Burns 29 (2003) 810–815 Serum cholesterol and triglycerides: potential role in mortality prediction Lars-P. Kamolz a,∗ , Harald Andel b , Martina Mittlböck c , Wolfgang Winter b , Werner Haslik a , Günther Meissl a , Manfred Frey a a Division of Plastic and Reconstructive Surgery, Department of Surgery, General Hospital Vienna, Medical School, University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria b Department of Anaesthesia and Intensive Care, Medical School, University of Vienna, Vienna, Austria c Department of Medical Computer Sciences, Medical School, University of Vienna, Vienna, Austria Accepted 24 June 2003 Abstract The present study was performed in order to evaluate the diagnostic usefulness of serial cholesterol and triglycerides measurements in patients with severe burns. One of the main objective was to find out if these parameters are clinically relevant to determine the morbidity of a burn patient and thereby the patient’s outcome. In 220 patients with thermal injuries, cholesterol and triglyceride concentrations were measured daily. Blood samples were drawn immediately upon admission and thereafter daily until patient’s discharge or death. For both parameters, a characteristic course was noted: in the group of non-survivors, a decrease of cholesterol prior to death was noted, while survivors, increased prior to discharge. The time courses of both groups (survivors–non-survivors) differed statistically significantly (P = 0.0068). An increase in triglycerides was observed in all non-survivors prior to death, but in the group of survivors triglycerides remained more or less unchanged. These time courses also had statistically significant differences (P = 0.0004). In our 220 patients, changes in cholesterol (P< 0.0001, hazard ratio 1.02) and triglycerides (P = 0.0008, hazard ratio 1.01) had comparable capability to predict the severity of a burn trauma and thereby its outcome than the established parameters in the treatment of burns (total body surface area burned, age, inhalation). We consider the serial measurements of cholesterol and triglycerides as clinically relevant to assess the morbidity of a patient and thereby to estimate the patient’s outcome. We think that these serial measurements provide useful information for the clinician treating patients with severe burns. © 2003 Elsevier Ltd and ISBI. All rights reserved. Keywords: Cholesterol; Triglycerides; Burn injuries; Morbidity; Outcome 1. Introduction The relationship between hypercholesterolemia and car- diovascular risk is largely accepted, but the risk related to low cholesterol levels is not well known. Hypocholes- terolemia occurs in a variety of acute and chronic diseases. Low CHOL concentrations have been reported in different pathologic conditions such as HIV infection, neoplasia, leukemia and lymphomas [1–13]. Alvarez and Ramos ob- served a decrease of CHOL and an increase of TG in septic patients [1]. Coombes et al. documented a fall of serum CHOL and an increase in TG following severe burn injuries [2]. Birke et al. demonstrated that the CHOL level fell Abbreviations: ABSI, abbreviated burn severity index; AUC, area under the curve; CHOL, cholesterol; TBSAB, total body surface area burned; TG, triglycerides ∗ Corresponding author. Tel.: +431-40400-6986; fax: +431-40400-6988. E-mail address: lars.peter.kamolz@univie.ac.at (L.-P. Kamolz). gradually after thermal injury and that these changes were proportional to the extent of this trauma [3,4]. More and more recent studies suggest a relationship be- tween low cholesterol levels and increased risk of premature death. Noel et al. showed that acquired hypocholesterolemia (CHOL < 120 mg/dl) is a common finding in hospitalized elderly patients [5]. They demonstrated that acquired low CHOL levels were associated with poor outcome. These pa- tients had more infectious and non-infectious complications as compared to patients with normal values. Gui et al. demonstrated a relationship between hypoc- holesterolemia and mortality in critically ill surgical patients [6]. Crook et al. observed the prevalence of low CHOL lev- els in a hospital population and also the causes and outcome of this condition [7]. They concluded that the prevalence of hypocholesterolemia is low in hospital population. They considered low CHOL levels as a possible useful predictor of death in hospital patients, but they mentioned that larger studies would be needed to confirm this. 0305-4179/$30.00 © 2003 Elsevier Ltd and ISBI. All rights reserved. doi:10.1016/S0305-4179(03)00196-7