Religiosity does not prevent cognitive declines: Cross-sectional and longitudinal evidence from the Survey of Health, Aging and Retirement in Europe

Abstract

Over the past hundred years, a plethora of studies on intelligence and religiosity associations predominantly yielded evidence for a meaningful negative relation between these two variables. However, effect strengths varied substantially between primary studies and it has been suggested that religiosity and intelligence associations change as people age, because religiosity may play a protective role for cognitive abilities in elderly individuals. Consequently, it has been suggested that negative intelligence and religiosity associations may decline in strength or even reverse signs as people age. Therefore, we examine here cross-sectional associations of self-reported religious behaviors and several measures of cognitive function (numeracy, verbal fluency, memory and a proxy of psychometric g) as well as their cross-temporal changes in respondents from 11 European countries and Israel aged 50+ years (N = 30,424) in three waves of the Survey of Health, Aging, and Retirement in Europe (SHARE). As expected, cognitive function scores were meaningfully negatively related to praying whilst associations with participation in religious services were trivial. Cross-lagged panel analyses yielded consistently negative, albeit small, effects of both intelligence on praying and of praying on intelligence. Multilevel random-intercept regressions showed tentative evidence for faster cognitive declines in more religious people for numeracy and g, but not for verbal fluency and memory. No conclusive evidence for a moderation by societal values of religiosity could be found. In all, our evidence shows a negative, non-trivial association between intelligence and religiosity in elderly participants which remains longitudinally robust. These findings corroborate the generality of the small negative intelligence and religiosity association.