Association of body mass index with the occurrence and multiplicity of postoperative pulmonary complications after cardiac surgery: a secondary analysis of the VIRS trial

Background The impact of BMI on postoperative pulmonary adverse events in cardiac surgery populations remains unclear. The aim of this study was to determine the relationship between BMI and postoperative pulmonary complications (PPCs) in patients undergoing cardiac surgery with cardiopulmonary bypass. Methods This was a secondary analysis of the Volatile anaesthesia versus total Intravenous anaesthesia in the caRdiac Surgery (VIRS) trial, including 1,322 patients who underwent elective cardiac surgery. Chinese BMI criteria were used to classify patients as underweight, normal weight, overweight, or obese. The primary outcome was the incidence of PPCs. The number of distinct PPC events per patient was also recorded to assess the multiplicity of complications. Multivariable binary logistic regression, ordinal logistic regression and restricted cubic spline analysis were used to assess the association between BMI and PPCs. Results Overall, 708 patients (53.5%) developed one or more PPCs after cardiac surgery. After multivariable adjustment, underweight status was associated with a higher risk of PPC occurrence (OR [95% CI], 1.89 [1.12–3.19]; P = 0.017) and greater PPC multiplicity (OR [95% CI], 2.56 [1.61–4.07]; P < 0.001), compared with normal weight. Overweight status was not significantly associated with either outcome. Obesity was not significantly associated with overall occurrence of PPCs (OR [95% CI], 1.27 [0.83–1.94]; P = 0.273), but was associated with greater multiplicity of PPCs (OR [95% CI], 1.52 [1.03–2.23]; P = 0.035). Continuous BMI showed a J-shaped association with PPC risk (P for nonlinearity < 0.001). Conclusions A nonlinear association exists between BMI and PPCs in cardiac surgery patients. Underweight status represents an important marker of postoperative pulmonary vulnerability, whereas obesity may be more closely associated with the coexistence of multiple PPCs than with their overall occurrence.