TRT Heart Safety: TRAVERSE Reassures, Clot Risk Remains

Testosterone replacement therapy did not raise the combined risk of cardiovascular death, heart attack, or stroke in the largest trial designed to answer that question. But the same trial found a more complicated safety picture: testosterone changed blood-cell patterns, and those changes tracked with venous thromboembolism.

That does not turn TRT into a cardiovascular disaster, nor does it make a normal CBC a crystal ball. It does mean that a reassuring average result should not erase individual clot risk—especially when hematocrit, neutrophils, or monocytes rise during treatment.

TRAVERSE passed its primary cardiovascular test

The TRAVERSE trial randomized 5,246 hypogonadal men ages 45 to 80 with pre-existing cardiovascular disease or high cardiovascular risk to transdermal testosterone gel or placebo. The treatment target was a serum testosterone concentration of 350 to 750 ng/dL. Participants were followed for a mean of 33.0 ± 12.1 months.

The primary endpoint was major adverse cardiovascular events (MACE): cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke. MACE occurred in 182 men receiving testosterone (7.0%) and 190 receiving placebo (7.3%). The hazard ratio was 0.96 (95% CI, 0.78 to 1.17), and the trial met its prespecified criterion for noninferiority (P<0.001).

In practical terms, this result argues against a large increase in heart attack or stroke risk from replacement-dose transdermal testosterone in men resembling the trial population. It does not prove that testosterone prevents cardiovascular events; the confidence interval includes both modest harm and modest benefit. Its finding is narrower and more useful: testosterone was not worse than placebo for the trial's primary MACE endpoint.

TRAVERSE also reported higher incidences of pulmonary embolism, atrial fibrillation, and acute kidney injury in the testosterone group. The published abstract does not provide the event counts, hazard ratios, or confidence intervals needed to quantify the pulmonary-embolism difference precisely. That missing precision matters because pulmonary embolism is a patient outcome, not a laboratory curiosity.

The blood-cell analysis adds a clot-risk caveat

A prespecified analysis of 5,204 TRAVERSE participants examined whether testosterone-related changes in circulating blood cells were associated with thromboembolic events. In the 2025 TRAVERSE leukocyte analysis, participants received testosterone gel or placebo for a mean 21.7 ± 14.1 months.

Compared with placebo, testosterone significantly increased neutrophil and monocyte counts and significantly reduced lymphocyte and platelet counts. The direction of these changes is established by the randomized comparison. The analysis did not provide the absolute size of the cell-count changes or their timing in the available results, so it cannot tell clinicians exactly when a count peaks or specify an evidence-based CBC schedule.

The event associations deserve attention but need careful language. For every 1-standard-deviation increase in testosterone-induced neutrophils, the odds of venous thromboembolism rose 32% (odds ratio 1.32; 95% CI, 1.01 to 1.73). For the same increase in monocytes, VTE odds rose 39% (odds ratio 1.39; 95% CI, 1.08 to 1.79). Both estimates remained statistically significant after adjustment for treatment assignment.

On-treatment absolute neutrophil and monocyte counts also predicted MACE: the odds ratios were 1.25 (95% CI, 1.12 to 1.40) and 1.18 (95% CI, 1.06 to 1.31), respectively. Baseline counts showed comparable associations with MACE. These are associations within the trial data, not proof that extra neutrophils or monocytes caused a clot or heart event. A formal mediation analysis has not quantified how much, if any, of the pulmonary-embolism difference was explained by the leukocyte changes.

The biology is plausible: neutrophils and monocytes can participate in coagulation and inflammation. Yet no human study described here directly shows that replacement-dose testosterone triggers neutrophil extracellular traps or monocyte tissue-factor activity. Plausibility is a reason to investigate, not a substitute for measurement.

Hematocrit is a separate pathway worth watching

Testosterone can also increase red-cell mass. In a retrospective cohort study, Ory and colleagues found that men who developed secondary polycythemia while receiving testosterone—defined as hematocrit above 52%—had higher risks of MACE and VTE during the first treatment year.

This study cannot establish that the elevated hematocrit caused those events. Unlike TRAVERSE, it was not randomized, and patients who develop polycythemia may differ from those who do not in ways that affect risk. Still, it identifies a clinically measurable risk marker with a biologically distinct rationale: a higher hematocrit can increase blood viscosity. A rising hematocrit and a rise in neutrophils or monocytes could coexist, but the studies do not quantify their combined effect.

This distinction also helps explain why broad statements about "TRT and clots" tend to fail. Risk may cluster in people who develop particular treatment-related changes rather than appearing evenly across every user.

Why the wider evidence does not deliver a simple yes-or-no answer

Meta-analyses of randomized trials have not shown a consistent universal association between testosterone therapy and VTE, as the Ory paper notes. That population-level null result fits TRAVERSE's reassuring primary MACE outcome. It can also coexist with subgroup risk: trials may lack enough clot events, follow participants for limited periods, or average together patients with very different baseline susceptibility.

The platelet finding adds another uncertainty. Testosterone lowered platelets in the TRAVERSE blood-cell analysis, a change that could theoretically reduce primary hemostasis. The available evidence does not reconcile that effect quantitatively with the neutrophil and monocyte findings. The higher pulmonary-embolism incidence in the testosterone arm shows why one should not assume the platelet decrease settles the net clotting question.

Evidence on endogenous hormones points in a different direction. A 2023 Mendelian-randomization study by Wang et al. reviewed prospective cohorts of 3,051 to 4,673 men followed for 17 to 21 years that found no consistent association between endogenous testosterone and VTE. This does not establish the safety of every testosterone prescription. It supports a narrower inference: risks observed with exogenous treatment may depend on dose, delivery, treatment-related blood changes, and the patient's underlying health rather than on testosterone concentration alone.

Who was studied—and who was not

TRAVERSE enrolled older hypogonadal cis men with established or elevated cardiovascular risk; the average BMI was about 35 kg/m². That is an important population, but it is not every person who uses testosterone. Obesity itself can raise neutrophil and monocyte counts through chronic low-grade inflammation, which may affect how well the leukocyte findings apply to leaner or metabolically healthier men.

The trial also excluded men with prior VTE and thrombophilia, including Factor V Leiden carriers. Those exclusions protect trial participants but leave a major evidence gap for the people most likely to ask whether testosterone could amplify a pre-existing clotting tendency.

A 2025 real-world cohort by Bonnet et al. examined testosterone therapy outcomes—including atrial fibrillation, VTE, cardiovascular events, and mortality—in hypogonadal cis men and trans men. Its inclusion of trans men broadens the evidence base beyond TRAVERSE. It does not make TRAVERSE estimates transferable without adjustment: trans men commonly begin treatment younger, have different baseline cardiovascular profiles, and may receive different dosing patterns.

Benefits still belong in the decision

Safety monitoring makes sense only in the context of why someone is considering therapy. In an RCT of 83 men age 65 or older with obesity, frailty, and low testosterone, testosterone added to intensive lifestyle treatment improved physical-function outcomes more than lifestyle treatment alone over six months. That result supports a potential functional benefit in a selected population; it does not establish a cardiovascular benefit.

Clinical guidance still distinguishes classic hypogonadism caused by identifiable disease from age-related declines in testosterone, where the benefit-risk decision remains less settled. The 2023 clinical review by Grossmann et al. recommends an individualized assessment rather than treating a low laboratory value as an automatic prescription.

What a cautious TRT discussion should include

For a patient considering TRT, the reasonable response to these findings is monitoring and context, not panic. A baseline CBC with differential can identify elevated neutrophil or monocyte counts before treatment. A repeat CBC with differential can detect treatment-related changes, although the published leukocyte analysis does not establish an optimal interval. Hematocrit also warrants monitoring, because values above 52% identified a higher-risk group in the Ory cohort.

Prior VTE, known thrombophilia, obesity, and elevated baseline innate-immune cell counts deserve particular clinical attention. TRAVERSE cannot provide a direct risk estimate for patients with prior VTE or inherited thrombophilia because it excluded them. Symptoms compatible with VTE or pulmonary embolism require prompt medical evaluation rather than a debate over whether a lab value "looks a little off."

TRAVERSE provides meaningful reassurance about major cardiovascular events in high-risk hypogonadal men using testosterone gel. Its leukocyte findings and the separate polycythemia evidence argue for individualized clot-risk assessment and blood-count monitoring during treatment.

The honest conclusion is conditional. Testosterone replacement passed its main cardiovascular safety test in the population studied. It also produced blood-cell changes associated with VTE and MACE, while pulmonary embolism occurred more often in the treatment arm. Those findings do not prove a single causal pathway, but they make "TRT is safe for everyone" a claim the evidence does not support.