The FDA’s testosterone heart warning is gone. Here’s what actually happened.

For about a decade, any man who looked up testosterone therapy ran into the same wall. Somewhere on the second page of the search results was a warning that testosterone raised the risk of heart attack and stroke, and somewhere in the conversation with their own doctor was a version of the same sentence. It shaped how the drug was prescribed, how insurers treated it, and how a lot of physicians talked about it.

That warning is no longer on the label. On February 28, 2025, the FDA told the makers of every testosterone product to take the cardiovascular-risk language out and to put the results of a large trial in. This is a plain-English account of how the warning got there in the first place, what the trial found, what the FDA actually changed — and, just as importantly, the two things it deliberately did not change, which almost nobody writing about this has mentioned.

The short answer, up front

The FDA reviewed the results of a trial it had asked for — the TRAVERSE trial — along with a separate set of required blood-pressure studies, and issued class-wide labeling changes. In the agency’s own words, led by the TRAVERSE results it recommended:

• “adding the results of the TRAVERSE trial to all testosterone products;”
• “retaining ‘Limitation of Use’ language for age-related hypogonadism; and”
• “removing language from the Boxed Warning related to an increased risk of adverse cardiovascular outcomes for all testosterone products.”

And, led by the blood-pressure studies, it required the addition of product-specific information on increased blood pressure, plus a new blood-pressure warning on any product that did not already carry one.

So the honest one-sentence summary is not “the FDA says testosterone is safe.” It is: the cardiovascular warning came off, a blood-pressure warning went on, and the limitation on treating age-related low testosterone stayed exactly where it was.

How the warning got there

In January 2014 the FDA issued a drug safety communication saying it was evaluating reports of stroke, heart attack and death in men taking approved testosterone products. In September 2014 it convened a joint meeting of two advisory committees on the question. That meeting produced two things: a recommendation for an industry-wide clinical trial to settle the safety question properly, and, in March 2015, a second safety communication announcing required labeling changes and urging caution about using testosterone for low levels due to aging.

Note the sequence, because it matters. The warning went on the label in 2015. The trial that was supposed to answer the question underlying it did not report until 2023. For eight years the label carried a conclusion the evidence base had not yet earned.

What the studies behind the warning actually said

This is the part Dr. Castellano spends a chapter on in his book, and it is the reason he kept treating men through the warning years rather than stopping. The observational studies that drove the 2014–2015 alarm did not hold up well, and you do not have to take a physician’s word for that — the published record says so on its own.

The 2013 JAMA study, and the correction nobody reads

The paper that started it — Vigen and colleagues, published in JAMAin November 2013 — reported an association between testosterone therapy and death, heart attack and stroke. It was reported everywhere. What was reported almost nowhere is that JAMAhad to publish a correction to it on March 5, 2014, titled “Incorrect Number of Excluded Patients Reported in the Text and Figure.”

The correction added three categories of excluded patients that the original publication had left out of its results and its figure entirely: 1,301 patients with missing or other coronary anatomy data, 128 patients who had testosterone prescribed after a heart attack, and 100 patients who were women. That last one is worth sitting with. It is a study about testosterone therapy in men, and the correction notice is where you learn that women had to be excluded from it.

An international group of medical societies, researchers and clinicians petitioned JAMA to retract the paper. It was not retracted. It is still cited.

The TOM trial — and what its own authors said about it

The other study that gets cited constantly is the TOM trial, published in the New England Journal of Medicinein 2010. It was stopped early because a safety monitoring board found significantly more cardiovascular adverse events in the testosterone group — 23 men, against 5 on placebo.

Read past the headline and the picture changes. TOM enrolled 209 men with a mean age of 74, with a high baseline prevalence of hypertension, diabetes, hyperlipidemia and obesity, and it used 100 mg of testosterone gel daily. And the authors’ own assessment of their own finding, in the paper itself, was this:

“The lack of a consistent pattern in these events and the small number of overall events suggest the possibility that the differences detected between the two trial groups may have been due to chance alone.”

A trial of 209 frail elderly men, whose authors said the signal may have been chance, became a load-bearing citation for a class-wide warning applied to every man on testosterone. It also fed the meta-analyses — and a meta-analysis inherits the weaknesses of what goes into it. Pooling does not clean data. It averages it.

The specialists said so in 2015. Nobody moved.

In October 2015 — months after the FDA labeling change — eighteen international experts in urology, endocrinology, diabetology, internal medicine and basic science met in Prague and voted on nine resolutions about testosterone deficiency and its treatment. All nine passed unanimously. A representative of the European Medicines Agency attended in a non-voting capacity. The resolutions were published in Mayo Clinic Proceedings in 2016. Two of them speak directly to this page:

“(6) there is no scientific basis for any age-specific recommendations against the use of T therapy in men; (7) the evidence does not support increased risks of cardiovascular events with T therapy…”

That was 2015. The label did not change for another ten years. This is the part worth understanding about how medicine actually moves: the specialists who read the primary studies reached the conclusion first, and the regulatory language followed a decade later, once a trial existed that it could point to.

TRAVERSE: the trial the FDA asked for

TRAVERSE was the industry-wide study that came out of the 2014 advisory committee meeting, and it was designed to be hard to argue with. Randomized, double-blind, placebo-controlled. 5,198 hypogonadal men, aged 45 to 80, all of whom had either a history of cardiovascular disease or multiple cardiovascular risk factors — roughly 69% had diabetes, 84% had hyperlipidemia and 93% had hypertension. Mean duration of therapy was about 22 months; mean follow-up was 33 months.

In other words, they did not test the drug on the healthiest available men. They tested it on the men most likely to have a cardiac event during the study — which is the correct way to look for a safety signal, and the reason the result carries weight.

One point of precision matters here, and it is an easy one to lose once a result gets summarized: TRAVERSE was a non-inferioritytrial. It was built to answer “is this worse?” and the answer was no. It was not built to answer “is this better for your heart?”, and it did not answer that. Anyone telling you testosterone protects the heart is going further than the trial went.

What the FDA changed — and what it kept

The 2025 action is easiest to understand as a trade. One warning came off. Another went on. And a third piece of language, the one that governs who the drug is approved for, was explicitly left alone.

The blood-pressure finding is real, and it deserves better than a footnote

It would be easy to write this whole story as a victory lap and leave the blood-pressure part out. That would be the same mistake the alarm years made, pointed in the opposite direction.

On the AndroGel labeling, the ambulatory monitoring study measured a mean increase of 1.9/1.3 mm Hg after 16 weeks — and 3.0/2.2 mm Hg in men who were already being treated for high blood pressure. Within TRAVERSE itself, mean systolic pressure rose 1.0 mm Hg from baseline over 36 months in the testosterone group while falling 0.5 mm Hg in the placebo group. The label now instructs prescribers to measure blood pressure periodically and says the product is not recommended for men with uncontrolled hypertension.

These are small numbers per man. They are not nothing across years, and they are exactly the kind of thing that gets missed when a man is prescribed testosterone by a telehealth questionnaire and then never physically seen again. A few millimetres of mercury is a monitoring problem, not a reason to withhold treatment — but it is only a monitoring problem if somebody is actually monitoring.

The same goes for the rest of what TRAVERSE reported. Clotting events were numerically higher in the testosterone group, and the trial also recorded higher rates of atrial fibrillation, non-fatal arrhythmias needing intervention, acute kidney injury and bone fracture. None of that overturns the headline result. All of it is a reason the prescription comes with a schedule of recheck visits rather than a refill button.

What this means if you are actually considering treatment

Four things follow from all of the above, and they are practical rather than political.

• The heart-risk objection is out of date. If a physician tells you testosterone will raise your risk of a heart attack, the current FDA-approved labeling no longer says that, and the trial the FDA itself called for did not find it. You are allowed to ask what evidence they are working from.
• Your blood pressure is now part of the conversation. It should be measured before you start and periodically after, and uncontrolled hypertension is a reason to fix that first. If a clinic never takes your blood pressure, that is a meaningful thing to notice about the clinic.
• “Not increased risk” is not the same as “no risk.” Testosterone is a prescription medication with real effects that need watching — red-cell concentration, estradiol, PSA, blood pressure. That is what the follow-up schedule exists for.
• None of this changes who should be treated. The case for treatment is still a genuine deficiency, confirmed on properly drawn bloodwork, in a man with symptoms that match. A softer warning is not a wider indication.

That last point is where a lot of the current marketing is going to go wrong over the next couple of years, and it is worth saying plainly: the FDA removing a cardiovascular warning is not the FDA endorsing testosterone for every tired forty-five-year-old. The limitation on age-related low testosterone is still printed on every label. What changed is that the specific fear that kept appropriate patients off an appropriate treatment no longer has a label to stand on.

Why Dr. Castellano writes about this at all

Because this is what the job looks like from the inside. Treating men with low testosterone through 2014 to 2024 meant making a decision every week about a warning whose underlying studies Dr. Castellano had read and did not find persuasive — and explaining that decision, honestly, to men who had read the headlines.

The point is not that the FDA was the villain. The agency asked for the trial, the trial got done, and when the result came back the agency changed the label. That is the process working, just slowly. The point is that a warning and the evidence for a warning are two different objects, and a doctor who can tell you which studies a recommendation rests on — and what those studies actually say when you open them — is a different proposition from one who can only repeat the recommendation.

That is the standard the whole practice runs on: an hour in the room, real bloodwork before any protocol, and a straight answer about safety that includes the parts that are still open.

Testosterone and the FDA — quick answers

Sources

1. U.S. Food and Drug Administration. FDA issues class-wide labeling changes for testosterone products. Drug Alerts and Statements, February 28, 2025. The source for every quoted FDA position on this page, including the removal of the cardiovascular language, the retained Limitation of Use, and the class-wide blood-pressure finding.
2. AndroGel 1% (testosterone gel) prescribing information, revised July 2025. FDA-approved labeling; the source for the TRAVERSE trial figures, the blood-pressure measurements, the venous thromboembolism rates, and the recorded removal of the Cardiovascular Risk warning.
3. Lincoff AM, Bhasin S, Flevaris P, et al. Cardiovascular safety of testosterone-replacement therapy.N Engl J Med. 2023;389(2):107–117. The TRAVERSE trial as published.
4. Basaria S, Coviello AD, Travison TG, et al. Adverse events associated with testosterone administration.N Engl J Med. 2010;363(2):109–122. The TOM trial, including the authors’ own statement that the difference may have been due to chance alone.
5. Correction: Incorrect number of excluded patients reported in the text and figure.JAMA. 2014;311(9):967. Correcting Vigen R, O’Donnell CI, Barón AE, et al. JAMA. 2013;310(17):1829–1836.
6. Morgentaler A, Zitzmann M, Traish AM, et al. Fundamental concepts regarding testosterone deficiency and treatment: international expert consensus resolutions.Mayo Clin Proc. 2016;91(7):881–896.

Drug labeling changes over time. The FDA positions described here were verified against the primary FDA documents above in August 2026; if you are reading this well after that, the current prescribing information for your specific product is the authority.