Defending Testosterone, Debunking the Myths

Abraham Morgentaler, MD


June 04, 2015

In This Article

Mistakes and Misrepresentation

Recent concerns about the CV risks of testosterone grew out of a study by Vigen and colleagues[3] published in 2013 in JAMA. The investigators analyzed records from 8709 men in the Veterans Affairs health system who underwent coronary angiography and had low testosterone levels. The primary findings were that the absolute rate of stroke, myocardial infarction (MI), and death was 25.7% among men who had received a testosterone prescription compared with 19.9% in the untreated group at 3 years after angiography.

These findings received enormous media attention and were repeated in a widely quoted accompanying editorial.[4] However they were incorrect, and the error was immediately noted by readers. The correct absolute rate of events (number of adverse events divided by the number of individuals) was lower by one half in the testosterone-treated group compared with the untreated group (10.1% vs 21.2%).[5]

In response, JAMA published a new version of the study in which the same reported values were said to represent "Kaplan-Meier estimated cumulative percentages with events."[6] Curiously, this little-known statistical methodology (stabilized inverse propensity weighting) counted events in the testosterone group as about three events, whereas events in the untreated group were counted as fewer than one event.[7] This questionable methodology effectively reversed the true results of the study.

In March 2014, JAMA published a second correction of several data errors involving more than 1000 individuals. Moreover, it was revealed that the "all-male" study population actually comprised nearly 10% women.[8] To date, 29 medical societies have called for the retraction of this article, arguing that the data are not credible.[9]

A second study published 2 months later, by Finkle and colleagues,[10] reported increased rates of nonfatal MI within a short period after receipt of a testosterone prescription compared with the previous 12 months. This analysis of an insurance data set suffered from such serious methodological concerns that the FDA's own conclusions were that " is difficult to attribute the increased risk for non-fatal MI seen in the Finkle study to testosterone alone...."[11]


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.
Post as: