Menopausal Hormone Therapy and Long-Term Mortality in a Danish Cohort
Key Takeaways
- Adjusted analyses did not indicate higher long-term all-cause mortality among women who used systemic menopausal hormone therapy.
- Overall cause-specific analyses did not show unequivocal differences in cardiovascular, cancer, or other-cause mortality between groups.
- Some stratified and sensitivity analyses suggested lower hazards in selected groups, but the authors interpreted those signals cautiously.
Researchers assembled a nationwide, registry-based cohort of women born between 1950 and 1977 who were alive and living in Denmark at age 45 years. Follow-up began on each woman’s 45th birthday, and 92,619 women were excluded because of contraindications, prior menopausal hormone therapy, or prior bilateral oophorectomy. The final cohort included 876,805 eligible women, of whom 104,086, or 11.9%, redeemed at least one systemic menopausal hormone therapy prescription. During a median 14.3 years of follow-up, with an interquartile range of 7.9 to 21.0 years, 47,594 deaths occurred. Death was the primary outcome, cardiovascular, cancer, and other-cause mortality were secondary outcomes, and the analysis focused on systemic rather than all menopause therapies.
All-cause mortality incidence rates were 54.9 per 10,000 person-years among users and 35.5 per 10,000 person-years among nonusers. After adjustment, that difference did not indicate higher overall mortality among women who redeemed systemic therapy. Researchers found no unequivocal overall differences in cardiovascular, cancer, or other-cause mortality, although cause-specific analyses ended on 31 December 2022. Shorter use showed a slight reduction in cardiovascular mortality and a minor increase in cancer mortality, but these patterns were not amplified or significant with longer use. Across outcomes, the overall mortality pattern did not show a clear excess signal for systemic therapy.
Duration-stratified adjusted hazard ratios were 1.01 for less than one year and 0.94 for 1 to 2.9 years of use. They were 0.90 for 3 to 4.9 years, 0.89 for 5 to 9.9 years, and 0.98 for 10 years or longer. Among stratified analyses, transdermal formulations had the lowest hazard at 0.85 (0.80 to 0.90), and initiation at age 52 years or older showed the lowest mortality. Women with bilateral oophorectomy at ages 45 to 54 who used systemic menopausal hormone therapy had 27% to 34% lower mortality hazard than women who did not, while sibling and as-treated estimates were 1.06 and 0.83. The authors did not present these heterogeneity signals as definitive and called for cautious interpretation.
The cohort was observational and nonrandomized, so the estimates cannot establish causality even after extensive adjustment. Exposure was defined from redeemed prescriptions rather than observed ingestion, and residual confounding could remain despite the main models and sibling analysis. The median age at the end of follow-up was 59.4 years, which limited visibility into possible later-life effects. These results also do not necessarily extend to low-dose vaginal estrogen, non-hormonal menopause therapies, or tibolone.
Overall, systemic menopausal hormone therapy was not associated with increased mortality, while subgroup findings warranted cautious interpretation and further scrutiny.