Hormone Replacement Therapy and dementia risk in postmenopausal women: UK Biobank findings highlight associations but stop short of causality

Summary

A large observational study using UK Biobank data examined whether hormone replacement therapy (HRT) could protect against dementia in postmenopausal women. Among 183,450 participants followed for an average of 13.3 years, 3,948 received a dementia diagnosis and 1,993 were diagnosed with Alzheimer’s disease. Women reporting HRT use for at least a year, or still using it at enrolment, had about a 10% lower relative risk of dementia after adjusting for age, education, smoking, deprivation, blood pressure and diabetes. A larger association appeared in the group classified as having undergone surgical menopause, with a 26% reduced dementia risk, though this group included both hysterectomy only and ovary removal, complicating interpretation. The paper emphasizes that association does not prove causation and notes several limitations inherent to observational studies.

Author: The Conversation

Overview

The article reports on a cohort analysis using UK Biobank data to explore whether hormone replacement therapy (HRT) influences dementia risk in postmenopausal women. While some analyses suggest a lower risk of dementia among HRT users, the authors stress that observational data cannot prove a protective effect and that findings should be interpreted with caution. The discussion covers the nuances of surgical versus natural menopause, potential confounding factors, and gaps in the data that limit causal inferences. NICE guidance is cited to place the findings in clinical context, which currently does not support using HRT solely to prevent dementia.

Key study design and findings

The study analyzed 183,450 postmenopausal women from the UK Biobank, followed for an average of 13.3 years. Dementia was diagnosed in 3,948 participants, including 1,993 cases of Alzheimer’s disease. Among women who reported HRT use for at least one year or who were still on HRT when they joined the study, dementia risk was about 10% lower relative to non-users after adjusting for several covariates. The reduction was more pronounced in women classified as having surgical menopause, with a 26% lower relative risk of dementia in this subgroup. There was no statistically clear association among women who experienced natural menopause.

The authors note several important caveats: the surgical menopause group combined hysterectomy and ovary removal, which can have markedly different hormonal consequences; there may be unmeasured differences between HRT users and non-users (education, health status, access to care) that could drive observed associations; and the study did not distinguish between estrogen-only therapy and estrogen with progestin, or between different dosages and routes of administration.

Context and limitations

The article places the findings within the broader, mixed evidence on HRT and cognitive outcomes. Earlier registry studies suggested hormonal therapy could be associated with higher Alzheimer’s risk with certain regimens, while estrogen-only therapy did not consistently show this pattern. The UK Biobank analyses add to the dialogue but cannot establish causality. The timing of HRT initiation also emerged as potentially important, with a weaker signal for dementia risk when started earlier in life, complicating any argument for early replacement as dementia prevention.

Genetic factors, specifically APOE ε4 status, were examined but did not show a definitive interaction with HRT effects on dementia risk in this dataset, echoing mixed results from prior research. Trials closer to menopause have not demonstrated consistent cognitive benefits from HRT, reinforcing that current trial evidence does not support using HRT to prevent dementia in typical menopause populations.

Clinical and policy implications

NICE guidance is cited to emphasize that HRT should not be offered to prevent dementia or cardiovascular disease, while acknowledging its established role in treating menopausal symptoms. The article underscores the need for careful consideration of those who undergo early surgical menopause, as loss of ovarian hormones might carry distinct health implications. Overall, the findings are not a directive to start or stop HRT specifically for dementia prevention, but they do justify further research, particularly into how timing, type of HRT, and surgical menopause interact with dementia risk.

Takeaways and future directions

The piece concludes that while observational findings generate hypotheses about potential protective associations, robust randomized trials or well-controlled prospective studies are required to determine whether HRT can meaningfully influence dementia risk. Future work should disentangle estrogen-alone versus combined regimens, assess dose and route of administration, stratify by menopausal type (surgical vs natural) and menopause timing, and examine genetic subgroups for differential effects.

Conclusion

In sum, the UK Biobank analysis contributes to the ongoing debate about HRT and dementia risk but stops short of providing evidence for a causal protective effect. Clinicians should continue to weigh HRT benefits and risks for menopausal symptoms while not leveraging it as a dementia prevention strategy at this time.