Longitudinal Changes in Testosterone Over Five Years in Community-Dwelling Men

Context: There are few population-based studies reporting longitudinal changes in total T, LH, FSH, and SHBG levels, and there is limited information on risk factors for their change. Objective: The objective of the study was to examine 5-year changes in serum T, LH, FSH, and SHBG levels among Australian men. Design: The study initially included a randomly selected, community-based cohort of 1588 men age 35 years or older at recruitment (mean age, 54 +/- 11 y) with available data at 2 visits. Men on medications known to affect, or with established pathology of, the hypothalamo-pituitary gonadal axis were excluded, leaving 1382 for analysis. Results: Mean baseline and follow-up T levels were 16.2 +/- 1.4 and 15.6 +/- 1.4 nmol/L, a change of -0.13 nmol/L/y. Annualized T changes were associated with obesity, being unmarried, and smoking at baseline, but not with diabetes, hypertension, or cardiovascular disease. T declined in men who had persistent depression or developed chronic disease, and it increased in men who were married, as compared to unmarried, at both time points. In the multivariate analysis, smoking cessation, development of central obesity (waist >= 100 cm), or generalized obesity (body mass index >= 30 kg/m(2)) resulted in T decreases of 0.36, 0.25, and 0.20 nmol/L/y, respectively. Quitting smoking, developing obesity, and having persisting depression were inversely related to SHBG change. Conclusions: An age-related decline in T levels is not inevitable but is largely explained by smoking behavior and intercurrent changes in health status, particularly obesity and depression.