Body mass index and hospital mortality in mechanically ventilated patients with acute hypoxaemic respiratory failure: A retrospective cohort study

Background An obesity paradox has been described in general and selected intensive care populations, whereby obese patients experience lower mortality than healthy-weight patients despite physiological disadvantages. Whether this paradox exists among a selected group of patients with acute hypoxaemic respiratory failure (AHRF) requiring mechanical ventilation remains uncertain. We examined the association between body mass index (BMI) and in-hospital mortality in a large, contemporary cohort of mechanically ventilated intensive care unit (ICU) patients with AHRF. Methods This retrospective cohort study used the ANZICS Adult Patient Database (2020–2024). We included unplanned adult ICU admissions requiring invasive mechanical ventilation with AHRF, defined as admission PaO 2 /FiO 2 <300 mm Hg. The primary outcome was in-hospital mortality, assessed using hierarchical logistic regression after adjustment for confounders. The nonlinear association between respiratory failure severity and mortality was modelled using restricted cubic splines. The secondary outcomes included organ support use, tracheostomy receipt, duration of ventilation, and ICU and hospital length of stay. Results Among 40 352 eligible patients, overall unadjusted in-hospital mortality was 21% (22.7% in healthy-weight patients). After adjustment, mortality was lower in preobese patients (BMI 25–29.9 kg/m 2 ; adjusted odds ratio [aOR] 0.92, 95% confidence interval [CI] 0.85–0.99), obesity class 1 (BMI 30–34.9 kg/m 2 ; aOR 0.85, 95% CI 0.79–0.93), and obesity class 2 (BMI 35–39.9 kg/m 2 ; aOR 0.88, 95% CI 0.79–0.98). The association weakened after adjustment and did not extend to obesity class 3 (BMI ≥40 kg/m 2 ). There was no significant interaction between BMI and respiratory failure severity (p = 0.08). Conclusion Among patients ventilated for AHRF, overweight and moderately obese BMI categories were associated with modestly lower in-hospital mortality. This association is conditional on the decision to intubate and is compatible with a modest degree of selection bias. The conclusions should be interpreted as a prognostic association within ventilated patients rather than the presence of obesity improving survival.