Does HRT Lower Dementia Risk? What the New Study Shows

Does hormone replacement therapy lower dementia risk?

A new study out of the UK Biobank is being covered everywhere right now, and the headline is a good one: women who used hormone replacement therapy had less dementia. In two decades of practicing medicine I have watched this particular question swing from “hormones protect the brain” to “hormones harm the brain” and back again, and each swing sent a generation of women to the wrong decision. So let me give you the finding straight, and then tell you exactly how much weight it can carry.

What the study actually found

Researchers followed 183,450 postmenopausal women for an average of 13.3 years — roughly 2.43 million person-years — and recorded 3,948 new cases of dementia. Women who had used hormone therapy for a year or more had about a 10 percent lower risk of dementia of all kinds, and press coverage has reported roughly a 16 percent lower risk for Alzheimer’s disease specifically. The association held after the researchers accounted for ethnicity, smoking, education and other factors.

That is a large, carefully handled dataset, and I do not want to wave it away. But the most interesting part of this study is not the headline number at all.

The benefit was not spread evenly, and that is the real finding

The protective association was concentrated in identifiable groups rather than showing up across the whole population. It clustered in women who started hormone therapy at roughly ages 46 to 56 — that is, near their actual menopause — and in women whose own estrogen exposure was lower, including women who went through surgical menopause and carriers of the APOE ε4 genetic variant.

In other words, the study is not saying hormone therapy protects women. It is saying that if there is a protective effect, it appears to belong to specific women under specific circumstances. That is a much more useful statement than the headline, and it is a much more honest one.

Why I will not call this proof

This is an observational study. It watches what happened to women who made their own choices about hormone therapy; it does not assign treatment. That opens three doors that statistical adjustment cannot fully close.

Healthy-user bias. Women who get prescribed hormone therapy tend to be healthier, better educated, and better connected to medical care in the first place. Adjusting for education and income narrows that gap. It does not erase it. Hormone use may partly be a marker for a woman who was already going to age well.
Confounding by indication. Women who take hormones usually have worse hot flashes and night sweats — and severe vasomotor symptoms are themselves independently linked to higher dementia risk. That can bend the result in either direction, which is an uncomfortable thing to sit with but an honest one.
It contradicts the randomized data. The Women’s Health Initiative Memory Study — a true randomized trial — found increased dementia risk when hormone therapy was started at age 65 or older. The new study does not overturn that. It coexists with it.

How both things can be true: the timing window

The way the field reconciles these results is called the critical window hypothesis, and this new study strengthens it rather than replacing it. The idea is that estrogen given to a brain that recently lost its own estrogen behaves very differently from estrogen introduced to a brain that has been without it for fifteen years and has already accumulated vascular damage.

The supporting evidence lines up reasonably well. The Cache County Study found roughly a 30 percent lower Alzheimer’s risk when hormone therapy began within five years of menopause. Pooled analyses put early initiation somewhere in the range of an 11 to 30 percent risk reduction, while initiation more than ten years past menopause looks neutral at best and harmful at worst. The Scientific American framing of a “window of opportunity” is a fair description of that pattern.

What is missing is the study that would settle it: a randomized trial of hormone therapy started at midlife and followed for decades. That trial is unlikely to ever be run, because the gap between menopause and a diagnosis of Alzheimer’s is roughly twenty years. We are going to be making this decision on imperfect evidence for the rest of my career.

What this changes in my office — and what it does not

It does not change who I prescribe hormones to. I do not put a woman on hormone therapy to prevent dementia, and neither should anyone else. The evidence is not there, and a clinic that markets it that way is getting ahead of the science in a way that should make you cautious about everything else they tell you.

What it does change is the timing conversation, and that conversation is genuinely urgent. I see women every month who are miserable with hot flashes, sleep loss, brain fog and vaginal dryness at 51, who have been told to wait it out, and who come back at 63 asking whether it is too late. Based on everything above, the window that matters most may be the one they were told to sit through.

So the decision stays where it has always belonged: your symptoms, your personal and family risk profile, your labs, and a real discussion of the known risks of hormone therapy. If a possible brain benefit is part of what makes the math work for you, fine — but it is a bonus line item, not the reason.

What is already proven for your brain

While we argue about a 10 percent association, several larger levers are sitting untouched in most people’s lives: blood pressure control in midlife, blood sugar control, treating hearing loss, physical activity, sleep, and staying socially engaged. None of them are as interesting as a hormone headline. All of them have better evidence behind them. We cover that ground on our disease prevention page.

Those levers work through a mechanism worth understanding: the aging brain builds compensatory circuitry, and every habit either adds to that scaffold or spends it. I walk through the science, and why midlife is the highest-leverage decade, on my page about preventing brain decline and the “brain cliff”.

References

Alzheimer’s & Dementia, 2026 — hormone replacement therapy and dementia risk in 183,450 postmenopausal women, identifying responsive subgroups in the UK Biobank.
JAMA, 2004 — Women’s Health Initiative Memory Study, dementia risk with hormone therapy begun at age 65 and older.
Neurology, 2012 — Cache County Study, hormone therapy timing relative to menopause and Alzheimer’s disease dementia.
Frontiers in Aging Neuroscience, 2023 — systematic review and meta-analysis of menopause hormone therapy and risk of Alzheimer’s disease and dementia.
Journal of Clinical Endocrinology & Metabolism, 2025 — menopause, menopausal hormone therapy and Alzheimer’s disease: current insights and future directions.
Journal of Clinical Investigation, 2026 — women’s midlife as the front line of Alzheimer’s prevention.