Extra Schooling Cut Dementia Incidence and Blunted Genetic Risk, Study Finds

A change in school-leaving law decades ago appears to have altered who developed dementia later in life, and it may have softened the effect of risky genes too.

That is the finding of a study in the Proceedings of the National Academy of Sciences by Barcellos, Carvalho, Langa, Nimmagadda and Turley. It is an observational study in people. It uses a quasi-experimental design rather than a randomised trial. The abstract available to us does not give the number of participants, so we cannot report it.

The researchers wanted causal evidence. According to the paper, there is little of it on what protects people against Alzheimer's disease and related dementias. Most earlier work simply notes that more educated people have less dementia. That could reflect other differences between people, such as income or health habits.

To get around this, the team used a compulsory schooling reform. A reform that raises the required years of school creates a sharp cutoff. People just on one side of it had to stay in school longer. People just on the other side did not. The two groups are otherwise similar, so the comparison works somewhat like an experiment. The method is called a regression discontinuity.

Dementia cases were identified from medical history, hospital records and death registries. The authors say this addressed concerns about selective sample attrition. That is the worry that people who drop out of a study over time are not typical, which can distort results.

The main result: according to the study, education reduces the incidence of Alzheimer's disease and related dementias. Incidence means the rate of new cases.

The second result concerns genes. Using molecular genetic data, the authors show that the reform weakened the relationship between genetics and dementia incidence. In other words, genetic risk mattered less for people affected by the reform. The authors conclude that genetic risk of dementia is not immutable and can be modified by social policy. Some limits follow from the design. The reform affected a specific group of people in a specific place and time, and the abstract gives few details. Results may not carry over to other populations or to today's schooling. We have only the abstract, so effect sizes and the size of the genetic change are not reported here.

Full-text details on effect sizes and sample, and whether similar results appear in other countries and reforms, and whether later-life learning shows any comparable effect.

Sources: Proceedings of the National Academy of Sciences of the United States of America (2026),