Investigating macro-level predictors influencing life expectancy in 217 countries and territories using World Bank data (2020–2024)

Life expectancy varies substantially across countries and territories, yet the macro-level factors driving this variation are not fully understood. Using World Bank data for 217 countries and territories (2020–2024), this study examined 17 macro-level predictors spanning demographic, economic, healthcare-system, environmental, behavioural, and education/technology/innovation domains to identify key determinants of life expectancy. Predictors with missing values were categorized by quartile (missing = "unknown"), and descriptive, univariate (by quartile), and multivariable linear regression analyses were performed in R (v4.6.1), with significance at p<0.05. Mean life expectancy across the 217 countries and territories was 74.0 ± 7.1 years (median 74.7; IQR 68.7–79.8). In univariate analysis, higher quartiles of urbanization, GDP, GNI, health expenditure, physician density, food insecurity, literacy, internet access, and research activity were associated with longer life expectancy, whereas higher PM2.5 exposure was associated with shorter life expectancy. The multivariable model showed strong fit (adjusted R² = 0.82), and six predictors were significantly associated with life expectancy: Female proportion, urbanization, physician density, suicide mortality, and hospital bed availability—with higher female proportion, suicide mortality, and food insecurity linked to shorter life expectancy and higher urbanization, physician density, and hospital bed availability linked to longer life expectancy. Macro-level predictors explained approximately 82% of the variation in life expectancy. The findings are largely consistent with prior evidence, although the country-level design may differ from individual-level associations; The inverse association with female proportion likely reflects residual confounding. Identifying these modifiable determinants offers evidence to support public health policy and interventions, and further research is needed to clarify the observed relationships.