Linking the Shingles Vaccine and Dementia Risk

Hi Phoenix friend,

In 2013, Wales drew a line through a single birthday. Nobody meant to run an experiment on the human brain. That line turned into the strongest causal-grade signal we have that a cheap, already-approved vaccine can lower dementia risk.

This is the full written synthesis of the shingles vaccine and dementia research. Prefer to watch? Here's the 24-minute deep dive on YouTube (studies on screen, full stories):

Introduction

This is not another "people who did X happened to live longer" story. The evidence here is built differently: a bureaucratic birthday cutoff accidentally created something researchers almost never get, a near-randomized test of what happens when one group of older adults gets the shingles vaccine and another doesn't.

Result: the vaccinated group got about 20% less dementia.

What follows gives you all of it: the numbers that hold up, where the evidence gets thin (including the gap that's personal for us carriers), and what to actually do on a Tuesday instead of waiting for the 2029 trial.

The Birthday Cutoff That Created a Near-Randomized Dementia Test

Wales, 2013. Adults born on or after September 2nd, 1933 were eligible for a free shingles vaccine. Born one day earlier, locked out for life.

Nobody designed this as a brain experiment. It was a budget and supply decision.

But it created something researchers almost never get: two groups of people separated by nothing except which side of a birthday they landed on. Same generation, same health history, same preventive-care habits across every measured dimension. Vaccine receipt jumped from 0.01% in people just one week too old to 47.2% in people just one week younger [1]. The only thing that differed was access to the shot.

The result: a 3.5 percentage-point drop in new dementia diagnoses over seven years, a 20% relative reduction (95% CI 6.5 to 33.4) [1].

The authors called it evidence "less vulnerable to confounding and bias" than the usual associational studies [1]. That framing matters. This is not an RCT. But the regression-discontinuity design means the two groups are effectively interchangeable on every measured characteristic except vaccine access. It is the closest thing to a coin-flip randomization you can engineer from real-world health records.

A vaccine people get to avoid a painful rash cut new dementia diagnoses by a fifth. By accident.

How the Finding Held Up: Australia and 100 Million People

One study in one country could be a fluke. So people went and checked.

Australia had its own birthday cutoff, a different date (November 2nd, 1936) and a completely different health system. The same regression-discontinuity method found the same pattern: shingles vaccine eligibility cut new dementia diagnoses by 1.8 percentage points over 7.4 years (95% CI 0.4 to 3.3, p=0.01) [4]. The JAMA authors called that "more likely to be causal" than ordinary associational evidence [4].

They also tested it against roughly 15 other common conditions and preventive screenings. The vaccine moved dementia. Nothing else [4]. If healthier people simply get more vaccines across the board, you'd expect the effect to spread everywhere. It didn't.

A 2.5-million Korean nationwide cohort (observational, so associational) found the live shingles vaccine tracked with about 25% lower Alzheimer's hazard (aHR 0.75, 95% CI 0.71 to 0.78) [7]. Then a meta-analysis pooled 21 studies covering 104 million participants [5]. Herpes zoster vaccination tracked with about 24% lower risk of any dementia and 47% lower risk of Alzheimer's specifically [5]. Of every adult vaccine analyzed, the shingles shot had the strongest Alzheimer's signal.

⚠️ CAVEAT: The large-cohort and meta-analysis numbers are observational. The Wales and Australia natural experiments carry the near-causal load. The big cohorts point the same direction but can't rule out healthy-vaccinee bias.

One more layer: this isn't one lab's pet result. The whole field is leaning the same direction.

Shingrix Looks Even Better, and the Benefit Spans the Full Disease Course

The live vaccine from those birthday experiments has mostly been replaced by Shingrix, a newer recombinant shot. The obvious question: does the new one do anything for the brain?

A US natural experiment used the October 2017 switch from live to recombinant as its dividing line. Shingrix tracked with a 17% increase in time lived free of a dementia diagnosis, translating to 164 extra days without a diagnosis in people who eventually developed dementia [3]. It also outperformed flu vaccines and Tdap on dementia risk [3]. Something specific about the shingles shot stood out.

The Welsh team also looked beyond new diagnoses at the full arc of disease. Two findings. The vaccine tracked with fewer mild cognitive impairment (MCI) diagnoses, down 1.5 percentage points over nine years (p=0.006) [2]. And among people already living with dementia, being eligible for the vaccine tracked with fewer deaths from dementia, down 8.5 percentage points over nine years (p=0.036) [2]. The deaths estimate carries wide confidence intervals from a smaller sub-group, so hold it loosely.

But the shape of it: benefit at the front of the disease, the middle, and near the end. That is not what a coincidence usually looks like.

Why a Rash Shot Might Touch Your Brain

Nobody has proven the mechanism. Two hypotheses are in play. Both are published, biologically plausible, and not yet confirmed in humans.

The first: if you had chickenpox, the varicella-zoster virus never actually left your body. It went dormant inside your nerve cells and stays there. With age, it can reactivate quietly, no rash, no obvious symptoms, acting as what a 2025 mechanism review calls a "renewable peripheral immune stressor" [8]. Repeated quiet reactivations may keep the brain's immune cells in a chronically primed, irritated state. The vaccine's job in this theory: suppress that reservoir and reduce what the same review calls "cumulative inflammatory tone" [8].

The second: trained immunity. The shot may give the innate immune system a broad anti-inflammatory reset, non-specific to the varicella-zoster virus, lowering inflammatory tone more generally [9]. Both mechanisms could be operating at once. The data can't yet separate them [8].

What both share: inflammation is the dial. And you can measure it.

High-sensitivity CRP is the most accessible systemic-inflammation proxy. Phoenix optimal: 0 to 0.5 mg/L (acceptable below 1.0). For APOE4 carriers, who tend to run hotter on neuroinflammation, this is a trackable window into the exact fire both theories are describing.

✅ ACTION STEP: Get an hs-CRP test. Phoenix optimal: 0 to 0.5 mg/L. Log and trend it in Phoenix Bloodwork.

Three Caveats the Headlines Don't Print

1. No carrier-specific data exists. Every number above, the 20%, the Australia replication, the 100-million meta-analysis, came from the general older population. Not one study broke the vaccine effect out by APOE4 status. The only study with any APOE4 data looked at the shingles disease, not the vaccine, and found the APOE4 interaction was "not consistent between women and men" [6]. We don't have a carrier-specific number. Anyone who claims one is guessing.

The other side: APOE4 runs hotter on neuroinflammation than non-carriers. The proposed mechanism, quieting inflammatory tone, is if anything more relevant for us. The gap isn't a no. It is an open question.

2. The sex pattern is real but unsettled. Several studies showed a stronger effect in women: 22% more diagnosis-free time for women versus 13% for men in the Shingrix data [3], and a similar lean in the Welsh data [1]. The Australian JAMA replication, using the same rigorous design, found "no evidence of a significant treatment effect heterogeneity by sex" [4]. If you're male, don't write yourself off a pattern that failed to replicate in one of the most carefully designed checks.

3. The protection fades. In the Korean cohort of 2.5 million, the protective effect attenuated over time and was weaker in smokers and drinkers [7]. A shot is a head start, not a force field. What you do with diet, sleep, exercise, and metabolic health in the years after determines how long that head start holds.

What to Do Right Now

You don't have to wait for the 2029 trial.

The shingles vaccine is already recommended for older adults to prevent shingles. That recommendation exists with or without any brain benefit. For most people in the right age band, this isn't an exotic intervention. It's already on the preventive care list.

A cost-effectiveness model (assumptions-based modeling, not a measured outcome) estimated vaccination could avert around 12% of dementia cases and pencils out as cost-effective even starting at age 50 with mild cognitive impairment [10]. File under encouraging, not proof.

The randomized trial the field needs? It's now being built. DAN-ZOSTER (NCT07485283) is recruiting around 162,000 people to test the recombinant shot with a pre-specified dementia endpoint. Results expected around 2029 [11].

Think about this like weighing odds. The downside of a vaccine already recommended for older adults, with a known safety profile, is low. The upside is the strongest near-causal dementia prevention signal we've ever had. For people in the eligible age band, that's a conversation worth having with your doctor now.

For timing and availability in your country: ask the Phoenix Community, not a YouTube video or a blog post.

Key Takeaways

💡 Quick-Start Protocol (This Week):

1. Check your shingles vaccination status. If you're in the recommended age band, ask your doctor specifically about the recombinant version (Shingrix). This is probably already on your preventive care list.

2. Track your hs-CRP. Both mechanistic theories point to inflammation as the shared dial. Phoenix optimal: 0 to 0.5 mg/L. Log and trend it in Phoenix Bloodwork.

3. Don't let the APOE4 gap paralyze you. No study has measured our specific number yet. That's an open question, not a no. No biological reason exists to think carriers are excluded.

4. Vaccination is a head start, not a finish line. The Korean data showed protection fades and erodes faster in smokers and heavy drinkers. The lifestyle dials still matter the decade after.

5. Watch DAN-ZOSTER (NCT07485283). The first randomized trial with a pre-specified dementia endpoint. Results expected 2029. The Phoenix Clinical Trials engine surfaces this and similar studies for you.

Track Inflammation, Get in the Trial Loop

If you carry APOE4 and you're tracking your dementia risk, Phoenix Bloodwork lets you log hs-CRP, set a goal, and watch the trend over time. Both mechanistic theories in this story point directly at that marker.

And because no study has tested the vaccine effect specifically in APOE4 carriers, the best move is to get counted in the data. Phoenix's Clinical Trials engine surfaces registered studies including DAN-ZOSTER (NCT07485283) and helps you find ones you're eligible for. Over 500 APOE4 carriers are already using it to stop being spectators in their own genetics.

If you are not part of the Phoenix Community yet, check us out: we are the largest community of APOE4 carriers actively tracking and optimizing our health via the Phoenix App. Join us here.

Cheers,
Kevin

Sources

Match each numbered citation in the article to the same number below. Select View source to open the original paper or trial record.

1. Eyting M, Xie M, Michalik F, Hess S, Chung S, Geldsetzer P. "A natural experiment on the effect of herpes zoster vaccination on dementia." Nature, 2025. View source

2. Xie M, Eyting M, Bommer C, Ahmed H, Geldsetzer P. "The effect of shingles vaccination at different stages of the dementia disease course." Cell, 2025. View source

3. Taquet M, Dercon Q, Todd JA, Harrison PJ. "The recombinant shingles vaccine is associated with lower risk of dementia." Nature Medicine, 2024. View source

4. Pomirchy M, Bommer C, Pradella F, Michalik F, Peters R, Geldsetzer P. "Herpes Zoster Vaccination and Dementia Occurrence." JAMA, 2025. View source

5. Maggi S, Fulop T, De Vita E, Limongi F, Pizzol D, Di Gennaro F, Veronese N. "Association between vaccinations and risk of dementia: a systematic review and meta-analysis." Age and Ageing, 2025. View source

6. Yeh TS, Curhan GC, Yawn BP, Willett WC, Curhan SG. "Herpes zoster and long-term risk of subjective cognitive decline." Alzheimer's Research & Therapy, 2024. View source

7. Oh J, Lee K, Yeo D, Cho J, Kim TH, Lee J, Lee H, Woo HG, Yon DK. "Herpes zoster vaccination and cognitive disorders in older adults." Alzheimer's & Dementia, 2026. View source

8. Huang X, Gu BJ. "Shingles vaccination and neuroimmune vulnerability." Trends in Neurosciences, 2025. View source

9. Ma YN, Karako K, Song P, Xia Y. "Can the herpes zoster vaccination be a strategy against dementia?" Drug Discoveries & Therapeutics, 2025. View source

10. Wu Y, Yao Y, Liu J. "Cost-Effectiveness Analysis of Recombinant Zoster Vaccine at Age 50 for Chinese Adults with Mild Cognitive Impairment: A Modelling Study." Vaccines (Basel), 2026. View source

11. DAN-ZOSTER (NCT07485283). ClinicalTrials.gov. [View source