Winter Temperature and Long-Term Mortality After Coronary Artery Bypass Grafting: A Multicenter Cohort Study

Background: Traditional cardiovascular risk models focus on patient-related clinical variables, while the impact of long-term environmental exposure remains insufficiently characterized in post-revascularization populations. Objective: To evaluate the association between environmental exposure and long-term mortality after coronary artery bypass grafting (CABG), and to determine whether integrated environmental measures provide additional prognostic value beyond established clinical risk factors. Methods: This retrospective multicenter cohort study included 1042 consecutive patients undergoing CABG with a median follow-up of 8.1 years. Regional environmental data were linked to individual patients. Multivariable Cox regression models were constructed using a hierarchical approach. To address collinearity among environmental variables, principal component analysis (PCA) was applied. Mean winter temperature was analyzed as a clinically interpretable proxy of overall environmental exposure. Results: During follow-up, 220 deaths (21.1%) occurred. Established clinical predictors of mortality included age, diabetes mellitus, peripheral vascular disease, and dyslipidemia. Individual environmental variables showed unstable associations due to collinearity. PCA identified a dominant environmental component explaining 82.0% of variance; however, its association with mortality did not reach statistical significance after adjustment (HR 1.17 per SD, 95% CI 0.98-1.39; p = 0.083). In contrast, higher mean winter temperature was independently associated with increased mortality (HR 1.24 per SD, 95% CI 1.05-1.48; p = 0.013) per 1 °C increase, with evidence of non-linearity. Conclusions: Environmental exposure represents a relevant component of long-term risk after CABG. While individual environmental variables are highly correlated and unstable, clinically interpretable measures such as winter temperature may provide practical support for risk assessment.