Study Design & Methodology
Researchers used a unique natural experiment in Wales to study the causal effect of live-attenuated herpes zoster vaccination (Zostavax) on dementia. They exploited a sharp eligibility cutoff where:
- Welsh adults born before September 2, 1933 were ineligible for the vaccine
- Those born on/after September 2, 1933 were eligible for at least 1 year
This created an abrupt change in vaccination rates (from 0.01% to 47.2%) between nearly identical populations, allowing researchers to bypass traditional confounding concerns through regression discontinuity analysis.
The study analyzed electronic health records of 282,541 adults in Wales, following them for 7 years after the vaccination program began on September 1, 2013.
Key Findings
- Dementia reduction: The zoster vaccine reduced new dementia diagnoses by 3.5 percentage points (95% CI: 0.6-7.1, p=0.019), representing a 20% relative reduction over 7 years.
- Sex differences: The protective effect was substantially stronger in women than men.
- Shingles reduction: The vaccine also reduced shingles diagnoses by 2.3 percentage points (37.2% relative reduction), consistent with clinical trial findings.
- Specificity: No effects were observed for other common health conditions, suggesting the finding was specific to dementia and shingles.
- Robustness: Results remained consistent across multiple analytical approaches, varying follow-up periods, and when confirmed in a separate population using death certificate data.
Potential Mechanisms
The researchers explored three possible mechanisms:
- Healthcare pathway changes: Ruled out as the primary mechanism through multiple analyses showing shingles episodes had minimal impact on healthcare utilization or diagnostic patterns.
- Reduction in varicella zoster virus (VZV) reactivations: Supported by evidence showing:
- Delayed emergence of dementia protection (>1 year after vaccination)
- Higher dementia incidence among patients with multiple shingles episodes
- Lower dementia incidence when shingles episodes were treated with antivirals
- VZV-independent immunomodulatory effects: Supported by:
- Stronger protective effect in women (consistent with known sex differences in vaccine off-target effects)
- Differential effects based on prior influenza vaccination
- Possible differences in efficacy between patients with and without autoimmune conditions
The authors suggest both viral and immunological mechanisms likely contribute to the observed effect.
Strengths & Limitations
Strengths:
- Natural experiment design reduces confounding bias
- Large sample size (282,541 adults)
- Consistent results across multiple analytical approaches
- Findings confirmed in separate population using different data (death certificates)
Limitations:
- Results primarily apply to adults aged 79-80
- Limited to 7-year follow-up period
- Some dementia cases may be underdetected
- Findings apply only to the live-attenuated vaccine (Zostavax), not the newer recombinant vaccine (Shingrix)
TL;DR
A natural experiment in Wales found that the herpes zoster vaccine reduced new dementia diagnoses by 20% over 7 years, with stronger protection in women. The effect likely operates through both reducing viral reactivations and broader immune system modulation. These findings suggest vaccination may be a cost-effective approach to preventing or delaying dementia.
Source: https://www.nature.com/articles/s41586-025-08800-x