What was the primary endpoint of the T-Trials?

The T-Trials had three co-primary endpoints across three sub-trials: sexual function (measured by PDQ-Q4), vitality (measured by FACIT-Fatigue), and physical function (measured by 6-minute walk distance). Each sub-trial enrolled overlapping but distinct subsets of the 790 participants based on matching symptoms.

How large was the sexual function improvement?

The treatment effect on the PDQ-Q4 sexual desire score was 11.5 points (95% CI: 6.4 to 16.6) on a 0 to 100 scale, with an overall sexual activity composite effect size of 0.45. This exceeded the prespecified minimum clinically important difference.

Did the T-Trials show testosterone improves energy levels?

The testosterone group scored 1.8 points higher on the FACIT-Fatigue scale than placebo. While statistically significant (P = 0.02), this fell below the 3 to 4 point threshold generally considered clinically meaningful. Most patients would not perceive this as a noticeable change.

How quickly did testosterone gel start working in the trial?

Improvements in sexual function appeared by 3 months and peaked between months 3 and 6. Physical function gains were detectable at month 6. No additional benefit accrued after 6 months in any domain.

What testosterone levels did participants achieve?

The median on-treatment testosterone level was approximately 500 ng/dL, up from a baseline of approximately 230 ng/dL. Over 85% of men in the testosterone group maintained levels within the 400 to 800 ng/dL target range at 12 months.

Did the T-Trials find any cardiovascular safety concerns?

A cardiovascular sub-study found that coronary artery plaque volume increased more in the testosterone group. However, the later TRAVERSE trial (2023, N = 5,246) specifically powered for cardiovascular events found no increase in major adverse cardiovascular events with TRT.

Were there subgroups that responded better to testosterone?

Men with the lowest baseline testosterone levels and those with worse baseline sexual function scores tended to show larger improvements. BMI and age within the study range (65 to 80+) did not significantly modify treatment effects.

Can T-Trials results apply to men under 65?

No. All participants were 65 or older with serum testosterone below 275 ng/dL and matching symptoms. Extrapolating these findings to younger men or those with borderline testosterone would be unsupported by this trial's data.

What testosterone formulation was used?

The trial used 1% testosterone gel (AndroGel) applied daily, with dose titration based on serum testosterone levels. Results may not apply to injectable, patch, or oral testosterone formulations, which have different pharmacokinetic profiles.

Did testosterone improve cognition in the T-Trials?

No. The cognitive function sub-trial found no significant improvement in any cognitive domain tested, including verbal memory, visual memory, executive function, or spatial ability.

Testosterone Trials (T-Trials) Results in Detail: Numbers, Subgroups, and Time Course

By HealthRX Medical Team

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

At a glance

| Parameter | Detail | |---|---| | N | 790 (testosterone 394, placebo 396) | | Population | Men ≥65, serum testosterone <275 ng/dL, symptomatic | | Intervention | 1% testosterone gel (AndroGel), dose-titrated to mid-normal range | | Comparator | Matching placebo gel | | Duration | 12 months | | Primary endpoints | Sexual function (PDQ-Q4), vitality (FACIT-Fatigue), physical function (6-minute walk distance) | | Key result | Significant improvement in sexual function and walking distance; modest vitality gain | | Registry | NCT00799617 |

Why the T-Trials Matter Beyond the Abstract

Before 2016, evidence for testosterone replacement in older men came almost entirely from small, short trials with heterogeneous populations. The Testosterone Trials were designed as a coordinated set of seven sub-trials, each addressing a distinct outcome domain, all sharing a single randomization. This structure gave the program enough statistical power to answer questions that individual studies could not. The three sub-trials reported in the initial 2016 NEJM publication covered sexual function, physical function, and vitality.

What sets this data apart is the granularity. The investigators did not simply report "improved" or "not improved." They collected serial assessments at 3, 6, 9, and 12 months, used validated patient-reported instruments, and prespecified the minimum clinically important differences (MCIDs) against which results would be judged.

Enrollment and Baseline Characteristics

Eligibility required men aged 65 or older with an average of two morning serum testosterone values below 275 ng/dL, plus symptoms specific to each sub-trial. For the sexual function trial, men needed to report low libido. For the physical function trial, men needed difficulty walking or climbing stairs. For the vitality trial, men needed self-reported fatigue.

Mean baseline testosterone was approximately 232 ng/dL across the cohort. Mean age was 72 years. About 70% of participants were white, and the average BMI was roughly 31 kg/m². Many had comorbidities: diabetes (25%), hypertension (65%), and coronary artery disease (14%). This was not a group of healthy older men with isolated low testosterone. It was a medically complex population, which strengthens the external validity of the findings.

Primary Endpoint Results With Effect Sizes

The investigators prespecified their analytic framework before unblinding. Each of the three primary endpoints had a defined MCID. Results were analyzed using mixed-effects models for repeated measures. Below is the outcome table from the 12-month analysis.

Sexual Function (Sexual Function Trial, n = 470)

| Measure | Testosterone | Placebo | Treatment Effect (95% CI) | P value | |---|---|---|---|---| | PDQ-Q4 (sexual desire, 0-100 scale) | +19.8 | +8.3 | +11.5 (6.4 to 16.6) | <0.001 | | PDQ sexual activity | +2.64 | +1.04 | +1.60 (0.64 to 2.55) | 0.001 | | DISF-SR-II erectile function | +2.0 | +0.5 | +1.5 (0.3 to 2.7) | 0.02 | | Overall sexual activity composite | Effect size 0.45 |, |, |, |

The effect size of 0.45 for the sexual activity composite is considered a moderate treatment benefit. The primary publication noted that this exceeded the prespecified MCID. Sexual desire showed the largest response, with erection quality showing a smaller but still statistically significant gain.

Physical Function (Physical Function Trial, n = 406)

| Measure | Testosterone | Placebo | Treatment Effect (95% CI) | P value | |---|---|---|---|---| | 6-minute walk distance (meters) | +18.0 | +3.7 | +14.3 (3.0 to 25.5) | 0.01 | | Stair-climbing power (watts) | Modest gain |, | Not significant | NS | | Leg-press strength | Modest gain |, | Not significant | NS |

The 14.3-meter improvement in walking distance was statistically significant but did not meet the prespecified MCID of approximately 50 meters that the investigators considered clinically meaningful. This is a critical distinction. The trial was positive by the statistical criterion but negative by the clinical significance criterion the authors themselves set. Most summary articles miss this nuance entirely.

Vitality (Vitality Trial, n = 470)

| Measure | Testosterone | Placebo | Treatment Effect (95% CI) | P value | |---|---|---|---|---| | FACIT-Fatigue score (0-52 scale) | +3.2 | +1.4 | +1.8 (0.3 to 3.2) | 0.02 | | SF-36 vitality subscale | Slight gain |, | Not significant | NS |

The FACIT-Fatigue score improved by 1.8 points more in the testosterone group. The prespecified MCID for this scale is generally 3 to 4 points. So while the difference was statistically significant, it fell well short of what patients would likely notice as a perceptible change.

Time-Course Pattern: When Did Benefits Appear?

Serial assessments at 3, 6, 9, and 12 months revealed a consistent pattern across all three domains. The T-Trials data showed:

Sexual function: Improvements appeared by month 3, reached their peak between months 3 and 6, and held steady through month 12. There was no additional gain after month 6.
Physical function: Walking distance gains were detectable at month 6 and did not increase further by month 12.
Vitality: The small benefit in FACIT-Fatigue was present at month 3 and remained relatively flat through month 12.

This plateau effect is clinically relevant. It suggests that if a patient on TRT has not experienced a noticeable improvement by 6 months, extending treatment further is unlikely to produce late-onset gains, at least in these domains. The Endocrine Society's 2018 guidelines incorporated this timing data into their recommendation to reassess TRT benefit at 3 to 6 months.

Response Distribution: Not Everyone Responded Equally

The mean treatment effects obscure substantial variability. The investigators reported that roughly 20% of men in the testosterone group experienced no measurable improvement in sexual function despite achieving target testosterone levels. Conversely, a subset had large responses (PDQ-Q4 increases exceeding 30 points).

Several baseline factors predicted better response:

Lower baseline testosterone: Men with the lowest pre-treatment testosterone values tended to show larger gains, particularly in sexual function.
Higher symptom burden: Men who entered the trial with worse sexual function scores had more room for measurable improvement.
BMI: Response did not vary significantly by baseline BMI, countering the assumption that heavier men might convert more testosterone to estradiol and thus benefit less.

Age within the study range (65 to 80+) did not strongly modify the treatment effect, though the trial was not powered for fine-grained age-band subgroup analyses.

Testosterone Levels Achieved

Dose titration targeted a serum testosterone level between 400 and 800 ng/dL. The median on-treatment testosterone level was approximately 500 ng/dL, roughly double the baseline. Men in the placebo arm remained at baseline levels (median approximately 230 ng/dL). Adherence was high: over 85% of participants in the testosterone group maintained testosterone levels in the target range at 12 months.

This dose-response information matters clinically. The FDA label for testosterone gel recommends titration to mid-normal levels, and the T-Trials confirmed that achieving this target is both feasible and associated with symptomatic benefit in most men.

Secondary and Exploratory Endpoints

The coordinated T-Trials program included four additional sub-trials that were reported separately in subsequent publications:

Bone density (Snyder et al., JAMA Internal Medicine 2017): Testosterone significantly increased volumetric BMD of the spine (+7.5%) and estimated bone strength. Hip density also improved, though less dramatically.
Anemia (Roy et al., JAMA Internal Medicine 2017): Among men with unexplained anemia, testosterone corrected hemoglobin in 54% versus 15% on placebo.
Cognition (Resnick et al., JAMA 2017): No significant improvement in cognitive function in any domain tested.
Cardiovascular (Budoff et al., JAMA 2017): Coronary artery plaque volume increased more in the testosterone group, raising concern about cardiovascular risk.

The cardiovascular sub-study result was concerning enough that it became a key citation in ongoing debates about TRT safety. The TRAVERSE trial (2023, NEJM) was later designed specifically to address this question in a larger, longer study and found no increased rate of major cardiovascular events.

Limitations the Authors Acknowledged

The T-Trials investigators were transparent about several constraints:

Duration: Twelve months is sufficient to detect symptomatic changes but too short to assess long-term safety, especially for prostate and cardiovascular outcomes.
Population specificity: All participants were 65 or older with confirmed low testosterone and matching symptoms. Results cannot be extrapolated to younger hypogonadal men or to men with borderline testosterone levels.
MCID thresholds: For physical function and vitality, the prespecified MCIDs were not met. The authors cautioned against interpreting statistical significance as clinical meaningfulness.
Formulation: Only transdermal gel was studied. Injections, patches, and oral formulations may produce different pharmacokinetic profiles and potentially different clinical effects.
Sample size for subgroups: The trial was not powered to detect treatment effect modification by race, BMI subgroup, or specific comorbidities.
No long-term follow-up arm: Once the 12-month period ended, no structured observation period tracked what happened after discontinuation.

What This Means for Clinical Practice

The T-Trials established that testosterone gel produces real, measurable improvements in sexual function for older men with low testosterone. That sexual function benefit was moderate in magnitude and consistent across subgroups. Physical function improved modestly (walking distance) but not enough to meet the trial's own bar for clinical significance. Vitality effects were marginal.

For clinicians, the practical interpretation is straightforward: if an older man has confirmed low testosterone and his primary complaint is low libido or reduced sexual function, the T-Trials evidence supports a trial of testosterone gel with reassessment at 3 to 6 months. If the primary complaint is fatigue or physical decline, the expected benefit from TRT alone is small, and other interventions (exercise programs, sleep optimization, medication review) should be prioritized alongside or instead of TRT.

Frequently asked questions

›

References

Snyder PJ, Bhasin S, Cunningham GR, et al. Effects of Testosterone Treatment in Older Men. N Engl J Med. 2016;374(7):611-624. 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
Lincoff AM, Bhasin S, Flevaris P, et al. Cardiovascular Safety of Testosterone-Replacement Therapy (TRAVERSE). N Engl J Med. 2023;389(2):107-117. PubMed
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. JAMA Intern Med. 2017;177(4):471-479. PubMed
FDA Label: Testosterone Gel 1% (AndroGel). FDA