Analysis of seasonality of multiple causes of death

Objective: In the light of the recurrent occurrence of temperature and precipitation anomalies, the question of the relationship between climate conditions and mortality becomes increasingly relevant. The central theme of this study is to analyze the differences in mortality seasonality for selected causes of death across climatic regions in the United States. The thesis focuses on multiple causes of death, which, as evidenced in respiratory diseases, are play important role in detecting seasonal specifics in mortality. Data and methods: The study is based on microdata of deceased individuals provided by the Centers for Disease Control and Prevention. Population counts are taken from the U.S. Census Bureau and basic temperature and precipitation data are provided by the National Centers for Environmental Information. Basic statistical methods are applied to test the significance of differences between regions (ANOVA), identify seasonality in death (and death rate) time series (χ2 tests, monthly "dummy" variables), reduce dimensionality (correspondence analysis), and estimate risks of coding multiple causes of death by season (logistic regression). Additionally, tools of demographic analysis are employed, including standardization, single-decrement and multiple-decrement life tables with hypothetical scenarios of summer and winter mortality regimes. Results: Statistically significant differences in both mortality and seasonality are observed among climatic regions, with the most pronounced seasonality found in cardiovascular and respiratory diseases. The multiple cause approach enhances the importance of these differences. If age-specific mortality rates followed the summer pattern throughout the year, life expectancy would increase by up to two years in certain regions, and this disparity is further magnified when disaggregated by causes of death. For other and cardiovascular diseases, the higher effect of seasonality can be attributed to states with lower life expectancy at birth. Season is significant determinant of coding of immediate causes of death, with odds of coding them increases up to 30% during winter and spring compared to summer. Overall, the main factors contributing to differences between climatic regions include (i) coding of contributing diseases, (ii) mortality from external causes, and (iii) generally poorer mortality conditions and higher seasonality in southeastern regions of the USA. Conclusions: Extensive mortality disparities exist among climatic regions in the United States. However, further research is necessary to understand the causal mechanisms between climate conditions and mortality.