Herpes Zoster Vaccination and Incident Dementia in Canada

Dr. Margarita Fedorova talks with Dr. Pascal Geldsetzer about how the herpes zoster vaccination may have neuroprotective effects, potentially delaying or preventing dementia.

Read the related article in The Lancet Neurology.

Disclosures can be found at Neurology.org.

Podcast Transcript:

Dr. José Merino: This is José Merino, editor-in-chief of the Neurology Family of Journals. The Neurology Podcast provides practical information to neurologists and other clinicians to help them provide better care for their patients. Thanks for listening and have a great week.

Dr. Margarita Fedorova: Hi, this is Margarita Fedorova and I'm a neurology resident at the Cleveland Clinic. Today, we're exploring a study that speaks to an idea that's been building for a few years now, that vaccination itself might be neuroprotective. I'm here with Pascal Geldsetzer, a physician and epidemiologist at Stanford University. Pascal is the author on a paper published in the February 2026 issue of Lancet Neurology titled "Herpes Zoster Vaccination and Incident Dementia in Canada: An Analysis of Natural Experiments." What makes the study different is how they got at the question. When Ontario started publicly funding the shingles vaccine, eligibility came down almost entirely to date of birth. People born a few weeks on one side of a cutoff could get the vaccine for free, and people born a few weeks on the other side could not. That left Pascal and his colleagues with two groups of older adults who should be alike in essentially every way, except their access to the vaccine. And they asked whether those groups went on to develop dementia in different rates. We wanted to bring Pascal to the podcast because this is a question our patients are already asking us in clinic, and because vaccination is one of the few things we can offer an older adult that might actually protect the brain. Pascal, welcome to the podcast.

Dr. Pascal Geldsetzer: Thank you very much for having me.

Dr. Margarita Fedorova: Let's start with a take-home message. If I'm rounding with a team tomorrow and I get to teach them one thing from your paper, one thing I want them to still remember six months from now, what should that be?

Dr. Pascal Geldsetzer: I think it is that for shingles vaccination and dementia, we now have a body of evidence from observational data that we're usually never able to obtain in terms of the fact that we are able to get at cause and effect in a much more plausible way than we are usually able to. And we can say here now that shingles vaccination may well have a delaying or preventing effect on the dementia disease process.

Dr. Margarita Fedorova: Before we get into your study, give us the background. We tend to think about dementia in terms of amyloid, tau, vascular risk, and lifestyle. Where do vaccines come into it? And is this really a shingles story or is it something broader about vaccination and brain health in older adults?

Dr. Pascal Geldsetzer: Yeah, so unfortunately we don't know exactly. And so of course, broadly speaking, it could be either through reducing reactivations of the chickenpox virus or varicella zoster virus, which we know is a neurotropic virus, which we know reactivates throughout life. And so reducing these reactivations and probably accompanying neuroinflammatory processes may well have benefits. But there may also be a broader potentially virus-independent immune mechanism at play here that may also apply to other vaccines than just shingles vaccination. So we don't really understand the mechanism yet. It's a very interesting and fascinating area of research and hopefully we'll learn more in the future.

Dr. Margarita Fedorova: We already knew from a lot of prior work that shingles vaccine was associated with a lower risk of dementia, but in the paper you were direct that those earlier studies can't really speak to cause and effect, and that's a really big part of why this one matters. What goes wrong when we simply compare vaccinated people to unvaccinated people and how does your design get around it?

Dr. Pascal Geldsetzer: So this gets to the fundamental problem that we usually have in observational data analysis in epidemiology, right? It's that those who get a certain intervention, who get a certain medication or who decide to adopt a certain lifestyle regimen, certain diet, or who decide to get vaccinated are just very different to those who don't. And especially in the case of vaccination, we know there's this healthy vaccinee bias whereby those who have better preventive health behavior, health motivations, et cetera, are the ones who go get vaccinated and they have a lower risk of dementia in any case regardless of vaccination. And we can try to adjust for these differences between these groups. And that's what we do in epidemiology with all sorts of fancy methods. But at the end of the day, what we're restricted to is the information that we have in the data sources that we're analyzing. And I just don't have very much information at all in electronic health record data, medical claims data, et cetera, on all these various health behaviors, health motivations, dietary behaviors, et cetera, et cetera. So it becomes really hard to know whether we are really truly able to adjust for all of these differences. What we are doing instead in these shingles vaccination natural experiment analyses is we're changing the comparison group to something that's much more close to what we have in a clinical trial. And of course the clinical trial is what we need to ensure that the comparison groups are comparable because we know all that's different about them is whether a coin landed on heads or tails. And here in our natural experiments, we know that all that's different about the groups that we're comparing is whether they happen to be born essentially by random chance a few days earlier or a few days later because our natural experiments are exploiting these date-of-birth-based eligibility rules for shingles vaccination. For instance, in Wales, in our first paper that we published in this series, we know that those who had their 80th birthday just prior to the start date of the shingles vaccination program, which just so happened to be September 1st, 2013, were ineligible and remained ineligible for life. Well, if you had your 80th birthday just after, you were eligible and we see a jump in your probability of ever getting the vaccine by just a week's difference in birth from 0% to 50%. So we have now this really sharp contrast just by the fact that you were born a few days earlier or a few days later. And that's what makes our comparison groups different and the effects that we're seeing much more likely to reflect true cause and effect rather than just a correlation.

Dr. Margarita Fedorova: And for this current natural experiment, what exactly did you find? Without talking too much about the numbers, can you tell us whether we think we're preventing dementia or delaying it with the vaccination?

Dr. Pascal Geldsetzer: We don't know exactly if we are preventing or delaying dementia. That's an active research area in my group to truly understand this. But what we're seeing is that these groups of people were born just on either side of this eligibility cutoff. Those who happen to be eligible because they were born just a little bit later have a lower probability of receiving a new dementia diagnosis in the future. It's hard to pinpoint the exact magnitude of the effect because our method really zooms in on people born just around this cutoff. And of course also various data limitations that we usually run into such as having knowledge on your exact va