microalbuminuria in hyperglycemic critically ill patients treated with insulin or metformin

microalbuminuria is thought to reflect the severity of inflammation-induced systemic vascular permeability. the present study investigated the effect of early administration of metformin or insulin on microalbuminuria in traumatized critically ill patients. between april 2006 and october 2007, thirty-one non-diabetics traumatized patients with systemic inflammatory response syndrome (sirs) and blood sugar (bs) >130 mg/dl at admission to icu (intensive care unit) of sina hospital (tehran, iran), were randomly assigned to receive intensive intravenous insulin (50 iu) or peroral metformin (1000 mg, twice daily) for three days. microalbuminuria to creatinine ratio (macr) and bs were measured during the three-day period. eight patients were excluded during the study and 23 remained for the evaluations. there was no statistically significant difference between two groups with respect to macr levels at admission and during the three-day period of treatment except for the time 6 and 48 h, that macr was higher in insulin group than that in metformin group (p < 0.05). metformin but not insulin reduced bs level significantly (p < 0.05). there was a significant positive correlation between bs and macr in both insulin (p < 0.05; r2 = 0.131) and metformin (p < 0.05; r2 = 0.127) groups. glasgow coma scale (gcs) and apache ii had significant correlation with macr in metformin treated patients (p < 0.05; r2 = 0.134 and p < 0.05; r2 = 0.149) while in insulin treated patients only the values of gcs had significant correlation with macr values (p < 0.05, r2 = 0.124). in conclusion, it was found that peroral metformin may be used instead of intravenous insulin in traumatized critically ill patients for lowering bs and macr. a predictive role for macr may be suggested although further studies with larger sample size of patients is required.