Low testosterone levels seem to be linked to a heightened risk of premature death from heart disease and all causes, suggests research published online in Heart.
The finding refutes received wisdom that the hormone is a risk factor for cardiovascular disease.
The researchers base their findings on 930 men, all of whom had coronary artery heart disease, and had been referred to a specialist heart centre between 2000 and 2002. Their heart health was then tracked for around 7 years.
On referral, low testosterone was relatively common. One in four of the men was classified as having low testosterone, using measurements of either bioavailable testosterone (bio-T) - available for tissues to use - of under 2.6 mmol/l or total testosterone (TT) of under 8.1 mmol/l.
These measures indicate clinically defined testosterone deficiency, referred to as hypogonadism, as opposed to a tailing off in levels of the hormone as a result of ageing.
During the monitoring period almost twice as many men with low testosterone died as did those with normal levels. One in five (41) of those with low testosterone died, compared with one in eight (12%) of those with normal levels.
The only factors that influenced this risk were heart failure (left ventricular dysfunction), treatment with aspirin or a high blood pressure drug (beta blocker) and low bio-T levels.
A low bio-T level was an independent risk factor for premature death from all causes and from heart disease, after taking account of other influential factors, such as age, other underlying health problems, smoking and weight.
Borderline levels of low total testosterone (15.1mmol/l) also increased the risk of an early death.
While high doses of testosterone found in anabolic steroids are harmful to health, the evidence suggests that low, rather than high, levels of the hormone, are associated with obesity, risky blood fats, and insulin resistance, all of which are risk factors for diabetes and heart disease, say the authors.
Men at high risk of these diseases may stand most to gain from testosterone replacement, they suggest.
An accompanying editorial points out that there is increasing interest in looking at the impact of testosterone replacement.
"There has been a marked increase in prescription of testosterone over recent years. While the long term cardiovascular impact of testosterone supplements in those with low levels remains to be demonstrated, accumulating evidence suggests there is a sound basis for examining this," write the authors.
But the authors warn that it is not without its risks, which include prostate cancer. And high testosterone in women boosts the risk of diabetes and cardiovascular disease.
Research into testosterone has taken a back seat for many years, in favour of looking at the impact of oestrogens on heart disease, they conclude, adding: "Recent data suggest that this important pathway warrants a lot more attention."