Shingles Vaccine and Dementia: Cheap Insurance the Science Is Starting to Back

Bottom line: a vaccine we already use for shingles may also delay or prevent dementia. Three natural experiments now point the same way: Wales, Australia and, newly reported, New Zealand. The evidence is strong enough to take seriously, not yet strong enough to call proven. The science has time to run. The economics says do it now. A two-dose course costs a few hundred dollars, and a single year of dementia in residential care costs about $88,000 in Australia.

What the science says

The hard part with vaccine and dementia studies is bias. People who get vaccinated tend to be healthier and more engaged with care anyway (the “healthy vaccinee effect”). The best recent work gets around this with natural experiments, where a birth-date cut-off decided who could get the jab.

  • New Zealand (conference report, November 2026): Dr Richard Haarburger, working with Stanford and Victoria University of Wellington, presented results at the Clinical Trials on Alzheimer’s Disease conference in Boston. They studied New Zealand’s April 2018 funded programme for ages 65 to 80, using a birthday cut-off around age 81. Uptake was 0.2% just outside the cut-off and 26% just inside. Over five years, the vaccine cut the chance of a dementia diagnosis by 37.6%, about 25 fewer cases per 1,000 people vaccinated. Effects looked stronger in women. This is a conference presentation, not a randomised trial, and I could not find a peer-reviewed paper yet. Treat it as preliminary.
  • Wales (Nature, 2025): Supply was limited in 2013, so the live vaccine (Zostavax) went to people who turned 80 just after the start date and was refused to those just before. Across more than 280,000 older adults, recipients were about 20% less likely to be diagnosed with dementia over seven years (3.5 percentage points). Uptake in the eligible group was about 47%. The effect was stronger in women.
  • Australia (JAMA, 2025): Same trick with the 1 November 2016 start of free Zostavax for ages 70 to 79. In 101,219 patients, 3.7% of eligible people were diagnosed with dementia versus 5.5% of ineligible people over 7.4 years. No clear sex difference.
  • Shingrix vs Zostavax (Nature Medicine, 2024): The newer recombinant vaccine was linked to 17% more diagnosis-free time, about 164 extra days among those later diagnosed.
  • Disease course (Wales, late 2025): Vaccination was linked to less mild cognitive impairment, and people already living with dementia were nearly 30% less likely to die from it.
  • Frail older adults (Annals of Internal Medicine, 2026): Among 509,926 people aged 66+ entering US skilled nursing facilities, Shingrix recipients had a 24% lower dementia diagnosis rate over four years.
  • Other vaccines: A meta-analysis of 17 studies and 1.86 million people found vaccinated adults had a pooled hazard ratio of 0.65 for dementia. Influenza (about 0.74) and shingles (0.69) were among the individually significant ones. Early RSV vaccine data show a similar signal.

Why it might work

  • The shingles virus (varicella-zoster) lies dormant in nerves for life. Reactivation causes inflammation and may injure blood vessels in the brain.
  • Vaccines may also have broader “off-target” effects on the ageing immune system. The adjuvant in Shingrix is one candidate, which would explain the RSV signal.
  • Mechanism is unresolved. Treat these as hypotheses.
Percentage reduction in cases of dementia if this risk factor is eliminated
Population attributable fraction of potential risk factors for dementia.

Why caution is still right

  • No randomised trial yet. Ireland’s health technology regulator (HIQA) reviewed ten vaccination studies in 2026. All showed 20% to 39% lower dementia risk, but it stressed that confounding cannot be ruled out.
  • The natural experiments used the old live vaccine. Whether Shingrix does the same is less certain.
  • Estimates vary: about 20% in Wales, about 38% in the preliminary New Zealand work. Big observational headline numbers (some above 50%) are probably inflated by bias.
  • It is unclear whether vaccination prevents dementia or just delays it past the follow-up window, and for how long.
  • A 33,600-person GSK trial in Finland, in adults aged 76 and over, is under way and will take years to report.

The economics: vaccine cost versus dementia care

  • Australia: Dementia cost the economy $14.25 billion in 2016 (about $35,550 per person), with 400,833 people affected. Projections reach $36.8 billion and 1.1 million people by 2056 (2016 dollars). Direct health and residential care for one person in residential aged care was estimated at about $88,000 a year.
  • New Zealand: About 83,000 people live with dementia, heading to nearly 170,000 by 2050. Alzheimers NZ projects around NZ$10.65 billion a year by 2050 in inflation-adjusted terms.
  • The vaccine: Shingrix privately costs up to A$560 in Australia and roughly NZ$600 to $800 in New Zealand for two doses (about £460 in the UK).

Here is a simple back-of-envelope test. If we assume the link is causal, the cost to avoid one dementia diagnosis is the vaccine price divided by the absolute risk reduction. I used the retail price, which overstates what governments pay, and a range of effects: 1.0 percentage point (low case), 1.8 (the Australian eligible versus ineligible gap), 2.5 (the preliminary New Zealand estimate) and 3.5 (the Wales estimate for recipients).

Absolute risk reductionPeople vaccinated per diagnosis avoidedCost at $560Cost at $800
1.0 points100$56,000$80,000
1.8 points56$31,111$44,444
2.5 points40$22,400$32,000
3.5 points29$16,000$22,857
Illustrative only. Assumes causation, retail prices, and ignores the shingles benefit itself.

Even the worst case here (about $80,000) is less than one year of residential care at the 2018 figure of $88,000, and that figure has only gone up. The realistic case is a few months of care. Three caveats: a “diagnosis avoided” may really be a diagnosis delayed (still valuable, since each year of delay saves a year of care); the effect may not hold for Shingrix; and none of this counts the avoided shingles and nerve pain, which is the main reason to vaccinate anyway.

For scale, Australia’s national modelling found a hypothetical 5% cut in new dementia cases among people 65+ would save $26.8 billion over 20 years and $120.4 billion by 2056 (constant 2016 dollars). A vaccine that delivers even a fraction of that is a serious public health tool.

The catch: HIQA notes that no national immunisation advisory group has put dementia into its cost-effectiveness models, and the UK committee (JCVI) judged last year that more research was needed before changing its advice. So today’s funding decisions mostly ignore the dementia benefit.

The science has time to run. The economics says do it now.

Two clocks are running, and they run at different speeds. This is my argument, not a finding from the studies.

  • The science clock is slow. Dementia develops over years, so a proper randomised trial needs years of follow-up. The Finnish trial of 33,600 people will not report soon. Regulators like HIQA and the UK’s JCVI are, reasonably, waiting for it.
  • The dementia clock is fast. New Zealand’s caseload is heading from about 83,000 to nearly 170,000 by 2050. Australia’s was 400,833 in 2016 and is projected at 1.1 million by 2056. Every year of waiting, another cohort ages past the age at which it could have been vaccinated.
  • The downside of acting is small. Shingrix already prevents shingles, which has a lifetime risk of about 1 in 3 in New Zealand. Protection lasts around 10 years. If the dementia effect turns out smaller than hoped, the shingles benefit remains. The dementia effect is a free option on top.
  • The downside of waiting is lopsided. If the effect is real, even at the cautious end of the table above, each year of delay forfeits avoided diagnoses that cannot be recovered later for those already past the age window.
  • Good decisions do not need certainty. Governments fund treatments every day on evidence weaker than a randomised trial when the cost is low and the potential gain is large. Waiting for proof is also a decision, and it has a price.

The fair counter-argument is Pharmac’s. Money is finite, and every dollar spent on widening shingles access is a dollar not spent on another medicine with firmer evidence. That is a real trade-off. But it argues for pricing the dementia option into the decision, which no agency has done yet, rather than ignoring it.

New Zealand versus Australia

AustraliaNew Zealand
Vaccine fundedShingrix on the National Immunisation Program since 1 Nov 2023Zostavax from 1 April 2018, with a catch-up for ages 66 to 80 until March 2020; Shingrix since 1 Dec 2022
Who gets it free nowEveryone 65+, First Nations people 50+, immunocompromised adults 18+People aged exactly 65, plus some immunocompromised adults 18+ (from July 2024)
People coveredAlmost 5 million, funded at $826.8 millionRoughly 55,000 turning 65 each year, plus about 15,000 immunocompromised in the first two years
Uptake2.5 million doses in year one; about 1 in 3 eligible 65+ had at least one dose76,646 doses from Dec 2022 to Aug 2024
Miss the windowNot a concern for 65+Pay about $600 to $800 privately

New Zealand’s 64 and 66 year olds miss out. In August 2026 Pharmac said it could not widen shingles access because of cost and competing priorities. Five widening proposals had positive clinical recommendations. Widening remains on its investment list, with no date. The irony: New Zealand’s 2018 catch-up programme is what produced some of the best dementia evidence, and the catch-up has long since closed.

Who makes these decisions

  • Australia: The TGA registers the vaccine. ATAGI gives clinical advice. The law requires a recommendation from the Pharmaceutical Benefits Advisory Committee (PBAC) before a vaccine can go on the National Immunisation Program. Government then approves, funds and lists it, so the Health Minister is the final political decision-maker.
  • New Zealand: Medsafe approves. Pharmac, a Crown agency working to a fixed budget, decides funding, advised by its clinical committees including the Immunisation Advisory Committee. Pharmac is not bound by that advice and ranks every request against other medicines and vaccines.
  • United Kingdom (for comparison): The JCVI advises and government decides. The NHS programme is being extended to people aged 60 and over from 2028, but JCVI has not yet counted dementia in its reasoning.

My read: the structural problem is the budget silo. Pharmac pays for the vaccine from a fixed medicines budget, but any dementia savings would fall to aged residential care, which sits elsewhere. A decision-maker judged on its own budget cannot see the full return. Australia’s larger, centrally funded program makes the trade-off easier, though it too has not formally priced the dementia effect.

What this means

  • Shingrix prevents shingles and its complications in the over-50s. That alone justifies it. Dementia is a possible bonus, not a promise.
  • If you are 50+ and have not had both doses, talk to your GP. Do not wait for the Finnish trial.
  • Let the science run, but do not make people wait for it. Governments should back more trials like the Finnish one and, in parallel, model the dementia effect now, with honest uncertainty ranges.
  • New Zealand’s “exactly 65” rule is hard to defend on the science or the economics, especially when New Zealand’s own data helped build the case.

References

  1. Donnelly L. Shingles vaccine cuts dementia risk by up to 40pc, study finds. The Telegraph, 17 November 2026. telegraph.co.uk (reports the Haarburger et al. New Zealand analysis presented at CTAD)
  2. Eyting M et al. A natural experiment on the effect of herpes zoster vaccination on dementia. Nature 2025. nature.com
  3. Wales study coverage: NBC News; Medicine Today (uptake figures)
  4. Pomirchy M et al. Herpes zoster vaccination and dementia occurrence. JAMA 2025. doi:10.1001/jama.2025.5013; results summary: Live Science
  5. Taquet M et al. The recombinant shingles vaccine is associated with lower risk of dementia. Nature Medicine 2024. PubMed
  6. Cardiff University. Shingles vaccine may help prevent and slow dementia (Dec 2025). cardiff.ac.uk
  7. Brown University. Shingles vaccine and dementia in skilled nursing facilities, Annals of Internal Medicine (June 2026). brown.edu
  8. Wu X et al. Adult vaccination as a protective factor for dementia: meta-analysis. Frontiers in Immunology 2022. frontiersin.org
  9. Gavi VaccinesWork. Eight vaccines linked to lower risk of dementia. gavi.org
  10. HIQA. Scoping report: herpes zoster and dementia (May 2026). hiqa.ie
  11. Petousis-Harris H. Is there something about the shingles vaccine that reduces dementia risk? University of Auckland, Sept 2026. auckland.ac.nz
  12. Brown L et al. The Economic Cost of Dementia in Australia 2016 to 2056 (NATSEM, for Alzheimer’s Australia). dementia.org.au
  13. Direct health and residential care costs of people living with dementia in Australian residential aged care (2018). Aged Care Guide summary
  14. Alzheimers NZ. New Zealand dementia data. alzheimers.org.nz
  15. Te Whatu Ora. Immunisation Handbook: zoster (lifetime risk of shingles). tewhatuora.govt.nz
  16. Australian Department of Health. Shingles vaccine now free for nearly 5 million Australians (about 10 years of protection); eligibility; immunisation policy and governance
  17. NCIRS. Shingrix uptake strong following NIP inclusion
  18. PBS. Role of ATAGI in requests to list vaccines
  19. Pharmac. Older people to benefit from funded access to shingles vaccine (2017); Shingles vaccine funding; August 2026 vaccine decision; Immunisation Advisory Committee record (June 2025)
  20. Medsafe. Summary of Shingrix adverse events (Dec 2024) (dose numbers)

Follow the serialisation. Free.

Energy on Mondays. The Debunk on Wednesdays. AI, food and health long reads on weekends. Delivered to your inbox.

Scroll to Top