- Design
- Systematic review and meta-analysis of randomised placebo-controlled trials
- Population
- 31 trials, 9,437 middle-aged and older men with low or low-normal testosterone
- Primary outcome
- Cardiovascular risk factors and events
- Effect
- LDL −6.6 mg/dL (95% CI −8.7 to −4.6); non-fatal arrhythmia OR 1.58 (1.21 to 2.05)
This meta-analysis pooled 31 randomised trials (9,437 men, middle-aged and older, with low or low-normal testosterone) lasting from 40 days to 36 months, with searches up to July 2024.
Testosterone replacement lowered total cholesterol by about 10 mg/dL, LDL by about 7 mg/dL and triglycerides by about 7 mg/dL against placebo, with no significant change in HDL. It lengthened time to 1 mm ST depression on treadmill testing by about 67 seconds and slightly reduced carotid intima-media thickness. Against that, non-fatal arrhythmias were more common (OR 1.58, 95% CI 1.21 to 2.05; 6,598 participants).
The favourable findings are all surrogate markers. The arrhythmia signal is a clinical event and is consistent with the atrial fibrillation excess reported in the TRAVERSE trial. For a cardiologist asked whether testosterone is safe in a patient with heart disease, the honest answer is that lipids may improve, rhythm risk may rise, and hard outcome data remain limited.
- Ask about testosterone use in men presenting with new atrial fibrillation or other arrhythmia.
- Before testosterone is started in a man with heart disease, discuss the possible arrhythmia risk.
- Do not count a modest LDL fall on testosterone as cardiovascular protection.
- Confirm hypogonadism biochemically before treatment is considered for cardiovascular reasons.
Why it matters
The case for testosterone's cardiovascular safety rests on surrogate markers while the one clinical signal points the other way.
Don't overread it
Arrhythmias were not a pre-specified primary outcome in most trials, and many trials were short; this is a signal, not a measured risk.
The statistics, in plain English
An odds ratio of 1.58 with an interval of 1.21 to 2.05 means arrhythmias were about one and a half times as likely on testosterone, and the interval does not reach 1, so chance alone is an unlikely explanation. The lipid changes are statistically clear but small; an LDL fall of 7 mg/dL is a fraction of what a statin achieves.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for cardiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free