A cardiovascular signal for timely hormone therapy, with caveats

Design

Target-trial emulation from a longitudinal multiethnic cohort (SWAN)

Population

2,737 women with vasomotor symptoms, CVD-free; 755 initiated hormone therapy

Primary outcome

Fatal and non-fatal cardiovascular events over 20 years

Effect

Overall aHR 0.78 (0.62-0.98); within 10 years 0.73 (0.58-0.93); Black women 0.51 (0.33-0.81)

Using 20 years of SWAN cohort data to emulate a sequence of target trials, investigators estimated cardiovascular outcomes among 2,737 women with vasomotor symptoms and no prior hormone therapy, of whom 755 started menopausal hormone therapy; 224 cardiovascular events occurred.

Overall, starting hormone therapy was associated with a lower cardiovascular risk (adjusted hazard ratio 0.78, 95% CI 0.62 to 0.98). The association was clearer when therapy began within 10 years of menopause (hazard ratio 0.73) than later (1.53, confidence interval spanning no effect), consistent with the timing hypothesis, and appeared stronger in Black women (0.51).

The authors are careful: this is observational target-trial emulation, open to residual confounding, and they state it should not be used to justify hormone therapy for cardiovascular prevention. The usable message is reassurance, not a new indication: for a woman with troublesome vasomotor symptoms close to menopause, the cardiovascular balance of symptom-driven hormone therapy looks favourable rather than alarming.

• Starting hormone therapy for vasomotor symptoms was linked to lower cardiovascular risk (hazard ratio 0.78).
• The signal was clearer within 10 years of menopause than later, supporting the timing hypothesis.
• The association appeared stronger in Black women.
• Do not start hormone therapy for cardiovascular prevention on this basis.
• Use it as reassurance when prescribing for symptoms near menopause.

Why it matters

It touches the daily decision of whether vasomotor symptoms near menopause are a reason to withhold hormone therapy.

Don't overread it

This is observational and the authors warn it should not support hormone therapy for cardiovascular prevention.

The statistics, in plain English

An observational hazard ratio of 0.78 whose interval nearly touches 1.0 is a modest, uncertain association; emulated target trials reduce but do not remove confounding, so this is weaker than a randomised result.

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.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.