After removing the results of any veteran who had died within a year of their second testosterone sample, then compensating for factors such as age, body mass index and illness, the researchers found that the mortality risk for men with low testosterone was 68 percent higher than those with normal levels of the hormone.