TRT and Longevity: What the Trials Actually Show

The one-sentence version

Testosterone replacement is a real treatment with real evidence behind it — for diagnosed hypogonadism, which is a specific condition confirmed by symptoms plus two low morning blood tests. There is no trial showing it extends life, no trial showing it slows aging, and the largest guideline in the field recommends against prescribing it for age-related decline alone¹. Almost everything sold as "testosterone optimization" lives in the gap between those two sentences.

What testosterone therapy is actually approved to do

The FDA-approved indication is replacement in men with low testosterone caused by a recognized medical condition — testicular, pituitary or hypothalamic. The label for testosterone cypionate carries an explicit limitation of use: safety and efficacy in men with age-related low testosterone have not been established².

The Endocrine Society's clinical practice guideline draws the same line and adds the diagnostic bar: diagnose hypogonadism only in men with symptoms and signs plus unequivocally low morning total testosterone confirmed on a repeat measurement, and do not prescribe on a single reading or on symptoms alone¹. A single afternoon test, which is what several telehealth intakes run on, is not a diagnosis.

What the big trials found

TRAVERSE (2023) — the safety question, answered. Just over 5,200 men aged 45–80 with hypogonadism and either existing cardiovascular disease or high risk were randomized to testosterone gel or placebo. Testosterone was noninferior to placebo for major adverse cardiac events³. This was the trial the field had been waiting on since the FDA's 2015 warning, and it is genuinely reassuring — for that population. It is a safety result, not a benefit result, and it says nothing about men whose levels are normal.

The Testosterone Trials (2016) — the benefit question, answered narrowly. In 790 men over 65 with genuinely low levels, testosterone produced a consistent improvement in sexual function, a smaller improvement in mood and depressive symptoms, and no significant benefit for vitality or walking distance⁴. Read that list again: the things it improved are real, and they are not longevity.

Fractures (2024) — the result nobody markets. A pre-specified TRAVERSE sub-study of 5,204 men found more clinical fractures in the testosterone group than in the placebo group⁵. It is an unexpected finding in a trial designed for cardiac safety, and it cuts directly against the "stronger bones" pitch that appears on clinic pages.

Why "optimization" is a different product from replacement

The marketing category that has grown up around TRT is not the treatment the trials studied. It targets men whose testosterone is within the reference range but toward the lower end, offers a number to raise rather than a symptom to treat, and measures success as the number moving.

Three things follow, and none of them are controversial:

1. 1.The trials do not cover it. Every result above was measured in men with diagnosed hypogonadism. Extending them to men with normal levels is an extrapolation, not evidence.
2. 2.The endpoints are surrogate. Serum testosterone is a biomarker. Nothing in the literature establishes that raising it in a man who is not deficient changes how long or how well he lives.
3. 3.The therapy is not trivially reversible. Exogenous testosterone suppresses the body's own production and impairs fertility, which the label states plainly². That is a manageable trade-off for a diagnosed deficiency and a substantial one for a lifestyle purchase.

This is the same shape as the hormone story in women, where two decades of confident practice was reversed by a single large randomized trial: the Women's Health Initiative found estrogen plus progestin increased the risk of coronary events, stroke and invasive breast cancer in healthy postmenopausal women, and the trial was stopped early⁶. Observational data had pointed the other way for years. Hormones are exactly the domain where mechanism and cohort studies have most often been wrong — the full account, including what the timing hypothesis did and did not rescue, is in HRT, menopause and longevity.

What actually predicts how long you live

If the goal is longevity rather than a lab value, the strongest human evidence in this area is not pharmacological.

Cardiorespiratory fitness shows a graded, dose-dependent association with survival across 122,007 patients — with no observed upper limit of benefit, and with the difference between the lowest-fit and the elite groups larger than the difference associated with smoking, diabetes or coronary artery disease⁷. Grip strength, measured across 139,691 adults in 17 countries, predicted all-cause and cardiovascular mortality more strongly than systolic blood pressure⁸.

Neither is a treatment you can buy, which is precisely why they are undersold. We cover them in VO2 max and longevity and grip strength and longevity.

If you are going to buy it anyway, buy it properly

A defensible testosterone purchase looks like this, and the checklist is short:

• •A real diagnosis first. Two morning blood tests, plus symptoms. Anyone willing to prescribe on one afternoon reading is skipping the step the guideline calls essential¹.
• •The price you can actually pay monthly. This category advertises twelve-month rates in large type and month-to-month rates in the asterisk. Our provider ranking grades every row on whether the published figure is the one you are charged.
• •A named prescriber and a named pharmacy. Both are knowable before you pay, and most of the field publishes neither.
• •Follow-up bloodwork inside the price, not sold separately — hematocrit and PSA monitoring are part of the standard of care, not an upsell.

The honest summary

Testosterone therapy treats hypogonadism, and for men who have it the evidence is decent: better sexual function, acceptable cardiovascular safety, an unexplained fracture signal worth discussing with a prescriber. For everyone else it is an intervention with a surrogate endpoint, a suppression risk, and no longevity evidence at all.

That is not an argument against it. It is an argument for knowing which of the two products you are being sold — and the graded provider ranking exists to make the second one legible.