Serum testosterone bioassay evaluation in a large male cohort

Abstract

Objective: To assess if a cell-based readout of androgen action in serum demonstrates a closer association with recognized classical parameters of androgen action in men than current measures of serum testosterone (T). Design: To develop, validate and utilize a mammalian cell-based assay to measure specifically bioactive T and determine if this measure is a physiologically relevant fraction of serum T. Measurements and participants: We have developed a specific serum T bioassay using human prostate cancer cells. A rapid 5-min exposure to 100% serum followed by serum withdrawal confers specificity of the assay to serum T and provides sufficient sensitivity to measure T in male serum samples. Matrix effects were experimentally discounted as a confounding issue. A total of 960 male serum samples from the Florey Adelaide Male Ageing Study (FAMAS) with previous comprehensive cohort data and serum measurements were utilized. Results: Bioassay T measurement in the 960 FAMAS serum samples returned a median of 10·7 nmol/l (1·7–45·4), and was most closely related to immunoassayed total T, but not immunoassayed bioavailable T or calculated free T. Immunoassayed total T demonstrated a positive association with isometric grip-strength (R2 = 0·127, P < 0·001), self-reported sexual desire (R2 = 0·113, P < 0·001) and erectile function (R2 = 0·085, P < 0·05) while bioassay T did not. Conclusions: While cellular bioassays offer a rapid and sensitive means of identifying the androgenic potential of complex environmental compounds, the utility of such assays in defining a clinically relevant fraction of serum T distinct from total T needs further investigation.