The triglyceride paradox: a retrospective analysis based on the MIMIC-Ⅳ database

Abstract

Abstract Background Previous studies suggest that sepsis patients often have elevated triglyceride levels due to various factors, and higher levels may indicate a poorer prognosis. However, few studies have investigated whether lower triglycerides are associated with a better prognosis. Methods The Medical Information Mart for Intensive Care IV (MIMIC-IV) database provided all the data. To assess the association between triglycerides and prognosis, we used logistic regression analysis (LR) and Cox proportional hazards models. We further controlled for confounders using propensity score matching (PSM). Results Inclusion criteria were met by a total of 804 patients with a mean triglyceride of 103. We found that patients had a higher risk of 30-day ICU mortality and 30-day in-hospital mortality when triglycerides were in the second percentile (74 mg/dL-103 mg/dL). Interestingly, this group of patients seems to benefit more from the use of atorvastatin. Conclusion The relationship between triglyceride levels and prognosis in patients with sepsis is complex. Our study indicates that a poor prognosis is often associated with triglyceride levels in the range of 74 mg/dL-103 mg/dL.