TRT for Men Over 50

TRAVERSE cleared testosterone on heart attacks and strokes. The same trial reported higher rates of atrial fibrillation, acute kidney injury and pulmonary embolism. At Longevity Benchmark, we review the published clinical trials to separate the measured health outcomes of testosterone replacement therapy (TRT) for men over 50 from commercial marketing claims. Bone mineral density rose 6.8% in one trial while clinical fractures rose 43% in a larger one, and the two findings have never been reconciled.

A man evaluating testosterone therapy after 50 encounters two contradictory arguments: that hormone replacement is essential preventive care, or that it inevitably triggers cardiovascular events. The published trials support neither extreme. Large placebo-controlled studies show defined physiological changes in men with diagnosed hypogonadism, alongside distinct boundaries where the therapy produces no measurable benefit.

Measured Rates of Age-Related Testosterone Decline

Testosterone concentrations decline gradually in aging men, but researchers disagree on whether the drop is a steady annual loss or a change restricted to late life. The Massachusetts Male Aging Study (Feldman et al., 2002) evaluated 1,709 men aged 40 to 70 at baseline and followed 1,156 participants across 7 to 10 years. That cohort showed a cross-sectional total testosterone decline of 0.8% per year of age, while the longitudinal within-subject decline reached 1.6% annually. Free and bioavailable testosterone fell by 2% to 3% each year, while sex hormone-binding globulin increased by 1.6% per year.

General health status influences this trajectory substantially. In the Massachusetts analysis, the absence of chronic illness, prescription medications, obesity, or excessive alcohol intake added 10% to 15% to androgen concentrations, leading the investigators to conclude that incident poor health accelerates the drop. Data from 890 men in the Baltimore Longitudinal Study of Aging (Harman et al., 2001) showed an average longitudinal decrease in total testosterone of 0.124 nmol/L per year (roughly 3.6 ng/dL). Based on total testosterone cut-offs, hypogonadal concentrations occurred in approximately 20% of men over 60, 30% over 70, and 50% over 80.

An individual-participant-data meta-analysis from the Androgens In Men Study (Marriott et al., Annals of Internal Medicine, 2023) contested this model by pooling cohorts in which testosterone was measured by mass spectrometry. The investigators documented a non-linear relationship with age: testosterone concentrations remained largely unchanged from age 17 to 70, with decline becoming evident only after age 70. This finding contradicts the traditional estimate of a steady 1% annual drop after age 30, and with it the andropause framing built on that estimate. The difference is likely explained by the superior specificity of mass spectrometry over older immunoassays and differing baseline health across cohorts. A companion analysis from the same programme (Yeap et al., 2024), covering 255,830 participant-years, saw higher all-cause mortality below 7.4 nmol/L (213 ng/dL) and higher cardiovascular mortality below 5.3 nmol/L (153 ng/dL). These are observational associations. Nothing in that paper tests whether raising the level lowers the risk.

Diagnostic Criteria and Guideline Thresholds

Major medical guidelines require repeated morning fasting laboratory tests combined with persistent clinical symptoms before confirming a diagnosis of testosterone deficiency. The American Urological Association guideline recommends using a total testosterone value below 300 ng/dL as a diagnostic cut-off, confirmed by two separate early morning blood draws. The AUA also issues a strong recommendation to measure serum luteinizing hormone (LH) to distinguish primary testicular failure from secondary hypothalamic or pituitary dysfunction.

Clinical guidelines do not share a single diagnostic threshold. The European Male Ageing Study evaluated 3,369 men aged 40 to 79 and identified only three sexual symptoms that had a genuine syndromic relationship with low testosterone: poor morning erections, low sexual desire, and erectile dysfunction. The European study defined late-onset hypogonadism as the presence of all three sexual symptoms alongside total testosterone below 11 nmol/L (320 ng/dL) and free testosterone below 220 pmol/L. Conversely, the Endocrine Society avoids setting a single universal numeric threshold, recommending confirmation through assay-specific reference ranges and equilibrium dialysis testing for free testosterone when total values sit near the borderline.

Target concentrations during treatment also vary across organizations. The AUA guideline directs clinicians to adjust dosing toward the middle tertile of the normal reference range, which it defines as 450 to 600 ng/dL, with follow-up monitoring every 6 to 12 months. The Endocrine Society advises aiming for mid-normal concentrations without prescribing an exact numerical interval. Our detailed guide on free versus total testosterone breaks down how protein binding alters these lab measurements.

Outcomes Measured in the Testosterone Trials

The national Testosterone Trials established that androgen therapy produces modest improvements in sexual function and correctable anaemia while failing to alter vitality, walking speed, or cognitive performance. Published in the New England Journal of Medicine (Snyder et al., 2016), the primary program enrolled 790 men aged 65 or older with symptomatic hypogonadism and confirmed baseline serum testosterone below 275 ng/dL. Participants received daily transdermal testosterone gel or placebo gel across 12 academic medical centres for exactly one year.

The primary efficacy domains produced distinctly uneven results. Sexual function showed the most reliable benefit: sexual activity scores increased significantly, erectile function scores improved by 2.64 points, and sexual desire scores rose by 2.93 points. In contrast, the physical function trial failed to meet its primary endpoint, as the proportion of men increasing their 6-minute walking distance by at least 50 metres did not differ significantly between groups. Walking distance showed a statistically significant difference of 6.69 metres only when investigators pooled all 790 participants across all sub-trials. Vitality assessments on the FACIT-Fatigue scale showed no statistically significant difference between active drug and placebo, and depressive symptoms improved by only 0.72 points on a 27-point depression inventory.

Cardiovascular Findings From the TRAVERSE Study

The largest randomized safety trial on testosterone therapy satisfied its non-inferiority margin for major cardiovascular events over an average exposure of under two years. The TRAVERSE trial (Lincoff et al., 2023) enrolled 5,246 men aged 45 to 80 with documented cardiovascular disease or multiple cardiovascular risk factors, clinical hypogonadal symptoms, and two consecutive fasting testosterone levels under 300 ng/dL. Participants received daily 1.62% transdermal testosterone gel adjusted to maintain levels between 350 and 750 ng/dL or matching placebo gel.

The trial design incorporated a broad non-inferiority safety margin. Non-inferiority required an upper limit of less than 1.5 for the 95% confidence interval of the hazard ratio for major adverse cardiovascular events (cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke). The composite primary endpoint occurred in 182 patients (7.0%) receiving testosterone and 190 patients (7.3%) receiving placebo, yielding a hazard ratio of 0.96 (95% CI 0.78 to 1.17, P<0.001 for non-inferiority). This confirmed that transdermal testosterone did not cause a 50% or greater relative increase in major cardiovascular events.

Trial duration and treatment retention represent important interpretive boundaries. Mean active treatment duration was 21.7 months, and mean total follow-up was 33.0 months, meaning participants received medication for less than two years in a clinical setting where therapy is often prescribed for decades. Discontinuation rates exceeded 61% in both treatment arms, and 18% of enrolled patients were lost to follow-up. An aggregate meta-analysis by Hudson et al. in Lancet Healthy Longevity (2022) pooled 35 trials covering 5,601 participants and similarly identified no short-term cardiovascular risk increase (odds ratio 1.07), but noted that the mean trial duration across the medical literature was only 9.5 months.

Observed Secondary Safety Signals

Secondary safety analyses in the TRAVERSE study documented higher observed incidences of heart rhythm disorders, venous blood clots, and renal impairment in men assigned to active treatment. The published trial report in the New England Journal of Medicine identified an increased incidence of atrial fibrillation (3.5% on testosterone versus 2.4% on placebo), acute kidney injury (2.3% versus 1.5%), and pulmonary embolism (0.9% versus 0.5%). Non-fatal arrhythmias requiring medical intervention occurred in 5.2% of men on testosterone compared with 3.3% on placebo.

The trial was statistically powered for its primary composite cardiovascular endpoint, not for these individual secondary outcomes, meaning the findings represent observed numerical patterns rather than proven causal mechanisms. A post-hoc analysis by Pencina et al. (2025) found that COVID-19 infection during the trial was strongly associated with acute kidney injury (hazard ratio 5.38). Among participants who contracted COVID-19, rates of kidney injury, atrial fibrillation, and venous thromboembolism showed no meaningful differences between the testosterone and placebo cohorts. Whether the kidney injury signal reflects a direct pharmacological action or an artifact of the viral pandemic remains an open scientific question.

Blood pressure changes on active therapy show consistent modest increases across trials. Per the AndroGel 1% prescribing information, mean systolic blood pressure rose by 1.0 mmHg at 36 months in the TRAVERSE trial while decreasing by 0.5 mmHg in the placebo arm, creating a net difference of 1.5 mmHg. In dedicated ambulatory blood pressure monitoring studies, 16 weeks of 1.62% gel increased mean systolic blood pressure by 3.0 mmHg and diastolic pressure by 2.2 mmHg in hypertensive men. Clinical reviews such as our summary of TRT side effects and risks outline the vascular implications of these sustained blood pressure shifts.

Structural Changes Versus Clinical Endpoints

Clinical trials on testosterone therapy have repeatedly exposed contradictions between surrogate biological markers and actual clinical outcomes. In bone health, the Testosterone Trials bone substudy (Snyder et al., 2017) demonstrated that one year of testosterone increased spine trabecular volumetric bone mineral density by 6.8% and estimated bone strength by 8.5%. However, the TRAVERSE fracture substudy (Snyder et al., 2024), which followed 5,204 men for a median of 3.19 years, found that clinical fractures occurred in 91 participants (3.50%) on testosterone versus 64 participants (2.46%) on placebo. This yielded a hazard ratio of 1.43 (95% CI 1.04 to 1.97), revealing that clinical fracture rates rose despite the predictable increase in bone mineral density.

A second divergence appeared between coronary plaque imaging and cardiovascular events. In the Testosterone Trials cardiovascular study (Budoff et al., 2017), 138 men underwent computed tomography angiography, which showed that testosterone significantly accelerated coronary artery plaque formation. Non-calcified plaque volume expanded by an estimated difference of 41 cubic millimetres compared with placebo over 12 months. Six years later, the TRAVERSE trial showed no excess of clinical myocardial infarctions or cardiovascular deaths over 33 months of observation. Surrogate imaging indicators and hard clinical endpoints have diverged in both skeletal and vascular systems.

Clinical Monitoring Guidelines and Threshold Disagreements

Medical organizations agree on the need for baseline blood testing before starting testosterone, but they maintain conflicting guidelines regarding hematocrit safety cut-offs. The official FDA prescribing information for AndroGel instructs clinicians to check hematocrit prior to initiation, re-evaluate at 3 to 6 months, and repeat testing annually thereafter, stopping treatment if red blood cell volume reaches an elevated concentration.

The definition of what constitutes an elevated hematocrit varies by medical body. The Endocrine Society defines drug-induced erythrocytosis as a hematocrit exceeding 54% (or 50% for men living at high altitude) and advises clinicians against starting therapy when baseline hematocrit sits above normal limits. In contrast, the American Urological Association guideline sets no numeric hematocrit threshold whatsoever, requiring only that clinicians measure baseline haemoglobin and hematocrit and inform patients of polycythaemia risks. Attributing the 54% threshold to the AUA is incorrect. Our focused guide to high hematocrit on TRT explains how plasma volume fluctuations affect these readings.

Prostate Safety and Trial Exclusion Criteria

Available trial evidence has not demonstrated an increased rate of high-grade prostate cancer among men treated with testosterone, but trial enrollment criteria intentionally excluded high-risk patients. In the TRAVERSE prostate safety substudy (Bhasin et al., 2023), adjudicated high-grade prostate cancer (Gleason score 4+3 or higher) occurred in 5 of 2,596 men (0.19%) on testosterone compared with 3 of 2,602 men (0.12%) on placebo, producing a hazard ratio of 1.62 (95% CI 0.39 to 6.77). Incidences of total prostate cancer, acute urinary retention, invasive prostate surgeries, and changes on the International Prostate Symptom Score did not differ significantly between study arms.

The study population was selected to minimize oncologic risk. TRAVERSE excluded all men with a baseline PSA exceeding 3.0 ng/mL or an International Prostate Symptom Score above 19, and the average baseline PSA of enrolled participants was only 0.92 ng/mL. The broad confidence interval for high-grade disease (0.39 to 6.77) leaves open possibilities ranging from a substantial risk reduction to a six-fold increase. The AUA guideline maintains that clinicians must inform hypogonadal men with a history of prostate cancer that existing medical evidence is inadequate to quantify the risk-benefit balance.

Regulatory Status and FDA Labeling Changes

In February 2025, the Food and Drug Administration issued class-wide labeling changes for prescription testosterone products, incorporating TRAVERSE trial findings while maintaining formal limitations of use. The FDA regulatory announcement directed manufacturers to remove language regarding general adverse cardiovascular risk from product warnings, add specific data from post-marketing ambulatory blood pressure trials, and integrate TRAVERSE primary cardiovascular outcomes into product packaging.

The FDA did not expand the indications for testosterone products to include normal aging. The revised prescribing information for AndroGel 1% retains its explicit Limitation of Use: safety and efficacy have not been established in men with age-related hypogonadism. Furthermore, the product Boxed Warning is dedicated solely to the risk of secondary exposure in women and children through dermal contact, not to cardiovascular events. European regulators reached a comparable stance earlier: in 2014, the European Medicines Agency completed a review finding no consistent evidence of cardiac risk with testosterone in hormone-deficient men, declining to place cardiovascular restrictions on the product class.

Unresolved Questions and Evidence Limits

Clinical trials have never evaluated the safety or efficacy of testosterone therapy in older men with normal baseline hormone levels, nor have studies established the health impact of treatment extending beyond three years. Every major trial, including TRAVERSE (mean active exposure 21.7 months) and the Testosterone Trials (12 months of exposure), required documented hypogonadism with repeated low morning blood measurements and clinical symptoms before enrollment.

Delivery formulations also limit how broadly trial results can be applied. The TRAVERSE trial investigated a single product formulation: transdermal 1.62% testosterone gel. Its findings cannot be directly applied to short-acting intramuscular injections, long-acting subcutaneous pellets, or oral testosterone undecanoate, which generate differing pharmacokinetic curves, peak hormone concentrations, and rates of red blood cell production. The trial record for TRT for men over 50 covers one gel, one to three years, and men who had two morning readings under 300 ng/dL. That is the boundary any claim made about it should sit inside.

Documented Reasons Not to Start

Both major guidelines list conditions under which testosterone should not be started at all, and most of them have nothing to do with cardiovascular risk. The Endocrine Society advises against initiating therapy in men with breast or prostate cancer, a palpable prostate nodule or induration, hematocrit above the upper limit of normal, severe lower urinary tract symptoms, uncontrolled heart failure, a myocardial infarction or stroke within the previous six months, thrombophilia, untreated severe obstructive sleep apnoea, or plans for fatherhood in the near term.

Sleep apnoea carries a second warning on the product label itself. The AndroGel prescribing information states that treating hypogonadal men with testosterone may potentiate sleep apnoea, especially in men with obesity or chronic lung disease. Fertility is the other reason a man over 50 may still be excluded: AUA Statement 16 requires the long-term effect on sperm production to be discussed with any man interested in future fertility, and Statement 23 states that testosterone should not be prescribed to men currently trying to conceive.

Related

• TRT in your 20s and 30s
• Signs of Low Testosterone
• TRT Side Effects and Risks
• High Hematocrit on TRT
• Free vs Total Testosterone
• Physician-Led Hormone Clinics