What Happened After the Testosterone Trials Ended, and Did the Benefits Last?

At a glance

Parameter	Detail
N	790 men aged 65+ with low testosterone (<275 ng/dL)
Intervention	Transdermal testosterone gel (AndroGel 1.62%) titrated to mid-normal range
Comparator	Placebo gel
Duration	12 months (with pre-specified substudies extending observation)
Primary endpoints	Sexual function (PDQ-Q4), vitality (FACIT-Fatigue), physical function (6-minute walk)
Key result	Significant improvement in sexual function and 6-minute walk distance; no vitality benefit on primary measure

The Original Trial in Brief

The Testosterone Trials enrolled 790 community-dwelling men 65 years and older across 12 U.S. academic centers between 2010 and 2014. Each participant had a confirmed serum testosterone below 275 ng/dL on two morning draws plus a relevant symptom (low libido, fatigue, or reduced mobility). The trial was coordinated by the University of Pennsylvania and funded by the National Institute on Aging, with Abbvie supplying testosterone gel but holding no role in analysis or publication.

Participants were randomized 1:1 to testosterone gel or matching placebo for one year. Testosterone doses were adjusted to maintain serum levels between 400 and 800 ng/dL. The trial was actually seven overlapping sub-trials (Sexual Function, Physical Function, Vitality, Cognitive Function, Anemia, Bone, and Cardiovascular), all sharing a single enrollment pool.

At 12 months, the primary publication reported that testosterone significantly improved sexual desire (PDQ-Q4 score difference 0.58, p <0.001), erectile function, and sexual activity. The 6-minute walk test improved modestly (an additional 6.4 meters, p = 0.04 by one analytic approach). Vitality, measured by FACIT-Fatigue, showed improvement only on a secondary measure (SF-36 vitality subscale), missing its co-primary endpoint.

What the Extension Substudies Found

Bone Mineral Density: The Clearest Durable Signal

The Bone substudy (N = 211) published in JAMA Internal Medicine 2017 showed that one year of testosterone gel increased volumetric bone mineral density (vBMD) at the spine by 7.5% and estimated bone strength by 10.8% compared to placebo, measured by quantitative CT. Trabecular vBMD at the hip increased by 3.2%.

These are not trivial magnitudes. For comparison, oral bisphosphonates typically produce 3-5% gains in areal BMD at the spine over 12 months. The T-Trials bone data suggested testosterone's skeletal effect in hypogonadal older men rivals or exceeds first-line osteoporosis agents in magnitude, though fracture endpoints were not powered.

No formal post-discontinuation bone follow-up was conducted within TTrials itself. However, observational data from testosterone cessation studies indicate that bone gains from TRT attenuate within 12-24 months of stopping. The clinical implication: testosterone's bone benefit likely requires ongoing treatment, similar to denosumab's rebound pattern.

Cardiovascular Substudy: The Early Warning

The Cardiovascular substudy (N = 170) used coronary CT angiography to measure non-calcified plaque volume. Published in JAMA 2017, it found that testosterone-treated men had a significantly greater increase in coronary artery plaque volume compared to placebo (estimated difference 41 mm³ vs. 14 mm³; between-group difference p = 0.003).

This was not a clinical events finding. No participant had a confirmed MI attributable to plaque progression during the trial. But the result was alarming enough to trigger an FDA-mandated large cardiovascular outcomes trial, which became TRAVERSE.

Cognitive Function: No Lasting Signal

The Cognitive Function substudy tested whether testosterone improved age-associated memory impairment. Results published in JAMA 2017 showed no significant improvement in verbal memory (the primary endpoint) or any other cognitive domain at 12 months. With no benefit detected during active treatment, there was no rationale for durability analysis.

Anemia Substudy: Erythropoietic Response

The Anemia substudy reported correction of unexplained anemia in 54% of testosterone-treated men vs. 15% on placebo. Hemoglobin rose by approximately 1 g/dL. This erythropoietic effect is pharmacologically expected and reversible upon discontinuation. Testosterone stimulates erythropoietin and directly activates erythroid progenitors; both pathways return to baseline within 3-6 months of cessation.

Regression to Baseline After Discontinuation

The T-Trials protocol did not include a formal off-treatment follow-up period. However, several lines of evidence indicate that symptomatic benefits were not durable:

Sexual function. The PDQ improvement was tightly correlated with on-treatment serum testosterone levels. Men whose levels fell below 300 ng/dL during dose titration showed attenuated sexual function gains. Observational registry data (European Male Ageing Study, Hypogonadism Registry) consistently show that sexual function returns to baseline within 3-6 months of TRT discontinuation.

Physical function. The 6.4-meter walk distance gain was modest and at the lower boundary of clinical significance. No mechanism (e.g., sustained muscle hypertrophy) would maintain this without ongoing testosterone exposure. Lean mass gains from TRT in older men (typically 1-2 kg) reverse within 6 months of stopping.

Vitality. Since the primary vitality endpoint was not met even during active treatment, durability is moot.

The overarching lesson: TRT in older hypogonadal men functions as replacement therapy (ongoing benefit during ongoing treatment), not as a restorative intervention that permanently resets physiology.

TRAVERSE: The Long-Term Safety Answer the T-Trials Could Not Provide

The coronary plaque finding from the T-Trials CV substudy made it impossible to prescribe testosterone to older men without resolving the safety question. The FDA required a large, event-driven cardiovascular outcomes trial.

TRAVERSE (Testosterone Replacement Therapy for Assessment of Long-Term Vascular Events and Efficacy Response in Hypogonadal Men) enrolled 5,246 men aged 45-80 with hypogonadism and pre-existing cardiovascular disease or high cardiovascular risk. Published in the New England Journal of Medicine in 2023, it found:

• Primary composite MACE (CV death, nonfatal MI, nonfatal stroke): hazard ratio 0.99 (95% CI 0.81-1.21). Non-inferior to placebo.
• No excess MI, stroke, or CV death individually.
• Confirmed higher rates of atrial fibrillation, pulmonary embolism, and acute kidney injury in the testosterone arm, though absolute rates were low.

TRAVERSE essentially cleared testosterone of the major cardiovascular harm signal that the T-Trials plaque data had raised. The non-calcified plaque progression seen in the CV substudy did not translate into excess clinical events over a median 33-month follow-up. This resolved, for most clinicians, the decade-long uncertainty about TRT and heart disease.

What the T-Trials Changed in Practice Guidelines

The Endocrine Society 2018 guideline update cited the T-Trials extensively. Key shifts attributable to TTrials data:

1. Acknowledged that men 65+ with confirmed hypogonadism can be offered TRT after individual risk-benefit discussion (previously guideline-level uncertainty).
2. Recommended discussing with patients that sexual function benefits require ongoing treatment.
3. Flagged the coronary plaque finding as requiring resolution (now resolved by TRAVERSE).
4. Did not endorse TRT for cognitive improvement, citing the negative cognitive substudy.

The American Urological Association 2018 guideline similarly cited TTrials as evidence that symptomatic low testosterone in older men is a treatable condition, not simply "aging."

Limitations the Authors Acknowledged

The original publication was explicit about several constraints:

• Duration. Twelve months cannot detect slow-developing harms (prostate cancer, cardiovascular events) or demonstrate whether benefits persist.
• Sample size. N=790 was powered for continuous symptomatic endpoints, not for rare clinical events.
• Selection bias. Participants were community-dwelling volunteers from academic centers, not representative of all older hypogonadal men.
• No discontinuation arm. The trial could not determine benefit durability.
• Plaque finding limitations. The CV substudy (N=170) was small and used a surrogate imaging endpoint, not adjudicated clinical events.

Clinical Translation: What This Means for Prescribers Today

The T-Trials plus TRAVERSE together establish that testosterone gel in older hypogonadal men:

• Reliably improves sexual desire, erectile function, and sexual activity while on treatment
• Produces a modest physical function benefit insufficient to justify TRT for mobility alone
• Does not improve cognition or prevent cognitive decline
• Builds meaningful bone density that likely requires continued treatment to maintain
• Does not cause excess MACE when prescribed to men with pre-existing CV risk (per TRAVERSE)
• Functions as chronic replacement therapy, not a one-time intervention

The FDA's updated labeling for testosterone products now reflects this evidence base, requiring disclosure of cardiovascular monitoring recommendations without a boxed warning for MACE.

Result Summary Table

Frequently asked questions

›Did the T-Trials have a formal extension phase beyond 12 months?

No. The protocol ended at 12 months for all seven substudies. There was no funded extension or off-treatment observation period. Post-trial durability data comes from separate observational cohorts and the later TRAVERSE trial.

›How long do sexual function improvements last after stopping testosterone?

Based on registry data and pharmacokinetic modeling, sexual function improvements from testosterone gel return to pre-treatment baseline within 3-6 months of discontinuation. Serum testosterone falls to pre-treatment levels within 1-2 weeks of stopping gel application.

›Was the coronary plaque finding from the T-Trials confirmed by later research?

The imaging finding was real but did not translate to clinical harm. TRAVERSE (N=5,246, median 33 months) showed a MACE hazard ratio of 0.99, indicating no excess cardiovascular events from testosterone treatment in men with pre-existing CV risk.

›Do bone density gains from testosterone persist after stopping treatment?

Likely not beyond 12-24 months. While the T-Trials did not measure post-discontinuation bone density, the mechanism (suppression of bone resorption plus stimulation of formation) is treatment-dependent. Analogous to denosumab, stopping testosterone removes the anti-resorptive signal.

›Why didn't the T-Trials detect a vitality benefit?

The primary vitality measure (FACIT-Fatigue) may have been too insensitive. A secondary measure (SF-36 vitality subscale) did show improvement. The authors noted that "vitality" is difficult to operationalize and the enrolled population may not have had sufficient fatigue at baseline to detect change.

›What testosterone levels were achieved in the treatment arm?

The gel dose was titrated to achieve serum testosterone between 400-800 ng/dL (mid-normal for young men). Median achieved level was approximately 500 ng/dL. Men who did not reach target levels had attenuated responses across all endpoints.

›Did TRAVERSE change prescribing guidelines for older men?

Yes. After TRAVERSE confirmed cardiovascular safety in 2023, both the Endocrine Society and AUA acknowledged that the prior cardiovascular concern (driven partly by the T-Trials plaque data) was resolved. TRT can be offered to older hypogonadal men with CV risk factors after shared decision-making.

›Is there evidence that longer TRT duration produces greater benefits?

Observational registries (Hypogonadism Registry, RHYME) suggest that sexual function and body composition benefits plateau by 6-12 months and remain stable with continued treatment. Bone density may continue accruing through 24-36 months. There is no evidence of increasing returns beyond 3 years.

›Were there any cancer signals in the T-Trials or its follow-up?

The T-Trials reported no prostate cancer excess at 12 months, but the trial was too short and small to detect this. TRAVERSE, over 33 months, found no significant increase in prostate cancer incidence (HR 0.95), though PSA monitoring remains standard of care per FDA labeling.

›What happened to the anemia correction after testosterone was stopped?

Hemoglobin returns to pre-treatment levels within 3-6 months of discontinuation. The erythropoietic effect of testosterone is pharmacologically maintained, not curative. This is consistent with the mechanism: testosterone stimulates EPO and directly activates erythroid progenitors, both of which cease without ongoing androgen exposure.

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. PubMed
2. 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
3. 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
4. Lincoff AM, Bhasin S, Flevaris P, et al. Cardiovascular Safety of Testosterone-Replacement Therapy. N Engl J Med. 2023;389(2):107-117. PubMed
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. PubMed
6. FDA Testosterone Products Labeling. AccessData