Background: Obesity is associated with overall higher mortality risk in the general population however, some authors have noted patterns to suggest a survival advantage in a certain clinical subpopulation. Obesity paradox is reported for stroke, myocardial infarction, heart failure, renal disease, and diabetes. Clostridium difficile infection (CDI) is the most common nosocomial infection in the United States. Some studies have linked CDI with the obesity paradox however, not all studies have consistently shown this finding. This study is to evaluate the association of obesity with in-hospital mortality in patients with CDI.

Methods: We used the 2016 National Inpatient Sample (NIS) database provided by the Healthcare Cost and Utilization Project (HCUP) that are weighted to optimize national estimates. HCUP is a source of healthcare databases developed through a Federal-State-Industry partnership and sponsored by the Agency for Healthcare Research and Quality. We queried the database for adults with CDI, obesity and excluded patients with malnutrition using the ICD-10 code. The primary outcome was mortality in CDI patients with and without a diagnosis of obesity. We used the logistic regression method to adjust for age, gender, race, Charlson index, economic status, hospital bed size, hospital region, teaching hospital status, urinary tract infection, pneumonia, acute coronary syndrome, inflammatory bowel disease, acute kidney injury, leukocytosis, hypoalbuminemia.

Results: There was a total of 360364 discharges associated with a diagnosis of CDI of which 90340 discharges had CDI as the primary diagnosis. After excluding the patients with a diagnosis of malnutrition, among a total of 89649 discharges, 7785 (8.6%) discharges with CDI as primary diagnosis had obesity. When compared with discharges without obesity, discharges with obesity had more females (71% vs 64%), relatively younger age (60.5 years vs 63.8 years) however, had a longer length of stay (5.39 days vs 5 days) with more mean hospital charge ($38078 vs $34171). There was a total of 55 death in obese cohort while 814 died in non-obese group. Using univariate regression analysis, we found that the unadjusted mortality was lower in patients with obesity compared to the non-obese patients, however, it wasn’t statistically significant [Odds ratio (OR) 0.71, 95% CI (0.38 – 1.33)]. After adjusting for age, gender, race, Charlson index, acute coronary syndrome, hospital location, teaching hospital status, presence of urinary tract infection, pneumonia, inflammatory bowel disease, acute kidney injury, leukocytosis, hypoalbuminemia using multivariate regression analysis revealed a similar result with decreased mortality among obese patients compared to non-obese patients, however, it was not statistically significant [OR 0.79 95% CI (0.41-1.53)].

Conclusions: Our study didn’t find a significant decrease in mortality among CDI patients with obesity compared to non-obese patients thus suggesting against the theory of ‘obesity paradox’. Additional studies at a larger scale would be beneficial to identify the real impact of obesity on this disease.