Abstract Background Food insecurity is a growing global issue driven by income inequality, food price fluctuations, and unequal access to essential resources. However, the interrelations among income distribution, real income, food prices, food insecurity, and health expenditure are not well understood, especially in terms of indirect and mediated effects. Materials and methods We used a longitudinal dataset to build vector autoregression models and applied the Toda–Yamamoto causality approach to examine direct and mediated pathways. The augmented Dickey–Fuller test assessed stationarity, and optimal lag lengths were selected using the Akaike information criterion. We used the K-means algorithm for income group classification and the Wald test for comparing findings across groups, based on data from 99 countries. Structural stability was tested using CUSUM test for parameter stability and CUSUMSQ test for variance stability of recursive residuals and Bai–Perron for multiple breakpoints. Results Income distribution and real income directly influenced food prices. In turn, food prices significantly impacted both food insecurity and health expenditure. In high-income countries, food insecurity was found to play a partial mediating role in the relationship between food prices and health expenditure. Importantly, in the global sample, income inequality mediated the relationship between food prices and food insecurity. The joint analysis of all variables revealed causal pathways that were not evident in isolated models. Conclusion These findings highlight the critical role of income inequality in worsening food insecurity and increasing health burdens. The observed mediation mechanisms also suggest that targeted, income group–specific interventions are needed to effectively mitigate the compounded impacts of food inflation on health systems. Addressing economic disparities, stabilizing food prices, and enhancing welfare systems could reduce both food-related and healthcare challenges. Future research should explore regional patterns and broader socioeconomic indicators to support sustainable policy design.
Günal et al. (Thu,) studied this question.