RESUMEN
BACKGROUND: Vulnerabilities of men and women to adverse health effects due to weather variability and climate change are not equal. Uganda was among the countries in the world most affected by extreme weather events during the last decade. However, there is still limited gendered empirical evidence on the links between weather variability and health and the possible pathways through which these health effects occur. Therefore, this study analyses the effect of weather variability on illness, and the extent to which water collection 'time burden' mediates the relationship between weather anomalies and illness among men and women of working age in Uganda. The study also quantifies the health inequalities to be eliminated if resources are equalized. METHODS: Socioeconomic, health and time use data were obtained from the World Bank Living Standards Measurement Studies - Integrated Surveys on Agriculture (LSMS -ISA), combined with high resolution remotely-sensed weather data. Two-parts and non-linear decomposition regression analysis were used on the national representative pooled dataset from the four household survey waves collected between 2009 to 2014, comprising a total of 22,469 men and women aged between 15 and 64 years. RESULTS: Empirical results show that low rainfall below the long-term mean increased the likelihood of illness by at least 8 and 6 percentage points for women and men, respectively. The indirect effect of low rainfall on illness through water access pathway was estimated at 0.16 percentage points in women. Decomposition results reveal that health inequalities among women and men would have been narrowed by 27-61%, if endowments were equalized. CONCLUSIONS: Strategies that promote women empowerment (such as education, labor force participation, access to financial services and clean water), health adaptation and time poverty reduction strategies (such as rain water harvesting and improved access to quality health care) would reduce gender-based health inequalities in Uganda despite changing climatic conditions.