FDA removed the testosterone heart warning. What TRAVERSE actually showed.

What the FDA actually did in February 2025

On February 28, 2025, the FDA announced a class-wide labeling change for every approved testosterone product — gels, patches, injections, pellets, oral testosterone undecanoate. The change had three parts.

First, the agency removed language from the Boxed Warning related to an increased risk of adverse cardiovascular outcomes. The boxed warning had been in place since 2015. It was the single most consequential restriction the FDA had ever placed on testosterone products and it shaped a decade of prescribing behavior.

Second, the FDA required all manufacturers to add the TRAVERSE trial results to their labels — so any clinician or patient reading the prescribing information would see the noninferiority data directly.

Third — and this is what gets lost in the relief — the FDA added warnings based on post-market ambulatory blood pressure monitoring (ABPM) studies showing that testosterone use modestly raises blood pressure. The boxed warning related to potential transfer to children and pregnant women from topical formulations remained. The "Limitation of Use" language flagging that testosterone is not indicated for age-related hypogonadism without a clinical hypogonadism diagnosis remained.

Net effect: the worst-feared cardiovascular language is gone. The label now reflects what TRAVERSE actually found — which is more nuanced than either side wanted.

The 2015 warning — and why it was always contested

The 2015 boxed warning was a response to two observational studies and one underpowered trial that, taken together, suggested testosterone replacement therapy might raise the risk of heart attack and stroke. The data was thin. The mechanism was unclear. Endocrinology and urology societies pushed back hard at the time, arguing the studies had methodological weaknesses — selection bias, confounding by indication, the trial that triggered the panic (Vigen et al., JAMA 2013) had data-handling errors that the journal had to correct.

But the FDA had to act on what was in front of it. The agency added the boxed warning and, more importantly, required manufacturers to fund a properly powered cardiovascular outcome trial. That trial became TRAVERSE.

It is worth saying clearly: the 2015 warning slowed TRT prescribing for a decade, restricted it to clearly hypogonadal men with documented low testosterone, and may have prevented some men from receiving treatment they would have benefited from. It also forced the field to generate the data needed to actually answer the question. TRAVERSE is the answer.

The label change is not a vindication of every TRT use case. It is a vindication of TRT specifically for hypogonadal men with cardiovascular risk — exactly the population TRAVERSE studied.

TRAVERSE: how the trial was built

TRAVERSE — formally, the Testosterone Replacement Therapy for Assessment of Long-term Vascular Events and Efficacy Response in Hypogonadal Men trial — was a multicenter, randomized, double-blind, placebo-controlled, noninferiority, event-driven study. It enrolled 5,246 men aged 45 to 80 with hypogonadism (defined by two morning serum testosterone measurements below 300 ng/dL) plus either pre-existing cardiovascular disease or high cardiovascular risk. Men were randomized to a 1.62% testosterone gel (titrated to keep serum testosterone in the normal range) or matching placebo gel. Mean follow-up was approximately 33 months.

The primary endpoint was MACE — major adverse cardiac events — defined as a composite of cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke. The trial was designed to test noninferiority: the question was not "does testosterone improve cardiac outcomes" but "does testosterone not make them worse." That framing matters. A noninferiority trial cannot prove benefit. It can rule out harm above a pre-specified margin.

Pre-specified secondary endpoints included atrial fibrillation, acute kidney injury, pulmonary embolism, venous thromboembolism, fracture rate, and prostate-specific outcomes. These secondaries are where the more interesting safety signals appeared.

The primary result: testosterone did not increase MACE

Testosterone-replacement therapy was noninferior to placebo for the primary MACE endpoint. The hazard ratio for MACE was approximately 0.96 with a 95% confidence interval that excluded the pre-specified noninferiority margin. In plain English: in a population of 5,246 hypogonadal men aged 45 to 80 with established cardiovascular risk, treating them with physiologic-range testosterone for a mean of 33 months did not raise their risk of heart attack, stroke, or cardiovascular death compared to placebo.

This is the finding that supported the FDA's removal of the cardiovascular boxed warning. It is also the finding that endocrinology had been arguing was the most likely outcome since 2015. The 2015 warning was based on weaker data and TRAVERSE was the higher-powered replication.

Two caveats worth stating clearly. First, the trial enrolled only hypogonadal men with cardiovascular risk — not healthy men, not men using TRT for performance enhancement, not men with supraphysiologic testosterone targets. The result does not extrapolate to those populations. Second, noninferiority does not mean superiority. TRAVERSE did not show TRT prevents heart attacks in hypogonadal men. It showed TRT does not cause them.

The atrial fibrillation signal that stayed on the label

Among the pre-specified secondary endpoints, the most clinically actionable was atrial fibrillation. The incidence of new-onset atrial fibrillation was higher in the testosterone group than the placebo group, with a hazard ratio approximately 1.35. The signal was strong enough that it appears in the post-2025 label and triggered cardiology commentary about the need for pre-TRT rhythm assessment.

Why might testosterone raise atrial fibrillation risk? Several plausible mechanisms. Testosterone modestly raises blood pressure (covered next). It increases hematocrit and blood viscosity. It can elevate sympathetic tone. All three contribute to atrial remodeling — the slow structural changes that precede atrial fibrillation. The signal in TRAVERSE was modest in absolute terms but consistent across pre-specified subgroups.

The clinical implication is not "do not use TRT." The implication is that men with a family history of atrial fibrillation, men with prior episodes of paroxysmal atrial fibrillation, or men with structural heart disease may warrant a more careful conversation about TRT and rhythm monitoring than the pre-TRAVERSE consensus suggested.

Blood pressure and the ABPM warning

Separate from TRAVERSE, post-market ambulatory blood pressure monitoring studies — the data the FDA specifically cited for adding new warnings in February 2025 — showed that testosterone use modestly raises 24-hour blood pressure. The increase averages a few mmHg in systolic pressure. The mechanism appears to involve hematocrit elevation, sodium retention via androgen receptors in the kidney, and possibly direct vascular effects.

A few mmHg sounds small. Across a population, across decades of TRT use, it is not trivial. Every 2 mmHg rise in systolic blood pressure is associated with roughly a 5% increase in stroke mortality and a 7% increase in ischemic heart disease mortality at the population level. The label change reflects this — the FDA's 2025 announcement specifically added new warnings about blood pressure based on the ABPM data, even as it removed the cardiovascular MACE warning.

The practical implication for any man on TRT: home blood pressure monitoring is no longer optional. Office readings every six months miss the issue. The post-2025 standard of care looks like baseline ABPM if available, home BP measurements at least weekly, and a low threshold to add or escalate antihypertensive therapy if the trend on TRT is upward.

What this means for hypogonadal men on TRT

For men who meet the FDA indication — confirmed hypogonadism with symptoms — the 2025 label change should reduce friction in the prescribing relationship. The boxed warning that had spooked some primary care physicians out of writing TRT scripts is gone. The TRAVERSE data sits on the label. Endocrinologists and urologists who had argued for a decade that the warning overstated the risk have been substantially vindicated.

The label change also reshapes the monitoring conversation. The cardiovascular question shifts from "are we going to give this man a heart attack" to "are we going to give him atrial fibrillation, raise his blood pressure, or push his hematocrit too high." All three are manageable with monitoring and dose titration. None requires stopping TRT in the absence of symptoms.

What this does not mean

The label change does not mean testosterone is approved for performance enhancement, body composition optimization, or anti-aging in men with normal baseline testosterone. The FDA preserved the "Limitation of Use" language: TRT is not indicated for age-related hypogonadism without a clinical diagnosis of hypogonadism. Eugonadal men using TRT for non-indicated purposes are outside the population TRAVERSE studied, and the safety data does not extrapolate to that use case.

The label change does not mean supraphysiologic dosing is safe. TRAVERSE used physiologic-range dosing aimed at restoring normal serum testosterone — typically 350 to 600 ng/dL during titration. Cycles aiming for 1,000 ng/dL or higher are not in the dataset. The TRT-as-performance-enhancement community sometimes cites TRAVERSE to justify higher doses; that is not what the trial showed.

The label change also does not eliminate prostate cancer concerns. Testosterone is still contraindicated in men with known or suspected prostate cancer, and PSA monitoring remains standard practice. The TRAVERSE prostate-specific findings showed no significant increase in clinically detected prostate cancer in the testosterone group, but the trial was not powered to definitively answer the long-term prostate risk question.

A monitoring framework after the label change

We do not write prescriptions. The following is a monitoring framework for discussion with a prescribing clinician — not a self-management protocol.