Body Mass Index and Mortality Among Hospitalized Patients with a COPD-Coded Diagnosis in MIMIC-IV

To examine associations between body mass index (BMI) and in-hospital, 30-day, 90-day, and 360-day all-cause mortality among hospitalized patients with a COPD-coded diagnosis.

This retrospective cohort study included 9228 patients from a MIMIC-IV analysis extract. BMI was classified as underweight, normal weight, overweight, or obesity. In-hospital mortality was analyzed using logistic regression and follow-up mortality using Cox regression. Restricted cubic splines, proportional-hazards diagnostics, landmark analyses, sensitivity analyses, and multiplicity correction were performed.

There were 387 in-hospital deaths and 585, 1038, and 1969 deaths by 30, 90, and 360 days, respectively. Compared with normal weight, Model 3 HRs for 90-day mortality were 1.523 (95% CI 1.178-1.969) for underweight, 0.809 (0.687-0.952) for overweight, and 0.789 (0.673-0.926) for obesity. Corresponding 360-day HRs were 1.374 (1.134-1.666), 0.742 (0.659-0.835), and 0.713 (0.635-0.801). Overweight and obesity were not significantly associated with in-hospital or 30-day mortality after full adjustment. Associations were nonlinear, and global proportional-hazards tests were significant. After Holm correction, the 90-day overweight comparison was no longer statistically significant.

BMI was associated with mortality, with the most consistent inverse associations for overweight and obesity at 360 days. These observational findings do not establish a survival benefit of obesity and remain susceptible to reverse causation and residual confounding.