Testosterone is best known as a sex hormone, but its levels also tend to track with general health. A large international analysis now suggests that men with very low testosterone are more likely to be diagnosed with cancer, and to die from it, in the years that follow. The finding comes with an important exception, and with an important warning against drawing the wrong conclusion.

The study was published on June 19, 2026 in The Lancet Healthy Longevity by the Androgens In Men Study (AIMS), an international collaboration led by Professor Bu Yeap at The University of Western Australia.

How the study worked

The team pooled data from 11 long-running cohort studies in Australia, the United States and Europe, covering more than 26,000 men. In each study, sex hormones were measured in a blood sample taken at the start, and the men were then followed for years or decades to see who developed cancer or died from it. The hormones measured included testosterone, dihydrotestosterone (DHT), sex hormone-binding globulin (SHBG) and luteinising hormone, which the pituitary gland releases to tell the testes to make testosterone.

For most analyses, the researchers used individual participant data, which lets them analyze every man’s results directly instead of relying on each study’s summary. They also adjusted for age, previous cancer, body mass index, smoking, alcohol, physical activity, blood pressure and diabetes.

What they found

  • Cancer death. Men in the lowest fifth of testosterone had an 18% higher hazard of dying from cancer than men in the highest fifth (hazard ratio 1.18, 95% confidence interval 1.04 to 1.34). Lower DHT showed a similar pattern. The risk began to rise when testosterone fell below about 8.6 nmol/L.
  • Cancer diagnosis. Men with very low testosterone were also more likely to be diagnosed with cancer, with risk rising below about 7.3 nmol/L.
  • Prostate cancer was different. Testosterone was not associated with prostate cancer risk. Instead, lower SHBG and lower luteinising hormone were linked to a higher risk of prostate cancer (hazard ratios of 1.28 and 1.45 for the lowest versus highest fifth). That analysis included five studies, 12,280 men and 918 prostate cancer cases.
  • Not a simple “more is better.” For SHBG and luteinising hormone, the lowest risk of cancer death appeared at mid-range levels, not at the highest.

For context, the researchers note that testosterone in healthy young men usually falls between 10 and 30 nmol/L and tends to be lower in older men. It is also influenced by factors such as body mass index, physical activity and conditions like diabetes.

Why this does not mean men should take testosterone

This is an observational study, so it shows an association, not cause and effect. Several explanations are possible, and the study cannot choose between them. Low testosterone could be a marker of poorer underlying health that also raises cancer risk. Hormone levels could play a more direct biological role. Or factors the researchers could not fully measure could explain part of the link. Each man’s hormones were also measured at one point in time, which may not reflect his levels over the following decades.

Professor Yeap was explicit on this point: there is no evidence yet that giving men testosterone would lower their cancer risk. What the findings suggest is that low testosterone may serve as a biomarker, a warning sign that could help identify men who might do worse if they develop cancer later. A hazard ratio of 1.18 is also a modest increase in relative risk, not a prediction for any individual man.

What men can take from it

The practical advice from the researchers is straightforward. A man who is found to have low testosterone should see a doctor for a full checkup, because it may reveal other health issues or risk factors worth addressing, even if there is no sign of testicular disease. They also framed a low result as a prompt to adopt healthier habits and keep up regular, broad health checks. As with any individual health question, this is a conversation for a doctor, not a conclusion to draw from a single study.

Funding listed for the study includes the Medical Research Future Fund, the Government of Western Australia and Lawley Pharmaceuticals.

This post is a summary based on the UWA press release and the published abstract. For full methods, statistics and limitations, see the original paper.

Sources