Geographical variation in dementia: systematic review with meta-analysis

Abstract

ABSTRACTBackgroundUnderstanding geographical variation of dementia could highlight important modifiable socio-environmental risk factors. A previous systematic review (2012) identified an increased risk of Alzheimer dementia in in rural living in High-Income Countries (HICs), with a dearth of studies in Low to Middle-Income Countries (L-MICs). We updated this review to examine geographical variations in dementia, to encompass the growing number of studies in this field.MethodsWe systematically reviewed the literature for cross-sectional or longitudinal observational studies that compared dementia incidence or prevalence between two or more geographical areas including rural and urban settings.We conducted a narrative synthesis of included papers. Where possible, we undertook meta-analysis, generating odds ratios for rural versus urban dementia prevalence and stratified the analysis by HICs and L-MICs.ResultsWe identified 38 relevant papers, encompassing approximately 98,502,147 people. Twenty-seven papers were included in the quantitative synthesis. Study methodologies varied widely. Dementia rates varied geographically (0.43-38.5%). Overall, rural living was associated with small increased odds of dementia (OR, 1.20, 95% CI 1.03-1.40;Pvalue = 0.0182). Stratification by HICs and L-MICs demonstrated further variation, with increased odds of dementia in rural areas in L-MICs but not HICs.ConclusionsThere is some evidence of geographical variation of dementia. Rural living was associated with small increased odds of dementia, with stratification showing evidence in rural areas of L-MICs but not HICs. We believe this has not been reported previously. Future research must consider life course geographical exposure and addressing heterogeneity in definitions of ‘rural’ and ‘urban.’What this study addsWe confirm that rural living (compared to urban living) is associated with a small increased odds of dementia (OR 1.20, 95%CI 1.03-1.40). We demonstrate for the first time that this is driven by increased odds of dementia in rural areas in Low to Middle-Income Countries (L-MICs) rather than High Income Countries (HICs), and that the odds of dementia were higher in urban areas in large studies in HICs. Future studies need to carefully consider study setting, method of dementia ascertainment, when exposures may occur, and risk of bias, to understand the role of environment and geography in dementia risk.