Heat-related illness and chronic obstructive pulmonary disease: causal inference using marginal structural models

BACKGROUND: Heat-related illness (HRI) may serve as a more reliable individual-level indicator of extreme heat exposure compared to ambient temperature, which has spatial autocorrelation problems. This study aimed to provide causal evidence for the association between HRI and the risk of chronic obstructive pulmonary disease (COPD). METHODS: A retrospective cohort study was conducted involving 1,392,876 Korean individuals who regularly participated in health screenings provided by the National Health Insurance Services. Using data from 2009 to 2021, both HRI and COPD were identified using primary diagnosis codes. Marginal structural models were employed to evaluate the causal relationship between HRI history and COPD risk. Standardized inverse probability of treatment and censoring weights were calculated based on time-dependent exposure and confounders. The risk of COPD was assessed according to the occurrence and frequency of HRI. RESULTS: Over an average follow-up period of 11.5 years, 5,173 participants (0.4%) experienced HRI at least once. One year after study entry, 26,919 (1.9%) were newly diagnosed with COPD. A history of HRI was associated with an elevated COPD risk ranging from 2.09 to 2.13, regardless of whether it occurred during the warm season (April to September) or throughout the entire year. Heat stroke, the most severe form of HRI, showed a stronger association than other HRI types. Older individuals, women, and rural residents were particularly vulnerable. CONCLUSIONS: These findings underscore the priority of thermal regulation during heat exposure for the prevention of COPD.