The TRT Trials

Testosterone therapy is one of the most heavily marketed medical products in the world — and, for once, one of the better studied. The modern record runs from the coordinated Testosterone Trials of 2016 through the 5,246-man TRAVERSE cardiovascular safety trial of 2023, with meta-analyses in between. This page goes outcome by outcome: what the therapy demonstrably does, what it demonstrably does not, and why every result carries the same footnote about who was actually enrolled.

The Trial Record in One Paragraph

The modern benchmark is the Testosterone Trials (Snyder et al., NEJM, 2016): seven coordinated, placebo-controlled trials in 788 men aged sixty-five and older with consistently low testosterone (under 275 ng/dL on repeated morning draws) and symptoms, treated with gel for one year. The results set the honest expectation bar for the whole category: sexual function improved consistently but modestly; the dedicated vitality trial's primary endpoint — fatigue — was not met; mood improved only in the subgroup that started depressed; physical function improved only in a pooled secondary analysis; bone density rose about 3 percent with no fracture data; anemia corrected in roughly half of anemic men. Meta-analyses and the later TRAVERSE safety trial refined the picture without overturning it. Everything on this page inherits one framing fact: these men had low testosterone plus symptoms. Nothing here transfers to a man with normal levels buying "optimization."

Outcome by Outcome: The Testosterone Trials

The table below is the granular ledger, with effect sizes described the way the papers describe them — modest, subgrouped, and sometimes null. Where the parent topic summarized, this page itemizes.

Read the ledger for its shape, not any single row: the reliable effects are sexual function, lean mass, and blood — the things testosterone directly controls. The things the marketing leads with — energy, drive, sharpness — are precisely the rows where the trials came up empty or subgrouped. The effect sizes sit in the same ballpark as what progressive resistance training does for muscle and strength without a prescription, without lab monitoring, and without the fertility cost — a comparison the muscle-aging topic develops in detail.

The Pattern: Benefit Tracks Baseline Deficiency

Across trials and meta-analyses, the same curve repeats: the lower the starting testosterone, the larger the response. Dose-response studies in healthy men showed lean-mass gains scale with the dose administered (Bhasin et al., JCEM, 2005), and trial subanalyses find sexual-function gains concentrated in men whose levels were genuinely low at baseline. The mechanism is unglamorous physiology: a hormone does more when there is a deficit to correct. The corollary is the one the industry omits: no randomized trial demonstrates a meaningful benefit in men whose testosterone is already normal. Normal-range "optimization" is a market position, not a treatment indication — and a clinic whose business model is prescribing to normal-range men is displaying its incentive structure in public.

TRAVERSE: The Safety Chapter

TRAVERSE randomized 5,246 men aged 45 to 80 with hypogonadism (symptoms plus two fasting testosterone values under 300 ng/dL) and pre-existing or high-risk cardiovascular disease to testosterone gel or placebo, with doses adjusted to a target of 350–750 ng/dL (Lincoff et al., NEJM, 2023). Treatment ran a mean of 21.7 months, follow-up 33. The primary result is genuine, useful reassurance: major cardiac events — cardiovascular death, heart attack, or stroke — occurred in 7.0 percent of treated men versus 7.3 percent of placebo, a hazard ratio of 0.96 that met the pre-specified non-inferiority bar. The secondary signals all leaned one direction, though each absolute difference was small:

Two honest glosses. First, the ceiling: two years of follow-up answers the short-term question well and the lifetime question not at all — the full risk accounting belongs to the risks and monitoring page, which also covers the earlier observational alarms (the 2013 claims of heart-attack risk) that TRAVERSE was built to settle. Second, the prostate footnote: prostate cancer appeared numerically less often in the testosterone arm, a finding the authors themselves cautioned not to over-read. An individual-patient meta-analysis across 35 studies adds a similar-grained reassurance: no increase in short-to-medium-term cardiovascular events, and deaths were fewer in treated men — six of 1,621 versus twelve in the placebo group (Hudson et al., Lancet Healthy Longevity, 2022). Small numbers, real signal, still not a lifetime of data.

What the Trials Cannot Tell You

• 🦴 Fractures. Bone density rose, but no trial has shown fewer broken hips — density is a surrogate, not the outcome that matters.
• 🕰️ Decades of exposure. The longest safety follow-up is measured in a few years; a 50-year-old starting therapy is committing to data that do not yet exist.
• 📈 Normal-range men. Every effect above was measured in hypogonadal men. For a man whose levels are fine, the trial record is silent — and silence is not support.
• 🧠 Cognition and "drive." The cognitive trials found little; the subjective "edge" anecdotes have no controlled evidence behind them.
• ⚖️ Whether any of it is worth it for you. Effect sizes are group averages; the individual decision — modest gains versus real monitoring and the fertility trade-off — is a clinician's conversation, not a trial's conclusion.

Questions, Answered Briefly

• ⚡ Will TRT fix my energy? The dedicated trial says no, on average: the vitality endpoint was missed. Fatigue has better-evidenced owners — sleep, ferritin, mood — worth ruling out first.
• 🔥 What improves most reliably? Sexual desire and activity, followed by lean mass and hemoglobin. If those are not the problems, expectations should drop accordingly.
• 💪 Isn't TRT just anabolic anyway? In part — roughly 1–2 kg of lean mass and small strength gains in hypogonadal men, effects comparable in size to a year of decent resistance training without the prescription.
• 🫀 Is it safe for the heart? Short term, TRAVERSE says non-inferior for major events with small directional signals elsewhere; long term is genuinely unknown. The risks page counts it all.
• 🧾 What if my levels are normal but I still want it? Then the trial record has nothing for you — no demonstrated benefit, unchanged risks. A clinic that prescribes anyway is telling you something about its business model.

Related Topics