Educational level has been regarded as one of the most widely accepted risk factors in the epidemiological studies for dementia, despite with discordant qualitative results. However, the dose-response relation between education and incident dementia was still unknown. To quantitatively evaluate the association between exposure level to high and low education and risk of dementia, we searched PubMed, EMBASE, and the Cochrane Library up to November 2014 and references of retrieved literatures. Specific prospective cohort studies, in which educational attainment was categorized into at least three levels, were included. Newcastle-Ottawa scale was used to assess the quality of included studies. Fifteen prospective cohort studies with 55655 for low education and eight prospective cohort studies with 20172 for high education were included. In the qualitative analysis, both low and high education showed a dose-response trend with risk of dementia and Alzheimer's disease (AD). In the quantitative analysis, the dementia risk was reduced by 7 % for per year increase in education (RR, 0.93; 95 % CI, 0.92-0.94; p for overall trend = 0.000; p for nonlinearity = 0.0643). Nonetheless, we did not find statistically significant association between per year decrease in education and dementia (RR, 1.03; 95 % CI, 0.96-1.10; p for overall trend = 0.283; p for nonlinearity = 0.0041) or AD (RR, 1.03; 95 % CI, 0.97-1.10; p for overall trend = 0.357; p for nonlinearity = 0.0022). Both low and high education showed a trend of dose-response relation with risk of dementia and AD. The dementia risk was reduced by 7 % for per year increase in education.
Keywords: Dementia; Education; Meta-analysis; Risk factor.