Introduction: Obesity is a well-established risk factor for cardiovascular disease and stroke. While a risk factor for the occurrence of these diseases, obesity has been shown to be unexpectedly protective in cardiac disease and ischemic stroke. Here we report the obesity paradox in patients with intracerebral hemorrhage (ICH) in the Nationwide Inpatient Sample (NIS).
Methods: Clinical data for adult patients admitted for ICH were extracted from the Nationwide Inpatient Sample (NIS). Body habitus was classified as nonobese (body mass index [BMI] < 30 kg/m2), obese (BMI ≥ 30 kg/m2 and ≤ 40kg/m2), or morbidly obese (BMI>40kg/m2). Multivariable logistic regression analyzed the association of body habitus with in-hospital mortality rate, discharge disposition, extended length of stay, tracheostomy or placement of an endoscopic percutaneous gastrostomy, and ventriculoperitoneal shunt placement. Covariates included patient demographics, comorbidities, insurance status, occurrence of a craniotomy/craniectomy, and hospital characteristics. Cox proportional hazard regression also analyzed the effect of body habitus on in-hospital mortality.
Results: In total, data from 61,324,882 patient discharges was screened from the NIS between 2007 to 2014. Of these, 99,212 patient discharges were 18 or older and admitted with atraumatic ICH as the primary diagnosis. In a multivariable logistic regression, both obese (0.67 [0.62, 0.76] p<.001) and morbidly-obese (0.85 [0.74, 0.97] p =0.02) patients were associated with decreased odds of in-hospital mortality as compared to those patients who were nonobese. This effect was not seen in other surrogate measures of outcome including non-routine discharge, extended hospitalization, tracheostomy/gastrostomy placement, and VP shunt placement. A Cox proportional hazards model also demonstrated that both obese (0.71 [-0.41, -0.26] p<.001) and morbidly obese (0.89 [-.021, -0.02] p<.001) patients had a significantly decreased risk of death during their hospital stay.
Conclusion: Obesity appears to be associated with decreased inpatient mortality in spontaneous ICH in a large national dataset even when controlling for demographic and clinical course factors.