Leptin in postmenopausal women: Influence of hormone therapy, insulin, andfat distribution

Whether use of hormone-replacement therapy (HRT) influences menopaase-relat ed changes in body weight is unclear. HRT may affect energy balance by infl uencing synthesis of the adipocyte-derived hormone leptin. The objectives o f this study were to: 1) identify factors influencing circulating leptin in postmenopausal women; 2) determine whether HRT influences serum leptin aft er adjusting for confounding factors; and, 3) identify potential independen t effects of HRT or leptin on resting energy expenditure (REE). Subjects we re 54 postmenopausal women, 45-55 yr old, 35 of whom used HRT (estrogen plu s progestin). Total and regional body composition and fat distribution were determined by dual-energy x-ray absorptiometry and computed tomography; fa sting serum leptin and insulin, by RIA; and REE, by indirect calorimetry. S tepwise multiple linear regression analysis indicated that serum leptin cou ld best be predicted from total fat mass, fasting serum insulin, and total lean mass [log leptin = 1.08.log fat mass) + (0.46 log insulin) + (-1.25.lo g lean mass) + 1.88; model R-2 = 0.78, P < 0.001]. Multiple linear regressi on analysis indicated that visceral fat was independently related to leptin (parameter estimate = 0.23, P < 0.05), after adjusting for sc abdominal fa t and leg fat, as well as lean mass and insulin. After adjusting for total fat mass, total lean mass, and fasting insulin, serum leptin did not differ between users and nonusers of HRT (21.7 +/- 1.0 vs. 20.2 +/- 1.3 ng/mL, P = 0.369, adjusted mean + se, respectively). Serum estradiol was inversely c orrelated with (adjusted) leptin in non-HRT users (r = -0.50), suggesting t hat ovarian senescence may lead to an increase in leptin. Multiple linear r egression analysis indicated that REE (adjusted for fat mass, fat-free mass , and ethnicity) was not associated with leptin (P = 0.298) or hormone use status (P = 0.999; 1323 +/- 31 vs. 1316 +/- 42 kcal/day, adjusted mean rt S E for users and nonusers, respectively). These results indicate that, in po stmenopausal women: 1) total fat mass, lean mass, and fasting insulin, but not HRT, are significant determinants of serum leptin; 2) visceral and sc f at contribute to serum leptin; and, 3) neither HRT nor leptin is independen tly related to REE.