Study Links Testosterone Therapy to Higher Cardiovascular Risk in Women

New Data Points to a Cardiovascular Safety Signal

Testosterone therapy in postmenopausal women is associated with a higher risk of incident cardiovascular disease (CVD). Researchers presented the findings at the 2026 Annual Meeting of The Menopause Society in San Diego. The analysis tracked over 6,300 propensity-matched women aged 45 and older across a mean follow-up period of nearly three years.

The composite cardiovascular outcome occurred in 3.8% of women using testosterone compared to 0.5% of women without exposure. That difference reflects an adjusted hazard ratio (HR) of 8.08, with a 95% confidence interval (CI) of 4.72 to 13.82. The outcome included coronary artery disease, cerebrovascular disease, and heart failure diagnosed between 30 days and five years after starting therapy.

The study team analyzed records from the TriNetX United States Collaborative Network. Investigators applied a 12-month washout period before the index date to exclude prior testosterone use, and they excluded all patients with a history of cardiovascular disease before the study began.

Why Women Take Testosterone Off Label

The Food and Drug Administration (FDA) has not approved any testosterone product specifically for women in the United States. Clinicians who recommend the hormone prescribe it off label, typically using low doses of male formulations compounded into creams or gels.

Clinical guidance supports testosterone therapy only for hypoactive sexual desire disorder (HSDD), a persistent and distressing loss of sexual desire that affects roughly 10% to 15% of postmenopausal women. The evidence remains insufficient to support using testosterone for other common midlife complaints, including fatigue, memory loss, weight gain, or bone density loss.

Higher doses of testosterone carry well-documented androgenic side effects such as acne, facial hair growth, thinning scalp hair, and voice deepening. Long-term cardiovascular safety data in female patients have remained scarce, leaving clinical decisions reliant on short-term trials.

How Researchers Matched Patient Groups

To reduce confounding, the investigators used one-to-one propensity score matching across several demographic and medical variables. Matching accounted for baseline age, smoking status, obesity, hysterectomy status, surgical menopause, premature ovarian insufficiency, and vasomotor symptoms.

The analysis also balanced baseline cardiovascular risk factors, including diabetes mellitus, essential hypertension, and dyslipidemia. In addition, the researchers matched patients for baseline systemic menopause hormone therapy use and vaginal estrogen therapy.

Sensitivity analyses showed a similar pattern when comparing women receiving testosterone combined with hormone therapy against women receiving hormone therapy alone. Lead author Dr. Xuezhi Daniel Jiang, professor of obstetrics and gynecology at Drexel University College of Medicine, noted that while absolute event rates remained low, the consistency across cohort definitions warrants further prospective investigation.

Study Limitations and Clinical Considerations

This research relies on an observational, retrospective database design rather than a randomized controlled trial. While propensity score matching addresses documented variables, retrospective electronic health record data cannot eliminate unmeasured confounding, differences in lifestyle habits, or unrecorded variations in product formulation and adherence.

The absolute cardiovascular event rate was 3.8% in the testosterone cohort over nearly three years of follow-up. While that rate is low in absolute terms, cardiovascular disease remains the leading cause of death among women, making any signal meaningful for clinical risk assessments.

Women using or considering off-label testosterone therapy can discuss these findings directly with their prescribing physician. A clinician can review individual cardiovascular risk profiles, check lipid and blood pressure markers, and weigh symptom distress against potential long-term risks.