Design: Prospective birth cohort analysis
Population: 1,356 adults born in 1946 (p-tau217) and 424 with amyloid PET
Primary outcome: Plasma p-tau217 at age 69–71
Effect: No notable associations; EBV and amyloid PET OR 2.84 (95% CI 0.95–8.52)

The 1946 British birth cohort measured antibodies against 14 pathogens, including herpes viruses, Epstein-Barr virus, cytomegalovirus and Helicobacter pylori, at age 60 to 64. About a decade later, 1,356 participants had plasma p-tau217 measured and 424 had amyloid PET.

There were no notable associations between serostatus, overall pathogen burden or recent immune activity and either outcome. Odds of amyloid positivity were numerically higher with Epstein-Barr virus and Toxoplasma seropositivity, but the confidence intervals included no effect. A few interactions with APOE ε4 and education were noted and are exploratory.

The infection hypothesis of Alzheimer disease has attracted attention and some patient interest in antivirals. This cohort, with well-characterised biomarkers in people of the same age, does not support it for these common exposures. It does not rule out effects of active or recurrent infection, or of shingles specifically.

Advise patients that past exposure to common infections has not been shown to cause Alzheimer pathology.
Do not order infection serology as part of a cognitive assessment.
Continue recommending vaccination, including shingles vaccine, on its own merits.
Treat reported interactions with APOE ε4 as hypotheses, not findings.

Why it matters

It weighs against a popular theory using biomarkers rather than diagnosis codes.

The statistics, in plain English

The Epstein-Barr odds ratio of 2.84 looks large, but its interval runs from 0.95 to 8.52, so it is compatible with no effect. With 14 pathogens and several outcomes tested, a few apparent associations would be expected by chance alone.