Neighborhood factors and the geography of type 2 diabetes in Malaysia: an ecological based geospatial modelling study

Type 2 diabetes (T2D) often exhibits long-standing disparities across populations. Spatial regression models can identify associations between local rates and observed neighborhood factors to inform timely, localized public health interventions. We identified area-level distributions of T2D rates across Malaysia and estimate associations between different neighborhood covariates and local T2D burden. We obtained aggregated counts of national level T2D cases data by administrative-districts between 2016 and 2020 and computed district-wise crude rates to correlate with district-level neighborhood demographic, socio-economic, safety, physical activity and resources access, and urbanization process indicators from various national sources and census data. We applied a simultaneous spatial autoregressive (SAR) modeling strategy to account for spatial autocorrelation and estimate risk factors for district-level logged T2D crude rates in Malaysia. District-level logged T2D crude rates were associated with the proportion of households living below 50% of the median income (β = 0.009, p = 0.003) and below the national poverty line (β = ˗0.011, p = 0.001), income inequalities (β = ˗1.778, p = 0.015), CCTV coverage per 1000 population (β = 0.060, p = 0.049), and perceived access to recreational parks (β = 0.007, p = 0.001) and to bowling centers (β = ˗0.003, p = 0.009). Area-level district-wise logged T2D crude rate estimates were associated with neighborhood socio-economic vulnerabilities, neighborhood safety, and neighborhood perceived access to physical activity facilities, after accounting for residual spatial autocorrelation via a SAR model.