Leptin and aging: Correlation with endocrine changes in male and female healthy adult populations of different body weights

Aging is associated with changes in plasma levels of several hormones. Ther e are conflicting reports on whether circulating leptin levels change durin g aging, the possible explanation for which is that alterations in adiposit y and body mass index (BMI) also occur. In this study we measured plasma le ptin and other hormonal parameters known to influence leptin in 150 men and 320 women of a wide age (18-77 yr) and BMI (18.5-61.1 kg/m(2)) range. Subj ects of each gender were separated into 2 groups of similar BMI, i.e. nonob ese (BMT, <30) and obese (BMI, >30), and treated separately. Statistical an alysis was performed, treating each group of subjects as a whole population or divided into age groups (<30, 30-50, and >50 yr). BMI-adjusted leptin l evels were progressively lower with increasing age in women, with a consist ent fall after menopause (-21%; P < 0.001); in men, leptin levels also tend ed to be lower in subjects more than 50 yr of age, but the reduction was no t significant. Multiple linear regression analysis, performed on subjects t reated either as a whole population or divided into obese and nonobese, sho wed that in both genders BMI and age were independent contributors of lepti n levels, and there was an inverse relationship between leptin and age in b oth obese (standardized coefficient beta = -0.25 in women and -0.23 in men; P < 0.01) and nonobese (-0.22 in women and - 0.20 in men; P < 0.05) subjec ts. The correlation of leptin and age with plasma levels of sex and thyroid hormones, GH, insulin-like growth factor I, PRL, and insulin was also eval uated. The variables that correlated with leptin were included in a multipl e regression model that included BMI and age. Testosterone in men (-0.43 in nonobese and - 0.19 in obese; P < 0.05) and estradiol in women (0.22 in no nobese and 0.24 in obese; P < 0.05) were important contributors to leptin l evels; also, dehydroepiandrosterone sulfate in obese women (-0.16) and sex hormone-binding globulin in obese subjects of both genders (0.15 in women a nd 0.19 in men) were significant determinants in the model. However, none o f the hormonal parameters abolished the negative correlation between leptin and age or the gender difference in leptin levels. In conclusion, our data show that in adult humans of different body weight;, serum leptin graduall y declines during aging; leptin reduction is higher in women than in men, b ut it is independent hom BMI and other age-related endocrine changes.