STUDY LOOKS AT CARDIOVASCULAR RISKS OF TESTOSTERONE REPLACEMENT THERAPY (NOV 2013)
Testosterone replacement therapy has become a highly commercialized treatment over the past several years. Whether or not testosterone deficiency is a pathologic condition is a
matter of debate. Testosterone levels decline as men age, and other factors including obesity and lack of exercise may play a role. Regardless of the cause, men are constantly being encouraged through
advertising to have their levels checked and treated if necessary.
The long-term risks of testosterone therapy are largely undefined. Estrogen replacement in postmenopausal women was a common practice until results from the Women's Health Initiative (WHI) trial
found a number of risks (see Women's Health Initiative results
for a discussion). The WHI was a very large, randomized controlled
trial, the likes of which are not likely to be repeated with testosterone replacement in men. Theoretical risks of testosterone therapy replacement include prostate cancer, benign prostatic hyperplasia,
sleep apnea, and cardiovascular disease.
A cohort study
published in the JAMA
evaluated the risk cardiovascular disease with testosterone replacement therapy
in a cohort of VA patients. PubMed abstract
- MAIN INCLUSION CRITERIA: Men in the Veterans Affairs (VA) system who underwent coronary angiography and had low total testosterone levels (< 300 ng/dl)
- MAIN EXCLUSION CRITERIA: Hematocrit > 50%; PSA > 4 ng/dl; patients who had previous testosterone therapy
- 8709 eligible patients were identified
- Testosterone was treated as a time-varying covariate
- All patients started in the No Testosterone cohort
- Once testosterone therapy was initiated, patients crossed-over to the Testosterone therapy cohort
- The statistical analysis involved a form of propensity score matching called stabilized inverse probability of treatment weighting
- This technique applies weights to individuals in each cohort and helps to balance covariates between cohorts.
Like propensity score matching, it is a method used to control for selection bias. Selection bias occurs when people who
are healthier or more likely to respond to an intervention are selected to receive the intervention.
- PRIMARY OUTCOME: Composite of all-cause mortality, heart attack, and ischemic stroke
- After an average follow-up of 27.5 months, the following was observed:
- The absolute rate of the composite outcome was 19.9% in the No Testosterone group and
25.7% in the Testosterone group. Absolute risk difference of 5.8% (95%CI: -1.4% to 13.1%).
- In the adjusted analysis, the risk of of the composite outcome was significantly greater in
the Testosterone group when compared to the No Testosterone group
(HR 1.29, CI 1.04 - 1.58, p-value= 0.02)
- The Testosterone group tended to be younger with a lower number of comorbidities. This strengthens
the results of the study because they would be expected to do better regardless of treatment.
Patients seeking testosterone replacement therapy should be
aware that the long-term risks of therapy have not been defined. While this study was observational, it was performed with
advanced statistical techniques that bolster its validity. Patients using or considering testosterone replacement therapy
should be aware that there is a good chance it will increase their risk of mortality, heart attack, or stroke.
- Study weaknesses:
- Observational study. No control over unmeasured confounders and hidden bias.
- Study used a medical database to determine outcomes. Medical databases have a number of shortcomings.
- Population being studied was a limited group of VA patients
- Study did not measure possible benefits of therapy
- Study strengths:
- Used advanced statistical techniques to limit selection bias
- Sensitivity analysis did not alter results
- The Testosterone group tended to be younger with a lower number
of comorbidities, but it still had worse outcomes