Does TRT cause heart problems? No extra heart attacks, more irregular rhythm

Testosterone replacement therapy did not increase heart attacks, strokes or heart-related deaths in the trials built to measure them [2][13]. Where TRT may still cause heart problems is the heart's rhythm, and that evidence is unresolved: the one trial designed around heart safety logged more irregular heartbeats among the men on testosterone [2].

That trial is TRAVERSE. It gave 5,246 men aged 45 to 80 either a daily testosterone gel or an identical gel containing no hormone, and every man had symptoms of deficiency, two fasting blood readings below 300 nanograms per decilitre (the unit a laboratory prints a testosterone reading in), and either established heart disease or a high risk of it [2]. Over an average of 33 months of follow-up, a heart attack, stroke or heart-related death reached 182 men on testosterone (7.0%) against 190 men on the hormone-free gel (7.3%) [2].

A separate pooling of the older evidence lands in the same place. The TestES synthesis combined 35 placebo-controlled trials covering 5,601 men, including the participant-by-participant records from 17 of them, and counted heart and stroke events in 120 of 1,601 men on testosterone (7.5%) against 110 of 1,519 on placebo (7.2%) [13]. Its authors also found no sign that treatment worsened blood pressure, blood fats or blood sugar markers [13], which answers the common worry that testosterone pushes blood pressure up.

The rhythm finding is the part nobody has pinned down. TRAVERSE logged more atrial fibrillation in the testosterone group than in the placebo group [2]. A 2024 review then pooled every placebo-controlled trial it could find, 8,126 men treated and 7,310 on placebo, and the excess showed up in TRAVERSE alone: across all the trials together the irregular-heartbeat count pointed upwards, but the range of plausible results still ran from slightly fewer events to roughly three times as many, which is another way of saying the data cannot separate an effect from chance [5]. The same review looked at population studies and found no link between a man's own natural testosterone level and his odds of developing atrial fibrillation once other risk factors were taken into account [5].

Two other problems were more common on testosterone in TRAVERSE over those same months: a blood clot in the lungs, and sudden kidney injury [2]. The question stayed open for a decade before that trial because no testosterone trial published until then had been designed or made large enough to count heart events at all [1].

Low testosterone and poor blood-vessel function travel together, which is part of why the arterial result surprised people. Among 58 non-smoking men, the main artery in the arm widened by 3.7% of its diameter when blood flow returned in middle-aged and older men with lower testosterone, against 5.7% in men the same age with higher testosterone [7]. That widening test is the standard way of checking whether the endothelium, the single-cell lining of every blood vessel, is doing its job.

The same study suggests why. Infusing vitamin C, which mops up the reactive molecules behind oxidative stress, lifted the widening to 5.3% in the low-testosterone men and changed nothing in the other two groups [7], so part of what looked like ageing was oxidative damage. Laboratory work fills in the chemistry: a testosterone shortage lowers production of nitric oxide, the gas that signals a blood vessel to relax, partly by raising a molecule called ADMA that blocks the enzyme making it [8].

Topping the hormone back up does not reliably reverse any of that. Sixty young men with congenital hypogonadism, a condition present from birth in which the testes cannot make testosterone, were treated for about six months: their blood pressure, weight and waist went up, their total and protective HDL cholesterol went down, and the ADMA that blocks nitric oxide rose rather than fell [11].

For the rhythm signal these papers offer no mechanism at all. A 2023 review reads the electrical evidence the other way, reporting that treatment reduces prolongation of the QT interval, the time the heart's electrical system takes to reset between beats, and concluding that testosterone restored to normal physiological levels does not threaten heart health [15].

The trial carrying most of the weight here tested one narrow group. TRAVERSE enrolled men aged 45 to 80 who had symptoms of deficiency, two fasting testosterone readings below 300 nanograms per decilitre, and existing heart disease or a high risk of it [2]. Nothing in it speaks to men whose testosterone is already in the normal range and who take it for muscle, mood or performance.

The clock is short, too. Men in TRAVERSE used the gel for about 22 months on average [2], and the TestES authors restricted their conclusion to the short to medium term [13].

The rhythm question cuts both ways, and more weakly than the headline numbers suggest. TRAVERSE was the only trial in the 2024 pooling that treated cardiovascular safety as its main objective [5], and trials not built to hunt for rhythm events will under-record them, so a pooled result compatible with no effect is softer evidence than it looks.

Who this evidence does not cover:

• Women. Testosterone's relationship with heart disease and overall death rates in women, older women in particular, has not been established, although studies of testosterone applied through the skin have not turned up adverse heart effects [4].
• Transgender men, whose cardiovascular risk runs higher than in the general population for reasons the usual risk factors do not fully explain [10]. In 18 transgender men starting testosterone, markers of platelet activation and inflammation drifted upwards over 52 weeks [9].
• Young men with a deficiency they were born with, where six months of treatment moved cardiometabolic markers the wrong way [11].
• Anyone after mortality figures: across 35 trials the TestES team found too few deaths to analyse [13].

A 2024 clinical summary written for family doctors still describes the evidence on heart events and death as conflicting [3].

Take a 60-year-old man with low morning testosterone on two blood tests, symptoms that bother him, and a stent in one coronary artery. He sits close to the population TRAVERSE actually recruited, so its arterial numbers are the most relevant he has: roughly 7 men in 100 had a heart attack, stroke or heart-related death over about three years, whether they were given the hormone or the hormone-free gel [2]. If he also has a history of atrial fibrillation, the rhythm data cannot tell him much, and that is the thing worth raising by name with the prescriber. Guidance written for family doctors has testosterone, the share of blood made up of red cells, and the prostate marker PSA checked before treatment starts and at least once a year afterwards in men aged 40 and over [3].