et al.’s (20) finding that low baseline leptin levels in a group of Pima Indians were associated with significant weight gain over an approximately 3-yr period. These and related findings (21) suggest that obesity, in some persons, may be attributable to hypoleptinemia, rather than to the more commonly suspected central and/or peripheral resistance to leptin (6). Until these potential abnormalities are better understood, it will be difficult to make predictions concerning the effects of dieting-induced reductions in leptin on long-term changes in body weight.
|
Footnotes |
|---|
Received May 27, 1997.
Revised September 12, 1997.
Accepted October 1, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B. Mokhlesi and A. Tulaimat Recent Advances in Obesity Hypoventilation Syndrome Chest, October 1, 2007; 132(4): 1322 - 1336. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. T. Haltia, A. Viljanen, R. Parkkola, N. Kemppainen, J. O. Rinne, P. Nuutila, and V. Kaasinen Brain White Matter Expansion in Human Obesity and the Recovering Effect of Dieting J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 3278 - 3284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W Krieger, H. S Sitren, M. J Daniels, and B. Langkamp-Henken Effects of variation in protein and carbohydrate intake on body mass and composition during energy restriction: a meta-regression 1 Am. J. Clinical Nutrition, February 1, 2006; 83(2): 260 - 274. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Mitchell, D T Armstrong, R L Robker, and R J Norman Adipokines: implications for female fertility and obesity Reproduction, November 1, 2005; 130(5): 583 - 597. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Doucet, M. Pomerleau, and M.-E. Harper Fasting and Postprandial Total Ghrelin Remain Unchanged after Short-Term Energy Restriction J. Clin. Endocrinol. Metab., April 1, 2004; 89(4): 1727 - 1732. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. A. Wittert, H. Turnbull, P. Hope, J. E. Morley, and M. Horowitz Leptin prevents obesity induced by a high-fat diet after diet-induced weight loss in the marsupial S. crassicaudata Am J Physiol Regulatory Integrative Comp Physiol, April 1, 2004; 286(4): R734 - R739. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Friedman A War on Obesity, Not the Obese Science, February 7, 2003; 299(5608): 856 - 858. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Hamerman Molecular-Based Therapeutic Approaches in Treatment of Anorexia of Aging and Cancer Cachexia J. Gerontol. A Biol. Sci. Med. Sci., August 1, 2002; 57(8): M511 - 518. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Rodrigues, R. B. Radominski, H. d. L. Suplicy, S. M. De Almeida, P. A. Niclewicz, and C. L. Boguszewski The Cerebrospinal Fluid/Serum Leptin Ratio during Pharmacological Therapy for Obesity J. Clin. Endocrinol. Metab., April 1, 2002; 87(4): 1621 - 1626. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. A Ricci, S. B Heymsfield, R. N Pierson Jr, T. Stahl, H. A Chowdhury, and S. A Shapses Moderate energy restriction increases bone resorption in obese postmenopausal women Am. J. Clinical Nutrition, February 1, 2001; 73(2): 347 - 352. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Doucet, S. St-Pierre, N. Alméras, P. Mauriège, J.-P. Després, D. Richard, C. Bouchard, and A. Tremblay Fasting Insulin Levels Influence Plasma Leptin Levels Independently from the Contribution of Adiposity: Evidence from Both a Cross-Sectional and an Intervention Study J. Clin. Endocrinol. Metab., November 1, 2000; 85(11): 4231 - 4237. [Abstract] [Full Text]
|
|
|
|
|
|
K. Hakala, B. Stenius-Aarniala, and A. Sovijarvi Effects of Weight Loss on Peak Flow Variability, Airways Obstruction, and Lung Volumes in Obese Patients With Asthma Chest, November 1, 2000; 118(5): 1315 - 1321. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. S. L. Thong, R. Hudson, R. Ross, I. Janssen, and T. E. Graham Plasma leptin in moderately obese men: independent effects of weight loss and aerobic exercise Am J Physiol Endocrinol Metab, August 1, 2000; 279(2): E307 - E313. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Devlin, S. Z. Yanovski, and G. T. Wilson Obesity: What Mental Health Professionals Need to Know Am J Psychiatry, June 1, 2000; 157(6): 854 - 866. [Abstract] [Full Text]
|
|
|
|
|
|
E. Doucet, S. St. Pierre, N. Alméras, P. Mauriège, D. Richard, and A. Tremblay Changes in Energy Expenditure and Substrate Oxidation Resulting from Weight Loss in Obese Men and Women: Is There an Important Contribution of Leptin? J. Clin. Endocrinol. Metab., April 1, 2000; 85(4): 1550 - 1556. [Abstract] [Full Text]
|
|
|
|
 |
|
 N. Hongu and D. S. Sachan Caffeine, Carnitine and Choline Supplementation of Rats Decreases Body Fat and Serum Leptin Concentration as Does Exercise J. Nutr., January 1, 2000; 130(2): 152 - 157. [Abstract] [Full Text]
|
|
|
|
 |
|
 L. Poretsky, N. A. Cataldo, Z. Rosenwaks, and L. C. Giudice The Insulin-Related Ovarian Regulatory System in Health and Disease Endocr. Rev., August 1, 1999; 20(4): 535 - 582. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. P. O'DONNELL, C. D. SCHAUB, A. S. HAINES, D. E. BERKOWITZ, C. G. TANKERSLEY, A. R. SCHWARTZ, and P. L. SMITH Leptin Prevents Respiratory Depression in Obesity Am. J. Respir. Crit. Care Med., May 1, 1999; 159(5): 1477 - 1484. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Gutin, L. Ramsey, P. Barbeau, W. Cannady, M. Ferguson, M. Litaker, and S. Owens Plasma leptin concentrations in obese children: changes during 4-mo periods with and without physical training Am. J. Clinical Nutrition, March 1, 1999; 69(3): 388 - 394. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Anderson and T. A. Wadden Treating the Obese Patient: Suggestions for Primary Care Practice Arch Fam Med, March 1, 1999; 8(2): 156 - 167. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
