— However, observational study should be hypothesis-generating only
by Rachael Robertson, Enterprise & Investigative Writer, MedPage Today
October 21, 2026 • 3 min read
Testosterone therapy for menopause was linked to heightened risk of incident cardiovascular disease, a large retrospective cohort study suggested.
Over 5 years, testosterone therapy during menopause showed a significantly elevated hazard for incident cardiovascular disease (HR 8.08, 95% CI 4.72-13.83), reported Xuezhi (Daniel) Jiang, MD, PhD, of Drexel University College of Medicine and Tower Health's Reading Hospital in Pennsylvania, and a colleague in a poster at the Menopause Societyopens in a new tab or window annual meeting.
This trend persisted at both 10-year (HR 11.50, 95% CI 6.78-19.50) and unrestricted (HR 5.92, 95% CI 4.09-8.56) follow-up intervals, Jiang reported.
"Our findings suggest that testosterone use is associated with a higher risk of cardiovascular events in postmenopausal women," Jiang told MedPage Today, emphasizing that the observational findings do not indicate a causal link.
"The findings also highlight an important distinction between the absence of a demonstrated cardiovascular safety signal in selected clinical trial populations and the establishment of long-term cardiovascular safety across broader populations," he said.
Stephen Burgess, PhD, a biostatistics professor at the University of Cambridge in England, noted that the 8-fold difference in cardiovascular risk seen in the study is concerning, especially since heart disease is relatively common.
"While this study cannot prove that testosterone treatment is the cause of this difference in risk, it does raise concerns, particularly as the authors have accounted for several other potential causes," he said in comments posted to the Science Media Centre websiteopens in a new tab or window.
There is insufficient evidence on whether testosterone therapy alleviates menopausal symptoms, but some research suggests it helps women with low libido. While there are no FDA-approved testosterone products for women, some doctors prescribe low doses of male formulations off-label.
There's also a lack of long-term data on cardiovascular safety and testosterone therapy for women. Jiang noted that most clinical trials on testosterone therapy for menopause have been limited by small sample size or short duration and were not designed to evaluate major cardiovascular events.
"We therefore wanted to investigate this question using a large, multicenter electronic health record database," he explained.
For this retrospective cohort study, he and a colleague used the TriNetX research network to look at postmenopausal women age 45 and older who were initiating testosterone therapy. These patients were compared to controls not exposed to the treatment. The mean follow-up time was 2.6 years.
Those with prior cardiovascular disease and other conditions that could complicate interpretation were excluded. After propensity score matching, there were 3,164 women in each cohort.
The primary outcome was composite incident cardiovascular disease, including coronary artery disease, cerebrovascular disease, and heart failure, in the 5 years following testosterone therapy initiation. Secondary analyses at 10 years and anytime after initiation were also conducted.
Overall, they found that at 5 years, cardiovascular events occurred in 3.8% of postmenopausal women exposed to testosterone therapy compared to 0.5% of unexposed, propensity-matched counterparts. Those relationships persisted at 10 years (5.3% vs. 0.5%) and in the unrestricted analysis (6.2% vs 1%).
A sensitivity analysis comparing women receiving testosterone plus menopausal hormone therapyopens in a new tab or window with women receiving menopausal hormone therapy alone showed that the former group had heightened cardiovascular disease risk at all three follow-up points (HR 6.73, 95% CI 5.00-9.05 at 5 years).
Jiang noted some limitations, including the retrospective study design and the possibility of misclassification and residual confounding. The mean follow-up time of 2.6 years was relatively short, and there were few cardiovascular events overall.
Importantly, TriNetX does not include information on the dose, formulation, route, delivery, or indication of testosterone. Jiang cautioned that these results shouldn't be generalized to all menopausal women receiving guideline-concordant physiologic-dose transdermal testosterone.
He noted that "the magnitude of the association should prompt further investigation rather than be interpreted as proof that appropriately prescribed testosterone causes cardiovascular disease."
Moreover, these findings should be validated with prospective studies, he said, and future research should parse out the individual cardiovascular outcomes.