Abstract

Context: The prevalence of frailty increases, whereas testosterone decreases, as men age. Low testosterone may be a risk factor for development of this syndrome. Objective: Our objective was to determine whether testosterone levels are associated with frailty. Design: We conducted a prospective cohort study. Setting and Participants: Between 2001 and 2004, frailty was assessed in 3616 community-dwelling men aged 70–88 yr. Frailty was reassessed in 1586 men aged 76–93 yr in 2008–2009. Main Outcome Measures: Frailty was assessed with the FRAIL scale, comprising five domains: fatigue, difficulty climbing a flight of stairs, difficulty walking more than 100 m, more than five illnesses present, or weight loss greater than 5%. Testosterone, SHBG, and LH were assayed at baseline. Free testosterone was calculated using mass action equations. Results: At baseline, 15.2% of men (n = 548) were frail (at least three deficits), increasing to 23.0% (n = 364) at follow-up. At baseline, each 1 sd decrease in total or free testosterone level was associated with increased odds of frailty [odds ratio (OR) = 1.23; 95% confidence interval (CI) = 1.11–1.38, and OR = 1.29; 95% CI = 1.15–1.44 for total and free testosterone, respectively]. Lower LH was associated with reduced odds of frailty (OR = 0.88; 95% CI = 0.81–0.95). Adjustments were made for age, body mass index, smoking, diabetes, social support, and other covariates. At follow-up, only lower free testosterone levels (OR = 1.22; 95% CI = 1.05–1.42) predicted frailty. Conclusions: Lower free testosterone was independently associated with frailty at baseline and follow-up. Randomized trials should explore whether testosterone therapy can prevent the development of frailty.