|Home | About | Journals | Submit | Contact Us | Français|
Despite guidelines recommending a daily fat intake of 0.7 to 1.5 g/kg/day in the ICU, subjects with hypertriglyceridemia >2 mmol/l (HTG) are common. As there are limited data on risk factors for HTG in the ICU, we aimed to determine the factors related with HTG in the ICU.
During 7 months, all consecutive patients staying ≥4 days in an adult ICU from a university hospital were enrolled. Patients eating regular meals or having an initiation or withdrawal of statin during their ICU stay were excluded. Peak values of triglycerides (TG) were collected and the relationships between log-TG and fat intake (g/kg/day) from nutritional (enteral and parenteral) and non-nutritional (propofol's emulsion) sources as well as propofol (mg/kg/day) were assessed using Pearson's correlation coefficients. Correlation was considered small for coefficients between 0.1 and 0.3 and medium for coefficients between 0.4 and 0.6. Nine groups of patients at risk of HTG were further compared with a control group (no risk factors for HTG) using Dunnett's test (significant if adjusted P < 0.05).
Of the initial 293 patients, 89 were excluded for eating regular meals or having an initiation or withdrawal of statin during their ICU stay. Of the remaining 204 patients, 79 (38.7%) had HTG, although guidelines for lipid intake were followed. Only three patients (1.5%) had a combined fat intake (enteral and parenteral) between 1.51 and 1.71 g/kg/day. Small positive correlations were observed with the intake of nutritional parenteral lipids (0.27), intake of all lipids given (0.20), intake of long-chain triglycerides (LCT) (0.15), and intake of parenteral LCT (0.20). Medium positive correlations were observed with the amount received of lipids administered with continuous propofol (0.40) and the amount of active principle of continuous propofol administered (0.42). In comparison with the control group (n = 81), patients with hepatic dysfunction (n = 6), pancreatitis (n = 14), severe insulin resistance (n = 2), sepsis (n = 32) and dyslipidemia without statin (n = 7) had significantly higher mean values of TG (all P < 0.05). Groups with cirrhosis and ascites (n = 3), diabetes mellitus (n = 11), chronic renal failure (n = 6) and patients with statin before and during hospitalization (n = 42) had similar levels of TG as the control group.
When guidelines for fat intake in the ICU are followed, modest fat intake does not seem to explain HTG. Our results show that the amount of propofol given (mg/kg/day) and some clinical factors might be correlated with HTG in the ICU.