Prognostic Implications of Serum Lipid Metabolism over Time during Sepsis

Abstract

Background: Despite extensive research and an improved standard of care, sepsis remains a disorder with a high mortality rate. Sepsis is accompanied by severe metabolic alterations.

Methods: We evaluated 117 patients with sepsis (severe sepsis [n = 19] and septic shock [n = 98]) who were admitted to the intensive care unit. Serum cholesterol, triglyceride (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), free fatty acid (FFA), and apolipoprotein (Apo) A-I levels were measured on days 0, 1, 3, and 7.

Results: Nonsurvivors had low levels of cholesterol, TG, HDL, LDL, and Apo A-I on days 0, 1, 3, and 7. In a linear mixed model analysis, the variations in TG, LDL, FFA, and Apo A-I levels over time differed significantly between the groups (p = 0.043, p = 0.020, p = 0.005, and p = 0.015, resp.). According to multivariate analysis, TG levels and SOFA scores were associated with mortality on days 0 and 1 (p = 0.018 and p = 0.008, resp.).

Conclusions: Our study illustrated that TG levels are associated with mortality in patients with sepsis. This may be attributable to alterations in serum lipid metabolism during sepsis, thus modulating the host response to inflammation in critically ill patients.

Figure 1

Flow chart of the inclusion and exclusion of patients in the study. A total of 179 patients were enrolled between August and December 2008, and 117 patients were included in the analysis. Patients who were being admitted for a second time (n = 4), those using steroids (n = 26), and those with a history of dyslipidemia (n = 13) or liver disease (n = 22) were excluded. Note: one patient had dyslipidemia and liver disease and two patients had liver disease and used steroids (≥15 mg/day) within the previous 7 days.

Figure 2

Time course of lipid levels in the survivor and nonsurvivor groups. Data for cholesterol (a), triglyceride (b), high-density lipoprotein (c), low-density lipoprotein (d), free fatty acid (e), and apolipoprotein A-I (f) are shown. Data were collected on the day of admission and days 1, 3, and 7 after admission. The circle and bar represent the mean value and standard error of the mean, respectively. Significant differences between the survivor and nonsurvivor groups are indicated by ∗.

Figure 3

Receiver operator characteristic (ROC) curves of triglyceride levels and Sequential Organ Failure Assessment (SOFA) scores to predict mortality. On the day of admission, the area under curve for triglyceride levels was 0.62 (sensitivity = 94.7%, specificity = 29.6, and p = 0.066), and that for the SOFA score was 0.63 (sensitivity = 21.2%, specificity = 96.9, and p = 0.019). In addition, adjusting for triglyceride levels with the SOFA score significantly improved the predictive accuracy of 28-day mortality (AUC = 0.716, sensitivity = 71.1%, specificity = 66.7%, and p < 0.001).