Body Mass Index Effect on Long-term Survival after Coronary Artery Bypass Surgery: A Competing-Risks Analysis.

Abstract

To investigate the presence of an obesity paradox on the long-term mortality of patients undergoing primary isolated coronary artery bypass surgery and to uncover whether any discrepancy found could be attributable to cardiovascular or noncardiovascular causes. Retrospective analysis of 5242 consecutive patients with body mass index (BMI) over 18.5 kg/m2 undergoing primary isolated coronary artery bypass surgery, performed from 2000 to 2015, in a Portuguese level III Hospital. The primary endpoint was long-term all-cause mortality. Secondary outcomes were long-term cause-specific mortality (cardiovascular and noncardiovascular). We fitted overall, and cause-specific hazard models, with BMI forced both as a categorical (using World Health Organization predefined cutoffs) and as a continuous variable. Follow-up was 99.7% complete. The median follow-up time was 12.79 years (interquartile range, 9.51 to 16.61). The cumulative incidence functions failed to uncover any difference in 15-year all-cause (log-rank test, p = 0.400), cardiovascular (Gray's test, p = 0.996), and noncardiovascular mortality (Gray's test, p = 0.305) between BMI categories. Likewise, extensive multivariable-adjusted Cox regression and cause-specific hazards models failed to demonstrate in-between category differences, with BMI forced as a categorical variable. On the other hand, using BMI as a continuous variable, the model identified the optimal BMI as between 25.8 and 30.3 kg/m2 (nadir around 28.9 kg/m2), albeit this was dependent on the definition of the reference value. In this longitudinal, population-level analysis of patients undergoing isolated primary coronary artery bypass grafting, we could not attest to any protective effect of obesity on long-term survival.