The women’s story: a promising signal in the vaccine-dementia evidence that hasn’t yet reached the clinic
Women face roughly twice the lifetime risk of dementia as men. They also appear to gain more from shingles vaccination — though the evidence is contested, the biology is still being worked out, and the signal isn’t yet stable enough for differentiated clinical guidance. This post looks at where the two stories intersect and why the intersection matters.
When the Welsh natural experiment was published in Nature in 2025, the headline finding was widely reported as a 20% relative reduction in dementia diagnoses — but that figure requires a careful reading. Eligibility for the vaccine was associated with an 8.5% relative reduction; the 20% estimate applies to people who actually received it, scaled to account for the fact that only 47% of those eligible took it up. Both numbers matter and neither is wrong, but they measure different things. What received even less attention than either figure was the sex subgroup analysis buried in the supplementary data: one analysis reported a stronger association in women than in men. A subsequent natural experiment in Ontario, published in Lancet Neurology in February 2026, found the same pattern — and its lead author explicitly addressed it. As this post will show, that finding is now present in two of three natural experiments, biologically plausible, and still not making it into clinical conversations.
The Shingrix (recombinant vaccine) observational data have shown a similar directional pattern. The biology of why women are disproportionately affected by dementia also gives the signal biological plausibility. Neither of those things makes it established — but together they make it worth taking seriously.
The baseline problem: women and dementia risk
The starting point isn’t the vaccine at all. It’s the fact that women have substantially higher lifetime dementia risk than men. The commonly cited figures are a lifetime risk of approximately 1 in 5 for women compared to 1 in 10 for men over age 65, and roughly two thirds of all Alzheimer’s cases occur in women (unadjusted prevalence — when adjusted for age, the gap is narrower, which reflects the longevity contribution discussed below). For a long time this disparity was attributed mainly to longevity — women live longer, and dementia rates rise sharply with age. That explanation accounts for much but not all of the difference.
There are at least four distinct biological mechanisms that independently increase women’s dementia risk, and they converge in ways that aren’t fully understood. It’s worth noting that the female excess in dementia is clearest for prevalence and lifetime risk — the picture for age-standardised incidence is more mixed, with the gap becoming pronounced mainly after age 80 and varying across cohorts. The four-factor framework below reflects established risk contributors rather than claims of universally higher incidence at all ages.
Women live longer on average, and dementia rates rise sharply after age 80 when the female/male gap widens. Longevity accounts for much of the difference in prevalence and lifetime risk. Age-standardised incidence is less consistently elevated across all age groups — the excess is clearest after 80 and varies across cohorts.
Carrying the APOE-ε4 allele is the strongest genetic risk factor for late-onset Alzheimer’s. In men, one copy roughly doubles risk. In women, the same allele appears to confer substantially higher risk — the sex interaction is most pronounced between ages 65 and 75, with some evidence that risk tends to equilibrate between sexes in older cohorts above 75. The mechanism isn’t fully established; hormonal modulation of APOE function is one candidate.
Oestrogen has neuroprotective effects on synaptic plasticity, cerebrovascular integrity, and pathways relevant to amyloid processing and neuroinflammation. As oestrogen levels fall during the menopausal transition, neuroinflammatory responses increase and brain metabolic efficiency declines. The mechanistic evidence suggests these changes create elevated vulnerability — though the link to downstream pathology is still being characterised. Subjective cognitive symptoms during the transition are reported by up to 60% of women, and objective testing confirms changes in verbal memory and executive function.
Women generally mount stronger innate and adaptive immune responses than men — a difference shaped by both hormonal and X-chromosomal biology. This confers advantages against many infections across life. In the context of neurodegeneration, however, the picture is more complex: a more activated immune system can contribute to neuroinflammation, and the post-menopausal loss of oestrogen’s modulatory effects may shift the balance toward immunopathology. Microglial activation patterns differ by sex, with female-specific signatures in Alzheimer’s pathology.
What the shingles vaccine studies show — and how certain we can be
Against this backdrop, the sex-disaggregated findings from the natural experiments carry particular weight. The studies aren’t looking at a random intervention in a random population — they’re looking at an intervention that targets VZV reactivation in a population where post-menopausal neuroinflammatory vulnerability is already elevated, and where the absolute burden of dementia is substantially higher. That context matters for interpreting the subgroup findings — and for understanding why the signal is worth taking seriously even though it’s not yet consistent across all study designs.
| Study | Design | Sex finding | What it means |
|---|---|---|---|
| Eyting et al., Nature 2025 Welsh natural experiment, Zostavax |
Regression discontinuity, quasi-experimental | Stronger in women | Exploratory subgroup finding — not the primary pre-specified analysis. Effect described as “far stronger” in women. High-quality causal design strengthens confidence in the overall effect, but the subgroup finding requires replication. |
| Pomirchy et al., Lancet Neurology 2026 Ontario natural experiment, Zostavax |
Regression discontinuity, quasi-experimental | More significant in women | ~230,000 older adults, two eligibility thresholds, 5.5 years. ~2 percentage point absolute reduction in dementia risk. Lead author Geldsetzer explicitly noted the protective effect was statistically more significant in women, hypothesising sex differences in antibody response and higher baseline risk for both shingles and dementia in women. Third independent natural experiment; first to show the sex signal in a non-Welsh population. |
| Pomirchy et al., JAMA 2025 Australian natural experiment, Zostavax |
Regression discontinuity, quasi-experimental | No significant sex heterogeneity | Independent replication — overall effect confirmed. Explicitly reports no significant sex heterogeneity. The Cell 2025 follow-up suggests this may reflect limited statistical power from smaller sample size and lower dementia incidence in the Australian data — though this does not resolve whether the sex differential is real. |
| Rayens et al., Nature Communications 2026 RZV (Shingrix) observational, US Kaiser Permanente |
Propensity-matched cohort | Stronger in women | 51% overall risk reduction (adjusted HR 0.49). Stratified analyses show the association was greater in females than males. Observational design — residual confounding cannot be excluded. The sex-differential direction is present but should be interpreted cautiously given the Australian null finding. |
| Taquet et al., Nature Medicine 2024 RZV (Shingrix) observational, US |
Propensity-matched cohort | Both sexes benefit | 17% reduction overall. Some analyses show greater female benefit; others do not reach statistical significance for the sex interaction. The overall effect is robust; the sex differential is less clearly established in this dataset. |
| Marra et al., Human Vaccines 2025 Systematic review and meta-analysis, 18 studies |
Meta-analysis of observational studies | No significant sex difference in pooled analysis | Subgroup analysis by sex found no statistically significant difference in pooled risk. This is the most important counterweight to the sex-differential claim. Pooling heterogeneous studies may mask real subgroup effects — but the absence of a statistically significant finding in the largest available synthesis demands honest acknowledgement. Absence of evidence is not evidence of absence, particularly for interaction effects in heterogeneous pooled analyses — but the burden of proof for a positive claim remains with the positive claim. |
Two of three natural experiments (Wales and Ontario) show a sex-differential signal; one (Australia) does not. Observational Shingrix data lean toward women; pooled meta-analysis finds no significant difference. The Ontario replication is notable — it uses two independent eligibility thresholds and a cross-provincial comparison, and its lead author explicitly attributes the sex signal to differences in antibody response and baseline risk. This is a signal worth watching — not a settled finding.
Why the sex differential might be real — four mechanistic candidates
The epidemiological pattern is suggestive enough to be worth taking seriously biologically. There are four primary reasons why women might gain more from shingles vaccination, and they’re not mutually exclusive. A fifth — the biological ageing pathway introduced above — is more speculative but worth noting.
Female sex is itself a risk factor for herpes zoster reactivation — the association is well replicated in epidemiological studies, though the mechanism isn’t fully established. If women reactivate VZV more frequently or more severely, they have more neurological damage to prevent, and vaccination has more work to do.
This is the simplest hypothesis: not that the vaccine works differently in women, but that there’s more of the thing it targets.
Even if the relative risk reduction from vaccination were identical in men and women, the higher absolute dementia risk in women means vaccination prevents more cases per thousand women vaccinated than per thousand men vaccinated. This is a purely arithmetic argument — and it’s robust regardless of how the biological mechanism debate resolves.
It also has a direct clinical implication. If baseline dementia risk is higher in women and the relative effect of vaccination is similar across sexes, all else equal the number needed to vaccinate to prevent one dementia case would be lower in women than in men. That inference isn’t a directly measured endpoint from the studies — it follows from combining the dementia epidemiology with the vaccine effect estimates — but it’s a robust arithmetic argument for treating shingles vaccination as particularly high-value for women’s brain health, even before any sex-differential in relative effect is established.
This is the most biologically interesting hypothesis. After menopause, falling oestrogen levels increase neuroinflammatory responses, impair brain glucose metabolism, and affect pathways relevant to amyloid processing. The downstream implications for pathology are still being characterised, but the direction of vulnerability is well supported. This creates a neurological environment that may be more vulnerable to the specific damage caused by VZV reactivation — vasculopathy, microglial activation, and neuroinflammatory cascades that Post 2 established as the likely pathway from VZV reactivation to dementia.
If post-menopausal neuroinflammation amplifies the brain damage from VZV reactivation, then preventing reactivation through vaccination has more to offer in this context than in men whose neurological environment hasn’t undergone the same transition.
Women mount stronger innate and adaptive immune responses to vaccines than men — generating higher antibody titres, stronger T-cell responses, and more robust immunological memory. This is consistently observed across multiple vaccine types and reflects both hormonal influences and X-chromosomal biology. The immune landscape is complex: the same biology that produces stronger responses also confers greater susceptibility to autoimmunity and immunopathology. Whether stronger immune responses to shingles vaccination translate into greater neuroprotective benefit specifically has not been directly tested — but it remains a plausible mechanistic link, particularly if the adjuvant-mediated effects explored in Post 2 contribute to the overall protective signal.
Why this finding hasn’t reached clinical practice
Given the mixed evidence picture shown in the table above, the absence of this signal from clinical messaging reflects appropriate conservatism rather than oversight. Two of three natural experiments now show a sex-differential signal, but one does not, pooled meta-analysis is null, and subgroup analyses in this literature are rarely powered to detect sex interactions reliably. There are three interlocking reasons why that’s unlikely to change until the evidence base strengthens further.
Subgroup signals are secondary findings, often underpowered and sometimes unstable. When the main finding is reported as a 20% reduction in dementia risk, that’s a population-averaged estimate from the primary analysis. The subgroup finding showing a stronger effect in women is in the supplementary material, described as exploratory, and rarely appears in the abstract or press release. Journalists and clinicians read the headline. More importantly, subgroup analyses in this literature are rarely pre-specified with sufficient power to detect sex interactions reliably — which is partly why the signal appears in some studies and not others.
Clinical guidelines don’t yet have stable enough evidence to stratify by sex for brain health indications. Shingles vaccination is recommended universally for adults over 50 without sex-specific considerations. That’s appropriate given the current evidence — a directionally consistent but statistically contested subgroup finding isn’t sufficient grounds for differentiated guidance. Guidelines are conservative for good reasons.
The biology of women’s dementia has been historically underfunded and understudied. The sex-differential in Alzheimer’s risk has been known for decades. The mechanisms driving it — APOE-ε4 sex interaction, post-menopausal neuroinflammatory vulnerability, microglial sex differences — have only recently attracted the investment they deserve. Without a robust mechanistic foundation, interpreting a sex-differential vaccine finding is harder even when the epidemiological signal is there.
Two of three natural experiments now show a sex-differential signal (Wales and Ontario); one does not (Australia). Observational Shingrix data lean the same way. Pooled meta-analysis finds no statistically significant difference. The biology is coherent. The right clinical posture is informed awareness — not differentiated guidance, and not dismissal.
What this means practically
The absolute benefit argument — that women’s higher baseline dementia risk means vaccination would be expected to prevent more cases per thousand women vaccinated, regardless of any sex-differential in relative effect — was laid out in the mechanism section above. It’s arithmetically robust and doesn’t require the contested relative benefit finding to hold.
What the biology adds is a plausible reason why the relative benefit might also be larger in women — post-menopausal neuroinflammatory vulnerability amplifying VZV-mediated damage, stronger vaccine immune responses, higher baseline reactivation risk. Plausible is the right word: none of these pathways has been directly tested in the context of the vaccine-dementia association. The research gap that most needs filling is an analysis of whether menopausal status and APOE-ε4 genotype moderate the vaccine-dementia relationship. That study hasn’t been done.
The research gap that most needs filling
The sex-differential finding in the natural experiments is exploratory by design — both Eyting and Pomirchy describe their sex subgroup analyses as secondary and note they weren’t powered specifically for these comparisons. What’s needed is a prospectively designed analysis examining whether menopausal status, APOE-ε4 genotype, and baseline neuroinflammatory markers moderate the vaccine-dementia association. That study hasn’t been done. Until it has, the sex differential remains an important signal that should drive clinical conversations — and further research — not a settled fact that should drive policy changes alone.
One further finding from this literature is worth noting, even though it sits slightly outside the prevention frame of this post. A December 2025 follow-up study from the same Welsh natural experiment group, published in Cell (Xie, Eyting et al.), found that shingles vaccination was associated with slower dementia progression among people already diagnosed — including lower dementia-related mortality during follow-up. Among the 7,049 Welsh patients with dementia at baseline, around half of unvaccinated individuals died from dementia during follow-up compared to around 30% of those who had been vaccinated. Because this evidence comes from a quasi-experimental analysis rather than a randomised therapeutic trial in people with dementia, it should be read as a promising signal rather than an established treatment effect. Researchers are continuing to investigate it. But it suggests the vaccine’s potential relevance to brain health may extend beyond prevention — a question that warrants proper randomised evaluation.
For women in their 50s approaching or going through the menopausal transition — when the neuroinflammatory vulnerability is establishing itself and the window for early intervention may be most meaningful — the shingles vaccine case is already strong on shingles grounds alone. The potential brain protection benefit, and the hint of therapeutic relevance, add further weight to a conversation that should be happening more often than it is.
Primary sources
- Eyting M et al. A natural experiment on the effect of herpes zoster vaccination on dementia. Nature. 2025;641:438–446. doi:10.1038/s41586-025-08800-x
- Xie M, Eyting M et al. The effect of shingles vaccination at different stages of the dementia disease course. Cell. 2025;188(25):7049–7064. doi:10.1016/j.cell.2025.11.007
- Pomirchy M, Chung S, Bommer C, Strobel S, Geldsetzer P. Herpes zoster vaccination and incident dementia in Canada: an analysis of natural experiments. Lancet Neurology. 2026;25(2):170–180. doi:10.1016/S1474-4422(25)00455-7
- Kim JK, Crimmins EM. Association between shingles vaccination and slower biological aging: evidence from a US population-based cohort study. Journals of Gerontology Series A. 2026;81(3):glag008. doi:10.1093/gerona/glag008
- Pomirchy M et al. Herpes zoster vaccination and dementia occurrence. JAMA. 2025;333(23):2083–2092. doi:10.1001/jama.2025.5013
- Rayens E et al. Recombinant zoster vaccine is associated with a reduced risk of dementia. Nature Communications. 2026. doi:10.1038/s41467-026-69289-0
- Taquet M, Dercon Q, Todd JA, Harrison PJ. The recombinant shingles vaccine is associated with lower risk of dementia. Nature Medicine. 2024;30:2777–2781. doi:10.1038/s41591-024-03201-5
- Marra F, Gomes K, Liu E, Khatri Vadlamudi N, Richardson K, Cragg JJ. Effects of herpes zoster infection, antivirals and vaccination on risk of developing dementia: a systematic review and meta-analysis. Human Vaccines & Immunotherapeutics. 2025 Dec;21(1):2546741. PMC full text · doi:10.1080/21645515.2025.2546741
- Price BR et al. Sex differences and the role of estrogens in the immunological underpinnings of Alzheimer’s disease. Alzheimer’s & Dementia: Translational Research. 2025. doi:10.1002/trc2.70139
- Ferretti MT et al. Sex differences in Alzheimer disease — the gateway to precision medicine. Nature Reviews Neurology. 2018;14:457–469.
- Fisher DW, Bennett DA, Dong H. Sexual dimorphism in predisposition to Alzheimer’s disease. Neurobiology of Aging. 2018;70:308–324.
- Nebel RA et al. Understanding the impact of sex and gender in Alzheimer’s disease: a call to action. Alzheimer’s & Dementia. 2018;14:1171–1183.
- Kawai K, Yawn BP. Risk factors for herpes zoster: a systematic review and meta-analysis. Mayo Clinic Proceedings. 2017;92(12):1806–1821. doi:10.1016/j.mayocp.2017.10.009
- Klein SL, Flanagan KL. Sex differences in immune responses. Nature Reviews Immunology. 2016;16:626–638.
Discover more from
Subscribe to get the latest posts sent to your email.
