Impact of body mass index on hospital mortality of old and oldest- old critically ill patients: a retrospective study

With advancing age, the Obesity Paradox undergoes changes. This study aims to investigate the applicability of the Obesity Paradox in elderly critically ill patients. A total of 16,050 elderly critically ill patients were included in this study. Linear regression analysis was employed to examine the relationship between BMI and ICU length of stay (LOS_ICU), hospital length of stay (LOS_Hospital), and hospital mortality in elderly critically ill patients. The correlation between BMI and hospital mortality of elderly critically ill patients was analyzed by logistic regression analysis and subgroup analysis with adjustment for covariates. The 65–80 years age group exhibited the lowest hospital mortality, LOS_ICU, and LOS_Hospital in the overweight group. In the > 90 years age group, the obese demonstrated the highest LOS_ICU and LOS_Hospital. For the elderly population, overweight and obese critically ill patients had a significantly lower risk of hospital mortality ( p < 0.05), while underweight patients had a significantly increased risk of hospital mortality ( p < 0.05). However, after Bonferroni adjustment, the association between BMI groups and hospital mortality showed no statistical significance (p adjust > 0.00125). In the 65–80 years age group, the odds ratios (ORs) (95% confidence intervals [CI]) for overweight and obese were 0.76 (0.63, 0.92) and 0.71 (0.59, 0.85). For individuals aged 80–90 years, the ORs (95% CI) for overweight and obese were 0.77 (0.61, 0.95) and 0.84 (0.65, 1.07). For individuals aged > 90 years, the ORs (95% CI) for overweight and obese were 1.46 (0.97, 2.18) and 1.00 (0.53, 1.86). While overweight and obesity are associated with a reduced overall risk of hospital mortality in elderly critically ill patients, this protective association appears to be attenuated in individuals aged over 90 years. However, these associations were not significant after Bonferroni correction for multiple comparisons, highlighting the need for validation through larger-scale longitudinal studies.