Biomarkers of alcohol consumption in patients classified according to the degree of liver disease severity

OBJECTIVE: In the search for optimal biomarkers of excessive drinking, only a few studies have been conducted to compare the relationships between ethanol consumption, liver status, and various laboratory markers of ethanol-induced diseases. MATERIAL AND METHODS: Concentrations of carbohydrate-deficient transferrin (%CDT and CDTect methods), serum sialic acid (SA), gamma-glutamyl transferase (gamma-GT), aspartate aminotransferase (ASAT), mean corpuscular volume (MCV), and a marker of fibrogenesis (PIIINP) were studied in 102 alcoholics with (n=59) or without (n=43) alcoholic liver disease. Controls were 34 healthy volunteers who were either social drinkers or abstainers. RESULTS: Although concentrations of all markers were significantly higher in the alcoholic patients than in the healthy controls, their diagnostic characteristics showed a considerable degree of variation. The %CDT, SA, and MCV showed the strongest correlations with the amount of recent alcohol intake. The presence of liver pathology notably influenced the results of CDTect, GT, ASAT, and PIIINP. In ROC analyses, the highest rates of diagnostic accuracy for detecting hazardous drinking were reached with GT (0.94), CDT (0.86), and SA (0.85), followed by MCV (0.79) and ASAT (0.77). Upon abstinence, the estimated times for normalization varied between 10 days (CDTect) and 25 days (GT). CONCLUSIONS: Our data suggest distinct differences in the clinical characteristics of biological markers of ethanol consumption. While the overall accuracy of CDT and GT appear to be highest in the detection of problem drinking, serum SA and PIIINP measurements are of further value when the effects of liver pathology and ethanol drinking need to be differentiated.