Shingles vaccine linked to lower dementia risk in 509,926 older adults

Earlier studies had already hinted at a link between shingles vaccination and lower dementia risk, but most of that work examined a live attenuated vaccine that is no longer available in the United States, or carried methodological limitations that made the signal hard to trust. Shingles arises when a virus already carried in the body reactivates, something that becomes more likely with age. Whether preventing that reactivation has any bearing on the brain is a question with enormous public health stakes. Older adults entering nursing care sit at the highest risk of dementia yet are rarely studied for vaccine benefits, and a proper randomised trial in that group would take years to deliver an answer.

The researchers built a cohort study from Medicare fee-for-service claims linked to nursing home electronic health records, using an approach designed to imitate the structure of a randomised trial. They included 509,926 beneficiaries aged 66 years or older who were admitted to a skilled-nursing facility between 1 January 2017 and 31 December 2022, had no diagnosed dementia and were eligible for the recombinant vaccine. Mean age was 79 years. People who received at least one dose in the facility, or within 12 months of admission if they had been discharged, were compared with those who received none, and everyone was followed for up to 4 years until a validated dementia diagnosis, Medicare disenrollment or death. The analysis adjusted for 57 baseline and time-varying characteristics.

Vaccination was uncommon: 8843 people, or 1.73% of the cohort, received at least one dose within 12 months of admission, and 87.0% of those doses were given after discharge rather than inside the facility. Over four years, dementia risk was 18.8% among those vaccinated compared with 24.6% among those not vaccinated, a difference of 5.8 percentage points, with a plausible range of 3.9 to 7.5 percentage points. Expressed as a ratio, risk was 0.76 times as high in the vaccinated group. The association was weaker in men and in people who had previously received the older live shingles vaccine. Negative control analyses, a check for bias, indicated that some residual confounding remained.

This extends the earlier observational signal to the recombinant vaccine that is actually in use, and does so in a population usually left out of such analyses. It does not settle causation. Because only 1.73% of this cohort was vaccinated, the people who got a jab may have differed systematically from those who did not, in health, mobility, family involvement or access to care, and the authors say their own bias checks detected traces of exactly that problem. The study was funded by the vaccine's manufacturer. Why the association weakened in men, and whether any of it would survive a randomised comparison, remain open.

For anyone weighing a shingles vaccine for themselves or an older relative, this adds to a growing but still indirect body of evidence that the decision might carry consequences beyond shingles itself. It is not a reason to expect the vaccine to prevent dementia, because a study that follows people rather than assigning treatment cannot separate the jab from the sort of person likely to receive it. The absolute numbers are worth holding onto: roughly one in five vaccinated and roughly one in four unvaccinated in this group were diagnosed with dementia within four years, so dementia was common either way. The useful takeaway is that the question is live and being taken seriously by researchers, not that it has been answered.