"T"errific or "T"errible? A Review of Testosterone Deficiency

How Common Is Testosterone Deficiency?
Metabolic effects of endogenous testosterone include increased bone strength, reduced fat mass, improved insulin and leptin sensitivity, reduced low-density lipoprotein cholesterol (LDL-C) and triglyceride levels, and an increase in high-density lipoprotein cholesterol (HDL-C). The concentration of free testosterone gradually declines as men age, due to decreased production as well as higher concentrations of sex hormone-binding globulin. One study found that the mean annual reduction in free testosterone levels was approximately 2%, although this reduction was attenuated among healthier men.^[1] The exact prevalence of testosterone deficiency (TD) is controversial. Many men have symptoms that might indicate TD. In a study of men at least 45 years old presenting to a primary care setting, the prevalence of a total testosterone level < 300 ng/dL was 38.7%.^[2] Less than 10% of these men were receiving testosterone treatment.

What Symptoms Can Indicate TD?
Symptoms of TD are nonspecific and include fatigue, loss of libido, hot flashes, depression, and sleep disturbance,^[3] although there is not necessarily a linear association between testosterone levels and symptoms. However, an association between testosterone levels and overall scores on a psychological health screening tool was not found in one population-based study of 3413 men.^[4] Nonetheless, men with TD in this study did have scores indicating higher levels of anxiety.

Consequences of TD
TD has been associated with obesity and higher rates of insulin resistance.^[5] A meta-analysis of 12 studies, while noting significant heterogeneity among studies, found an increase in both all-cause and cardiovascular mortality, particularly among older men.^[6] A more recent study found a U-shaped association between testosterone levels and mortality among men.^[7]

Beyond the effect of TD on individual patients, the larger consequences of this association are staggering. In a study which assumed a conservative prevalence of TD of 13.4% among middle-aged and older men in the United States, the additional number of cases of diabetes attributable to TD over a 20-year period was 1.1 million.^[8] The health consequences of TD were estimated to cost the US economy between $190 billion and $525 billion over 2 decades.

Testosterone and Mental Health
Studies examining the effect of testosterone replacement therapy on mood, well-being, libido, and dysthymia have varied in methodologic rigor, characteristics of the sample population, and outcomes studied. A meta-analysis involving 862 men with low-normal testosterone levels found small improvement in erectile function with no improvement in libido or satisfaction.^[12] In aggregate, research has documented improvements in libido and erectile dysfunction and possible improvements in mood. However, not all research has supported the efficacy of TT in improving psychological or quality-of-life outcomes.

Testosterone and Body Composition
Despite decreases in fat mass, at the end of this randomized controlled trial, 47.8% in the testosterone group vs 35.5% in the placebo group had the metabolic syndrome (P = .07).^[13] Another study of men with low testosterone and metabolic syndrome or type 2 diabetes treated with 2% testosterone gel vs placebo for 1 year found improvements in insulin resistance, hemoglobin A1c, lipid profiles, and libido. These changes, however, were not accompanied by a decrease in fat mass or waist circumference.^[14] A meta-analysis of 8 clinical trials involving 365 men examined the effects of testosterone on bone health. Injectable testosterone, but not other formulations, was found to be weakly effective in increasing bone mineral density; there was insufficient data to determine whether that translated to a decrease in fractures.^[15] Most trials lasted less than 1 year and there was substantial heterogeneity. Taken together, studies did demonstrate improvements in fat mass and possible increases in muscle power.

Effects of TT on the Prostate
A 2010 systematic review and meta-analysis by Fernández-Balsells and colleagues^[16] evaluated the adverse effects of TT. The methodologic quality of the 51 included studies varied from low to medium, with follow-up periods from 3 months to 3 years. TT was not associated with a significant difference in the risk for prostate biopsy or prostate cancer. A more recent review in 2014 reached a similar conclusion and went beyond this previous report.^[17] It found no significant association between the incidence of prostate nodules, prostate biopsy, or prostate cancer among different formulations of TT, and the duration of treatment failed to significantly change this conclusion.

Testosterone and Cardiovascular Disease
This same meta-analysis^[16] raised concern about CV risk associated with TT. Therapy was found to have no significant effect on mortality or on prostate or CV outcomes. This study followed the early discontinuation of a small trial of 209 community-dwelling men (average age, 74 years) with multiple comorbidities after 23 cardiovascular events were recorded in the TT group -- a much higher number than the 5 that occurred in the placebo group.^[18] Despite recommendations by the Data and Safety Monitoring Board to halt the trial because of this higher rate of adverse events, the researchers were careful to note that the small size of the trial and the unique population studied prevented drawing broader inferences about the safety of TT.

VA Study: Cardiovascular Risk
A number of subsequent studies have further raised concerns about CV risk with TT. A Veterans Affairs cohort study^[19] assessed the association between TT and all-cause mortality, myocardial infarction (MI), or stroke among male veterans with multiple comorbidities undergoing coronary angiography. The researchers further sought to determine whether this association was modified by underlying coronary artery disease. The use of TT in men was found to be associated with an approximately 30% higher risk for death, MI, or ischemic stroke. In an interview with heartwire, senior investigator Dr. Michael Ho (Veterans Affairs Eastern Colorado Health Care System) was careful to note that although an association was found, "it's not causal, given the observational nature of the study."

Systematic Review: Cardiovascular Risk
A systematic review and meta-analysis of placebo-controlled randomized trials published in 2013 found an odds ratio of 1.54 (95% CI, 1.09-2.18) for CV events associated with TT.^[20] Half of the trials were sponsored by pharmaceutical companies and the effect of therapy varied with the source of funding, but not with baseline testosterone levels. A higher risk for CV events was found in trials not funded by the pharmaceutical industry (8% in nonsponsored trials vs 4% in sponsored trials).

Cohort Study: Cardiovascular Risk
Early in 2014, results from the largest cohort study to date provided more fodder for concerns about CV risk with TT in men with and without pre-existing cardiac disease.^[21] In all subjects, the post-/pre-prescription acute nonfatal MI rate ratio for TT prescription was 1.36 (95% CI, 1.03-1.81). Among all men aged 65 years and older, there was a 2-fold increase in the risk for MI in the 90 days after filling an initial TT prescription. The risk declined to baseline in the 91-180 days after initial TT prescription among those who did not refill their prescription. Risk was similarly increased in younger men with pre-existing diagnosed heart disease. The researchers concluded: "Given the rapidly increasing use of testosterone therapy, the current results, along with other recent findings, emphasize the urgency of the previous call for clinical trials adequately powered to assess the range of benefits and risks suggested for such therapy."

Mechanisms for CV Risk With TT
The results of these analyses suggest that an association between TT and adverse CV events does indeed exist, despite several limitations with some of the research. There are a number of plausible biological mechanisms that could potentially explain this increase. But there is also some evidence that low endogenous testosterone levels may also be positively associated with cardiovascular events.^[24]

Putting It All Together
What is the practicing clinician to do with this conflicting and highly concerning data? In 2010 the Endocrine Society produced a practice guideline for the use of TT.^[25] This guideline is designed in large part to curb the overuse of exogenous testosterone. Only symptomatic men should be evaluated for possible TD, and TD should not be diagnosed without 2 morning testosterone levels that are unequivocally low. Serum testosterone levels should be reevaluated 3-6 months after the initiation of testosterone treatment, with the goal of achieving serum testosterone levels in the mid-normal range. TT should be prescribed only for men who truly need it, and these men require close follow-up to ensure that treatment goals are met and to avoid overtreatment. Caution is warranted in recommending TT to men at high CV risk.

TT: Lots of Choices
Testosterone preparations available in the United States include oral, buccal, topical, and injectable agents. Available data indicate that all products may be equally effective and associated with similar side-effect profiles. The US Food and Drug Administration (FDA) is now officially investigating the potential that FDA-approved testosterone products increase the risk for serious adverse CV outcome. Hard on the heels of that decision, the Endocrine Society issued a statement calling for large-scale studies of the risks and benefits of testosterone treatment in older men with declining levels of the hormone.