Published June 17, 2026 Updated August 4, 2026 8 min read

# The Shingles Vaccine and Dementia: A Surprising Prevention Signal for APOE4 Carriers

## The short version

Stay current on shingles vaccination for the usual reason: it prevents shingles. Several strong real-world studies suggest an added dementia benefit, including in APOE4 carriers. Treat that as a welcome extra, not a guarantee.

By the OutliveAPOE4 editorial team. [How we research & source](/methodology).

A shingles shot lowered new dementia diagnoses by about 20% over seven years, and the result came from one of the cleanest study designs medicine ever gets by accident. If you carry APOE4 and you spend your energy chasing every lever that might protect your brain, this is an unusual one: cheap, already recommended for your age group, low risk, and quite possibly sitting in your medical chart as overdue.

The headline first landed as one of those associations that is easy to wave away. Healthier people get more vaccines, and healthier people get less dementia, so maybe the shot was just a marker for the worried well. What changed the conversation is that two different research groups found ways to strip that bias out, and the protection survived.

## What the natural experiments showed

The strongest evidence comes from a 2025 *Nature* study led by Markus Eyting and Pascal Geldsetzer at Stanford, built on a quirk of how Wales rolled out the older live shingles vaccine, Zostavax. Eligibility was set by an exact birthdate. People born just before a cutoff were not offered the vaccine; people born just after were. A person one week younger was suddenly far more likely to get vaccinated, and in every other way the two groups were essentially identical.

That accident does something a normal observational study cannot. It mimics a coin flip, which is the entire point of a randomized trial, and it neutralizes the “healthy people get vaccinated” problem because nobody chose their own birthday. Researchers call this a regression-discontinuity design, and it is about as close to causal proof as you can get without running a trial.

The result: people made eligible for the vaccine were 3.5 percentage points less likely to be newly diagnosed with dementia over the next seven years, a 20% relative reduction. The effect was stronger in women than in men. It held up when the team checked it against the same kind of birthdate rollouts in England, Australia, New Zealand, and Canada.

A second 2024 study in *Nature Medicine* from Maxime Taquet’s group at Oxford used a different natural experiment: the moment health systems switched from the old live vaccine to the newer recombinant one, Shingrix, in 2017. Comparing roughly 207,000 people who got one shot or the other, the Shingrix group lived about 17% more of the following six years free of a dementia diagnosis, which worked out to 164 extra dementia-free days among those who eventually were diagnosed. So the newer vaccine looked even better than the older one that drove the Welsh result.

## The APOE4 question, with a real answer

The natural-experiment studies did not break their results down by genotype, so for a while the honest carrier answer was “probably, but we cannot point to your number.” A 2025 analysis in *Alzheimer’s & Dementia* by Jacqueline Davitte and colleagues closed that gap. Using the UK Biobank and FinnGen cohorts, they matched vaccinated and unvaccinated adults by age, sex, and APOE-e4 status, then asked whether carriers benefited.

They did. In APOE4 carriers, shingles vaccination was tied to a 25% lower risk of dementia (hazard ratio 0.75, 95% confidence interval 0.63 to 0.90). Non-carriers got a larger benefit, a 41% reduction (HR 0.59, 0.48 to 0.73). Both groups clearly came out ahead. The plain read for you is that the protection is real for carriers, somewhat smaller than for non-carriers, and still well worth having. This is observational data rather than a randomized trial, so weigh it alongside the quasi-randomized Welsh result rather than on its own.

## Why a shingles shot would touch the brain at all

The leading idea starts with what shingles actually is. The chickenpox virus, varicella-zoster, never fully leaves your body. It hides in your nerves for decades and can reactivate as shingles when your immune defenses dip. The hypothesis is that each reactivation stokes a little fire of inflammation in and around the nervous system, and that chronic, smoldering neuroinflammation is one of the engines that drives Alzheimer’s. Tamp down the reactivations, the thinking goes, and you turn down a fire that was feeding the disease. That viral-inflammation thread runs through a lot of what makes the [APOE4 brain](/topics/how-apoe4-affects-the-brain) more vulnerable.

There is a second, more provocative possibility. A 2025 study in *npj Vaccines* compared two unrelated AS01-adjuvanted vaccines, the Shingrix shingles shot and a newer RSV vaccine, across about 437,000 adults. Both were tied to lower dementia risk in the following 18 months: 29% for the RSV vaccine and 18% for the shingles vaccine. Since these vaccines protect against completely different viruses but share the same adjuvant (AS01, the ingredient that revs up the immune response), the authors floated that the adjuvant itself might “train” the immune system in a way that protects the brain, independent of which virus it targets. That is a genuinely interesting lead, not settled science. The short 18-month window and the fact that the comparison cannot fully rule out other differences between groups mean it should be read as a hypothesis to test, and a recent published critique argues the data cannot pin the effect on the adjuvant specifically.

Hold both ideas loosely. The mechanism is not proven. What is solid is the outcome: across very different designs, vaccinated people got less dementia.

## Dementia rates after shingles vaccination

| Study | Design | What it found |
| --- | --- | --- |
| Eyting/Geldsetzer, *Nature* 2025 | Welsh birthdate cutoff (regression discontinuity, ~quasi-randomized) | 3.5 percentage points (20% relative) lower dementia over 7 years; stronger in women |
| Taquet, *Nature Medicine* 2024 (~207,000 people) | Live-to-Shingrix switch (natural experiment) | Shingrix: 17% more dementia-free time, ~164 extra dementia-free days over 6 years |
| Davitte, *Alzheimer’s & Dementia* 2025 (UK Biobank + FinnGen) | Matched cohort, by APOE-e4 status | Carriers HR 0.75 (0.63-0.90), ~25% lower; non-carriers HR 0.59, ~41% lower |
| AS01 adjuvant, *npj Vaccines* 2025 (~437,000 people) | Matched cohort, 18-month window | Shingles 18% and RSV 29% less dementia; shared-adjuvant hypothesis |

## Strong real-world signal, not a dementia-prevention trial

None of these is a randomized trial designed to prevent dementia, and that is the honest ceiling on the evidence. The Welsh design is the closest thing to one, which is why it carries the most weight, but the rest are cohorts that can still hide confounders no statistician can fully erase. Bring the usual skepticism here, the kind worth applying to any health headline ([reading a study like a skeptic](/topics/reading-a-study-like-a-skeptic) walks through how). The RSV/adjuvant signal rests on just 18 months of follow-up, far too short to call durable. And the absolute reductions, while meaningful, are modest: a few percentage points off a person’s risk, not a wall against the disease. A randomized vaccine-versus-placebo dementia trial would settle it, and until one reports, this stays strong, convergent, real-world evidence rather than proof.

## How to use the vaccine now

This is the rare item on the prevention list where the action is genuinely easy, because you may already qualify on standard grounds.

-   **Check whether you are due.** Shingrix is recommended for most adults 50 and older, and for younger adults with weakened immune systems. If you are in that group and have not had it, this is a low-regret box to tick. It is a two-dose series.
-   **Ask for the recombinant vaccine (Shingrix) specifically.** It is the one tied to the larger dementia signal, and it is what most health systems now use anyway.
-   **Frame it correctly with your doctor.** You are not asking for an experimental brain treatment. You are getting a routine, recommended vaccine that happens to carry a promising bonus signal. That is a short, easy conversation.
-   **Keep it in proportion.** A shingles shot is a small, cheap lever, not a substitute for the heavy hitters. The strongest prevention evidence still comes from stacking the basics: see the [FINGER trial](/topics/finger-trial-and-lifestyle) (the Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability, a two-year multidomain lifestyle trial), [exercise](/topics/exercise-and-apoe4), and [blood pressure control](/topics/blood-pressure-and-brain-health). Think of the vaccine as a free add-on while you do the real work.

## Common questions

**Which vaccine, the old one or the new one?** Shingrix, the recombinant vaccine. The older live vaccine (Zostavax) drove the original Welsh finding, but the head-to-head comparison favored Shingrix, and it is the standard option now regardless.

**How much did carriers benefit compared with everyone else?** In the genotype-matched analysis, vaccinated carriers had about 25% lower dementia risk. The benefit was a bit smaller than in non-carriers, but it was real and statistically solid.

**I am under 50. Should I get it now for my brain?** The dementia data come from older adults, and the routine recommendation generally starts at 50 (earlier if your immune system is compromised). There is no good evidence yet to vaccinate young, healthy carriers early specifically for brain protection. Revisit it when you reach eligibility.

> The shingles vaccine is shaping up to be one of the easiest, lowest-regret moves on a carrier’s list: already recommended, low risk, and backed by unusually strong real-world signals that include people with your genotype. Talk to your doctor about whether you are due.

## Sources & further reading

1.  [Eyting M, Geldsetzer P, et al. A natural experiment on the effect of herpes zoster vaccination on dementia. Nature, 2025](https://pubmed.ncbi.nlm.nih.gov/40175543/)
2.  [Taquet M, et al. The recombinant shingles vaccine is associated with lower risk of dementia. Nature Medicine, 2024](https://pmc.ncbi.nlm.nih.gov/articles/PMC11485228/)
3.  [Davitte J, et al. Associations between herpes zoster vaccination and dementia risk in APOE-e4 carriers and non-carriers. Alzheimer's & Dementia, 2025](https://pmc.ncbi.nlm.nih.gov/articles/PMC12725106/)
4.  [Lower risk of dementia with AS01-adjuvanted vaccination against shingles and RSV infections. npj Vaccines, 2025](https://pmc.ncbi.nlm.nih.gov/articles/PMC12198376/)

## Related deep dives

-   [Reading a study like a skeptic: a carrier’s media-literacy guide APOE4 headlines swing from "cure" to "doom" weekly. A practical toolkit for reading health studies and news so you can tell real signal from hype, and act on the right things.](/topics/reading-a-study-like-a-skeptic)
-   [HSV-1, cold sores, and Alzheimer’s risk in APOE4 carriers The cold-sore virus interacts with APOE4 specifically, raising Alzheimer's odds. Carriers ask if antivirals help. A 2026 randomized trial finally tested it. Here is the honest answer.](/topics/hsv1-cold-sores-and-alzheimers)
-   [How APOE4 affects the brain APOE4 influences how the brain clears amyloid, handles tau, manages inflammation, regulates neural activity, and maintains its blood vessels. A plain-language tour of the leading mechanisms.](/topics/how-apoe4-affects-the-brain)
-   [APOE4 and Alzheimer’s risk: what the numbers actually mean Relative risk, absolute risk, and age of onset: how to read the scary statistics about APOE4 and Alzheimer’s, with real ranges and the hopeful part that gets buried.](/topics/apoe4-and-alzheimers-risk)