Spatial spillover effects of area-level socioeconomic factors on life expectancy in Japan: an ecological study

Area-level socioeconomic status is a well-established determinant of geographical disparities in life expectancy. However, limited attention has been paid to spatial spillover effects, whereby socioeconomic conditions in neighbouring regions influence health outcomes. This study aimed to estimate the direct and spatial spillover effects of socioeconomic factors on life expectancy in Japan and to explore possible mechanisms underlying the observed spillover patterns.

Life expectancy at birth by sex at the municipal level in Japan for 2020 was the outcome variable. A spatial Durbin error model was used to estimate the direct and spatial spillover effects of ten regional socioeconomic factors, along with six control variables, on life expectancy. To ensure robustness, six spatial weight matrices were used. The results were compared with those obtained from a non-spatial linear regression model.

Moran's I values for the residuals of the non-spatial model were statistically significant, indicating spatial autocorrelation. The unemployment rate and the proportion of individuals with no high school diploma showed negative direct and spillover effects, suggesting that being surrounded by regions with employment instability and low educational attainment is associated with lower life expectancy. Taxable income per capita showed no statistically significant spillover effects.

The findings indicate that socioeconomic conditions in neighbouring regions, in addition to those within a region, are associated with life expectancy. The observed spillover effects for employment and education support the role of collective resources in shaping regional health. These results indicate the need to incorporate interregional socioeconomic contexts into public health strategies to address geographical disparities in health.