Abstract
Low educational attainment is recognized as a modifiable risk factor for dementia. Despite the commonly accepted notion that greater educational attainment confers lower dementia risk, few family-based studies have investigated the causal bases for the association. Using data from seven twin samples from Sweden, Denmark, Australia, and the US participating in the IGEMS (Interplay of Genes and Environment in Multiple Studies) consortium (N = 60,027, 10.92% with dementia), we tested whether twins who achieve higher education than their co-twins have lower risk of dementia. The primary analysis applied a multilevel between-within regression framework, supported by descriptive statistics of within-pair differences. Results confirmed an overall association between educational attainment and dementia risk, such that individuals with higher educational attainment had less likelihood of developing dementia (phenotypic regression coefficient = -0.68, p <.0001). Within twin pairs, however, twins who achieved greater education than their co-twins did not uniformly show lower dementia risk (within-family regression coefficient = -0.07, p =.0983, while between-family regression coefficient = -0.98, p <.0001). Taken together, the pattern of results shows that the effect of educational attainment on dementia risk is largely attributable to genetic influences in common to educational attainment and dementia, although there are also contributions from environmental influences shared between members of the same family. Results were similar in men and women. These findings add to the literature by using a co-twin control design to address possible reasons that low educational attainment is associated with greater dementia risk.