Testosterone Conundrum: New Study Yields Clues

November 22, 2013

A new study in older men, published online November 20 in the Journal of Clinical Endocrinology & Metabolism, has shown that those with testosterone levels in the mid-range had the lowest rates of death from any cause.

This U-shaped association of testosterone with mortality differs from that seen in prior research, said lead author Bu Beng Yeap MBBS, FRACP, PhD, from the University of Western Australia, Perth.

"There was no benefit of having a high-normal testosterone level. Therefore, an optimal rather than a high testosterone level predicts [best] survival," Dr. Yeap told Medscape Medical News.

Trying to untangle all of this information on testosterone and its metabolites is difficult because different studies have provided conflicting results, particularly with regard to testosterone replacement therapy, Dr. Yeap conceded. While some research has suggested lower mortality in men treated with testosterone as opposed to controls, others have found the opposite.

There is particular concern about the use of testosterone-replacement therapy in older men with comorbidities. A study published in the Journal of the American Medical Association last month, for example, found a 30% higher risk for adverse cardiovascular events, including death, in a Veterans Administration cohort of men with multiple comorbidities undergoing coronary angiography treated with testosterone, compared with those who did not receive it (JAMA. 2013;310:1829-1836). And a National Institutes of Health study of testosterone supplementation in older men was terminated early in 2010 as a result of an excess of cardiovascular events in the treatment arm.

Asked to comment on the new findings, Richard Quinton, MBBChir, MD, FRCP, an endocrinologist from Newcastle University, United Kingdom, told Medscape Medical News: "A clue emerges from this study that indicates peak survival associated with the 50 to 75 serum testosterone centile, greater than that of men in the highest centiles. Serum T levels in the various unsuccessful testosterone-intervention studies would have been pushed well into (and indeed

If serum T is a biomarker for health, why is having the very highest levels not quite as good as having mid-high levels?

"If serum T is a biomarker for health, why is having the very highest levels not quite as good as having mid-high levels? This may reflect the ability of testosterone to induce erythrocytosis, which in older men can predispose to vascular thrombosis," he postulated.

"A similar signal emerges from men and women on renal-replacement therapy receiving erythropoietin [EPO] therapy; although EPO is a great treatment for fatigue resulting from anemia of end-stage renal disease, boosting erythrocytosis too much increases adverse cardiovascular outcomes. Aiming for mid-range serum T levels is sensible, but in fact the top priority in monitoring testosterone-replacement therapy is: 'Don't let the hematocrit rise!' "

Clarifying T Link With Outcomes Will Help Inform Trials

The conflicting results with regard to the risks and benefits of testosterone-replacement therapy are of great concern, given that the market for testosterone treatment is increasing exponentially, with many older men taking it without a clearly established indication. This is especially true in the United States, where direct-to-consumer advertising is driving the demand, as well as campaigns targeted at physicians by companies marketing testosterone. The phenomenon has also been observed elsewhere (eg, the United Kingdom, due to online access to such commercials).

"There have been no randomized trials of testosterone therapy with the prespecified end points of cardiovascular events or mortality, and such necessarily large studies would pose major logistic challenges," Dr. Yeap and colleagues observe in their paper.

However, clarifying associations of testosterone with health outcomes independently of conventional risk factors for ill health will help in the design of interventions to properly ascertain the benefits and risks of testosterone therapy, they note.

Hence, in this current study, they analyzed the mortality rate in a group of 3690 community-dwelling men aged 70 to 89 years of age in Western Australia. They divided the men into 4 groups based on their endogenous testosterone levels.

Plasma total T, dihydrotestosterone (DHT) — a metabolite of testosterone — and estradiol (E2) were assayed using liquid chromatography tandem mass spectrometry in early-morning samples collected from 2001 to 2004. Deaths to December 2010 were ascertained by data linkage.

U-Shaped Curve for Testosterone and Death in Older Men

There were 974 deaths (26.4%), including 325 from ischemic heart disease (IHD). Men who died had lower mean baseline T (12.8 vs 13.2 nmol/L; P = .013), DHT (1.4 vs 1.5 nmol/L; P = .002), and E2 (71.6 vs 74.0 pmol/L, P = .022).

After adjustment for other risk factors — including age, overweight, and other confounding factors including education, smoking, body mass index (BMI), waist-to-hip ratio, hypertension, dyslipidemia, diabetes, creatinine, and prevalent cardiovascular disease and cancer — testosterone and DHT were associated with all-cause mortality.

Cumulative mortality was highest in those with total testosterone in the lowest quartile, with the second-highest rate seen in men with testosterone in the highest quartile Those with testosterone in the middle two quartiles at baseline had the lowest incidence of death from any cause (testosterone quartile Q2 vs Q1, adjusted hazard ratio [HR], 0.82 [P = .033]; Q3 vs Q1, HR, 0.78 [P= .010]; and Q4 vs Q1, HR 0.86 [P= .05]).

Interestingly, higher levels of DHT were associated with lower mortality from IHD (Q3 vs Q1, HR 0.58 [P = .002]; Q4 vs Q1, HR 0.69 [P = .026]).

"Ours is the first study that I am aware of that demonstrates conclusively that higher DHT level is associated with lower ischemic heart disease mortality. DHT may be an important biomarker for ischemic heart disease in older men," Dr. Yeap noted.

Estradiol was not associated with either all-cause or IHD mortality.

"Older men who had testosterone in the middle range survived longer than their counterparts who had either low or high levels of the hormone," Dr. Yeap reiterated. "Having the right amount of testosterone and higher levels of DHT might indicate these men are in better health overall, or it could help them maintain good health as they grow older," he speculated.

Treat to Mid-Normal Levels, Rather Than High End?

But "in view of the controversy over possible adverse effects of testosterone therapy, particularly in older men who have limited mobility or comorbidities, the benefits vs risks of testosterone should be carefully considered," Dr. Yeap stressed.

We need rigorous randomized clinical trials to test whether testosterone supplementation in men with low-normal as opposed to unequivocally low levels would improve health outcomes.

"We need rigorous randomized clinical trials to test whether testosterone supplementation in men with low-normal as opposed to unequivocally low levels would improve health outcomes. Pending such studies, we should adhere to existing consensus guidelines when considering testosterone therapy in men."

Testosterone therapy should "be considered [only] in men who have symptoms and signs of androgen deficiency and who have been found to have unequivocally low levels of testosterone measured early in the morning using an accurate assay and confirmed at least once," he told Medscape Medical News. This is consistent with guidance from professional organizations — for example, those of the US Endocrine Society, he noted.

And "in older men, we should consider treating to mid-normal levels of testosterone rather than raising testosterone levels to the high end of the reference range," he concluded.

The authors have reported no relevant financial relationships, nor has Dr. Quinton..

J Clin Endocrinol Metab . Published online November 20, 2013. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.