Testosterone, sex hormone-binding globulin, and frailty in older men

To determine whether testosterone (T) levels are associated with frailty or its components.

Population-based cohort study conducted in three waves (T1: 1987-1989, T2: 1995-1997, T3: 2002-2004).

Communities in the Boston, Massachusetts, area.

Six hundred forty-six men aged 50 to 86 at T(3) with complete data on frailty components and hormone measurements.

The frailty phenotype was defined as the presence of three or more of the following: weight loss, exhaustion, low physical activity, slowness, and weakness. Men were classified as frail (> or = 3 components), intermediate (1-2 components), and nonfrail (0 components). Whether total and free T or sex hormone-binding globulin (SHBG) levels were associated cross-sectionally with frailty and with degree of frailty was determined. Potential confounders such as age, chronic disease, lifestyle factors, diet, and physical activity were considered.

No association was observed between total or free T and the frailty phenotype after adjusting for confounders. Conversely, a significant association was observed between SHBG and frailty phenotype with an adjusted odds ratio of 1.25 (95% confidence interval=1.06-1.46) per 10-nM increase in SHBG levels. Associations between hormones and degree of frailty were similar to those for overall frailty. Of frailty components, grip strength and physical activity, but not exhaustion, slow walking, or weight loss, were associated with total T levels, whereas SHBG was related to weight loss, exhaustion, and physical activity.

Total and free T levels were not associated with frailty phenotype, but SHBG was. Furthermore, T and SHBG levels were associated with some, but not all, components of frailty. Therefore, T trials in older men should focus on men experiencing decreases in strength.