An analysis of health records from Wales found that older adults who received a shingles vaccine were about 20% less likely to be diagnosed with dementia over seven years than their unvaccinated peers. The research, which took advantage of an age-based rollout of the vaccine as a natural experiment, also suggests potential benefits for people already living with dementia and indicates stronger effects in women.
In Wales, a national shingles vaccination program launched on September 1, 2013, created what researchers describe as a natural experiment.
Under the policy, people became eligible for a single-dose, live-attenuated shingles vaccine (Zostavax) once they reached a defined age threshold. In practice, this meant that individuals born on or after September 2, 1933 were eligible to receive the vaccine from September 1, 2013, while those who were even slightly older were not routinely offered it. This sharp age cutoff allowed scientists to compare groups of older adults who were very similar in age but differed in their access to the vaccine.
Stanford Medicine–led researchers analyzed anonymized health records from more than 280,000 adults in Wales born between 1925 and 1942, focusing on those aged 71 to 88 at the start of the program and without dementia at that time. They tracked health outcomes for seven years. Because vaccine uptake among those eligible was roughly 50%, while uptake among those ineligible was close to zero, the team could estimate the effect of receiving the vaccine.
According to Stanford Medicine’s report on the work, vaccination cut the incidence of shingles by about 37% over seven years, a reduction similar to that seen in clinical trials. By 2020, when many participants were around 86 or 87 years old, about one in eight had been diagnosed with dementia. Those who received the shingles vaccine were about 20% less likely to develop dementia than those who did not.
"It was a really striking finding," said Pascal Geldsetzer, MD, PhD, assistant professor of medicine at Stanford and senior author of the study. "This huge protective signal was there, any which way you looked at the data." The researchers reported that people on either side of the vaccine eligibility cutoff were similar in measured characteristics such as education, common chronic conditions and receipt of other routine vaccines, suggesting that differences in dementia risk were unlikely to be explained by obvious health or socioeconomic factors alone.
The main analysis of dementia risk and shingles incidence was published April 2, 2025, in the journal Nature. A follow-up study by the same group, published December 2, 2025, in Cell, extended the observation period to nine years and examined additional outcomes, including mild cognitive impairment and progression among people already diagnosed with dementia.
In the longer-term analysis, vaccinated individuals were less likely to be diagnosed with mild cognitive impairment compared with those who remained unvaccinated. Among 7,049 Welsh seniors who already had dementia at the start of the vaccination program, receiving the shingles vaccine was linked to a substantially lower risk of dying from dementia over the following nine years: roughly half of unvaccinated individuals with dementia died from the disease during follow-up, compared with about 30% of those who were vaccinated.
"The most exciting part is that this really suggests the shingles vaccine doesn't have only preventive, delaying benefits for dementia, but also therapeutic potential for those who already have dementia," Geldsetzer said in a Stanford Medicine news release, while cautioning that the findings are based on observational data.
Across both studies, the apparent protective effect of the shingles vaccine against dementia was stronger in women than in men, a pattern the researchers suggest could relate to sex differences in immune responses or in dementia biology. Women are known to mount higher antibody responses to many vaccines, and shingles itself is more common in women.
The vaccine studied is a live-attenuated formulation targeting varicella-zoster virus, the pathogen that causes chickenpox. After a childhood infection, the virus remains dormant in nerve cells and can reactivate decades later as shingles, particularly in older adults or people with weakened immune systems. The researchers and outside experts say one plausible explanation for the observed dementia benefit is that preventing viral reactivation reduces inflammation and other forms of damage in the nervous system, though broader immune effects of the vaccine may also play a role.
Similar associations between shingles vaccination and reduced dementia risk have now been reported in other large health-records studies using data from England, Australia, New Zealand and Canada, according to Stanford Medicine and other summaries of the research. However, those analyses are also observational and cannot on their own prove cause and effect.
The live-attenuated vaccine examined in the Welsh program has since been largely replaced in many countries by a newer recombinant vaccine, Shingrix, which is more effective at preventing shingles. Geldsetzer and colleagues have called for a randomized clinical trial using the older live-attenuated vaccine to test directly whether it can delay or slow dementia, and say further work is needed to determine whether the newer recombinant vaccine has similar or greater cognitive benefits.
While the findings have drawn attention for suggesting that a widely used vaccine might substantially reduce dementia risk and potentially slow its progression, experts emphasize that the results, though unusually robust for an observational design, still need to be confirmed in randomized trials before the vaccine can be recommended specifically as a dementia intervention.
