On Sensitivity, Specificity, and the Influence of Various Parameters on Ethyl Glucuronide Levels in Urine—Results From the WHO/ISBRA Study


  • WHO/ISBRA Investigators: K.M. Conigrave, M. Dongier, H. Edenberg, C.J.P. Eriksson, M.L.O.S. Formigoni, B.F. Grant, A. Helander, P.L. Hoffman, K. Kiianmaa, T. Koyama, L. Legault, T-K Li, M. Monteiro, T. Methuen, T. Saito, M. Salaspuro, J.B. Saunders, B. Tabakoff, S. Tufik, J.B. Whitfield, F.M. Wurst

Friedrich Martin Wurst, MD, Psychiatric University Hospital, University of Basel, Wilhelm Klein Strasse 27, CH 4025 Basel, Switzerland; Fax: ++41 61 325 5583; E-mail: friedrich.wurst@pukbasel.ch.


Background: Ethyl glucuronide (EtG), a direct ethanol metabolite, seems to meet the need for a sensitive and specific marker for monitoring recent alcohol consumption in different settings. Our aim was to study sensitivity, specificity, and the influence of various parameters on EtG levels in urine.

Patients and Methods: Urine samples for a total of 453 patients (373 male, 80 female) were statistically analyzed. The mean age was 37.1 years (median 36, SD 12.59), body mass index was 24.7, total ethanol consumed last month was 1817.66 g (each median), and 80 patients reported cannabis use within the last 30 days. Determination of EtG was performed with a liquid chromatography-tandem mass spectrometry method with deuterium-labeled EtG as internal standard.

Results: For EtG in urine, a good correlation was found with other state markers and days of sobriety. In a regression analysis, age, gender, marijuana use, kidney disease, and total grams of ethanol consumed last month were the variables that significantly influenced EtG levels in contrast to race, smoking, body mass index, cirrhosis of liver, age began drinking regularly, packs of cigarettes smoked last month, and total body water. Furthermore, in a receiver operating characteristic curve analysis to distinguish between nondrinkers and individuals sober >4 days versus individuals drinking in the recent 4 days, area under the curve was 0.834. At a cutoff of 0.145 mg/liter, sensitivity was 83.5% and specificity 68.3%. A receiver operating characteristic curve was calculated for lifetime alcohol abuse or dependence against those who had never been abusers or dependent. In this case, subjects were either never dependent or lifetime dependent, but those currently dependent were excluded. The resulting area under the curve was 0.694. At a cutoff of 0.145 mg/liter, sensitivity was 73.8% and specificity 60.3%. For those with a self-reported sobriety of less than 24 hr, the area under the curve was 0.899, sensitivity was 90.8%, and specificity was 76.5% at a cutoff of 0.435 mg/liter when we calculated nondrinkers and light drinkers against heavy drinkers and drinkers needing treatment. Cannabis-using patients showed significant differences with regard to almost all state markers when compared with nonconsuming subjects.

Conclusions: Age, gender, marijuana use, kidney disease, and total grams of ethanol consumed last month should be taken into consideration when interpreting results of EtG in urine. Sensitivity and specificity seem promising. Cannabis use can be regarded as an indicator for other serious mental problems in alcohol-using subjects.