Obesity Paradox in a Cohort of 3863 Consecutive Patients Undergoing Cardiac Surgery

P. Berretta¹, O. Bifulco², G. Michele³, J. Alfonsi³, V. Fontana⁴, B. Buratto³, R. Silvano⁵, P. Giorgio. Malvindi³, M. Di Eusanio^1
1Cardiac Surgery Unit, Lancisi Cardiovascular Center, Polytechnic University of Marche, Ancona, Marche ²Cardiac Surgery Department, University of Padova, Ancona, Marche ³Lancisi Cardiovascuar Center, Polytechnic University of Marche, Ancona, Marche ⁴Lancisi Cardiovascuar Center, Ancona, Marche ⁵Lancisi Cardiovascular Center, Polytechnic University of Marche, Ancona, Marche

Purpose: Conflicting results exist on the “obesity paradox” in cardiac surgery, with few reports evaluating the long-term association between BMI and mortality. We investigated the relationship between body mass index (BMI) and operative and long-term outcomes in a large population operated on a tertiary-care center.

Methods: The study population included 3863 consecutive patients who underwent coronary artery bypass grafting, aortic or mitral valve surgery, and ascending aorta/aortic root surgery between 2016 and 2023. Patients with aortic dissection, active endocarditis and combined operations were excluded. Patients were divided into five BMI categories: underweight < 18.5 kg/m2, normal weight 18.5 to 24.9 kg/m2 (used as reference), overweight 25 to 29.9 kg/m2, obese 30 to 34.9 kg/m2 and morbidly obese >34.9 kg/m2. Perioperative outcomes and 5-year mortality were compared using logistic and Cox multivariable regression to investigate potential confounders (age, sex, diabetes, dialysis, COPD, cerebrovascular disease, peripheral vascular disease, NYHA III-IV, reoperation, urgent/emergent status and type of operation) on the relationship between BMI and adverse outcomes. An adjusted survival curve was constructed using the Cox model.

Results: Of the patients, 34.4% were normal weight (n=1329), 1.2% underweight (n=45), 43.2% overweight (1667), 17% obese (n=658) and 4.2% morbidly obese (n=164). A reverse J-shaped association was found between BMI and late mortality with their respective hazard ratios (table 1 and figure 1A). Despite adjustment for multiple confounders, overweight and obese patients were associated with a reduced risk of late mortality compared with normal weight (HR 0.71, 95%CI: 0.55-0.72, and HR 0.82, 95%CI: 0.67-0.97, respectively) (table 1). Adjusted 5-year survival was higher for overweight and obese individuals with respect to the other BMI categories (figure 1B).
Unadjusted and adjusted logistic models showed that overweight and obese patients had a significant reduced risk of transfusions (obese OR 0.69, 95%CI 0.54:0.91; morbidly obese OR 0.49, 95%CI 0.29:0.84) and intensive care unit (ICU) length of stay (overweight  -0.49, 95%CI -0.85: -0.18; obese  -0.53, 95%CI -0.94: -0.16) but an increased risk of wound complications (overweight OR 1.71, 95%CI 1.13:2.61; obese OR 1.97, 95%CI 1.21:3.22; morbidly obese OR 2.71, 95%CI 1.32:5.52) compared with normal weight (table 1).

Conclusion: Findings from our study support the “obesity paradox” for overweight and mildly obese patients. When compared with normal weight group, overweight and obese categories were associated with improved 5-year survival, lower risk of transfusions and shorter ICU length of stay. Conversely, increased BMIs showed a higher incidence of wound complications.

Identify the source of the funding for this research project: none