The William Finkel et. al. retrospective study on the relationship between testosterone therapy in non-fatal myocardial infarction is certainly misleading, especially when reading the abstract. However, after reviewing this study, the cases of non-fatal myocardial infarctions (MI) were less in the testosterone group as shown on Table 3. The number of cases of non-fatal MI (695) were less under all the all ages when compared to after testosterone therapy, where there were 152 cases of non-fatal MI. Similarly, there were less cases of non-fatal MI for the groups less than 65 years old and over 65 years old. However, the outcome of the study was based on rates per 1000 persons per year, rather than the actual cases of non-fatal MI, which was less when testosterone was used. If the statistics were preformed with the actual numbers of subjects having non-fatal MI, then the outcome most likely would be favorable.
Dr. Abraham Morgentaler, Associate Clinical Professor in the Department of Urology at Harvard University, wrote the following regarding the statistical results of the study: “It is possible that the men’s heart attacks in this study were caused by their underlying medical problems not testosterone… most heart attacks occurred in the first 90 days after a prescription was written. It is unlikely that heart attacks could develop in such a short period of time.” (Featured interview: USA Today, January 29, 2014)
An important recent study about the benefits of the use of testosterone was not mentioned in this study. Dr. Shores and her colleagues published "Testosterone treatment and mortality in men with low testosterone" in 2012 in the Journal of Endocrinology and Metabolism 2012 Jun;97(6):2050-8. The results showed that the group that received testosterone had less mortality compared to the untreated group. In regards to William Finkel’s retrospective study, the type of testosterone replacement was not standardized, blood levels were not evaluated, therapy goals were not set and the use of rates of myocardial infarction per 1000 persons per year was used instead of evaluating the actual number of subjects having MI lead to a misleading conclusion. I do agree that a larger non-pharmaceutical-sponsored prospective study is needed to settle this recent debate.
Return to February 2014 e-Journal