ENDO: Low Testosterone May Be All-Cause Mortality Marker

SAN DIEGO, June 5 -- Men older than 50 with androgen deficiency are at a greater risk for all-cause mortality than their peers with age-appropriate testosterone, reported investigators here.
Among nearly 800 men followed for an average of 18 years, those in the lowest third of endogenous testosterone levels had a 33% greater risk for death from any cause than men in the highest third, reported Elizabeth Barrett-Connor, M.D., of the University of California at San Diego, and colleagues.
The difference in mortality risk between in the lowest and highest tertiles of testosterone could not be explained by smoking, drinking, physical activity level, diabetes, or cardiovascular disease, the investigators said at the Endocrine Society meeting.
But the association between androgen levels and death was attenuated when the investigators controlled for metabolic syndrome, and was eliminated when they also controlled for inflammatory cytokines, the authors reported.
"The study is only the second report linking deficiency of this sex hormone with increased death from all causes, over time, and the first to do so in relatively healthy men who are living in the community," said co-author Gail Laughlin, Ph.D., also of UCSD.
The data emerged from a substudy on endogenous testosterone of men enrolled in the Rancho Bernardo Heart and Chronic Disease Study, a longitudinal follow-up study now in its 35th year.
The investigators looked at 794 men from the ages of 50 to 91 years who had had serum testosterone measured at baseline (from 1984 to 1987). The men were followed through July 2004 or until their deaths. The authors defined androgen deficiency as lower than 250 ng/dL, generally agreed to be the lower limit of normal.
They found that among the 143 men in the cohort who met the definition of metabolic syndrome (waist circumference greater than 40 inches, low HDL, high triglycerides, hypertension, hyperglycemia), testosterone levels were 22% lower than that of others in the cohort (P<0.001). In addition, in men with the metabolic syndrome, the lower the testosterone levels the higher the levels of the inflammatory markers plasma interleukin-6 (IL-6) and high-sensitivity C-reactive protein (CRP ) (P for both <0.01).
There were 538 deaths during an average 18 years of follow-up. In unadjusted data and in an analysis adjusting for age, body mass index, and waist girth there were no differences in baseline testoserone levels (geometric means: 300 ng/dL versus. 298 ng/dl, respectively, P=0.84).
But when the authors conducted Cox regression analyses using testosterone tertiles, they noticed a low threshold effect of testosterone on mortality. Specifically, in an analysis adjusted for age, BMI, waist girth, and lifestyle choices, the 229 men with levels below 250 ng/dL had an odds ratio for death of 1.33 (95% confidence interval, 1.10 to 1.62) compared with men in with higher testosterone levels.
"This association was independent of diabetes and prevalent cardiovascular disease, but was attenuated by adjustment for the metabolic syndrome, and eliminated by further adjustment for IL-6 and high-sensitivity CRP levels," the authors reported.
When they further adjusted for serum creatinine or weight loss as markers of overall health status, the results did not change, and the addition of either total or bioavailable levels of estradiol had only minimal effects on results.
In multivariate logistic regression models that included age, BMI, lifestyle characteristics, the metabolic syndrome, and IL-6 and CRP levels, only the metabolic syndrome were associated with odds of low testosterone. Men with the metabolic syndrome had a three-fold risk for lower testosterone (odds ratio 3.05, 95% CI 1.88 to 4.95) and men who had one or more drinks per day compared with less or none had a nearly two-fold risk for low testosterone (odds ratio 1.95, 95% CI 1.32, to 2.86).
The authors cautioned that the results do not suggest a protective effect of exogenous testosterone on mortality risk.
"We are very excited about these findings, which have important implications, but we are not ready to say that men should go out and get testosterone to prolong their lives," said Dr. Barrett-Connor. "We're not ready to take this to the prescribing pharmacist."
"Conventional wisdom is that women live longer because estrogen is good and testosterone is bad. We don't know. Maybe the decline in testosterone is healthy and comes with older age. Maybe the decline is bad and is associated with chronic diseases of aging."