Dementia Prevention

The evidence that dementia risk is modifiable: partly through a known list of lifestyle and medical risk factors, and partly through one unexpected intervention, the shingles vaccine. Also covers the brain-health habits that recur across that evidence. Research notes, not advice; population, design and effect size stay attached to each claim.

The trend behind the headlines

• Forecast: dementia cases may triple to 153m by 2050 (Global Burden of Disease projection, Lancet Public Health 2022) as populations age.
• Observed: in rich countries the age-adjusted incidence has fallen sharply over recent decades. The forecasts generally assume that decline stops; The Economist's Leader argues the new evidence says it need not. (Identification of the underlying studies is the vault's; the Leader names none.)

Modifiable risk factors (the Lancet Commission list)

The 2024 Lancet Commission on dementia prevention, intervention and care (Livingston et al.) identifies 14 modifiable risk factors which together account for an estimated ~45% of dementia cases worldwide (the 2020 edition listed 12 factors at ~40%; 2024 added high LDL cholesterol and untreated vision loss). The Economist Leader draws its examples from this list without naming it:

• Medical: treating hearing loss, depression, high blood pressure, high cholesterol.
• Habits ("Brain health habits"): healthy diet, more exercise, keeping the brain active, drinking less.
• Caveat the Leader makes explicit: these need years of adherence before the brain benefits, and getting people to adopt and keep habits is hard. The Commission's 45% is a population-attributable fraction from observational data, not a trial result.

The shingles-vaccine finding

• Effect size: a course of the shingles vaccine is associated with about 20% lower risk of dementia.
• Design and population: the strongest evidence is a natural experiment in Wales (Eyting et al., Nature, 2025, ~280,000 people): eligibility for the live-attenuated Zostavax vaccine depended on being born on or after 2 September 1933, so people either side of the cutoff are otherwise comparable. Over ~7 years, eligibility cut new dementia diagnoses by roughly 3.5 percentage points absolute, ~20% relative, more in women. Quasi-experimental (regression discontinuity), not randomised. An Australian replication (Queensland, JAMA 2025) used the same design.
• Which vaccine: most evidence concerns the older live vaccine; the current Shingrix (recombinant protein, cannot cause infection) "may be at least as powerful" per a US health-record comparison (Taquet et al., Nature Medicine, 2024: lower dementia-diagnosis rate over 6 years vs Zostavax).
• Mechanism (debated): either dormant varicella-zoster virus damages or inflames the nervous system and vaccination suppresses it, or vaccination broadly primes the immune system against other dementia-implicated pathogens.
• What is unproven: causation (no RCT); the optimal age; whether boosters are needed; Shingrix vs Zostavax head-to-head. The Leader calls for randomised trials.

The policy gap

• Shingles itself hits 20–30% of unvaccinated people; vaccination is cost-effective from about age 55 on shingles alone (several studies, per the Leader).
• Britain lowered eligibility from 70 to 65 in 2023 and has been slow to reach the late-60s cohort; 17 of 27 EU members recommend the vaccine, several only for over-65s.
• Cost of a two-dose Shingrix course: £320 ($430) to the NHS; ~$270 to US federal programmes; ~£460 UK retail.

Pharmaceutical prevention (adjacent)

Eli Lilly is testing donanemab (anti-amyloid) in people without symptoms to see whether early treatment delays or prevents Alzheimer's: the drug-industry version of the prevention bet. See Eli Lilly Is Reinventing the Pharma Business (Economist) and, for the broader "act on the system, not the symptom" framing, GLP-1s and The Brain May Be About to Have Its Ozempic Moment (Economist).

Open questions for this vault

• A second independent source on the shingles-dementia association (the Nature and Nature Medicine papers themselves, or a trial registration) before a thesis forms.
• Whether any of the 14 risk factors has randomised evidence of dementia reduction (the Commission's estimates are observational).

Sources

• A No-Brainer for Protecting Your Brain (Economist) — 2026-07-11, Leaders