Education and dementia: a meta-analytic study.

Considerable controversy exists about the role of education in the risk of dementia. Individual studies have not been conclusive so far. To examine the hypothesis that lower education is associated with a higher risk of dementia, we carried out a meta-analysis. Observational studies published as of October 2005 that examined the association between education and risk of dementia were systematically reviewed. Relative risks (RRs) and odds ratios were extracted from cohort and case-control studies. We first compared the risk of dementia in subjects with high level of education with the risk of dementia in those with low educational level. In a subsequent analysis, we compared the risk of persons with high education with the risk of subjects with education level other than high (medium, low). We weighted log RRs for cohort studies or odds ratios by the inverse of their variances. Nineteen studies were included in our meta-analysis (13 cohort and 6 case-control studies). RRs for low versus high education level were: Alzheimer's disease (AD) 1.80 (95% CI: 1.43-2.27); non-AD dementias, 1.32 (95% CI: 0.92-1.88), and all dementias 1.59 (95% CI: 1.26-2.01). For low and medium versus high education level, the RRs were: AD 1.44 (95% CI: 1.24-1.67); non-AD 1.23 (95% CI: 0.94-1.61), and all dementias 1.33 (95% CI: 1.15-1.54). These results confirm that low education may be a risk factor for dementia, especially for AD.