Longitudinal changes in sex hormones in a population of older community dwelling men

Context Based on cross-sectional studies and short longitudinal studies, levels of sex hormones (SH) change with age in men. However, there are limited data about SH in men over 80, and few longitudinal studies >10 years. Objective The aim of this study was to characterize SH concentrations in older men, and assess long-term changes. Methods We studied 477 community dwelling men, ages 78-97 years, who participated in the Osteoporotic Fractures in Men (MrOS) study. Medical history was obtained via clinical visits and questionnaires. Serum sex steroid levels (testosterone, estradiol, estrogen, dihydrotestosterone (DHT)) were assayed using liquid chromatography/mass spectrometry, free testosterone by equilibrium dialysis, and sex hormone binding globulin (SHBG) by radioimmunoassay. In 215 men, SH levels were compared to levels obtained 14.0±0.65 years earlier. Results In cross sectional analyses, mean total testosterone levels were 413.9±180.0, with considerable variation among participants. In 21.4%, total T levels were <275 ng/dL. Similarly, directly measured free T levels were low (<6.6 ng/dL) in 37.6%. Total testosterone levels were modestly correlated with directly measured free T, estradiol, estrone, DHT and SHBG, and weakly negatively associated with BMI and smoking. In longitudinal analyses (n=215), average total testosterone, estradiol and estrone level decreased, while SHBG increased, but changes were highly variable. Among those with total T >275ng/dL at baseline, T levels decreased to <275 ng/dL in 20.5%. Conclusions In old men there was a broad range of total testosterone levels, and a significant fraction were in a range considered hypogonadal. In longitudinal analyses, although there was considerable interindividual variation, gradual declines in mean total testosterone continued in the 8th through 10th decades of life.