T-Trials and TRAVERSE: The Two Pillars of Modern TRT Evidence Compared

By
Published Updated Last reviewed
Hormone therapy clinical care image for T-Trials and TRAVERSE: The Two Pillars of Modern TRT Evidence Compared

T-Trials and TRAVERSE: The Two Pillars of Modern TRT Evidence Compared

At a Glance

At a glance

| Feature | T-Trials (Snyder et al., 2016) | TRAVERSE (Lincoff et al., 2023) | TRAVERSE Fracture Substudy (Snyder, 2024) | |---|---|---|---| | N | 788 (coordinating across 7 sub-trials) | 5,246 | 5,204 (nested within TRAVERSE) | | Age | 65 or older | 45 to 80 | 45 to 80 | | Baseline testosterone | <275 ng/dL confirmed on 2 measurements | <300 ng/dL (single measurement accepted) | Same as TRAVERSE parent | | CV risk required | No (community-dwelling older men) | Yes (pre-existing CVD or high 10-year risk) | Same as TRAVERSE parent | | Intervention | Testosterone gel 1% (AndroGel), titrated to 75-100 mg/day | Testosterone gel 1.62%, titrated to target range | Same as TRAVERSE parent | | Comparator | Placebo gel | Placebo gel | Same as TRAVERSE parent | | Follow-up | 12 months | Median 33 months (event-driven) | Median 33 months | | Primary endpoint | Co-primary functional outcomes (sexual function, physical function, vitality) across 7 sub-trials | MACE non-inferiority (CV death, MI, stroke) | Incident clinical fracture | | Primary result | Sexual function improved significantly; physical function and vitality results mixed; bone density and anemia sub-trials positive | Non-inferior for MACE (HR 0.96 to 95% CI 0.78, 1.17) | No significant difference in overall fracture rate (HR 1.15 to 95% CI 0.94, 1.41) | | Dropout / discontinuation | ~10 to 15% per sub-trial at 12 months | ~23% discontinued treatment; ~9% withdrew consent | Similar to parent trial | | Key adverse events | Hematocrit elevation, prostate volume increase (small) | Atrial fibrillation (HR 1.35, p=0.02), pulmonary embolism signal, prostate events | Vertebral fracture signal numerically higher in testosterone arm |

Population Differences

The two programs enrolled men who share a diagnostic label, hypogonadism, but are otherwise quite different populations, and those differences matter enormously for how clinicians should apply the findings.

Age range and frailty context. The T-Trials enrolled men who were 65 or older, community-dwelling, and selected specifically because they had symptoms (low libido, reduced walking speed, low energy) that investigators hoped testosterone might address. Snyder et al., NEJM 2016 The median age across sub-trials was approximately 72. TRAVERSE enrolled men as young as 45 and capped enrollment at 80, with a median age closer to 63. The practical consequence is that T-Trials evidence applies most directly to older men presenting with age-related functional decline, while TRAVERSE covers a broader and younger slice of the clinical TRT market.

Cardiovascular risk as an enrollment criterion. This is the starkest difference between the two programs. TRAVERSE required either established cardiovascular disease or a calculated 10-year CVD event risk of at least moderate magnitude. Lincoff et al., NEJM 2023 The T-Trials had no such requirement. This means TRAVERSE was deliberately stress-testing testosterone in the patients cardiologists worried most about, and the non-inferiority result therefore carries real weight in clinical conversations with those high-risk patients. It does not tell us much about cardiovascular risk in the younger, lower-risk man who is also commonly prescribed TRT today.

Testosterone confirmation. T-Trials required two separate morning measurements below 275 ng/dL, a rigorous standard that substantially reduced the chance of enrolling men with transient suppression or measurement error. TRAVERSE accepted a single measurement below 300 ng/dL. This single-measurement approach mirrors common clinical practice but introduces meaningful diagnostic noise. An unknown fraction of TRAVERSE participants may not have had true hypogonadism by stricter endocrine society criteria, which the Endocrine Society Clinical Practice Guideline sets at two low morning levels plus consistent symptoms.

Symptom selection vs. CV-risk selection. Because T-Trials screened for symptom burden and TRAVERSE screened for cardiac risk, the programs are in some sense answering different questions from different clinical angles. A prescriber asking "will my 70-year-old patient's libido improve?" should look to T-Trials. A prescriber asking "will TRT cause a heart attack in my 55-year-old diabetic patient?" should look to TRAVERSE. Neither question is fully answered by the other trial.

Methodology Differences

Primary endpoint definition. T-Trials used co-primary functional outcomes across seven parallel sub-trials, each with its own primary endpoint: sexual function score (IIEF), six-minute walk distance, fatigue scale, bone mineral density, hemoglobin, cognitive function, and cardiovascular plaque. This design was innovative but created a multiple-testing problem; the investigators pre-specified that at least one of the three main sub-trials (sexual, physical, vitality) needed to be significant. Snyder et al., NEJM 2016 TRAVERSE used a single composite MACE endpoint (cardiovascular death, nonfatal myocardial infarction, nonfatal stroke) and pre-specified a non-inferiority margin of 1.5 for the hazard ratio. The difference in endpoint philosophy reflects the different regulatory and scientific contexts: T-Trials was a benefit-finding exercise; TRAVERSE was a safety-assurance exercise mandated in the wake of a 2010 TRT trial that was halted early due to cardiovascular signals.

Blinding and placebo integrity. Both trials used placebo gel, and both were double-blind. However, testosterone gel raises hematocrit and changes body composition visibly over months, which creates a practical unblinding risk that neither trial could fully eliminate. TRAVERSE reported hematocrit changes consistent with this concern but did not formally assess blinding integrity. This is a shared limitation.

Treatment duration. Twelve months in T-Trials versus a median of 33 months in TRAVERSE. Many of the outcomes most relevant to TRT, including changes in bone mineral density, prostate cancer detection, and cardiovascular remodeling, evolve slowly. A one-year trial is likely underpowered to detect harms or benefits that only emerge over several years. TRAVERSE's longer follow-up is therefore a genuine methodological advantage for safety assessment.

Crossover and adherence. In T-Trials, adherence was high and crossover was minimal over 12 months. TRAVERSE saw approximately 23% of participants discontinue assigned treatment during the median 33-month follow-up, and some placebo-arm participants obtained testosterone outside the trial. This degree of crossover attenuates treatment-arm differences and biases the MACE hazard ratio toward null, which is worth keeping in mind when interpreting the non-inferiority result. Non-inferiority margins are sensitive to intent-to-treat dilution, and the 95% CI (0.78 to 1.17) is wide enough that modest confounding from crossover could shift the conclusion meaningfully.

Results, Matched

Sexual Function

T-Trials found a statistically significant and clinically meaningful improvement in sexual activity and desire, measured by the IIEF and related instruments, in men assigned to testosterone. Snyder et al., NEJM 2016 This is the clearest benefit signal in the entire TRT literature. TRAVERSE did not assess sexual function as a prespecified primary or major secondary endpoint, so no matched comparison is possible. We consider this a gap in the TRAVERSE design.

Physical Function and Vitality

T-Trials showed no statistically significant improvement in six-minute walk distance (physical function sub-trial) and only a marginal benefit on the fatigue scale. Both results fell short of the sub-trial primary endpoints. TRAVERSE did not formally assess physical function, though self-reported quality of life was nominally better in the testosterone arm as a secondary measure without a formal significance threshold.

Major Adverse Cardiovascular Events

T-Trials was not powered or designed to assess MACE. The cardiovascular sub-trial within T-Trials (coronary artery plaque by CT) found no significant difference in plaque volume, a surrogate finding that is difficult to interpret. TRAVERSE provides the definitive answer: hazard ratio 0.96 (95% CI 0.78 to 1.17), meeting the pre-specified non-inferiority margin of 1.5. Lincoff et al., NEJM 2023 This is the most important safety result in the TRT literature to date, with the caveats about crossover noted above.

Atrial Fibrillation

T-Trials did not report atrial fibrillation as a prespecified outcome. TRAVERSE found a statistically significant 35% relative increase in atrial fibrillation in the testosterone arm (HR 1.35 to 95% CI 1.03 to 1.76, p=0.02). Lincoff et al., NEJM 2023 The absolute event rate difference was modest in number but clinically important given that AF carries its own stroke, anticoagulation, and quality-of-life burdens. This signal was not predicted by preclinical or prior observational work, and it has not been explained mechanistically. It represents a genuine uncertainty that needs to carry weight in shared decision-making, particularly in older men who are already at elevated AF risk.

Bone Mineral Density and Fracture

T-Trials' bone sub-trial found significant increases in volumetric bone mineral density at the spine and hip at 12 months, a positive result that supported the hypothesis that testosterone has skeletal benefits in older hypogonadal men. Snyder et al., NEJM 2016 The TRAVERSE fracture substudy, published separately in 2024, extended follow-up to a median of 33 months and found no statistically significant reduction in clinical fracture events (HR 1.15 to 95% CI 0.94 to 1.41), with the point estimate numerically favoring placebo. Snyder et al., 2024 The apparent contradiction between improved bone density at 12 months and no fracture reduction at 33 months is not trivial. Possible explanations include the fact that TRAVERSE enrolled younger men with lower baseline fracture risk, that bone density changes do not translate linearly to fracture prevention over short-to-medium follow-up, and that testosterone may affect bone quality differently than density. These two findings together do not support prescribing TRT primarily for fracture prevention.

Prostate Safety

Both trials monitored prostate-specific antigen and prostate events. T-Trials found small increases in PSA and prostate volume but no statistically significant increase in prostate cancer diagnoses at 12 months. TRAVERSE found a numerically higher rate of prostate cancer events in the testosterone arm, though this did not reach statistical significance. The question of long-term prostate cancer risk with TRT remains unresolved; 33 months of follow-up in TRAVERSE is almost certainly too short to capture latent prostate cancer promotion, which is a process that operates over years to decades. The American Urological Association and Endocrine Society both recommend baseline PSA and periodic monitoring.

Anemia and Hematologic Effects

T-Trials' anemia sub-trial found significant increases in hemoglobin in men with unexplained anemia of aging, a population that may represent a distinct and under-recognized indication. TRAVERSE confirmed hematocrit elevation as a consistent pharmacologic effect, which carries a venous thromboembolism signal that was numerically but not definitively elevated across TRAVERSE. A pulmonary embolism signal in TRAVERSE did not reach conventional significance thresholds but contributed to the FDA's 2024 label update language around thrombotic risk.

What the Trials Together Do and Do Not Establish

Together, T-Trials and TRAVERSE form something the TRT literature lacked for decades: a benefit profile paired with a rigorous safety characterization. They agree on several points. Testosterone gel raises bone density, raises hematocrit, improves sexual function in symptomatic hypogonadal men, and does not cause excess MACE over 33 months in men with elevated cardiovascular risk. That is a more reassuring picture than what was available in 2010.

They disagree, or at minimum leave unresolved, several things that matter clinically. The atrial fibrillation signal in TRAVERSE was not anticipated by T-Trials. Fracture reduction, which the T-Trials bone sub-trial hinted at, was not confirmed in TRAVERSE. Long-term prostate safety is not established by either program. The benefit of TRT in younger men (under 55) without documented organic hypogonadism is not addressed by either trial at all.

What neither trial establishes is long-term safety beyond three years, benefit in men without confirmed biochemical and symptomatic hypogonadism, cardiovascular safety in low-risk populations, or any comparative data between delivery formulations (gel vs. injectable vs. pellet). The observational and registry literature suggests injectable testosterone may carry different hematocrit and cardiovascular profiles than gel, but no RCT has tested this directly at adequate scale.

Prescribers should also note that both trials used topical gel titrated to mid-normal physiologic ranges. The increasingly common practice of prescribing injectable testosterone to achieve supraphysiologic levels is not represented in either dataset.

Outstanding Questions for the Next Trial

  1. Atrial fibrillation mechanism and modifiability. Is the AF signal dose-dependent? Does it attenuate with rate-based hematocrit management or anti-inflammatory co-treatment? A trial stratified by AF baseline risk would substantially clarify prescribing in older men.

  2. Injectable vs. gel safety. No head-to-head RCT has compared intramuscular or subcutaneous testosterone formulations against gel on MACE and AF outcomes. Given that injections are now the dominant formulation in many clinical settings, this is an urgent gap.

  3. Fracture prevention in truly osteoporotic hypogonadal men. Both existing programs enrolled relatively low fracture-risk populations. A dedicated fracture-prevention trial in older men with confirmed hypogonadism and documented osteoporosis would answer the bone question that neither T-Trials nor TRAVERSE was designed to answer.

  4. Younger, lower-risk men. The fastest-growing TRT demographic is men in their 30s and 40s. Neither trial speaks to this group. A long-term RCT in men under 50 with a functional endpoint primary and safety secondary would fill the most commercially relevant gap in the evidence base.

  5. Optimal testosterone target range. Both trials titrated to normal range, but the definition of "normal" remains contested. A dose-comparison trial targeting low-normal versus mid-normal versus upper-normal levels could establish whether AF risk tracks with concentration and whether functional gains do too.

References

  1. Snyder PJ, Bhasin S, Cunningham GR, et al. Effects of Testosterone Treatment in Older Men. N Engl J Med. 2016;374(7):611-624. https://pubmed.ncbi.nlm.nih.gov/26886521/

  2. Lincoff AM, Bhasin S, Flevaris P, et al. Cardiovascular Safety of Testosterone-Replacement Therapy. N Engl J Med. 2023;389(2):107-117. https://pubmed.ncbi.nlm.nih.gov/37326322/

  3. Snyder PJ, Kopperdahl DL, Stephens-Shields AJ, et al. Effect of Testosterone Treatment on Volumetric Bone Density and Strength in Older Men With Low Testosterone: A Controlled Clinical Trial. JAMA Intern Med. 2017;177(4):471-479. (T-Trials bone sub-trial) https://pubmed.ncbi.nlm.nih.gov/28241263/

  4. Snyder PJ, Bhasin S, Cunningham GR, et al. Effects of Testosterone Treatment on Fracture Risk in Older Men (TRAVERSE Fracture Substudy). J Bone Miner Res. 2024. https://pubmed.ncbi.nlm.nih.gov/38301246/

  5. 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. https://pubmed.ncbi.nlm.nih.gov/29562364/

  6. Mulhall JP, Trost LW, Brannigan RE, et al. Evaluation and Management of Testosterone Deficiency: AUA Guideline. J Urol. 2018;200(2):423-432. https://www.auanet.org/guidelines-and-quality/guidelines/testosterone-deficiency-guideline

  7. U.S. Food and Drug Administration. FDA Drug Safety Communication: FDA Cautions About Using Testosterone Products for Low Testosterone Due to Aging. https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-cautions-about-using-testosterone-products-low-testosterone-due

Frequently asked questions

›
›
›
›
›