¹Department of Epidemiology, Harvard School Public Health, Boston, MA, USA

²Departments of Public Health and Community Medicine, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, Republic of China

³Department of Biostatistics, Harvard School Public Health, Boston, MA, USA

⁴Department of Nutrition, Harvard School Public Health, Boston, MA, USA

⁵Department of Laboratory Medicine, Children's Hospital, Boston, MA, USA and Department of Pathology, Harvard Medical School, Boston, MA, USA

⁶Division of Biological Science, Harvard School Public Health, Boston, MA, USA

⁷Channing Laboratory, Department of Medicine, Harvard Medical School and Brigham and Women's Hospital, Boston, MA, USA

Correspondence to: N-F Chu, Department of Community Medicine, Tri-Service General Hospital, National Defense Medical Center, PO Box 90048-509, Nei-Hu, Taipei, Taiwan, Republic of China. E-mail: chuepi@ndmctsgh.edu.tw

OBJECTIVE: Leptin, a primarily adipose tissue-derived protein product of the obesity (ob) gene, is an important regulator of energy metabolism. The strong association between body fat mass and elevated circulating leptin levels in humans suggests that leptin resistance, rather than leptin production, may contribute to the development of obesity and associated disorders. The purpose of this study is to evaluate the relationship between circulating plasma leptin levels and regulation of body weight over time among US men.

DESIGN: Four-year prospective study.

SUBJECTS: A total of 247 men from the Health Professionals Follow-up Study, who at baseline (1994), were 47-64 y of age, were free of cardiovascular diseases, diabetes mellitus and malignant neoplasmas, and completed a detailed lifestyle questionnaire. In addition, all participants completed a follow-up questionnaire in 1998.

MEASUREMENTS: Baseline plasma leptin levels and 4-y weight change.

RESULTS: At the start of follow-up, men in the highest quintile for plasma leptin (mean=12.1 ng/ml) weighed more, were less physically active, and had higher circulating insulin levels than men in the lowest quintile (mean=2.7 ng/ml). After adjustments for baseline age, weight, height, smoking status, alcohol intake, and physical activity, each 10 ng/ml increase in plasma leptin concentration was associated with a 1.68 kg (95% CI 0.14-3.18 kg) weight gain over the 4-y follow-up period. The observed association between leptin level and weight gain was limited to men with a baseline body mass index (BMI) of 25 kg/m², in whom a 10ng/ml higher baseline leptin was associated with a 2.45 kg (95% CI 0.73-4.18-kg) weight gain. Further adjustments for baseline total energy intake, plasma insulin and soluble tumor necrosis factor receptors levels did not appreciably alter these results. Plasma insulin level was not independently associated with subsequent weight gain.

CONCLUSION: These results suggest that elevated plasma leptin concentrations among overweight men may be a marker of leptin resistance and subsequent weight gain.

International Journal of Obesity (2001) 25, 346-353

leptin; weight gain; men

Introduction

Obesity defines a condition of excessive adiposity and predisposes the affected individuals to a cluster of pathological states including cardiovascular disease, diabetes mellitus, stroke and other chronic diseases.^{1,2,3,4,5,6,7} Obesity has become an increasingly important global public health problem, especially among developed countries. In the US, the prevalence of overweight (BMI25 kg/m²) has risen from 47.9 to 57.6% in men and from 37.3 to 45.7% in women between 1960 and 1994.⁸

The etiology of human obesity is complicated and poorly understood, however, positive energy balance over time is believed to be the primary cause.^9,10,11 In recent years, identifications of several new genes in rodent and human obesity have intensified the efforts to understand the mechanisms underlying disturbed energy balance. One such molecule, leptin, is primarily an adipose tissue protein product of the obesity (ob) gene. Leptin is a multifunctional polypeptide with important effects on energy metabolism.^{12,13,14,15,16,17,18,19,20,21} In animal models, leptin deficiency results in a complex syndrome with severe obesity, insulin resistance and reproductive dysfunction that is reversed by exogenous leptin administration.^13,14,15,22 Leptin is also a significant regulator of body weight in humans.^23,24 However, leptin deficiency either at the level of ligand or receptor appears to be an extremely rare cause of obesity in humans.^25,26

Instead, in most human populations, leptin concentrations are highly positively correlated with the body fat amount, mass and percentage and with body mass index.²⁷ This suggests that human obesity is associated with a state of resistance to the effects of leptin,^28,29,30 a model reminiscent of insulin action in type 2 diabetes. Hyperleptinemia, as hyperinsulinemia for insulin resistance, may be a marker of leptin resistance in humans.

Although current evidence suggests that leptin resistance, rather than leptin deficiency, is more common in human obesity.^27,31,32,33 Furthermore, it has been difficult to establish that leptin resistance precedes obesity or is a consequence thereof.³⁴ The limited prospective data on leptin levels and weight change in humans are inconsistent. Differences between studies may be due to variations in study size, follow-up period, and completeness of assessment of energy intake and physical activity.^35,36,37,38

To address these issues, we conducted a 4-y prospective study to evaluate the baseline plasma leptin levels and body weight change among US male health professionals, and tested the hypothesis that the association between baseline plasma leptin and weight gain over time may differ for normal-weight and overweight subjects as a result of relative leptin resistance in the latter subjects.

Materials and methods

The Health Professionals Follow-up Study

The Health Professionals Follow-up Study (HPFS) is a prospective study designed primarily to investigate the association between diet and chronic disease among men. At baseline in 1986, 51 529 male US dentists, pharmacists, veterinarians, podiatrists and osteopathic physicians, who were 40-75 y of age, completed a detailed questionnaire assessing average dietary intake, lifestyle characteristics, and medical history. Every 2 y, participants have been recontacted to ascertain newly diagnosed disease and re-assess risk factors. In 1993-1995, in addition to the biennial questionnaire information, 18 225 participants provided venous blood samples. The distributions of anthropometric, dietary and lifestyle characteristics were similar between the two populations who did and did not return a blood sample.

Research design

From the 18 225 men who returned a blood sample between 1993 and 1995, we excluded 8922 men who had not completed questionnaire information on diet, cigarette smoking, alcohol consumption and physical activity from 1986 to 1994. Since several chronic disease conditions may influence weight change, we also excluded 208 men with cardiovascular disease, diabetes, gastric or duodenal ulcer, liver disease and cancer (except nonmelanoma skin cancer). From the remaining men, we randomly sampled 468 subjects (47-83 y of age) based on seven clusters defined by their self-reported alcohol consumption pattern (eg none, light, moderate, heavy, binge, etc). We also excluded men with missing or invalid weight information for 1998. Finally, due to the complexities of body weight change among older populations (changes are due to bone loss and muscle wasting, as well as to changes in body fat), we included only those that were 47-64 y of age in 1994 (n=247).

Exposure and outcome measurements

Anthropometric measurements: Each participant was asked to report his height to the closest inch at baseline and his current weight in pounds at baseline and on each biennial questionnaire. The validity of the self-reported anthropometric measures in this cohort is described elsewhere.³⁹ We calculated body mass index (BMI) as the ratio of body weight to body height squared (kg/m²) In accordance with the new guidelines,⁴⁰ we defined overweight as a BMI of 25 kg/m².

Lifestyle characteristics: At baseline and on each biennial follow-up questionnaire, participants reported their current smoking status. Current smokers were divided into two subgroups according to the number of cigarettes smoked per day (1-14 or 15). Alcohol intake was calculated by summing the consumption frequency (as reported by participants on the semiquantitative food-frequency questionnaire (SFFQ) and average alcoholic content of beer, red wine, white wine and spirits. The reproducibility and validity of self-reported alcohol consumption have been evaluated in detail within the HPFS. Intakes of alcohol reported over the previous year by SFFQ were highly correlated with intake assessed by two 1 week diet records (r=0.86).⁴¹

The biennial questionnaire, which contains questions about the types and frequency of common leisure time physical activities undertaken, measures the average amount of time spent weekly at four sedentary activities and 10 specified activities during the past year. Metabolic equivalents for the task (METs) are defined for each type of physical activity as a multiple of the metabolic equivalent for sitting quietly for 1 h. Further details on physical activity assessment in this cohort are presented elsewhere.⁴² In a validation study of this questionnaire, vigorous activity assessed by the questionnaire was correlated with resting pulse (r=-0.45) and post-exercise pulse (r=-0.41).⁴²

Dietary information: Data on average nutrient intake were derived from the SFFQ administered in 1994. Consumption of more than 135 nutrients and non-nutrient dietary factors was calculated from the questionnaire, whose reproducibility and validity have been evaluated in detail within the HPFS. The correlation of values for nutrients estimated on the SFFQ and those measured during 2 weeks of diet recording ranged from 0.37 for polyunsaturated fat to 0.92 for vitamin C with supplements (average=0.65) after adjustment for total energy and for within-person variation in intake.⁴³

Measurements of biochemical variables: From April 1993 through August 1995, we received, aliquoted and stored in liquid nitrogen 18 225 blood samples from study participants. To reduce extraneous between-person variation, we requested fasting blood samples. However, if subjects were not fasting, a questionnaire was requesting information on the date and time the sample was drawn and the time elapsed since the preceding meal. Blood samples were collected in three 10 ml liquid EDTA blood tubes, placed on ice packs stored in styrofoam containers, and returned to the laboratory by overnight courier. More than 95% of samples arrived within 24 h after venipuncture. Upon arrival, each sample was centrifuged and aliquoted for storage in liquid nitrogen (-150°C). Fewer than 15% of samples were slightly hemolyzed, and very few were moderately hemolyzed (<3%), lipemic (<1%), or not cooled upon arrival (<0.5%). We found excellent stability of plasma leptin⁴⁴ and soluble tumor necrosis factor receptor (sTNF-R) levels (unpublished observations).

We measured plasma leptin concentrations by radioimmunoassay (RIA), using a commercial kit (Linco Research, St Charles, MO) with antibody raised to highly purified recombinant human leptin.⁴⁵ The inter- and intraassay coefficients of variation (CVs) were 8.3 and 3.4%, respectively. Plasma insulin concentrations were measured by RIA method (Linco Research, St. Charles, MO) which allows accurate assessment with little or no proinsulin cross-reactivity and a CV of <10%. Plasma sTNF-R concentrations were measured by ELISA methods (R&D Systems, Minneapolis, MN for sTNF-R1 and Genzyme Diagnostics, Cambridge, MA for sTNF-R2) that permit accurate assessment with no TNF-, TNF-, or other interleukin cross-reactivity and a CV of <10%.

Statistical analyses

Age-adjusted anthropometric measures, dietary and lifestyle characteristics, and biochemical parameters were presented as mean values according to the quintile distribution of plasma leptin levels. We calculated Spearman correlation coefficients between anthropometric measures and biological parameters to insure validity of inference without assumptions of normality. In multivariate linear regression analyses, we evaluated plasma leptin levels in 1994 as an independent linear predictor of 4-y weight change All three models controlled for age (years), height (inches), and baseline (1994) weight (pounds). Model 2 also controlled for cigarette smoking (never, past, and current smoking of 1-14 or 15 cigarettes/day), alcohol drinking (average amount of alcohol consumption in quintiles), physical activity (average METs/week), and hours since last meal before blood draw. Since insulin and TNF- are important regulators of adipocyte biology and may be associated with leptin expression and production,^46,47,48 for model 3, we further controlled for plasma insulin, sTNF-R1 (pg/ml) and sTNF-R2 (pg/ml) levels, and total energy intake (kcal/day) in 1994. We used regression diagnostics to determine the influence of outliers. Because tests for nonlinearity were not significant, we present only the regression coefficients from linear models. After excluding potential outliers (±3 ´ interquartile range of the leptin distribution) from the primary analyses, the results were not appreciably different from those presented below. In the multivariate linear regression models, we used the robust variance model to insure validity of inference without invoking assumptions of normality by PROC MIXED in SAS (SAS Institute, Cary, NC).

Results

General characteristics of study subjects

The study subjects were divided into five quintiles of plasma leptin level. At baseline, men in the highest leptin quintile (mean=12.1 ng/ml) weighed more, were less physically active, and had higher circulating insulin and sTNF-R levels than men in the lowest quintile (mean=2.7 ng/ml; Table 1).

Association of plasma leptin levels with weight gain

The correlations of body weight with plasma leptin, insulin, sTNF-R1 and sTNF-R2 levels are shown in Table 2. Both plasma leptin and plasma insulin levels were positively correlated with body weight in 1994 and 1998. However, the association was limited to overweight men (BMI25 kg/m²). Plasma leptin level was the only variable that positively correlated with weight change during the 4 y of follow-up (r=0.202) and this correlation was also limited to overweight men. Plasma sTNF-R levels were positively correlated with age and sTNF-R1 levels were positively correlated with body weight in 1994 and 1998 in normal weight men.

To determine whether the association between leptin and subsequent weight gain was independent of other predictors of weight change, we conducted analyses in multivariate linear regression models of 4-y weight change (Table 3). After adjustment for age, baseline weight and height, cigarette smoking, alcohol drinking, and physical activity, increasing leptin level was significantly associated with weight gain ( = 0.17, 95% CI 0.01-0.32). In other words, a 10 ng/ml higher baseline leptin level was associated with a 1.68 kg weight increase. To determine whether this association differed by baseline weight status, we stratified our regression model by baseline BMI (<25 and 25 kg/m²). Plasma leptin level was a strong significant positive predictor of 4 y weight gain, 2.45 (95% CI 0.73-4.18) kg per 10 ng/ml increase in leptin among overweight men, but not among normal-weight men. Adjustment for total energy intake and plasma insulin, sTNF-R1 and sTNF-R2 levels at baseline did not appreciably alter these results (Table 3). Furthermore, plasma insulin level was not independently associated with 4-y weight gain.

Because smoking may affect leptin levels^48,49 and introduce random error, we conducted a subanalysis excluding current smokers. Among overweight men in this analysis, a 10 ng/ml higher baseline leptin level was associated with a 4-y weight gain of 2.72 kg (95% CI 0.86-4.63). Among normal-weight men, the same increase in leptin level was associated with only a marginal weight change (-0.27 kg/ng/ml, 95% CI-3.59-3.04 kg/ng/ml). To ensure that our results were not an artifact of choosing an arbitrary cut-off point for obesity, we conducted further analyses, stratifying subjects into four groups according to baseline BMI (<23, 23-24.9, 25-27.9, and 28 kg/m²) (Table 4). Although the study's power to test for statistical significance within each category of BMI was limited, we did find an increasing trend in regression coefficients from 0.07 for the lowest BMI group to 0.51 for the highest BMI group, which strongly supports the resistance hypothesis to leptin activity in obese subjects (ie hyperleptinemia is a marker of leptin resistance and associated with 4-y weight gain among overweight men; Figure 1).

To confirm our results, we divided the study subjects into two groups based on their weight change status (weight change <0.5 vs 0.5 kg/4-y). Except for baseline plasma leptin levels, there was no significant difference in age, body weight in 1994 and 1998, and BMI in 1994 (Table 5).

Discussion

In this prospective study of 45 to 64-y-old US male health professionals, we found a significant positive correlation between plasma leptin concentrations, weight, and 4-y weight change. The observed association between leptin and weight gain was limited to men with a baseline BMI of 25 kg/m², where a 4-y weight gain of 2.45 kg (95% CI 0.73-4.18 kg) for each 10 ng/ml increase in plasma leptin concentration was observed. Further control for plasma insulin and sTNF-R levels did not appreciably alter these results, and none of the variables including baseline plasma insulin was independently associated with weight gain. These data suggest that plasma leptin level is a significant positive predictor of weight change among overweight men.

Although this study was prospective, it had several limitations. Single measurement of plasma leptin, sTNF-R1, sTNF-R2 and insulin levels may be susceptible to substantial variation over a short interval. An ideal biochemical indicator would reflect long-term or cumulative exposure.⁵⁰ In a subsample of 82 men with plasma leptin measurements taken 4 y apart, we found an intra-class correlation of 0.74 and almost identical mean values (6.48 vs 6.54 ng/ml), the implication is that fluctuation of leptin levels over time is replicable.⁴⁴ Therefore, any error in leptin levels due to intraindividual variation or sample handling is likely to be small. Furthermore, to the extent that the error is not associated with weight change, the results reported here will be biased towards the null. Errors in self-reported anthropometric measures, dietary intake and lifestyle characteristics are also likely to be random and unrelated to biological markers. In a previous validation study of a subpopulation of 123 men from the HPFS, we found a strong correlation (r=0.97) between self-reported and technician-measured weight. Although men tend to underreport their true weight by 2.3 pounds,³⁹ systematic underreporting would not affect our results predicting weight change.

We recognize that the HPFS cohort does not represent a random sample of US men. Therefore, the plasma leptin concentrations, dietary patterns and other lifestyle characteristics of this age group may not reflect those of the general population. However, the homogeneity of the cohort with respect to educational attainment and socioeconomic status reduces the likelihood that other extraneous factors have biased our results. Because the ranges of plasma leptin levels (from 1.5 to 23.4 ng/ml) and weight changes (from -10 to+15 kg/4-y) are quite broad and are similar to what has been reported in other male populations,³⁸ the biological effects of leptin in this population can probably be generalized to all men of this age range.

In most human populations, plasma leptin concentrations are highly correlated with body fat amount, mass, and percentage, and with BMI.^{17,27,28,51,52} Both ob mRNA levels in white adipose tissue and circulating leptin levels are higher in obese than in nonobese subjects; this observation suggests that leptin resistance, rather than leptin deficiency, may be a cause of human obesity.^27,31,32,33 However, it is difficult to prove that leptin resistance precedes rather than follows obesity.³⁴ Epidemiological evidence linking leptin to weight change is inconsistent, our study differs in design and conclusions from several previous reports that have found no positive association between leptin and weight gain.^{35,36,37,38,53}

Lindroos et al³⁷ evaluated 49 control subjects (average BMI 40.7 kg/m²) from an obesity intervention study in Sweden, seeking an association between serum leptin levels and 4-y weight change among women with and without a parental history of obesity. A higher leptin level was associated with weight loss among women with no history of parental obesity. It is difficult to compare these results with those of our study because Lindroos and colleagues measured leptin in subjects before initiating conventional weight-reducing treatment. Substantial changes in diet or physical activity among morbidly obese subjects may increase leptin sensitivity and thereby lead to weight loss.

In a study of 36 Pima Indians, Ravussin et al³⁶ reported an inverse association between leptin and 3-y weight gain. This result is contrary to our findings, but, as in the study from Sweden, the average participant was morbidly obese in this population (mean BMI, 36.5 kg/m²). In fact, these investigators specifically selected non-diabetic, morbidly obese participants. Obesity and weight gain in this highly selected population may be more strongly determined by genetic factors than by the influence of leptin or leptin resistance.^54,55,56

In a prospective study of 180 non-diabetic native Americans, Haffner et al³⁸ examined the association between leptin and weight gain during 3.25 y of follow-up. The average BMI in this population (25.3 kg/m² among men and 27.2 kg/m² among women) was comparable to that in our population. These researchers found no difference among the baseline leptin levels of those whose weight increased, stayed the same, or decreased. Likewise, we found no significant association between leptin and weight gain (Table 2) in the total population. Only when we limited our analysis to overweight men did an association between leptin and weight gain become apparent.

In conclusion, we found that higher baseline leptin levels were associated with weight gain during a 4-y follow-up period among overweight males. These results should be confirmed in women and in other larger populations. If true, our results suggest that obesity-related leptin resistance (as hyperleptinemia at baseline) may be a marker of subsequent weight gain. However, it is simplistic to assume that human obesity is mediated by leptin resistance alone; the mechanisms involved in the development of obesity are complicated and are influenced by energy balance, environmental and social influences, dietary and physical activity patterns, energy regulation, and individual/biological susceptibility. Further studies of leptin resistance and of factors that can modify leptin levels in overweight individuals are necessary to better understand the etiology and management of obesity.

Acknowledgements

This study is supported by research grant HL35464, CA55075 and AA11181. Dr Chu's work is supported by a Research Award from the National Defense Medical Center, Taiwan.

1 Garrow JS. Health implications of obesity. In: Obesity and related diseases. Churchill Livingstone: London, 1988, pp 1-16.

2 Gidding SS. A perspective on obesity. Am J Med Sci 1995; 310: (Suppl 1) S68-71, MEDLINE

3 Bray GA. Health hazards of obesity. Endocrinol Metab Clin Nutr Am 1996; 25: 907-919,

4 Kannel WB, Wilson PW. An update on coronary risk factors. Med Clin Nutr Am 1995; 79: 951-971,

5 Rimm EB, Stampfer MJ, Giovannucci E et al. Body size and fat distribution as predictors of coronary heart disease among middle-aged and older US men. Am J Epidemiol 1995; 141: 1117-1127, MEDLINE

6 Huang Z, Willett WC, Manson JE et al. Body weight, weight change, and risk for hypertension in women. Ann Intern Med 1998; 128: 81-88, MEDLINE

7 Golay A, Felber JP. Evolution from obesity to diabetes. Diabetes Metab 1994; 20: 3-14,

8 Flegal KM, Carroll MD, Kuczmarski RJ, Johnson CL. Overweight and obesity in the United States: prevalence and trends, 1960-1994. Int J Obes Relat Metab Disord 1998; 22: 39-47, MEDLINE

9 Ravussin E, Lillioja S, Knowler WC et al. Reduced rate of energy expenditure as a risk factor for body-weight gain. New Engl J Med 1988; 318: 467-472, MEDLINE

10 Bouchard C, Tremblay A, Despres JP et al. The response to long-term overfeeding in identical twins. New Engl J Med 1990; 322: 1477-1482, MEDLINE

11 Grundy SM. Multifactorial causation of obesity: implications for prevention. Am J Clin Nutr 1998; 67: (Suppl) 563S-572S, MEDLINE

12 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, MEDLINE

13 Pelleymounter MA, Cullen MJ, Baker MB et al. Effects of the obese gene product on body weight regulation in ob/ob mice. Science 1995; 269: 540-543, MEDLINE

14 Halaas JL, Gajiwala KS, Maffer M et al. Weight-reducing effects of the plasma protein encoded by the obese gene. Science 1995; 269: 543-546, MEDLINE

15 Campfield 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, MEDLINE

16 Spiegelman BM, Flier JS. Adipogenesis and obesity rounding out the big picture. Cell 1996; 87: 377-389, MEDLINE

17 Zimmet P, Hodge A, Nicolson M et al. Serum leptin concentration, obesity, and insulin resistance in Western Samoans cross sectional study. Br Med J 1996; 313: 965-969,

18 Weigle DS. Leptin and other secretory products of adipocytes modulate multiple physiological functions. Ann Endocrinol 1997; 58: 132-136,

19 Flier JS. Clinical review 94: What's in a name? In search of leptin's physiologic role. J Clin Endocrinol Metab 1998; 83: 1407-1413, MEDLINE

20 Friedman JM. Leptin, leptin receptors, and the control of body weight. Nutr Rev 1998; 56: s38-46, MEDLINE

21 Auwerx J, Staels B. Leptin. Lancet 1998; 351: 737-742, MEDLINE

22 Halaas JL, Boozer C, Blair-West J, Fidahusein N, Denton DA, Friedman JM. Physiological response to long-term peripheral and central leptin infusion in lean and obese mice. Proc Natl Acad Sci USA 1997; 94: 8878-8883, Article MEDLINE

23 Caro JF, Kolaczynski JW, Nyce MR et al. Decreased cerebrospinal-fluid/serum leptin ratio in obesity. a possible mechanism for leptin resistance. Lancet 1996; 348: 159-161, Article MEDLINE

24 Caro JF, Sinha MK, Kolaczynski JW, Zhang PL, Considine RV. Leptin: the tale of an obesity gene. Diabetes 1996; 45: 1455-1462, MEDLINE

25 Maffei M, Stoffel M, Barone M et al. Absence of mutations in the human OB gene in obese/diabetic subjects. Diabetes 1996; 45: 679-682, MEDLINE

26 Carlsson B, Lindell K, Gabrielsson B et al. Obese (ob) gene defects are rare in human obesity. Obes Res 1997; 5: 30-35, MEDLINE

27 Considine RV, Sinha MK, Heiman ML et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. New Engl J Med 1996; 334: 292-295, MEDLINE

28 Hassink SG, Sheslow DV, de Lancey E, Opentanova I, Considine RV, Caro JF. Serum leptin in children with obesity, relationship to gender and development. Pediatrics 1996; 98: 201-203, MEDLINE

29 Haffner SM, Stern MP, Miettinen H, Wei M, Gingerich RL. Leptin concentrations in diabetic and nondiabetic Mexican-Americans. Diabetes 1996; 45: 822-824, MEDLINE

30 Haffner SM, Gingerich RL, Miettinen H, Stem MP. Leptin concentrations in relation to overall adiposity and regional body fat distribution in Mexican Americans. Int J Obes Relat Metab Disord 1996; 20: 904-908, MEDLINE

31 Hamilton BS, Opaglia D, Kwan AY, Deitel M. Increased obese mRNA expression in omental fat cells from massively obese humans. Nature Med 1995; 1: 953-956, MEDLINE

32 Lonnqvist F, Arner P, Nordfors L, Schalling M. Overexpression of the obese (ob) gene in adipose tissue of human obese subjects. Nature Med 1995; 1: 950-953, MEDLINE

33 Maffei M, Halaas J, Ravussin E et al. Leptin levels in human and rodent measurement of plasma leptin and ob RNA in obese and weight-reduced subjects. Nature Med 1995; 1: 1155-1161, MEDLINE

34 Arch JR, Stock MJ, Trayhurn P. Leptin resistance in obese humans: does it exist and what does it mean? Int J Obes Relat Metab Disord 1998; 22: 1159-1163, MEDLINE

35 Niskanen LK, Haffner S, Karhunen LJ, Turpeinen AK, Miettinen H, Uusitupa MI. Serum leptin in obesity is related to gender and body fat topography but does not predict successful weight loss. Eur J Endocrinol 1997; 137: 61-67, MEDLINE

36 Ravussin E, Pratley RE, Maffei M et al. Relatively low plasma leptin concentrations precede weight gain in Pima Indians. Nature Med 1997; 3: 238-240, MEDLINE

37 Lindroos AK, Lissner L, Carlsson B et al. 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, MEDLINE

38 Haffner SM, Mykkanen LA, Gonzalez CC, Stern MP. Leptin concentrations do not predict weight gain the Mexico City Diabetes Study. Int J Obes Relat Metab Disord 1998; 22: 695-699, MEDLINE

39 Rimm EB, Stampfer MJ, Colditz GA, Chute CG, Litin LB, Willett WC. Validity of self-reported waist and hip circumferences in men and women. Epidemiology 1990; 1: 466-473, MEDLINE

40 National Heart Lung, and Blood Institute. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. The evidence report. NHLBI: Bethesda, MD, 1998,

41 Giovannucci E, Colditz G, Stampfer MJ et al. The assessment of alcohol consumption by a simple self-administered questionnaire. Am J Epidemiol 1991; 133: 810-817, MEDLINE

42 Chasan-Taber S, Rimm EB, Stampfer MJ et al. Reproducibility and validity of a self-administered physical activity questionnaire for male health professionals. Epidemiology 1996; 7: 81-86, MEDLINE

43 Rimm EB, Giovannucci EL, Stampfer MJ, Colditz GA, Litin LB, Willett WC. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. Am J Epidemiol 1992; 135: 1114-1126. [Discussion 1127-1136.], MEDLINE

44 Chu NF, Makowski L, Hotamisligil GS, Rimm EB. Stability of human plasma leptin concentrations with 36 hours following specimen collection. Clin Biochem 1999; 32: 87-89, MEDLINE

45 Ma Z, Gingerich RL, Santiago JV, Klein S, Smith CH, Landt M. Radioimmunoassay of leptin in human plasma. Clin Chem 1996; 42: 942-946, MEDLINE

46 Cusin I, Sainsbury A, Doyle P, Rohner-Jeanrenaud F, Jeanrenaud B. The ob gene and insulin. A relationship leading to clues to the understanding of obesity. Diabetes 1995; 44: 1467-1470, MEDLINE

47 Mantzoros CS, Moschos S, Avramopoulos I et al. Leptin concentrations in relation to body mass index and the tumor necrosis factor-alpha system in humans. J Clin Endocrinol Metab 1997; 82: 3408-3413, MEDLINE

48 Mantzoros CS, Liolios AD, Tritos NA et al. Circulating insulin concentrations, smoking, and alcohol intake are important independent predictors of leptin in young healthy men. Obes Res 1998; 6: 179-186, MEDLINE

49 Wei M, Stern MP, Haffner SM. Serum leptin levels in Mexican Americans and non-Hispanic whites association with body mass index and cigarette smoking. Ann Epidemiol 1997; 7: 81-86, MEDLINE

50 Hunter D. Biochemical indicators of dietary intake. In: Willett W (ed). Nutritional epidemiology 2nd edn. New Oxford University, press, New York, 1998, 174-243.

51 Rosenbaum M, Nicolson M, Hirsch J et al. Effects of gender, body composition, and menopause on plasma concentrations of leptin. J Clin Endocrinol Metab 1996; 81: 3424-3427, MEDLINE

52 Muller N, O'Brien P, Nair KS. Disruption of the relationship between fat content and leptin levels with aging in humans. J Clin Endocrinol Metab 1998; 83: 931-934, MEDLINE

53 Wing RR, Sinha MK, Considine RV, Lang W, Caro JF. Relationship between weight loss maintenance and changes in serum leptin levels. Horm Metab Res 1996; 28: 698-703, MEDLINE

54 Howard BV, Bogardus C, Ravussin E et al. Studies of the etiology of obesity in Pima Indians. Am J Clin Nutr 1991; 53: 1577S-1585S, MEDLINE

55 Reed DR, Ding Y, Xu W, Cather C, Green ED, Price RA. Extreme obesity may be linked to markers flanking the human OB gene. Diabetes 1996; 45: 691-694, MEDLINE

56 Bouchard C. Genetics of human obesity: recent results from linkage studies. J Nutr 1997; 127: 1887S-1890S, MEDLINE

Figure 1 Four-year (1998-1994) body weight change in relation to baseline plasma leptin concentrations (1994) among subjects with different body mass indexes (kg/m²).

Table 1 Baseline characteristics of 247 study subjects by quintile of plasma leptin concentrations in 1994

Table 2 Spearman correlation coefficients for 1994 plasma leptin, insulin, and sTNF-R levels vs age, body weight, and weight change among normal weight and overweight subjects

Table 3 Four-year (1998-1994) body weight change in relation to baseline plasma leptin concentrations (1994) among normal-weight and overweight study subjects

Table 4 Four-year (1998-1994) body weight change in relation to baseline plasma leptin concentrations (1994) among subjects with different body mass indexes (n=230)

Table 5 Characteristics of study subjects based on 4- y body weight change status