The longitudinal association between progressive temporal change in sexual (dys)function and mortality in older men.
Community-dwelling men aged 70 years and older from the Concord Health and Ageing in Men Project were assessed at baseline (2005–2007, n = 1,705), 2-years follow-up (n = 1,367), and 5-years follow-up (n = 958). Self-reported sexual function (erectile function and sexual activity) using standardized questions were analyzed by generalized estimating equations to examine the longitudinal prediction of mortality according to change in sexual function across three time-points.
Men reported to have erectile dysfunction increased from 64% to 80%, and to be sexually inactive increased from 56% to 59% over the course follow-up. In univariate analyses, erectile dysfunction (hazard ratio: 2.02, 95% confidence interval [CI]: 1.45–2.81) or having no sexual activity (hazard ratio: 2.31, 95% CI: 1.82–2.93) at baseline predicted increased mortality over the subsequent 7 years. Models adjusted for multivariate and major reproductive hormones had negligible impact on mortality prediction, but neither statistically significantly predicted mortality after adjusting for depression. Similarly, change in erectile dysfunction over time was associated with mortality over 7 years in univariate (odds ratio: 1.69, 95% CI: 1.34–2.14) and multivariate analysis, including hormones, but not after adjusting for depression (odds ratio: 1.24, 95% CI: 0.95–1.62). Change in sexual activity was associated with mortality over 7 years in univariate analysis (odds ratio: 2.37, 95% CI: 1.33–4.20) but not after adjusting for age (odds ratio: 1.45, 95% CI: 0.79–2.64).
Our analyses suggest sexual dysfunction was not an independent risk factor of, but rather may be a biomarker for, all-cause mortality in older men.