The Flu Shot and Alzheimer's: Can a Common Vaccine Cut Dementia Risk by More Than Half?

A growing stack of research suggests that vaccines already on pharmacy shelves — flu shots and shingles vaccines — may offer significant protection against Alzheimer's disease. The latest study, published in Neurology in April 2026, reports that adults 65 and older who received the high-dose influenza vaccine had a nearly 55% lower risk of developing Alzheimer's compared to those who received the standard dose [1][2]. If the association is real, it would represent one of the most consequential — and cheapest — tools ever identified for reducing dementia risk. But the history of Alzheimer's prevention is littered with promising findings that collapsed under scrutiny, and several critical questions remain unanswered.

The Study: What Was Actually Found

The research, led by Paul Schulz, MD, a professor of neurology at McGovern Medical School at UTHealth Houston, and Avram Samuel Bukhbinder, MD, a clinical fellow at Massachusetts General Hospital, analyzed health records from approximately 165,000 to 200,000 adults aged 65 and older [1][3]. It was a retrospective cohort study — meaning researchers looked backward through existing medical claims data rather than enrolling patients in a prospective trial.

The central finding: high-dose influenza vaccination was associated with a roughly 55% reduced risk of Alzheimer's disease over a follow-up period of approximately two years [2][3]. This builds on a 2022 study by the same team showing the standard-dose flu vaccine was linked to a 40% risk reduction over four years [4]. The high-dose vaccine, marketed as Fluzone High-Dose, contains four times the antigen of the standard formulation [2].

The protective effect was stronger among women — a finding that aligns with the broader epidemiology of Alzheimer's, which disproportionately affects women [1][3].

Not Just the Flu Shot: Shingles Vaccines Show Similar Patterns

The influenza findings do not exist in isolation. Multiple studies have linked the recombinant zoster vaccine (Shingrix), used to prevent shingles, to reduced dementia risk.

Source: Multiple peer-reviewed studies (2022-2026)

Data as of Apr 7, 2026CSV

A 2024 study using electronic health records from more than 174,000 adults across 107 U.S. health systems found that two doses of the recombinant zoster vaccine were associated with a 51% reduction in dementia risk among adults 65 and older [5][6]. A separate analysis of 436,788 propensity-score matched individuals found that AS01-adjuvanted vaccines — the adjuvant system used in Shingrix — were associated with reduced 18-month dementia risk [7].

The strongest evidence for any vaccine-dementia link comes from a natural experiment in Wales. When the UK introduced a shingles vaccination program in September 2013, eligibility was determined by exact date of birth: people born before September 2, 1933 were permanently ineligible, while those born on or after that date could receive the vaccine [8]. This near-random assignment created conditions resembling a natural randomized trial. Among 296,603 individuals, the percentage receiving the vaccine jumped from 0.01% among those one week too old to 47.2% among those just one week younger. Over seven years of follow-up, vaccination reduced the probability of a new dementia diagnosis by 3.5 percentage points (95% CI: 0.6–7.1, p = 0.019), corresponding to a 20% relative reduction [8]. The effect was again stronger in women.

A 2025 systematic review and meta-analysis published in Age and Ageing consolidated these findings, reporting that herpes zoster vaccination was associated with a relative risk of 0.53 for Alzheimer's disease — meaning roughly a 47% reduction in risk [9].

Absolute Risk: Putting the Numbers in Perspective

A 55% relative risk reduction sounds dramatic, but the absolute numbers tell a more measured story. Alzheimer's incidence among adults over 65 in the United States runs approximately 10–15 new cases per 1,000 person-years [10]. A 55% relative reduction applied to a base rate of, say, 12 per 1,000 would mean roughly 5.4 fewer cases per 1,000 vaccinated individuals — meaning about 185 people would need to receive the high-dose vaccine to prevent one Alzheimer's case over the study period.

By contrast, the Wales natural experiment — which used a more rigorous design — found an absolute reduction of 3.5 percentage points over seven years [8]. That translates to roughly 29 people needing to be vaccinated to prevent one dementia diagnosis. The gap between these numbers reflects differences in study design, follow-up duration, and the specific vaccines studied.

Why Would a Flu Shot Affect the Brain?

Several biological hypotheses have been proposed, none yet proven.

Neuroinflammation reduction. Severe influenza and other infections trigger systemic inflammation, which can cross the blood-brain barrier and accelerate neurodegenerative processes. The high-dose vaccine, by providing stronger protection against influenza infection, may reduce the frequency and severity of these inflammatory insults [2][11]. A 2022 review in Frontiers in Immunology noted that vaccines may reduce dementia risk by preventing infections that would otherwise cause neuroinflammatory cascades damaging to neurons [12].

Viral reactivation and the pathogen hypothesis. For shingles vaccines specifically, a leading theory involves the varicella-zoster virus (VZV), which remains dormant in nerve cells after chickenpox. Reactivation events — which the shingles vaccine prevents — may promote neuroinflammation that contributes to dementia. Some researchers have also linked herpes simplex virus-1 (HSV-1) to Alzheimer's pathology, suggesting that immune stimulation against related herpesviruses may have cross-protective effects [11][13].

Trained immunity and immune modulation. Vaccines and their adjuvants may reprogram aspects of innate immunity through a process called trained immunity, shifting the immune system toward a Th1-skewing profile that reduces pro-inflammatory microglial priming in the brain and improves clearance of protein aggregates like amyloid-beta [7][11].

Not amyloid-targeting. Crucially, none of these vaccines were designed to target amyloid-beta plaques or tau tangles — the proteins central to the dominant amyloid hypothesis of Alzheimer's. This is a distinct approach from drugs like lecanemab and the now-withdrawn aducanumab, which directly attack amyloid deposits [14].

The Amyloid Drug Track Record: Context for Skepticism

The history of Alzheimer's therapeutics provides good reason for caution. Over two decades, the drug development success rate for Alzheimer's treatments stood at just 2% for phase II and III trials [14]. Billions of dollars were spent on amyloid-targeting therapies that failed to show meaningful clinical benefit.

Aducanumab was controversially approved by the FDA in 2021 despite skepticism from the agency's own advisory committee; it was pulled from the market just 31 months later after Medicare declined to cover it without additional efficacy evidence [14][15]. Lecanemab, approved in 2023, showed a statistically significant but clinically modest 27% slowing of cognitive decline — an absolute difference of 0.45 points on an 18-point scale — while nearly doubling the rate of brain swelling and hemorrhages compared to placebo (17.3% vs. 9%) [15][16]. The European Medicines Agency rejected lecanemab, concluding the benefits did not justify the risks [17].

Against this backdrop, a cheap, widely available vaccine showing a 55% risk reduction naturally attracts both excitement and skepticism.

The Steelman Case for Skepticism

Eric Topol, MD, of Scripps Research, offered a direct assessment: "The evidence is not nearly as strong as the four natural experiments with shingles vaccines" [2]. His skepticism highlights several specific concerns:

Healthy-user bias. People who get vaccinated tend to be healthier overall — they exercise more, eat better, have higher education levels, and engage more with the healthcare system. These same behaviors independently reduce dementia risk. Disentangling the vaccine effect from the "healthy vaccinee" effect is the central methodological challenge [12][18]. One study found a dose-response relationship showing a 14% higher dementia rate after one vaccination and gradually decreasing with successive shots — but similar patterns appeared with hip fracture and cancer outcomes, suggesting residual confounding rather than a true vaccine effect [18].

Observational design limitations. The UTHealth flu vaccine study used retrospective claims data, which can misclassify diagnoses and lacks information on important confounders like education, socioeconomic status, and mortality [2][3]. Claims-based studies cannot control for variables they don't measure.

Short follow-up. The roughly two-year follow-up period for the high-dose flu vaccine finding is brief for a disease that develops over decades. Longer follow-up is needed to determine whether the effect persists, grows, or attenuates [2].

No independent replication. The 55% finding comes from one research group building on their own prior work. Independent replication by teams without ties to the original investigators has not yet been published [2].

The Wales natural experiment partly addresses healthy-user bias because eligibility was determined by birth date, not individual health-seeking behavior [8]. But even that study cannot fully exclude all confounders, and its 20% relative risk reduction is considerably smaller than the 55% reported for the flu vaccine.

Who Was Studied — and Who Was Left Out

The UTHealth study included men and women aged 65 and older [1][3]. However, the published news coverage does not detail the racial or ethnic composition of the cohort. This is a significant gap given the disparities in Alzheimer's burden.

Black Americans face an Alzheimer's incidence rate of approximately 22.9 per 1,000 person-years for women and 21.5 for men, compared to 16.4 and 15.5 for White Americans [19]. A higher proportion of African Americans (37%) carry the APOE ε4 allele — the strongest known genetic risk factor for late-onset Alzheimer's — compared to 20–23% among other groups [19][20]. Female APOE ε4 carriers face 1.5 times higher Alzheimer's risk and more amyloid-beta plaques than male carriers [20].

Whether the observed vaccine effects hold across racial groups, APOE ε4 carriers, and lower socioeconomic populations remains unknown. If health systems act on these findings without addressing access disparities, the benefits could flow disproportionately to populations already better served by preventive care.

Source: Lancet Global Health / Alzheimer Disease International

Data as of Apr 7, 2026CSV

The Scale of What's at Stake

Approximately 58 million people worldwide lived with dementia as of 2020, a figure projected to reach 83 million by 2030 and 152 million by 2050 [21]. In the United States alone, 7.2 million Americans aged 65 and older live with Alzheimer's dementia [10]. The global macroeconomic burden runs into the hundreds of billions of dollars annually [21].

Even a modest absolute reduction in incidence applied at the population level — through a vaccine that already exists and costs a fraction of amyloid-targeting drugs — could prevent millions of cases over the coming decades.

Source: OpenAlex

Data as of Jan 1, 2026CSV

Academic interest in the vaccine-Alzheimer's connection has surged, with more than 13,000 papers published on the topic since 2011 and a peak of over 2,000 publications in 2023 [22].

Policy Implications and Access Questions

If confirmatory evidence mounts, the policy implications are substantial. The high-dose influenza vaccine is already recommended for adults 65 and older by the CDC, and Shingrix is recommended for adults 50 and older [1][5]. Adding dementia prevention to the rationale for these vaccines would not require new products — it would require updated messaging and potentially expanded access programs.

However, supply constraints are already an issue. Some older adults receive standard-dose flu vaccines when high-dose supplies run short in certain regions [3]. If evidence solidifies that the high-dose version offers meaningfully greater dementia protection, ensuring equitable supply would become a public health priority.

The cost differential is also relevant. A high-dose flu shot costs roughly $50–70 at most pharmacies, while the two-dose Shingrix series runs approximately $300–400 [5]. Compare this to lecanemab's $26,500 annual price tag plus monitoring costs, or donanemab at $32,000 per year [15]. If vaccines offer even a fraction of the protection these observational studies suggest, the cost-effectiveness ratio would be extraordinary.

What Comes Next: The Regulatory and Scientific Path Forward

There is currently no pathway for the FDA to approve influenza or shingles vaccines with an Alzheimer's prevention indication based on observational data alone. The standard regulatory process for a new vaccine indication requires a three-phase clinical trial program, beginning with small safety studies and culminating in large Phase III randomized controlled trials [23]. For a disease with a decades-long prodromal period, such trials would need to follow thousands of participants for years — a logistically and financially enormous undertaking.

A more realistic near-term path would involve large pragmatic trials or analyses of existing national health databases in countries with comprehensive vaccination and health records — such as the UK, Denmark, or South Korea — to replicate the observational findings with stronger controls for confounding [8][9]. The Wales natural experiment model could potentially be replicated using birth-date eligibility cutoffs in other vaccine programs.

Preregistered studies testing the vaccine-dementia hypothesis prospectively are needed. Without them, the 55% figure remains an association — striking, but unproven as causal.

The Bottom Line

The accumulating evidence linking routine vaccines to reduced Alzheimer's risk is among the most intriguing findings in dementia research. The 55% risk reduction reported for the high-dose flu vaccine is the latest and largest such estimate, but it comes from an observational study design that cannot establish causation. The strongest causal evidence comes from the Wales shingles vaccine natural experiment, which found a smaller but still meaningful 20% reduction with a design that largely controls for healthy-user bias.

The biological plausibility is there: infection prevention reduces neuroinflammation, and immune modulation may help the aging brain clear toxic protein aggregates. But plausibility is not proof. Independent replication, longer follow-up, diverse study populations, and ideally randomized trial data are needed before any clinical guidelines change.

What makes this line of research unusual in the Alzheimer's field is that the interventions already exist, are widely available, and have well-established safety profiles. If the findings hold up, the gap between evidence and action could close faster than for any novel therapeutic. If they don't, the vaccines remain valuable for their original purposes. The risk of pursuing this research further is low. The cost of ignoring it could be measured in millions of preventable cases.