Background: Serum carbohydrate deficient transferrin is a marker of chronic alcohol consumption; it increases above normal in healthy individuals after a daily alcohol intake of more than 60 g/d for more than 2 weeks. The influence of liver disease itself on carbohydrate deficient transferrin levels has not been sufficiently established.
Methods: We investigated serum levels of carbohydrate deficient transferrin in 196 consecutive patients admitted to our Gastroenterology and Hepatology Unit and correlated this parameter with the patients' statements about alcohol intake during the previous 2 weeks and with other markers of chronic alcohol consumption.
Results: In our patient population, carbohydrate deficient transferrin had the best overall performance with respect to sensitivity (88%), specificity (82%), and negative predictive value (98%), as compared to other markers, although specificity was much lower than previously reported in patients without liver disease. In the group of patients with liver disease, sensitivity and specificity were 90% and 73%, respectively, and in patients without liver disease, 80% and 88%. The negative predictive value was excellent (96% for patients with liver disease and 99% for patients without liver disease).
Conclusions: Thus, in a patient with a negative interview for chronic alcohol abuse and normal carbohydrate deficient transferrin level, alcohol is unlikely to be the cause of liver disease, and further investigations to establish the etiology of liver disease are warranted. An increased carbohydrate deficient transferrin level, however, cannot be regarded as reliable evidence for chronic alcohol abuse in patients with liver disease.