Recent studies demonstrated significantly higher serum leptin concentrations in females as compared with males, even after correction for differences in body fat mass. The aim of our study was to measure serum leptin concentrations in a large group of obese children and adolescents to determine the possible role of sex steroid hormones on both leptin serum concentrations and production in human adipocytes. Obese girls were found to have significantly higher leptin concentrations than boys at the same degree of adiposity (25.2 Ϯ 14.1 vs. 17.2 Ϯ 12.6 ng/ml, P Ͻ 0.001). In a multiple regression analysis with age and body mass index (percent body fat) as fixed variables, it turned out that testosterone had a potent negative effect on serum leptin in boys, but not in girls. In vitro experiments using newly developed human adipocytes in primary culture showed that both testosterone and its biologically active metabolite dihydrotestosterone are able to reduce leptin secretion into the culture medium by up to 62%. Using a semiquantitative reverse transcriptase-PCR method, testosterone was found to suppress leptin mRNA to a similar extent. These results suggest that, apart from differences in body fat mass, the higher androgen concentrations in obese boys are responsible for the lower leptin serum concentrations compared with obese girls. ( J. Clin. Invest. 1997. 100:808-813.) Key words: androgens • gender difference • human adipose tissue • leptin

The aim of this study was to investigate the regulation of leptin expression and production in cultured human adipocytes using the model of in vitro differentiated human adipocytes. Freshly isolated human preadipocytes did not exhibit significant leptin mRNA and protein levels as assessed by reverse transcriptase (RT)-polymerase chain reaction (PCR) and radioimmunoassay (RIA). However, during differentiation induced by a defined adipogenic serum-free medium, cellular leptin mRNA and leptin protein released into the medium increased considerably in accordance with the cellular lipid accumulation. In fully differentiated human fat cells, insulin provoked a dose-dependent rise in leptin protein. Cortisol at a near physiological concentration of 10(-8) mol/l was found to potentiate this insulin effect by almost threefold. Removal of insulin and cortisol, respectively, was followed by a rapid decrease in leptin expression, which was reversible after readdition of the hormones. These results clearly indicate that both insulin and cortisol are potent and possibly physiological regulators of leptin expression in human adipose tissue.

No abstract

The aim of the present study was to develop an equation for the prediction of total body water (TBW) from bioelectrical impedance analysis (BIA) in obese children and adolescents before and after weight reduction. In 146 obese subjects with a mean age of 12.7 +/- 3.0 y (5.5-17.8 y), TBW was measured by using deuterium dilution as well as the resistance index (RI; ht2/resistance) using BIA before and after weight loss. Initially, the RI correlated well with measured TBW (r2 = 0.92, P < 0.001). A multiple-regression analysis using forward stepwise selection of the variables RI, sex, age, weight, height, and waist-hip ratio revealed that the equation TBW = 0.35 x RI + 0.27 x age + 0.14 x weight - 0.12 predicts most accurately individual values of TBW before weight loss (adjusted r2 = 0.96, SEE = 1.9 L) with a mean error of predicted TBW of 1.40 +/- 1.38 L. This equation was validated in 1000 random samples (bootstrap-sampling method), giving a mean r2 of 0.95. During the weight-reduction program, which included an energy-restricted diet and an extensive exercise program, the patients lost 7.7 +/- 3.2 kg, leading to a small decrease in TBW of 0.4 +/- 1.5 L. When the developed prediction equation was applied to the data after weight loss, an r2 value of 0.94 between measured and calculated TBW and a mean error of 2.18 +/- 1.89 L was obtained. Validation of the equation in 1000 random samples after weight loss again gave a mean r2 value of 0.95. Individual changes in predicted TBW correlated only weakly with those of measured TBW (r = 0.21, P < 0.05). Thus, individual TBW values before and after weight loss can be predicted by BIA with acceptable accuracy by using the developed equation. However, prediction of small individual changes in TBW during weight loss is not possible by BIA.