Toward Health.
In your inbox.

Testosterone Therapy: Benefits, Heart and Prostate Safety, and What We Monitor

When low testosterone and symptoms persist, what does treatment accomplish, and how safe is it? We examine the randomized evidence on heart safety, prostate risk, and what clinicians must monitor.

In our earlier article, Low-Carbohydrate Diets and Testosterone: What the Evidence Shows, we explained why obesity, insulin resistance, weight loss, sleep, and nutrition belong in a testosterone evaluation. Weight loss can raise testosterone in men with obesity-related low levels, while carbohydrate restriction alone does not reliably do so. That raises the next question: when symptoms and low testosterone persist, what does treatment accomplish, and how safe is it?

Low testosterone can affect sex drive, sexual activity, blood counts, muscle, fat distribution, and how a man feels day to day. Yet many men who might benefit from treatment have been told that testosterone inevitably causes heart attacks or prostate cancer. Those claims do not describe what the largest randomized trials found. Other concerns, including rising blood pressure and hematocrit, deserve real attention. The answer is to identify the right patient, aim for a physiologic testosterone level, and follow what happens.

Randomized trials are especially useful here: they assign men to testosterone or placebo and measure what changes. They help distinguish treatment effects from the effects of obesity, illness, aging, or other factors that can accompany low testosterone. This article focuses on that interventional evidence.

What can treatment actually improve

Sexual desire and activity. In a TRAVERSE randomized substudy, 1,161 men with low testosterone and low libido received testosterone gel or placebo. Testosterone improved sexual desire, sexual activity, and overall hypogonadal symptoms through two years. It did not significantly improve erectile function compared with placebo. The earlier Testosterone Trials also found a moderate benefit for sexual function. A man whose main problem is erectile dysfunction may need evaluation and treatment for vascular disease, medication effects, or other causes as well.

Anemia. Testosterone stimulates red-cell production. In the TRAVERSE Anemia Study, 815 men began with anemia. At six months, anemia had resolved in 41.0% receiving testosterone versus 27.5% receiving placebo; at 12 months, the figures were 45.0% versus 33.9%. Other causes of anemia still require evaluation. This is a tangible benefit for a man with both symptomatic low testosterone and anemia. It also explains why we monitor hematocrit: the same biological effect can raise it too far.

Muscle and fat. Testosterone can improve body composition. In a 56-week randomized trial of 100 men with obesity and low testosterone, both groups followed a diet program. Compared with placebo, testosterone produced 2.9 kg more fat loss and a 3.4 kg advantage in lean mass preservation or regain. These are between-group differences; they do not mean every treated man gained 3.4 kg of muscle. Testosterone is no substitute for resistance training or adequate nutrition, and more lean mass does not automatically mean a large improvement in strength or physical performance. It can be useful as part of a broader plan for a man who has a real testosterone deficiency.

Bone density. A randomized Testosterone Trial found that spine trabecular volumetric bone density increased 7.5% with testosterone versus 0.8% with placebo after one year. Estimated spine trabecular bone strength increased 10.8% versus 2.4%. But bone scans are not fractures: in the larger TRAVERSE fracture study, clinical fractures occurred in 3.50% of men assigned testosterone versus 2.46% assigned placebo over a median 3.19 years. We should not prescribe testosterone as a proven fracture-prevention treatment. A man with osteoporosis needs his bone health evaluated on its own merits.

Mood and energy. A 2024 randomized TRAVERSE analysis found modest improvements in mood and energy, including among men with substantial depressive symptoms. It did not improve sleep quality or cognition. The earlier Testosterone Trials, involving 790 older men, found moderate sexual benefits and small mood benefits, but no significant improvement in the primary vitality outcome. A separate cognitive trial found no memory benefit. Treatment can help some men feel better; it is not an established treatment for dementia or a replacement for depression care.

Insulin resistance and metabolic health. There is positive interventional evidence, although results vary by population and treatment. In TIMES2, 220 men with low testosterone and type 2 diabetes and/or metabolic syndrome were randomized to testosterone gel or placebo. Testosterone reduced HOMA-IR, a measure of insulin resistance, by approximately 15% at six months and 16% at 12 months compared with placebo.

In the two-year T4DM trial, 1,007 men with central obesity and impaired glucose tolerance or newly diagnosed diabetes received a lifestyle program plus injectable testosterone or placebo. Among men with available two-year glucose tests, 12% in the testosterone group versus 21% in the placebo group met the diabetes threshold. These men had low-normal testosterone without pathological hypogonadism. A hematocrit safety trigger above 54% occurred in 22% versus 1%, an important tradeoff with that regimen.

In contrast, the TRAVERSE diabetes substudy did not show better glycemic control or less progression to diabetes. Testosterone can improve body composition and insulin resistance in some settings, but a diabetes-prevention benefit has not been consistently demonstrated. We treat confirmed symptomatic deficiency while addressing nutrition, activity, sleep, and weight alongside it.

Does testosterone cause heart attacks or strokes

The most informative study is TRAVERSE. It randomized 5,246 men, ages 45 to 80, with symptoms, two fasting testosterone results below 300 ng/dL, and established cardiovascular disease or elevated cardiovascular risk to testosterone gel or placebo. Major cardiovascular events (cardiovascular death, nonfatal heart attack, or nonfatal stroke) occurred in 7.0% receiving testosterone and 7.3% receiving placebo (hazard ratio 0.96, 95% CI 0.78-1.17). Average treatment lasted about 22 months and average follow-up about 33 months. This is strong evidence against the claim that appropriately dosed testosterone routinely causes heart attacks or strokes in men like those studied. It is not proof of cardiovascular benefit, or a lifetime safety result for every formulation and dose.

TRAVERSE also fits a broader randomized evidence base. A 2022 individual-participant meta-analysis obtained data from 17 trials involving 3,431 men. In trials reporting cardiovascular events, rates were 7.5% with testosterone versus 7.2% with placebo, without a significant difference. Average treatment duration in the individual-data studies was only about 9.5 months, making this supporting evidence for short- to medium-term safety.

What do artery-imaging studies show

The three-year TEAAM randomized trial enrolled 308 men aged 60 or older with low or low-normal testosterone. Testosterone did not significantly accelerate coronary calcium or carotid artery wall-thickness progression compared with placebo.

A smaller 2017 randomized coronary CT study, however, found a greater increase in noncalcified plaque after one year: an adjusted between-group difference of 41 mm3. These studies measured different aspects of atherosclerosis. Neither was large enough to establish heart-attack risk; TRAVERSE provides the stronger evidence for actual major cardiovascular events.

Which heart concerns remain legitimate

Some concerns originated in randomized research. The 2010 TOM trial enrolled 209 older men with mobility limitations and substantial chronic illness. It stopped early after cardiovascular-related adverse events occurred in 23 testosterone-treated men versus five receiving placebo. Its broad adverse-event category differed from TRAVERSE's prespecified heart-attack, stroke, and cardiovascular-death outcome. It remains a reason to consider frailty, dose, and the population being treated.

There were also signals in TRAVERSE that we take seriously. In TRAVERSE, atrial fibrillation occurred in 3.5% versus 2.4%, nonfatal arrhythmia warranting intervention in 5.2% versus 3.3%, acute kidney injury in 2.3% versus 1.5%, and pulmonary embolism in about 0.9% versus 0.5%, testosterone versus placebo. These findings warrant attention to symptoms and baseline risk; they do not erase TRAVERSE's major-event result.

The FDA removed its boxed-warning language about increased major cardiovascular events in 2025 after reviewing TRAVERSE. In the same action, it required warnings that testosterone products can raise blood pressure. We measure blood pressure before treatment and during treatment, particularly if a patient has hypertension or cardiovascular disease. We also ask about palpitations, chest symptoms, leg swelling, and new shortness of breath rather than assuming a normal testosterone level answers every safety question.

What about PSA and prostate cancer

An early randomized prostate-biopsy study assigned 44 men with symptomatic low testosterone to injections or placebo for six months. Although blood testosterone increased into the normal range, investigators found no significant treatment-related change in prostate tissue testosterone or dihydrotestosterone, or in measured markers of cell proliferation. This helps challenge the assumption that normalizing blood testosterone automatically overstimulates prostate tissue. A small six-month study cannot establish cancer safety by itself.

The TRAVERSE prostate-safety trial followed 5,204 men with structured PSA monitoring and adjudicated outcomes. High-grade prostate cancer was found in 5 of 2,596 men (0.19%) receiving testosterone and 3 of 2,602 (0.12%) receiving placebo; there was no statistically significant difference in high-grade or overall prostate cancer, urinary retention, prostate surgery, or new medication for urinary symptoms. Testosterone can cause a small early PSA rise. A PSA rise calls for repeat testing and evaluation, not an automatic cancer diagnosis and not automatic dismissal.

The small number of cancers matters: a nonsignificant difference is reassuring but does not prove identical risk. What did this study leave unanswered? It excluded men with prior prostate cancer, suspicious prostate findings, PSA above 3.0 ng/mL, or severe lower urinary tract symptoms. Follow-up was also too short to settle the development of cancers that may take many years to appear. A man with a history of treated prostate cancer needs an individual discussion involving his urologist; the TRAVERSE result cannot simply be applied to him. In June 2026, FDA requested further testosterone-label revisions concerning age-related low testosterone, prostate cancer, and benign prostatic hyperplasia. Those were requested changes, so a clinician should check the current label for the specific product when prescribing.

Who is a candidate, and what do we monitor

We start with symptoms and two separate morning testosterone measurements. We consider free testosterone or SHBG when the total result is difficult to interpret, and LH, FSH, prolactin, medication history, sleep apnea, obesity, and other possible causes when indicated. We ask about fertility plans before starting: external testosterone can suppress sperm production. Someone without symptoms or without confirmed low testosterone should not assume that a higher number will improve his health.

Low testosterone is not always permanent hypogonadism. Obesity, sleep apnea, certain medications, inadequate nutrition, and other medical conditions can suppress testosterone production. Identifying and addressing reversible contributors is part of treating the underlying problem.

Before treatment we document symptoms and goals, blood pressure, a complete blood count and hematocrit, and prostate history with PSA when appropriate to age, preference, and risk. We review cardiovascular disease, prior blood clots, urinary symptoms, untreated sleep apnea, and medications. We choose a formulation and dose aimed at a normal physiologic range, then recheck testosterone at a time appropriate to the formulation, along with hematocrit, blood pressure, symptoms, and adverse effects. A practical follow-up plan is reassessment within about three months, again during the first year, and every six to 12 months once stable, with earlier checks when clinically indicated. This is a clinical monitoring approach; trials have not established that one fixed testing schedule is best for everyone. A hematocrit reaching 54% calls for holding or adjusting therapy and evaluating contributing causes. A confirmed, concerning PSA change or new prostate finding calls for urologic evaluation; in TRAVERSE, a confirmed PSA increase of more than 1.4 ng/mL in the first year triggered referral.

If testosterone normalizes without meaningful symptom improvement, we reassess the diagnosis and whether continuing treatment is worthwhile.

For long-term treatment, we keep looking at the whole patient, not just testosterone. More frequent laboratory review may be useful when doses change, levels fluctuate, hematocrit rises, or symptoms evolve. We may use cardiovascular testing or prostate imaging when a patient's history, exam, PSA, or other findings justify it. Routine coronary CT, carotid imaging, or prostate MRI has not been shown to make testosterone treatment safer merely because it is repeated on a schedule. Imaging should answer a clinical question.

Our earlier testosterone article explains why we address weight, nutrition, sleep, and other reversible contributors while evaluating the cause of a low result. When low testosterone and meaningful symptoms remain, treatment deserves a serious discussion of its benefits. The best randomized evidence does not support treating heart attacks and prostate cancer as inevitable consequences. It also does not support ignoring blood pressure, hematocrit, arrhythmia, clots, fertility, or the prostate over years of care. Individualized treatment and close follow-up are how we protect the upside without pretending there are no tradeoffs.

For an earlier clinical perspective on metabolic contributors, watch the selected short interview with Dr. José Carlos Souto. That 2020 discussion predates the later trials reviewed here and does not address their findings.

For related reading, see our guide to hormone care, our discussion of why coronary calcium can change a cardiovascular risk conversation, and our heart and metabolic screening options.