Men over the age of 40 with low levels of testosterone have an increased risk of dying, according to research.
Testosterone helps to build muscle
US scientists studied the relationship between testosterone levels and death in 858 military veterans aged over 40.
In the Archives of Internal Medicine, they report men with low hormone levels had an 88% increased risk of death compared with men with normal levels.
But another expert said the study was not representative of the general population and contained weaknesses.
Scientists at the University of Washington and Veteran's Association (VA) Puget Health Care System, both based in Seattle, studied male veterans who had had their testosterone levels measured at least twice between 1994 and 1999.
Their health records were then tracked until 2002.
The researchers found 53% of the men had normal testosterone levels, 28% had fluctuating levels where the average was classified as normal (termed equivocal levels), and 19% had low levels.
Over the course of the study, the researchers found 20.1% of men with normal levels had died, compared with 24.6% of men with equivocal levels and 34.9% of men with low levels.
Increased death risk
After adjusting the figures to take into account factors such as age, body mass index and other illnesses, the researchers found men with low testosterone levels were 88% more likely to die than those with normal levels.
The researchers then removed the men who had died within one year of taking the second testosterone sample from their calculations, as critical illnesses can cause testosterone levels to dramatically decrease.
They still found men with low levels had a 68% increased mortality risk.
Molly Shores, the author of the study, wrote with colleagues: "The persistence of elevated mortality risk after excluding early deaths suggests that the association between low testosterone and mortality is not simply due to acute illness.
"Large prospective studies are needed to clarify the association between low testosterone levels and mortality."
They pointed out their study might not be representative because it focused on military veterans, who had a higher risk of death than the normal population.
But Pierre Bouloux, professor of endocrinology from the Royal Free Hospital in London, said while testosterone and mortality had been linked, the study contained several weaknesses and could not be extrapolated to the general population.
"This is a VA - Veteran's Administration - study. The poorest people in the US go to the VA, who come from a very low socio-economic status - which is known to have an impact on mortality.
"This makes it very difficult to know if you can extrapolate these results to the generality of the population.
"It also seems very odd to me that the mortality amongst all of the men is so high but the problem is that the study does not even tell us what the causes of the deaths are," he added.
"The fact that it is also retrospective study rather than prospective is also not helpful. As the authors themselves state, to pursue any association, you really need a prospective study."