What does testosterone therapy tell us about all-cause mortality?

The largest randomised trial on this topic, the TRAVERSE study involving more than 5,000 men with testosterone deficiency and elevated cardiovascular risk, found no significant difference in mortality or major cardiovascular events between testosterone gel and placebo. The composite endpoint (cardiovascular death, heart attack, stroke) occurred in 7.0% of the testosterone group compared with 7.3% in the placebo group. Testosterone was therefore no worse than placebo, but not clearly better either. The study was partly funded by the manufacturer of the product used.

A large systematic analysis of 35 randomised trials, with a combined total of more than 5,600 participants, confirms this picture: no demonstrable difference in cardiovascular events between testosterone and placebo. The total number of deaths across all these studies was too small to draw a firm conclusion about mortality alone. Testosterone did, however, improve quality of life and sexual function.

Two other studies present a different picture, but are less convincing. A registry study of 737 men with nearly ten years of follow-up found that men treated with a long-acting testosterone injection survived significantly less often, but this was observational research: because participants were not randomly assigned, other factors may explain the difference. A narrative review article with a clearly pro-therapy stance claimed that studies with adequate dose monitoring show a survival benefit, but that article does not quantify this and is therefore less suitable as evidence.

One point deserves special attention: in the TRAVERSE study, the testosterone group more frequently experienced atrial fibrillation, acute kidney failure, and pulmonary embolism. These were not fatal outcomes in the study, but they are not side effects to be ignored either. Monitoring of blood values, haematocrit, and PSA is therefore recommended during use.