The Relationship Between Four Measures of Religiosity and Cross-National Variations in the Burden of Dementia.

Background Several researchers have identified a possible protective effect of religiosity on the risk of dementia. Specific aspects of religiosity may be associated with this attenuation of risk, and it may be partially mediated through an effect on depressive symptoms or social support. However, this effect has only been demonstrated in selected cohorts to date. Methods This study was based on a cross-national analysis of associations. Correlations between World Health Organization estimates of the burden of dementia and four survey-derived measures of religiosity were examined across 101 countries, while controlling for estimates of late-life depression and social capital. Results Specific aspects of religiosity, such as attendance at religious services (Pearson's r= -0.57), daily prayer (r = -0.58), and perception of religion as very important (r = -0.65), were associated with lower national levels of Alzheimer's and other dementias (p< 0.01 for all correlations). This effect was partially mediated through an inverse relationship between religiosity and depression, but remained significant even after controlling for it and on multivariate analyses (β = -0.38 to -0.57, p< 0.01 for all measures). There was no evidence for a mediating effect of social capital. Conclusions Specific religious beliefs and practices may have a protective effect on dementia risk at the population level. These may involve group effects that require further study, such as reductions in depression in the elderly, or may involve beneficial effects on the stress response and cellular ageing in vulnerable individuals; however, the latter cannot be inferred with certainty from a group-level analysis. These results are consistent with earlier research and suggest a potential role for religious-based preventive strategies at the population level.