Series: The Heart Scare That Almost Killed TRT
Subtitle: How a 2010 study, a flawed meta-analysis, and a media panic nearly derailed testosterone therapy for a generation
Post: 2 of 5
Tags: TRT, Vigen 2013, JAMA, FDA warning, cardiovascular risk, data errors, meta-analysis
Word Count: ~1,150
If the 2010 Basaria study was a yellow caution flag, what happened in 2013 and 2014 was a flashing red siren — loudly alarming, and based, it turned out, partly on a significant data error that would have been caught by a careful second look.
The story of how a flawed observational study and a subsequent FDA warning shaped testosterone prescribing policy for nearly a decade is a case study in how medical policy can run ahead of the evidence. It is also, in a more charitable reading, a case study in how regulators respond when there is enough smoke to warrant reaching for the extinguisher, even before locating the fire.
The 2013 Vigen Study: Anatomy of a Controversy
In November 2013, JAMA published a study by Dr. Rebecca Vigen and colleagues at the University of Texas Southwestern Medical Center titled “Association of Testosterone Therapy with Mortality, Myocardial Infarction, and Stroke in Men with Low Testosterone Levels.” The paper was a retrospective observational analysis of Veterans Affairs patients who had undergone coronary angiography — meaning they had already been evaluated for coronary artery disease.
The study examined 8,709 men with low testosterone levels. Of these, 1,223 received testosterone therapy and 7,486 did not. The analysis compared absolute rates of adverse cardiovascular events over a follow-up period.
The headline finding: testosterone therapy was associated with a 29% increase in the absolute rate of adverse events — death, heart attack, and stroke combined.
The paper received enormous media attention. It was published, once again, in JAMA. It appeared rigorous. The sample size dwarfed the 2010 Basaria trial. For a field already nervous about cardiovascular risk after 2010, this looked like confirmation.
The problem was that the data was wrong.
The Errors That Slipped Through Peer Review
Within weeks of publication, a group of researchers and clinicians began identifying serious problems with the Vigen analysis. The most glaring: the study had inadvertently included women in a study explicitly described as being about men.
An analysis of the published dataset found that 5.5% of the patients listed as male in the study were, based on the data, almost certainly female. Some patients in the testosterone therapy group had received treatments inconsistent with male patients — including progesterone therapy. Others appeared to have bilateral orchidectomies listed in their records as baseline conditions not actually present. The denominator was, in other words, not quite what the paper said it was.
A second problem: some men counted as having “received testosterone therapy” in the analysis had never actually received it. They were categorized in the treatment group based on having a prescription written, not based on having filled it or taken it. This is a significant distinction in a retrospective analysis, where treatment assignment drives the entire comparison.
A letter signed by over 200 researchers, clinicians, and scientists — including some of the most prominent figures in andrology and endocrinology — was sent to JAMA requesting a correction. JAMA published a partial correction in which the proportion of women was acknowledged, but the paper was not retracted and the core conclusions were not formally reversed.
The Concurrent FDA Response
The regulatory machinery, meanwhile, had been set in motion.
In January 2014, the FDA convened an advisory panel to review the cardiovascular evidence for testosterone therapy. The panel met without the benefit of resolved controversy about the Vigen study’s accuracy. By June 2014, the FDA had required manufacturers of all testosterone products to add language to their labels warning about potential cardiovascular risks.
The FDA label change was significant in two ways. First, it imposed a general warning — not a black box warning, but a “General Warning” — that would now accompany every testosterone product on the market, regardless of formulation or indication. Second, and perhaps more consequentially for the field, the FDA also clarified that testosterone products were approved only for men with hypogonadism due to an underlying medical condition — hypogonadism caused by organic pathology affecting the hypothalamic-pituitary-testicular axis or the testes themselves — rather than for the broader category of age-related testosterone decline.
This language shift was subtle but meaningful. It drew a regulatory distinction between, say, a man with Klinefelter syndrome whose testosterone was deficient as a direct result of a diagnosable condition, and a 55-year-old man whose testosterone had declined over time as part of normal aging. The former clearly met the approved indication. The latter was now in a regulatory gray zone.
The Medical Community Splits
The result was a field at war with itself.
On one side: urologists, endocrinologists, and men’s health physicians who argued that the Vigen data was compromised, that observational studies of this design couldn’t establish causation, and that the practical effect of label changes and media coverage was to deny treatment to men with documented deficiency and real symptoms.
On the other: cardiologists, regulators, and some endocrinologists who argued that the precautionary principle demanded restraint until well-powered prospective trials established safety, and that the history of medicine was littered with treatments that had seemed reasonable until they weren’t.
Between 2014 and 2018, testosterone prescription volume in the United States declined substantially from its early-2010s peak, according to data published in JAMA Internal Medicine. The decline was particularly notable in primary care settings, where physicians without specialized endocrinology training were most likely to be influenced by FDA label language and headline-driven coverage.
A Note on What “Observational” Means
One reason the Vigen data problems mattered so much is that observational studies — studies where researchers look backward at medical records and compare outcomes across groups who received different treatments — are inherently susceptible to something called confounding by indication. Men who are sicker are less likely to be prescribed testosterone therapy, precisely because their physicians are more cautious. Men who are prescribed testosterone are, on average, healthier and more functional than untreated men with similar testosterone levels.
This creates a paradox: observational studies of TRT tend to make testosterone look either better or worse than it actually is, depending on how the groups were selected. The Vigen study was particularly susceptible to this because the patient population — VA patients who had undergone coronary angiography — was already cardiovascularly compromised.
What the field actually needed, and what the FDA had been calling for, was a properly designed, prospective, randomized controlled trial with adequate statistical power to test cardiovascular safety directly.
That trial was coming. It would take several more years to design, fund, enroll, and complete. It would eventually answer the question definitively.
But not before the damage — to prescribing habits, to patients who went undertreated, and to the reputation of a legitimate therapy — had already been done.
Next up: Post 3 — “What We Actually Already Knew: The Evidence That Low Testosterone Itself Was a Cardiovascular Risk Factor.”
