Effects of testosterone on sexual function in men: results of a meta-analysis

The role of androgen decline in the sexual activity of adult males is controversial. To clarify whether sexual function would benefit from testosterone (T) treatment in men with partially or severely reduced serum T levels, we conducted a systematic review and meta-analysis of placebo-controlled studies published in the past 30 years. The aim of this study was to assess and compare the effects of T on the different domains of sexual life.

A comprehensive search of all published randomized clinical trials was performed in MEDLINE, the Cochrane Library, EMBASE and Current Contents databases.

Guided by prespecified criteria, software-assisted data abstraction and quality assessed by two independent reviewers, a total of 17 randomized placebo-controlled trials were found to be eligible. For each domain of sexual function we calculated the standardized mean difference relative to T and reported the results of pooled estimates of T treatment using the random effect model of meta-analysis. Heterogeneity, reproducibility and consistency of the findings across studies were explored using sensitivity and meta-regression analysis.

Overall, 656 subjects were evaluated: 284 were randomized to T, 284 to placebo (P) and 88 treated in cross-over. The median study length was 3 months (range 1-36 months). Our meta-analysis showed that in men with an average T level at baseline below 12 nmol/l, T treatment moderately improved the number of nocturnal erections, sexual thoughts and motivation, number of successful intercourses, scores of erectile function and overall sexual satisfaction, whereas T had no effect on erectile function in eugonadal men compared to placebo. Heterogeneity was explored by grouping studies according to the characteristics of the study population. A cut-off value of 10 nmol/l for the mean T of the study population failed to predict the effect of treatment, whereas the presence of risk factors for vasculogenic erectile dysfunction (ED), comorbidities and shorter evaluation periods were associated with greater treatment effects in the studies performed in hypogonadal, but not in eugonadal, men. Meta-regression analysis showed that the effects of T on erectile function, but not libido, were inversely related to the mean baseline T concentration. The meta-analysis of available studies indicates that T treatment might be useful for improving vasculogenic ED in selected subjects with low or low-normal T levels. The evidence for a beneficial effect of T treatment on erectile function should be tempered with the caveats that the effect tends to decline over time, is progressively smaller with increasing baseline T levels, and long-term safety data are not available. The present meta-analysis highlights the need, and pitfalls, for large-scale, long-term, randomized controlled trials to formally investigate the efficacy of T replacement in symptomatic middle-aged and elderly men with reduced T levels and ED.