What was the primary finding of the TRAVERSE trial?

TRAVERSE found that 1.62% testosterone gel was non-inferior to placebo for major adverse cardiovascular events (MACE) in hypogonadal men aged 45-80 with elevated cardiovascular risk, with a hazard ratio of 0.96 (95% CI 0.78-1.17).

Why is the 1.5 non-inferiority margin considered permissive?

A margin of 1.5 means the trial would accept up to a 50% increase in MACE as "non-inferior." While common in post-marketing cardiovascular safety trials, this threshold allows clinically significant excess risk to pass as acceptable. The observed upper CI bound of 1.17 is narrower than 1.5, but still does not exclude a meaningful MACE increase.

Does TRAVERSE apply to injectable testosterone?

No. TRAVERSE used topical gel only. Injectable testosterone cypionate and enanthate produce different pharmacokinetic profiles with supraphysiological peaks. Cardiovascular safety of injectables has not been tested in an equivalently powered RCT.

What was the pulmonary embolism signal in TRAVERSE?

The testosterone arm showed a higher rate of pulmonary embolism compared to placebo (HR 1.92 to 95% CI 1.12-3.32). This was a prespecified secondary endpoint but was not adjusted for multiple comparisons, so it should be interpreted cautiously while prompting further study.

Were men under 45 included in TRAVERSE?

No. The trial enrolled men aged 45-80. Younger hypogonadal men receiving TRT were not represented, limiting the generalizability of the findings to that growing clinical population.

How long were participants followed in TRAVERSE?

The median follow-up was approximately 33 months, with a maximum of about 4 years. For a cardiovascular outcomes trial, many experts consider this insufficient to detect slow-developing risks like plaque progression or late thrombotic events.

Who funded the TRAVERSE trial?

AbbVie, the manufacturer of AndroGel (1.62% testosterone gel), funded the trial. An independent data safety monitoring board and an academic statistical center provided oversight, but the funder participated in trial design and data collection.

Did TRAVERSE exclude men prone to polycythemia?

Yes. Men with baseline hematocrit above 48% were excluded. Since polycythemia is a known TRT side effect linked to thromboembolic risk, this exclusion may have removed the patients most vulnerable to cardiovascular complications from the safety analysis.

Should clinicians stop prescribing TRT based on TRAVERSE limitations?

Not necessarily. TRAVERSE provides the strongest available evidence that topical TRT does not cause a large MACE increase in older high-risk men. The limitations call for informed consent, ongoing hematocrit monitoring, and caution with extrapolation to untested populations or formulations, not avoidance of TRT altogether.

What did the TTrials coronary plaque finding mean for TRAVERSE interpretation?

The Testosterone Trials (TTrials) found increased noncalcified coronary plaque volume after 12 months of testosterone. This raises the possibility that longer TRT exposure could promote subclinical atherosclerosis, an outcome TRAVERSE's 33-month median follow-up and MACE-only endpoint may not capture.

Honest Criticisms and Limitations of the TRAVERSE Trial

By HealthRX Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

At a glance

| Parameter | Detail | |---|---| | N | 5,246 | | Intervention | 1.62% testosterone gel (AndroGel), dose-titrated to 350-750 ng/dL | | Comparator | Placebo gel | | Duration | Median follow-up ~33 months | | Primary endpoint | First occurrence of MACE (CV death, nonfatal MI, nonfatal stroke) | | Key result | HR 0.96 (95% CI 0.78-1.17); non-inferiority margin 1.5 met |

Why This Trial Exists

For over a decade, the FDA flagged testosterone replacement therapy (TRT) as a potential cardiovascular hazard. A 2010 TOM trial in frail older men was stopped early for excess cardiac events, and retrospective database studies in 2013-2014 added fuel. The FDA mandated a label warning in 2015 and required a large, randomized cardiovascular outcomes trial. TRAVERSE was that trial: the first adequately powered RCT designed to test whether TRT raises MACE risk in men with hypogonadism and preexisting or high risk of cardiovascular disease.

The headline result, non-inferiority with a hazard ratio of 0.96 (95% CI 0.78-1.17), was broadly reassuring. But reassurance is not the same as a clean bill of health. Below is a structured examination of what TRAVERSE does and does not settle.

The HealthRX 6-Domain Limitation Framework for TRAVERSE

We assess TRAVERSE through six lenses: enrollment bias, follow-up adequacy, statistical design choices, generalizability, conflict-of-interest transparency, and post-publication critique. Each domain carries a rating: LOW concern, MODERATE concern, or HIGH concern.

1. Enrollment Bias, MODERATE Concern

TRAVERSE required men aged 45-80 with either established atherosclerotic cardiovascular disease (ASCVD) or elevated CV risk factors (hypertension, dyslipidemia, diabetes, CKD, or a high Framingham score). This deliberately selected a sick population, which is appropriate for a safety trial. The problem is subtler than simple selection bias.

Who was excluded. Men with recent acute coronary syndrome (<3 months), NYHA Class III-IV heart failure, or hematocrit above 48% were excluded. Polycythemia is one of TRT's most common dose-limiting effects, and men prone to hematocrit spikes, the very patients most likely to experience thrombotic events on testosterone, were filtered out before randomization. The FDA prescribing information for testosterone warns explicitly about polycythemia-related thromboembolic events. By excluding these men, TRAVERSE tested TRT in a population already screened for lower thrombotic susceptibility.

Testosterone thresholds. Entry required two morning total testosterone levels <300 ng/dL. But 300 ng/dL is an arbitrary cutoff. Many men who receive TRT in clinical practice fall in the 300-400 ng/dL range, a gray zone that TRAVERSE did not address. This matters because the benefit-risk calculus may differ when baseline testosterone is mildly versus severely low.

Symptom requirements. Men needed at least one hypogonadal symptom, but the symptom list was broad (fatigue, low libido, decreased energy). These are nonspecific complaints common in the aging male population, meaning the enrolled cohort may have included men with functional hypogonadism from obesity or comorbidity rather than true primary or secondary hypogonadism. Whether TRAVERSE results apply equally to men with organic pituitary or testicular disease remains uncertain.

2. Follow-Up Duration, MODERATE to HIGH Concern

The median follow-up was approximately 33 months, with a maximum of about 48 months. For a cardiovascular safety trial, this is short.

Atherosclerosis is slow. Plaque progression, arterial stiffness changes, and remodeling effects of chronic androgen exposure operate over years. The Testosterone Trials (TTrials) found increased coronary artery plaque volume by CT angiography after just 12 months of testosterone, raising the possibility that longer exposure could shift MACE curves in a direction TRAVERSE was not powered to detect.
Event rates were lower than expected. The protocol assumed a placebo-arm MACE rate high enough to generate roughly 256 primary events. The actual accrual of events was slower, leading to protocol-mandated extensions. Lower-than-expected event rates reduce statistical power to detect modest excess risk, even if it exists.
No post-cessation follow-up window. TRAVERSE assessed events during active treatment and a brief post-treatment period. Cardiovascular effects of chronic TRT (vascular remodeling, erythrocytosis-driven viscosity changes) could manifest after discontinuation. No structured observation window addressed this.

3. Statistical Design Choices, MODERATE Concern

TRAVERSE used a non-inferiority design with a hazard ratio margin of 1.5. This means the trial would declare TRT "non-inferior" as long as the upper bound of the 95% confidence interval for MACE did not exceed 1.5.

| Design Parameter | TRAVERSE Choice | Context | |---|---|---| | Non-inferiority margin | HR < 1.5 | Allows up to 50% excess MACE risk | | Observed HR | 0.96 | Point estimate near null | | 95% CI upper bound | 1.17 | Comfortably below 1.5 | | Alpha (one-sided) | 0.025 | Standard for NI trials |

A 1.5 margin is within the range used by other CV safety trials (the FDA's 2008 diabetes guidance uses 1.3 for pre-approval, 1.8 for post-approval). But it means a true 30% or even 40% MACE increase could still result in a "non-inferior" verdict. The observed upper bound of 1.17 is more comforting than the protocol-specified margin, but the confidence interval does not exclude a clinically meaningful 17% MACE increase.

Put differently: TRAVERSE can reassure us that TRT does not double MACE risk. It cannot rule out a 10-15% increase in cardiovascular events, a magnitude that would matter at the population level given how widely TRT is prescribed. The Endocrine Society's 2018 guidelines noted the need for definitive safety data. TRAVERSE improved the evidence base substantially but did not deliver the tight confidence interval needed to declare testosterone cardiovascularly benign.

4. Generalizability Gaps, HIGH Concern

Several populations commonly receiving TRT in practice were underrepresented or absent from TRAVERSE.

Injectable testosterone. TRAVERSE used topical gel (1.62% testosterone, the AndroGel formulation). Injectable testosterone cypionate and enanthate produce supraphysiological peaks and wider pharmacokinetic swings. These formulations account for a large share of real-world TRT use and may carry different cardiovascular risk profiles. TRAVERSE findings should not be directly extrapolated to injectable regimens.
Younger men. The minimum age was 45. Men in their 30s and early 40s receiving TRT for confirmed hypogonadism, a growing clinical population, were not studied.
Men without elevated CV risk. Every TRAVERSE participant had preexisting ASCVD or multiple risk factors. The trial cannot speak to MACE risk in healthier hypogonadal men, though one could argue baseline risk modifies absolute (not necessarily relative) hazard.
Race and ethnicity. The enrolled population was predominantly White (~70%). Black men, who have higher baseline cardiovascular event rates and may metabolize testosterone differently, represented roughly 16% of participants. Hispanic/Latino enrollment was limited. Subgroup analyses by race were underpowered.
Duration of prior TRT use. Some enrollees had used TRT before randomization. Prior exposure could condition the cardiovascular system in ways that bias results in either direction.

5. Conflict of Interest and Funding, MODERATE Concern

TRAVERSE was funded by AbbVie, the manufacturer of AndroGel. AbbVie employees participated in trial design, data collection oversight, and had access to the data. The statistical analysis was performed by an academic coordinating center (Cleveland Clinic), and an independent data safety monitoring board oversaw the trial. These are standard safeguards.

Still, certain design choices benefit the sponsor's commercial interests:

A non-inferiority design (rather than superiority for harm) sets a lower bar for a favorable result.
The 1.5 margin is permissive enough that moderate excess risk would still yield a "safe" label.
Using gel rather than injectables tested the sponsor's own product while leaving competing formulations unaddressed.
The trial was mandated by the FDA, meaning AbbVie had a regulatory obligation and a financial incentive to produce a reassuring result. This does not imply misconduct, but it is context every reader should hold.

Several editorials, including an accompanying NEJM editorial, noted these tensions while acknowledging the trial was well-conducted within its design parameters.

6. Post-Publication Commentary, What the Letters Surfaced

After publication, letters to the editor and invited commentaries raised additional points:

Adherence. Treatment adherence in the gel arm declined over time, a common problem with topical testosterone. If men randomized to testosterone were undertreated for portions of the trial, the safety signal would be diluted. Per-protocol analyses (as-treated populations) were presented but carry their own biases.
Secondary endpoints. TRAVERSE found a nominally significant increase in pulmonary embolism and a trend toward more atrial fibrillation in the testosterone arm. These were prespecified secondary endpoints but not adjusted for multiplicity. The PE signal (HR 1.92 to 95% CI 1.12-3.32) warrants scrutiny in future studies.
Hematocrit management. Dose reductions for hematocrit rises above 54% were protocol-mandated. This active safety management is appropriate in a trial but does not reflect real-world practice, where monitoring is inconsistent. The trial's safety profile may therefore be more favorable than what clinicians see in routine care.
Prostate safety. While not the primary focus, TRAVERSE reported no significant increase in prostate cancer incidence. Some commentators noted the follow-up was too short to assess prostate outcomes meaningfully, given the disease's long natural history.

What TRAVERSE Settles and What It Does Not

| Question | TRAVERSE Answer | Confidence | |---|---|---| | Does TRT with gel roughly double MACE risk? | No | High | | Does TRT with gel carry zero excess CV risk? | Unclear | Low | | Is injectable TRT safe? | Not tested | N/A | | Is TRT safe in men under 45? | Not tested | N/A | | Does TRT increase PE risk? | Signal present (HR 1.92) | Moderate | | Is TRT safe beyond 4 years? | Unknown | N/A |

Clinical Translation

For the prescribing clinician, TRAVERSE provides a defensible basis to prescribe topical testosterone to hypogonadal men aged 45+ with cardiovascular risk factors, provided hematocrit is monitored and baseline thrombotic risk is assessed. It does not provide blanket safety assurance for all TRT formulations, all patient populations, or indefinite treatment durations. Clinicians should discuss these gaps with patients as part of informed consent.

The AUA/Endocrine Society guidelines continue to recommend individualized risk assessment. TRAVERSE supports, but does not replace, that approach.

Frequently asked questions

›

References

Lincoff AM, Bhasin S, Flevaris P, et al. Cardiovascular Safety of Testosterone-Replacement Therapy. N Engl J Med. 2023;389(2):107-117. PubMed
Budoff MJ, Ellenberg SS, Lewis CE, et al. Testosterone Treatment and Coronary Artery Plaque Volume in Older Men With Low Testosterone. JAMA. 2017;317(7):708-716. PubMed
Bhasin S, Brito JP, Cunningham GR, et al. Testosterone Therapy in Men With Hypogonadism: An Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2018;103(5):1715-1744. PubMed
Testosterone gel (AndroGel) prescribing information. U.S. Food and Drug Administration. FDA Label
Nissen SE. Testosterone Replacement Therapy and Cardiovascular Risk. N Engl J Med. 2023;389(2):172-173. PubMed
Basaria S, Coviello AD, Travison TG, et al. Adverse Events Associated with Testosterone Administration. N Engl J Med. 2010;363(2):109-122. PubMed