Shingles (herpes zoster) vaccination is linked to a lower risk of dementia in two natural experiments that exploited birth-date eligibility cutoffs: in Wales, receiving the live vaccine cut new dementia diagnoses by about 20% over 7 years, and an Australian rollout showed a similar pattern. A large US record study found the recombinant vaccine associated with 17% more dementia-free time than the live vaccine. These designs are much stronger than ordinary observational studies, but no randomized trial has tested dementia as an outcome.
Summary drafted from this entry’s sources; editorial review pending.
01Key questions
Does the shingles vaccine prevent dementia?
It may. In Wales, comparing people just old enough to qualify for the live zoster vaccine with people just too old showed that receiving it cut new dementia diagnoses by about 20% over 7 years, a natural experiment that largely rules out healthy-user bias (Eyting et al., Nature 2025).1
Is this just healthier people getting vaccinated?
That is the usual problem, and the reason these studies matter. Eligibility depended on an exact birth date, so people on either side of the cutoff were otherwise alike. A similar cutoff analysis in Australia found the same direction of effect (Pomirchy et al., JAMA 2025).2
Which vaccine looks better, live or recombinant?
In US health records, the recombinant vaccine was associated with 17% more time lived without a dementia diagnosis than the older live vaccine (Taquet et al., Nature Medicine 2024). This is observational, not randomized.4
What would change this rating?
A randomized trial with dementia or cognitive decline as a prespecified outcome, or replication of the natural-experiment design in more countries with the recombinant vaccine.
Shingles (herpes zoster) is a reactivation of the chickenpox virus that sits dormant in nerve tissue for life. Since 2023 a string of natural experiments has suggested that vaccinating against it also prevents or delays dementia. These studies are far harder to dismiss as confounding than ordinary observational work, because who got the vaccine was decided by an arbitrary birthday cut-off rather than by health-conscious choice. No randomized trial has tested dementia as an outcome yet.
02Why the evidence is unusually strong for a non-trial
Most "vaccine users have less dementia" findings suffer from the healthy-user effect: people who get vaccinated also exercise, see doctors and take their pills. The shingles studies sidestep this. In Wales, people born on or after 2 September 1933 became eligible for the live zoster vaccine; people born a week earlier never did. Uptake jumped from 0.01% to 47.2% across that one-week boundary, while everything else about the two groups should be the same. Comparing them is close to a randomized experiment (a regression discontinuity design).
03The main findings
| Study | Design | Result |
|---|---|---|
| Eyting et al., Nature 2025 (Wales) | Birthdate cut-off, live vaccine, 7-year follow-up | Receiving the vaccine cut new dementia diagnoses by 3.5 percentage points (95% CI 0.6 to 7.1), a 20% relative reduction; stronger in women. Confirmed with dementia deaths on death certificates in England and Wales (PMID 40175543) |
| Pomirchy et al., JAMA 2025 (Australia) | Birthdate cut-off, 101,219 primary-care patients | Eligibility lowered the probability of a new dementia diagnosis over 7.4 years by 1.8 percentage points (95% CI 0.4 to 3.3); eligibility did not change uptake of other preventive care (PMID 40267506) |
| Xie et al., Cell 2025 (Wales) | Same natural experiment, later disease stages | Fewer mild cognitive impairment diagnoses and, among people already living with dementia, fewer deaths due to dementia; no single dementia type drove the effect (PMID 41338191) |
| Taquet et al., Nature Medicine 2024 (US records) | Live-to-recombinant switchover, recombinant vaccine | Compared with the live vaccine, the recombinant vaccine (Shingrix) was associated with 17% more dementia-free time over 6 years, about 164 extra diagnosis-free days in those later diagnosed; it also beat influenza and Tdap vaccines as comparators (PMID 39053634) |
Swipe table →
The Welsh and Australian studies used the live vaccine (Zostavax), now discontinued in many countries. The recombinant vaccine evidence comes from a different, weaker design (an observational comparison around a product switch), so whether today's vaccine protects as much is less certain, though the direction is the same.
04Why it might work
Two explanations are on the table, and they are not exclusive. One is viral: the varicella-zoster virus and other neurotropic herpesviruses may reactivate in or near the brain and drive inflammation, and preventing reactivation removes that insult. The other is immunological: vaccines can have broad off-target effects that train innate immunity. The stronger effect in women, who also mount stronger vaccine immune responses, is consistent with either. See Stealth pathogens and Immune function for the wider infection-and-dementia picture.
05Limitations and open questions
- No randomized trial yet. The natural experiments are as good as observational evidence gets, but a regression discontinuity estimate applies most cleanly to people right at the age cut-off (around 79-80 years).
- Diagnosis, not pathology. Outcomes are recorded diagnoses and death certificates, not amyloid scans or cognitive testing, so some effect could reflect timing of diagnosis. The death-certificate replication argues against that being the whole story.
- Effect size is moderate. A 20% relative reduction over 7 years is large for a population intervention but means most vaccinated people who would get dementia still do.
- Which vaccine, at what age. Live-vaccine data cannot be transferred directly to the recombinant vaccine, and no study tells us whether vaccinating at 50 differs from vaccinating at 80.
06Where it sits on the Heretic scale
Promising. Replicated quasi-experimental evidence in two countries plus a supportive recombinant-vaccine analysis, but no randomized trial with dementia as the endpoint. The vaccine itself is already recommended for older adults for shingles prevention in many countries, so the dementia question is about added benefit, not a new risk.
Further reading: detailed protocols
07Connections
08Sources
- Eyting M, Xie M, et al. (2025). "A natural experiment on the effect of herpes zoster vaccination on dementia." Nature 641:438-446. PMID 40175543 . DOI 10.1038/s41586-025-08800-xPubMed 40175543 · DOI 10.1038/s41586-025-08800-x
- Pomirchy M, et al. (2025). "Herpes Zoster Vaccination and Dementia Occurrence." JAMA 333(23):2083-2092. PMID 40267506 . DOI 10.1001/jama.2025.5013PubMed 40267506 · DOI 10.1001/jama.2025.5013
- Xie M, Eyting M, et al. (2025). "The effect of shingles vaccination at different stages of the dementia disease course." Cell 188(25):7049-7064. PMID 41338191 . DOI 10.1016/j.cell.2025.11.007PubMed 41338191 · DOI 10.1016/j.cell.2025.11.007
- Taquet M, et al. (2024). "The recombinant shingles vaccine is associated with lower risk of dementia." Nat Med 30(10):2777-2781. PMID 39053634 . DOI 10.1038/s41591-024-03201-5PubMed 39053634 · DOI 10.1038/s41591-024-03201-5
09Entry history
Latest content update.
Heretic informs and never prescribes. Where amounts appear, they describe what published studies used. Talk to a qualified clinician before acting on anything here.