They noted that in a model that included late-life comorbidities, the association between education and dementia was slightly weakened.
"People with low education also more often had stroke, myocardial infarction, diabetes, and more depressive symptoms," they wrote. "Thus, partly the effect of education could be mediated via these factors, or low education might lead to frailty and thereby increase the risk of both dementing and other late-life disorders."
The large population sample and high-degree of retention (72% of the original sample) are both strengths of the study, the authors said, but they did note a few limitations.
For example, the analyses are based on a sample of participants who survived until the 1998 reexamination, but "it is known that persons with low education also have higher mortality. If those who died between the examinations also were more often demented, then our results would underestimate the true association between education and dementia," they wrote.