After TRAVERSE: What the Largest Testosterone Safety Trial Actually Proved — And What It Didn't

For a decade, prescribing testosterone felt like a liability. The FDA slapped a black box warning on every testosterone product in 2015, citing cardiovascular risk. Clinicians retreated. Patients went untreated. The question — does testosterone therapy cause heart attacks? — hung over every prescription.

Then came TRAVERSE. And in 2026, the answer is clearer than it has ever been. But clearer is not the same as complete.

The Trial That Changed Everything

TRAVERSE (Testosterone Replacement therapy for Assessment of long-term Vascular Events and efficacy ResponSE in hypogonadal men) was the trial the field needed. Published in the New England Journal of Medicine in 2023, it was the first adequately powered, placebo-controlled cardiovascular outcome trial of testosterone therapy ever conducted.

The design was built for the question. Men aged 45–80 with hypogonadism (serum testosterone <300 ng/dL) and pre-existing cardiovascular disease or elevated cardiovascular risk were randomized to daily transdermal 1.62% testosterone gel or placebo. Testosterone doses were adjusted to maintain levels between 350–750 ng/dL.

The primary endpoint: first occurrence of a composite of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke (MACE). The trial was designed for noninferiority — to prove testosterone was no worse than placebo.

The Answer

TRAVERSE met its primary endpoint. Testosterone was noninferior to placebo for MACE. The hazard ratio was 0.73 (95% CI: 0.52–1.03) — a point estimate that actually favored testosterone, though the confidence interval crossed 1.0.

In plain language: no excess heart attacks, strokes, or cardiovascular deaths. In men at high cardiovascular risk. Over nearly three years.

No increased prostate cancer. No worsening of lower urinary tract symptoms. The specter that had haunted testosterone therapy for a decade was, by the largest and most rigorous evidence, unfounded.

The Secondary Signals — And the COVID Twist

But TRAVERSE was not a clean sweep. Several secondary outcomes were higher in the testosterone group:

These signals prompted concern. But then came a finding from the Zitzmann PaTeR position statement (Andrology, 2026) that may reframe everything: COVID-19 confounding.

TRAVERSE enrolled patients from May 2018 through February 2022 — straddling the pandemic. COVID-19 was associated with substantially increased risk of MACE, venous thromboembolism, atrial fibrillation, and acute kidney injury in the TRAVERSE cohort. When COVID infection status was accounted for, "the relation of such events to TTh was no longer visible."

In other words, the concerning secondary signals may be pandemic artifacts — not testosterone effects.

Two Expert Panels, One Dataset, Two Tones

The same data produced two very different expert responses, and the contrast reveals how medicine processes uncertainty.

The PaTeR panel (Zitzmann, Rastrelli, Murray et al.) takes the measured view: TRAVERSE resolves the MACE question but leaves secondary signals that deserve investigation. They emphasize individualized treatment, organic-vs-functional distinction, and careful monitoring.

The Androgen Society (Morgentaler, Dhindsa, Dobs et al.) takes the advocacy stance: the question is answered. Testosterone therapy does not increase cardiovascular risk. Period. They cite 19 meta-analyses reaching the same conclusion and argue the field should stop relitigating settled science.

Both are right, in their way. The MACE question is answered. But what TRAVERSE can't tell us is as important as what it can.

The Meta-Analytic Consensus

TRAVERSE does not stand alone. Multiple independent meta-analyses now converge on the same conclusion:

García-Becerra et al.'s 2026 IJIR analysis also assessed prostate cancer risk: OR 0.88 (95% CI: 0.52–1.51) for prostate cancer events and OR 1.13 (0.39–3.26) for clinically significant prostate cancer — neither significant.

The Corona meta-analysis deserves special attention. With 106 RCTs and over 15,000 patients, it found no significant MACE difference. The arrhythmia signal (MH-OR 1.61, 95% CI 0.84–3.08) and atrial fibrillation signal (MH-OR 1.44, 0.46–4.46) were not statistically significant across the full dataset — consistent with TRAVERSE's secondary findings being noise or confounding rather than true effect.

The FDA's Evolution

The regulatory story tracks the evidence in real time:

In February 2025, the FDA removed the cardiovascular risk language from the black box warning on testosterone products. Then in December 2025, an FDA expert panel recommended three dramatic changes: expanding indications to include functional/age-related hypogonadism, removing the prostate cancer contraindication, and removing Schedule III controlled substance designation.

FDA Commissioner Martin Makary suggested that "dogma" about testosterone risks may have been curtailing its appropriate use. Over 2,000 public comments were submitted by the February 2026 deadline. The review is ongoing.

What TRAVERSE Still Can't Tell Us

The MACE question is answered for transdermal testosterone gel in older, high-risk men. But the gaps are significant:

Formulation gap: TRAVERSE used only transdermal gel. Injectable testosterone (cypionate, enanthate, undecanoate) — the most commonly prescribed formulations globally — produces supraphysiological peak-and-trough pharmacokinetics that may carry different cardiovascular risk. No MACE-powered RCT exists for any injectable formulation.

Population gap: The TRAVERSE population — mean age 63.3, mean BMI 35, ~70% with type 2 diabetes — is not representative of the younger, leaner men who increasingly seek testosterone therapy. No outcome data exist for men under 45 or men without cardiovascular risk factors.

Severity gap: With a median baseline testosterone of 227 ng/dL, TRAVERSE enrolled men with moderate deficiency. Men with severe hypogonadism (<150 ng/dL) — who may benefit most — were underrepresented.

Subgroup void: No published subgroup analysis by baseline testosterone level exists. We don't know if men at 150 ng/dL respond differently from men at 280 ng/dL in terms of cardiovascular risk.

Dropout problem: Over 60% of participants in both arms dropped out before study completion. This is not unusual for long testosterone trials, but it limits statistical power for secondary outcomes and long-term safety conclusions.

The Uncomfortable Asymmetry

Here is perhaps the most important finding from my research: while TRT now has a dedicated cardiovascular safety RCT, zero cardiovascular outcome trials exist for any alternative treatment.

No MACE-powered RCT for clomiphene. None for enclomiphene. None for hCG. None for kisspeptin.

The 2025 SERM meta-analysis (Hohl et al.) pooled 10 RCTs with 819 patients — all with hormonal endpoints. The authors explicitly noted their analysis was "underpowered with regard to safety endpoints." No serious adverse events were observed, but absence of evidence is not evidence of absence.

This creates a paradox: the treatment with the most safety data (TRT) gets the most scrutiny, while alternatives operate in an evidence vacuum. Clinicians who prescribe SERMs or hCG over TRT for "safety" are making an assumption the evidence does not support — the alternatives may well be safer, but we genuinely do not know.

The Hidden Benefit

Lost in the cardiovascular debate is a TRAVERSE finding that deserves more attention: testosterone therapy produced a 22.5% reduction in progression from prediabetes to type 2 diabetes (p=0.029) among men without baseline diabetes. This metabolic benefit is biologically plausible — testosterone improves insulin sensitivity, reduces visceral fat, and restores metabolic signaling — and could translate to meaningful cardiovascular risk reduction over decades.

It also aligns with the broader evidence I've covered in previous articles: the metabolic trap between insulin resistance and low testosterone is bidirectional, and breaking it — whether with weight loss, GLP-1 receptor agonists, or appropriately prescribed testosterone — produces cascading benefits.

What This Means for Men with Secondary Hypogonadism

TRAVERSE was designed to answer the cardiovascular question, and it did. But it was not designed to guide treatment selection in secondary hypogonadism, and its limitations here are substantial.

The PaTeR panel's distinction between organic and functional hypogonadism matters. For men whose low testosterone stems from a identifiable pituitary or hypothalamic lesion, TRT is straightforward replacement therapy. For men whose low testosterone reflects obesity, metabolic syndrome, sleep apnea, or polypharmacy, the first intervention should address the root cause — not replace the end product.

This is the tension at the heart of modern andrology: TRAVERSE makes TRT safer to prescribe, which is good. But it also risks accelerating the trend of treating functional hypogonadism with testosterone instead of addressing why the axis failed in the first place.

The goal shouldn't be a normal testosterone number. The goal should be a functioning HPG axis.

The Bottom Line

TRAVERSE answered the question that mattered most: testosterone therapy does not cause heart attacks in high-risk men. The 2015 black box warning was wrong, and the FDA has corrected course.

But resolving the cardiovascular question exposes everything else we don't know. We don't know about injectables. We don't know about younger men. We don't know about severe hypogonadism. And we know nothing about the cardiovascular safety of the alternatives that millions of men are already using.

The field's next challenge is not whether to prescribe testosterone — it's knowing when not to, and what to prescribe instead, and whether those alternatives are actually safer or just less studied.

The signal started in the brain. The answer, as always, is more complicated than the question.