An extra year of secondary school lowered dementia diagnoses in a large UK study, the National Institutes of Health said Oct. 23. The effect was largest in people with higher genetic risk of Alzheimer’s disease and related dementias.
The NIH news release describes a paper in the Proceedings of the National Academy of Sciences titled “Education Reduces Dementia Incidence and Mitigates Genetic Risk.” Silvia Barcellos, a health economist at the University of Wisconsin-Madison, is the lead author. Researchers at the University of Southern California also worked on it, and National Institute on Aging grants paid for it, according to the release.
How the Study Tested Education and Dementia
People with more education have long had lower rates of dementia, but that pattern does not show that schooling caused the difference. Wealthier or healthier families may keep children in school longer. The researchers used a law to get around that problem.
A 1972 law in England, Wales and Scotland raised the minimum school-leaving age from 15 to 16 for students born after Sept. 1, 1957. The release says the change raised the number of students enrolled until age 16 by 14%. A child born a day before the cutoff could leave at 15, and a child born a day after could not. The team compared people on each side of that date, a design economists call a regression discontinuity.
The data came from more than 100,000 participants in the UK Biobank, a long-running health study of British adults, the release says. Dementia diagnoses came from medical histories, hospital records and death registries. The main comparison used diagnoses made before December 2022, when the oldest people affected by the law turned 65.
What the Study Found About Dementia Diagnoses
The release says the extra year of school “reduced AD/ADRD incidence over the following 50 years by .88%.” It does not say whether that is a percentage point or a relative change, and it does not give a confidence interval.
An earlier version of the work gives numbers that are easier to check. Barcellos and four coauthors posted a January 2025 working paper with the National Bureau of Economic Research. The authors of that paper write that the law made affected students about 0.14 years more likely to be in school, and that people it affected were 0.2 percentage points less likely to be diagnosed with dementia by age 65. The working paper and the NIH release use different wording, and the release does not say which numbers come from the published version.
The baseline matters. Before the law, the working paper puts the share of people diagnosed by age 65 at about 0.46%. That is roughly 4.6 people per 1,000. A drop of 0.2 percentage points equals about two fewer diagnoses per 1,000 people, or a 43% relative decline, both calculations from the working paper’s figures. The paper’s own chart shows values near 0.42% and 0.25%, a gap of about 0.17 points, according to the National Bureau of Economic Research’s Bulletin on Health.
The working paper also scales the result to one full year of school. That estimate is about one percentage point, which the authors call “implausibly large”. They say their 95% confidence interval rules out reductions smaller than 0.23 points. The paper also reports that diagnoses came later in the group affected by the law.
How Education Changed the Effect of Genetic Risk
The paper builds a genetic risk score from a 2019 Alzheimer’s genome study and from the APOE gene, which has a variant tied to higher risk. APOE dominates the score. In the top half of the score, 58% of people carry at least one APOE E4 variant, compared with 0.02% in the bottom half.
Before the school-leaving cutoff, a one standard deviation rise in the score raised dementia incidence by 0.24 percentage points, the working paper says. After the cutoff, it raised incidence by 0.06 points. The gap in incidence between the top and bottom halves of the score shrank from 0.18 points to 0.02 points, a reduction of about 89%, a calculation from those two figures.
Richard Hodes, director of the National Institute on Aging, said in the release that dementia risk is determined by factors beyond one’s genetic makeup. The working paper reads the result as evidence that social policy can change how much genetic risk matters.
Why Health and Income May Explain the Result
The authors do not claim to know the mechanism. The working paper reports that people affected by the law were 2.2 percentage points more likely to live in a household earning at least 31,000 pounds a year. They were 0.45 points less likely to have diabetes and 0.59 points less likely to have had a stroke or heart attack. Among people in the high genetic risk group, obesity fell by 1.64 points and diabetes by 0.65 points.
The effects on those measures in the low-risk group were smaller and not statistically significant. The authors describe these findings as suggestive. The release says the law was also tied to better socioeconomic and health outcomes that may help explain the result.
Why the Age 65 Cap Limits the Findings
Few people in the sample had reached the ages when dementia is common. The working paper says cumulative incidence was under 1% because the cohort was so young. The release says the authors urge caution because participant age was capped at 65, and that the gaps between groups could grow or shrink as the cohort ages. The working paper expects the gap to widen.
The sample also does not mirror the general public. The working paper says UK Biobank participants tend to have higher socioeconomic status and better health than the general population, and most are of European ancestry. The result applies to people who left school at 15 and were kept in school by the law, and effects could differ at other education levels.
The working paper estimates that at least 84.5% of coded dementia diagnoses in the Biobank were true cases, and that about 94% came from hospital records. Misclassification is possible, but the authors argue false positives would pull the estimate toward zero.
What is Known About Dementia Prevention
The National Institute on Aging says on its dementia information page that “we don’t yet know for certain what, if anything, can prevent dementia.” It lists older age, family history, heavy long-term alcohol use and head injuries among risk factors, and says dementia is not a normal part of aging.
The working paper opens with the scale of the problem: 131.5 million dementia cases worldwide are predicted by 2050, and the paper puts the annual U.S. cost at $305 billion. The release says researchers have long struggled to prove that education causes lower risk, which is why the 1972 law matters to the design. Because the participants were capped at 65, the final size of the effect is not yet known. The working paper says the follow-up period will matter as dementia becomes more common in the cohort.
