The Hypothalamic-Pituitary-Testicular Axis in Exceptionally Old Men

Introduction: As life expectancy continues to increase and more men reach the extremes of age, it is important to understand the physiology of the aging hypothalamic-pituitary-testicular (HPT) axis and its role in health. While prior studies primarily focused on men younger than age ninety, we studied a unique cohort enriched for men with exceptional longevity to characterize the age-related trends in male sex-hormones, the etiology of the observed changes in the HPT axis, and its relationship with metabolic dysfunction and survival at the extremes of age. Methods: This is a cross-sectional study of community-dwelling Ashkenazi Jewish men (n = 427), age range 50-106 years. Longitudinal follow-up for vital status was conducted for men age ≥ 88 at enrollment (n = 86). Measurements included serum total testosterone (TT) by LC/MS, LH, SHBG, lipids, glucose and BMI. Free testosterone (FT) was calculated according to Vermeulen et. al. A change-point linear regression model was applied to describe the age trend of TT. Multivariable linear regression adjusted for comorbidities tested the associations between metabolic parameters and TT. The association between survival and TT was evaluated with the age-adjusted Cox proportional hazards model. Age-specific cutoffs for TT and LH were used to define primary and secondary hypogonadism. Results: The change point model was a significantly better fit for the data compared to the straight-line model (p = 0.004), indicating that TT significantly declines after age 88 years. Men age < 88 years had higher average TT (401 ± 162 vs. 278 ± 178 ng/dL, p < 0.001), FT (6.3 ± 2.0 vs. 3.3± 2.1 pg/mL, p < 0.001), and lower LH (4.3 [3.0 - 6.1] vs. 14.6 [7.2 - 25.5] mIU/mL, p < 0.001), compared to men age ≥ 88 years. The prevalence of primary and secondary hypogonadism was 2% and 11%, respectively, in men age < 88 years, and 30% and 11%, among men age ≥ 88 years (p < 0.001). A multivariable linear regression analysis revealed interactions between age, dichotomized at the change-point of 88 years, and metabolic parameters. Models stratified at age 88 demonstrated an inverse association between TT and BMI (p = 0.02), serum triglycerides (p = 0.007), and random glucose levels (p = 0.02) among men age < 88; whereas a positive association was noted between TT and HDL cholesterol (p = 0.009) in this group. In men age ≥ 88 years, TT was not associated with any of the metabolic parameters or overall survival. Conclusions: Low testosterone in men with exceptional longevity is largely a result of primary testicular failure that occurs around age 88 and is accompanied by preserved hypothalamic-pituitary response with no associated metabolic dysfunction or effect on survival. This is in contrast to younger men, whose low T typically results from hypothalamic-pituitary dysfunction and is associated with metabolic derangements.