Shingles Vaccine Deflects Dementia in Aussies, Too

The association between vaccination against the herpes zoster (HZ) virus, which causes shingles, and dementia risk continues to gain traction. Earlier this month, Pascal Geldsetzer and his team at Stanford University, California, reported that HZ vaccination lowers dementia risk in a cohort in Wales (Apr 2025 news). They have now corroborated these findings in a second population, strengthening the long-debated idea that immunomodulation may effectively reduce or delay neurodegeneration.

  • New study strengthens the link between Herpes zoster vaccination and reduced risk for dementia.
  • Study design minimizes bias.
  • How the protection works is unclear.

As in the previous paper, first author Michael Pomirchy and colleagues exploited a public health policy, this time in Australia, that went into effect in November of 2016. The policy demarcated HZ vaccine eligibility based on an age cutoff—those 79 or younger on November 1, 2016, were deemed eligible, while those even a day past 80 were not. In their sample of 101,219 patients, the researchers found no discontinuities across this eligibility threshold in diagnoses of common diseases, risk factors for dementia, or uptake of health services besides the HZ vaccination. This supported their assumption that the two samples are comparable in characteristics that could influence dementia diagnosis, HZ vaccine uptake being the one exception. Indeed, there was an abrupt jump in the probability of HZ vaccine receipt from 6.5 percent to 30.2 percent between the ineligible and eligible groups, respectively.

Ruth Itzhaki, University of Oxford, who championed work linking viral activation to dementia, lauds this study’s experimental approach. “Previous shingles vaccine studies were subject to various confounding factors, including the likelihood of major differences in healthcare between the groups examined. The approach of Geldsetzer and colleagues is novel in the way it selects the groups, which differ merely by a minute age difference,” Itzhaki wrote to Alzforum (comment below).

The new paper reports that HZ vaccination eligibility decreased the probability of a new dementia diagnosis during the follow-up period of 7.4 years by 1.8 percentage points. According to the authors, the difference is significant within such a large sample, and it is comparable to the 1.3 percentage point reduction in dementia risk in the shingles-vaccine-eligible group in Wales. In contrast to the analysis of the Welsh cohort, the new paper does not account for actual vaccination. In Wales, this reduced the likelihood of dementia by 20 percent.

“Being able to replicate these findings in different populations, different health systems settings, at different date-of-birth eligibility cutoffs, and for different dementia-related outcomes is extremely exciting because it drastically reduces the chance of bias or a fluke finding,” Geldsetzer wrote to Alzforum (comment below).

HZ, and other viruses, including polio, can remain dormant in the body for years, waiting to wreak havoc upon reactivation. Alzheimer’s researchers who want to understand the biological mechanism of the effect say these findings necessitate deeper inquiry. Paul Schulz, University of Texas Health Houston, wrote, “Besides scientific curiosity, this is vital because if we know the answer, we may be able to bypass the vaccines and invoke their mechanism to reduce AD risk—perhaps even more than vaccines reduce it.” (Comment below.) –Anna Bright

Anna Bright is a Ph.D. student in New York City.

Comments

    • User Profile Image Ruth Itzhaki
      Universities of Manchester and Oxford
    • Posted:

    The recent article by Geldsetzer and colleagues in Nature using Welsh data and this one using Australian data provide a very useful addition to the many studies showing the protective effect of the live-attenuated herpes zoster vaccine (Zostavax), and of the recombinant vaccine (Shingrix) against dementia.

    These papers follow the first one published in 2021 by me and epidemiologist colleagues (Lophatananon et al., 2021), and soon afterwards by many others (PubMed lists 22 articles including reviews and comments). All show protection with these vaccines. Incidentally, shingles vaccines might be the only type, of the various  vaccines tried against diverse infectious disease, in which all the studies agree in finding a protective effect against dementia.

    However, previous shingles vaccine studies were subject to various confounding factors, including the likelihood of major differences in healthcare between the groups examined. The approach of Geldsetzer and colleagues is novel in the way it selects groups that differ merely by a minute age difference, and also in finding a sex difference (noted also by Lophatananon et al.  post-publication). 

    The latter remains unexplained. In discussing possible explanations for the main result, Geldsetzer and colleagues mention the possibility that VZV reactivations and consequent neuroinflammation cause reactivation of dormant HSV1 present in brain. This would be consistent with findings in a three-dimensional brain model that show HSV1 reactivation after either of two known major risks for AD—infection in general (VZV—Cairns et al., 2022, and HHV6—unpublished HHV6 data), and head injury (Cairns et al., 2025). Reactivation of HSV1 and its reduction after vaccinations against diverse types of infection was proposed in 2002 (Itzhaki and Dobson) in a response to the novel findings of Verreault et al., 2001).

    All these studies provide further strong support for the development of a vaccine against HSV1.

    References:

    Lophatananon A, Mekli K, Cant R, Burns A, Dobson C, Itzhaki R, Muir K. Shingles, Zostavax vaccination and risk of developing dementia: a nested case-control study-results from the UK Biobank cohort. BMJ Open. 2021 Oct 8;11(10):e045871. PubMed.

    Cairns DM, Itzhaki RF, Kaplan DL. Potential Involvement of Varicella Zoster Virus in Alzheimer's Disease via Reactivation of Quiescent Herpes Simplex Virus Type 1. J Alzheimers Dis. 2022;88(3):1189-1200. PubMed.

    Cairns DM, Smiley BM, Smiley JA, Khorsandian Y, Kelly M, Itzhaki RF, Kaplan DL. Repetitive injury induces phenotypes associated with Alzheimer's disease by reactivating HSV-1 in a human brain tissue model. Sci Signal. 2025 Jan 7;18(868):eado6430. Epub 2025 Jan 7 PubMed.

    Itzhaki RF, Dobson CB. Alzheimer's disease and herpes. CMAJ. 2002 Jul 9;167(1):13. PubMed.

    Verreault R, Laurin D, Lindsay J, De Serres G. Past exposure to vaccines and subsequent risk of Alzheimer's disease. CMAJ. 2001 Nov 27;165(11):1495-8. PubMed.

    View all comments by Ruth Itzhaki

  2. There have been many analyses in electronic health record and medical claims data, which have found that receiving a given vaccine is correlated with a reduced risk of dementia in the future. These studies have all compared those who go get vaccinated with those who do not. The critical limitation of this research is that we know that the health behaviors of those who get vaccinated are different from those who do not, and we have very little, if any, information on these behaviors in electronic health record or medical claims data. For instance, we do not know about dietary behaviors or physical activity levels. We therefore don't know if we're merely looking at correlations or actual cause and effect.

    Our studies take a fundamentally different approach and, thus, provide a different level of evidence. In medicine, to prove that an intervention, like a medication or a vaccine, works, you need a randomized trial. You get, say, a thousand study participants and for each you throw a coin to assign them to getting the vaccine or no vaccine. We then know that, on average, the vaccine and the no-vaccine group should be similar to each other, so are good comparison groups.

    Our study takes advantage of a similar scenario. In Wales, when they rolled out the shingles vaccine, they said that if you had your 80th birthday just prior to the start date of the program, you are ineligible and you remain ineligible for life. While if you had your 80th birthday just after, you were eligible for at least one year. We see in our data, that just a one-week difference across this date-of-birth cutoff means that you go from essentially no one getting vaccinated to about half of the population getting vaccinated. Just like a randomized trial! We have a vaccine-eligible and a vaccine-ineligible group for which we know that they should be on average similar to each other, and therefore good comparison groups, because all that's different about these two groups is if they were born a few days earlier or a few days later. In Australia, we had the same situation at a different point in time (November 1, 2016, instead of September 1, 2013) and thus at a different date-of-birth eligibility cutoff (November 2, 1936, versus September 2, 1933).

    We see large protective effects not only in Wales and Australia, but also in other countries that rolled out the vaccine in a similar way. Being able to replicate these findings in different populations, different health systems settings, at different date-of-birth eligibility cutoffs, and for different dementia-related outcomes is extremely exciting because it drastically reduces the chance of bias or a fluke finding.

    If the shingles vaccine really prevents or delays dementia, then this would be a hugely important finding for clinical medicine, population health, and research. What we really want to do as a next step in this research is a clinical trial. I'm currently trying to raise funds to conduct such a trial from private and philanthropic foundations. We would ideally like to trial a version of the old live-attenuated vaccine, which is the vaccine for which we have generated our compelling body of evidence. But this vaccine is no longer being manufactured in the U.S. or Europe.

    View all comments by Pascal Geldsetzer

  3. Our group has been studying the effects of vaccines on AD risk for five to seven years. We waited a long time to publish the first study because we wanted to be sure about its results. It demonstrated that six annual flu vaccines are associated with a 40 percent reduction in AD risk for four years, and it goes out to eight years if we take a shorter baseline phase (Bukhbinder et al., 2023).

    We wondered whether this finding was specific for the influenza vaccine, so we investigated three other vaccines and found similar results, though slightly different efficacy rates.

    Geldsetzer et al., 2025, and Pomirchy et al. here used different techniques than we did. The took advantage of idiosyncrasies in the Welsh and Australian health systems to find controls matched for age and other factors. Those are both different samples than were in our American database. We did propensity score matching on many similar and some different parameters.

    It is very reassuring to see that both papers came to the same conclusions as we did in 2023. There are now numerous replications of the efficacy of vaccinations in reducing AD risk. We are delighted each time another group, using different databases and different analytic techniques, comes to the same conclusion we did, as it gives us more confidence in our own results. The developing weight of evidence is progressively supporting the hypothesis that vaccinations are potentially an inexpensive, readily available, low-risk way of reducing AD risk.

    The finding that several vaccines reduce AD risk is critical, because we still cannot stop AD or replace lost brain cells. In our clinic, we use donanemab, lecanemab, and experimental therapies every day to slow the course of AD. It is great to have these agents. But prevention remains vital.

    Many interesting and important questions remain. Are vaccine effects synergistic? If they all work through the same mechanism, they may not be. If they work through different mechanisms, their effects on reducing AD risk may be additive. In addition, if “immune conditioning” is important, as per Richard Lathe’s suggestion (see below), then getting more vaccinations may “keep the immune system in shape.” So, even if they worked through the same mechanism, more frequent vaccination could be valuable.

    How do vaccines reduce AD risk? Beyond scientific curiosity, this is vital because if we know the answer, we may be able to bypass the vaccines and invoke their mechanism to reduce AD risk—perhaps even more than vaccines reduce it. For example, it may be some combination of cytokines that are released during vaccination that influence CNS inflammation that then reduces AD risk. Perhaps we can give people those cytokines directly and at higher doses to reduce AD risk. Right now, the reduction associated with multiple influenza vaccines requires giving multiple annual vaccinations. Perhaps we could get greater AD prevention rates, more rapidly, by supplying the intermediate agent at higher doses at closer intervals?

    I’d like to comment on the hypotheses raised by Pomirchy et al. regarding how vaccinations may reduce AD risk. Regarding the idea of fewer episodes of shingles, hence fewer visits to doctors, this is an important potential explanation for less AD in the vaccinated group, i.e., they did not see their doctor unless they had shingles, so their AD went undiagnosed. The counterargument is that they saw their doctor to get vaccinated. Interestingly, then, vaccination could, in theory, be associated with lesser or greater AD detection. In addition, only 47 percent of eligible persons in the Geldsetzer study were vaccinated. That is an important problem in studying vaccine effects. Who got the vaccine? Were they the ones who see their doctors regularly, so they got vaccinated? But then might they be more likely to have AD recognized, rather than less likely?

    In our studies, we attempt to address concerns about medical exposure by requiring that the vaccinated and unvaccinated groups have the same number of annual doctor visits. We also attempt to balance the treatment of common disorders in each group, like hypertension and hyperlipidemia. We hypothesize that if those disorders are being equally treated, then study subjects must be seeing their doctors for BP and lipid assessments and medication prescriptions.

    There is also the “healthy survivor/patient effect.” This posits that those who get vaccinated do so because they see their doctors more, follow their doctor’s advice to get vaccinated, and get other dementia risk factors treated, including hypertension, high cholesterol, etc. For this reason, some are understandably skeptical that it is the vaccines and not a healthier lifestyle that is leading to reduced AD risk.

    To investigate that, we are looking at the effects of different vaccine doses. In that way, everyone has received vaccines to avoid the heathy-patient effect. We are about to submit this publication.

    Regarding the hypothesis of reduced reactivation of dormant virus, it is reasonable that vaccines could work by simply reducing the risk of contracting the diseases against which they protect. Some diseases increase the risk of AD, so decreasing disease burden could protect from AD. We agree. We are addressing this in an animal model of AD where they are raised in a vivarium, get none of these infections.

    Regarding the hypothesis that vaccines generate a widespread immune response that lowers inflammation, independently of any effect on the virus, it seems plausible that peripheral vaccines could work via increasing or decreasing central immune activity, i.e., in the brain. There is precedent for increasing CNS immune activity leading to improved AD outcomes. We believe that we take advantage of that mechanism every day in clinic with donanemab, lecanemab, and experimental therapies. Those antibodies appear to turn on CNS inflammation leading to phagocytosis of plaques and amyloid intermediaries, hence reducing plaque numbers and improving outcomes.

    The alternate hypothesis here is that vaccinations lower immune activity. That could be valuable, because central immune activation is thought to also contribute to producing neuronal dysfunction and cell death in AD. Certain peripheral cytokines/interleukins do, in fact, activate pro- and anti-inflammatory central microglia, astroglia, etc., in the brain. It is possible that vaccinations would activate peripheral cytokines that inhibit brain inflammation. Along those lines, we have shown that peripherally administered stem cells lead to reduced brain inflammation in mouse models of AD and PD, and these are associated with improved clinical outcomes. We don’t know what stem cells release, but we hypothesize that it is diffusible factors, like cytokines, that inhibit CNS inflammation.

    So it is possible that peripheral immune activity inhibits brain immune activity, but the mechanism is yet to be explicated. Of course, it is also possible that vaccinations increased CNS immune activity as per Lathe’s hypothesis below. In his comment /news/research-news/zapping-zoster-virus-dodging-dementia#comment-55081 Richard Lathe notes another hypothesis, i.e., that of “nonspecific immune boosting,” or "trained immunity." This was elegantly put forward by Mihai Netea, Peter Aaby, essentially saying that that specific microbial agents, and/or the molecules they release, provoke beneficial changes in the immune system that can last for months to years or more (Netea et al., 2020; Aaby et al., 2023).” 

    Inflammaging is an important concept about reduced immune reactivity with aging that may be relevant for a number of human diseases. It is possible that vaccinations, then, by “exercising” the immune system may lead to an increase in helpful immune activity.

    The question is how this increased immune activity in AD would be helpful. As noted above, increased immune activity can have a negative impact on AD; for example, with UTIs, sepsis, surgery, broken bones, etc. Nonetheless, this hypothesis may be correct in ways that we don’t yet understand. To address both Geldsetzer’s third hypothesis and Lathe’s above, we are studying CNS immune activity in mouse models of AD after immunizations.

    References:

    Bukhbinder AS, Ling Y, Harris K, Jiang X, Schulz PE. Do vaccinations influence the development of Alzheimer disease?. Hum Vaccin Immunother. 2023 Aug 1;19(2):2216625. Epub 2023 Jun 8 PubMed.

    Netea MG, Domínguez-Andrés J, Barreiro LB, Chavakis T, Divangahi M, Fuchs E, Joosten LA, van der Meer JW, Mhlanga MM, Mulder WJ, Riksen NP, Schlitzer A, Schultze JL, Stabell Benn C, Sun JC, Xavier RJ, Latz E. Defining trained immunity and its role in health and disease. Nat Rev Immunol. 2020 Jun;20(6):375-388. Epub 2020 Mar 4 PubMed.

    Aaby P, Netea MG, Benn CS. Beneficial non-specific effects of live vaccines against COVID-19 and other unrelated infections. Lancet Infect Dis. 2023 Jan;23(1):e34-e42. Epub 2022 Aug 26 PubMed.

    View all comments by Paul Schulz

  4. Capitalizing on the unique vaccination patterns in the U.K. and Australia, both teams demonstrate a causal connection between herpes zoster (VZV) vaccination availability and reductions in new dementia diagnoses. Both studies elegantly eliminated the most confounding variables by examining a tight subsection of the population surrounding the eligible birth dates for vaccination.

    These findings demonstrate the positive impact of VZV vaccinations and align with our recent findings on infection, immune response, and AD pathology. We have shown that Aβ is an antimicrobial peptide capable of suppressing herpes simplex 1 viral (HSV1) infection (Eimer et al., 2018). Reduction in VZV via vaccination could quell the Aβ-driven antimicrobial response and inflammation through reduction in VZV and/or decrease in HSV1, which is known to be reactivated by VZV.  According to the antimicrobial protection hypothesis (Moir et al., 2018), viruses entering the brain can instantly nucleate monomeric Aβ into amyloid aggregates (extracellular traps) owing to the role of Aβ as an antimicrobial (host-defense) peptide. By protecting against viral infections, vaccines may prevent microbial-driven beta-amyloid deposition and downstream induction of tauopathy and inflammation. Alternatively, viral infections in the brain may also directly trigger inflammation, a key feature of AD neuropathology, which  would be attenuated by vaccination.

    While the exact mechanism of action for how VZV vaccinations reduce risk for AD remains up for debate, these new corroborating studies present important advancements in examining and supporting the potential role of neuroinfectious pathogens like VZV in AD. These and previous vaccine studies support the potential for early prevention of AD through vaccinations and antimicrobials and warrants further investigation of other vaccines and pathogens.

    References:

    Eimer WA, Vijaya Kumar DK, Navalpur Shanmugam NK, Rodriguez AS, Mitchell T, Washicosky KJ, György B, Breakefield XO, Tanzi RE, Moir RD. Alzheimer's Disease-Associated β-Amyloid Is Rapidly Seeded by Herpesviridae to Protect against Brain Infection. Neuron. 2018 Jul 11;99(1):56-63.e3. PubMed.

    Moir RD, Lathe R, Tanzi RE. The antimicrobial protection hypothesis of Alzheimer's disease. Alzheimers Dement. 2018 Dec;14(12):1602-1614. Epub 2018 Oct 9 PubMed.

    View all comments by William Eimer View all comments by Rudy Tanzi

  5. Reproducibly, persons with amyloid pathology but no tau accumulation have lower CSF levels of key markers of innate immunity (e.g., TREM-2) when compared with age-matched amyloid-negative individuals (Suárez-Calvet et al., 2019; Meyer et al., 2018). One interpretation of this finding is that diminished innate immune activation may be permissive to amyloid accumulation. Might the shingles vaccines act as a sort of adjuvant, promoting development of an innate immune response to amyloid fibrils? Immunologists have known for decades that challenge with a protein antigen usually requires a particulate adjuvant to provoke an immune response, beginning with innate immune activation (e.g., Murphy K and Weaver C, Janeway's Immunobiology, 9th edition, pp. 37 et seq.). We should consider whether viral vaccines or other innate immune-enhancing approaches could reduce the occurrence of AD by preventing initiation of the Alzheimer process with amyloid oligomers.

    References:

    Suárez-Calvet M, Morenas-Rodríguez E, Kleinberger G, Schlepckow K, Araque Caballero MÁ, Franzmeier N, Capell A, Fellerer K, Nuscher B, Eren E, Levin J, Deming Y, Piccio L, Karch CM, Cruchaga C, Shaw LM, Trojanowski JQ, Weiner M, Ewers M, Haass C, Alzheimer’s Disease Neuroimaging Initiative. Early increase of CSF sTREM2 in Alzheimer's disease is associated with tau related-neurodegeneration but not with amyloid-β pathology. Mol Neurodegener. 2019 Jan 10;14(1):1. PubMed.

    Meyer PF, Savard M, Poirier J, Labonté A, Rosa-Neto P, Weitz TM, Town T, Breitner J, Alzheimer’s Disease Neuroimaging Initiative, PREVENT-AD Research Group. Bi-directional Association of Cerebrospinal Fluid Immune Markers with Stage of Alzheimer's Disease Pathogenesis. J Alzheimers Dis. 2018;63(2):577-590. PubMed.

    View all comments by John Breitner

  6. It is challenging to interpret the significance of multiple studies that seemingly show that various vaccines, including against herpes zoster and influenza, reduce the subsequent risk of dementia. If confirmed to be true, then this could have major implications for dementia research as well as for vaccine policy. The risk, of course, is that known confounders such as the healthy vaccinee effect are behind these observations, rather than the vaccines actually reducing dementia risk. Ideally, what is needed to confirm that this effect is real is a large randomized placebo-controlled trial. Additionally, or as an alternative, demonstration that such findings can be replicated in an animal model of dementia, thereby allowing delineation of the exact mechanisms involved.

    Of course, if it is true that the reason for the vaccines reducing dementia is an effect on chronic viral infection in the CNS, then the animal models used would need to recapitulate such a chronic viral infection. But while an effect on chronic herpes virus infection in the CNS might be an explanation for the effects of the zoster vaccine on dementia, what then would be the mechanism whereby an influenza vaccine could reduce dementia, with no evidence of any chronic CNS influenza infection?

    If the postulated effect is just nonspecific innate immune activation, then arguably not even a vaccine should be needed as just an innate immune activator such as a vaccine adjuvant may be needed. Also, do such findings have any significance for the Alzheimer’s vaccines in development, including with our collaborators at IMM, that directly target the abnormal aggregates of Aβ and tau protein that cause neuronal damage leading to dementia?

    Overall, even if real,  the absolute effect size on dementia seen with the zoster vaccine appears modest, and hence unlikely to compare to the large effect size we have seen in our animal models with our specific dementia vaccines. Hence, while further investigation of these interesting observations that infectious disease vaccines may have modest effects on reducing dementia are worthy of further study, even if confirmed it is unlikely that these will impact to any major degree the ongoing urgent need for vaccines targeting the specific pathways driving dementia pathology.

    View all comments by Nikolai Petrovsky

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Zapping the Zoster Virus, Dodging Dementia

Further Reading

Primary Papers

  1. Pomirchy M, Bommer C, Pradella F, Michalik F, Peters R, Geldsetzer P. Herpes Zoster Vaccination and Dementia Occurrence. JAMA. 2025 Apr 23; Epub 2025 Apr 23 PubMed.

Annotate

To make an annotation you must Login or Register.