Phosphatidylethanol as an objective measure of heavy drinking days in a clinical trial for alcohol use disorder.

AIMS: This study evaluated the accuracy of phosphatidylethanol (PEth), a direct biomarker of alcohol consumption, in distinguishing individuals with and without heavy drinking days, and sought to establish an optimal cutoff using data from a recent clinical trial in alcohol use disorder (AUD). METHODS: Data from a 26-week randomized, double-blind, placebo-controlled clinical trial of 127 individuals with AUD were analyzed. A total of 447 blood PEth samples were compared with self-reported heavy drinking days (≥60 grams of alcohol/day for men and ≥48 g/day for women) via Timeline Follow back method and weekly alcohol logs. Spearman correlations between PEth and self-reports were calculated for the past 7, 14, 21, and 28 days at each study visit. Periods with the strongest correlations (past 14-21 days at the Week 4 follow-up) were used for receiver operating characteristic (ROC) analyses. RESULTS: Among 127 participants, 63 completed the 26-week follow-up. Moderate-to-strong correlations were observed, strongest at Week 4 (Rho = 0.65, 95% CI, 0.52-0.75, P < .001). ROC analysis at Week 4 showed high discrimination between participants with ≥1 heavy drinking days versus no heavy drinking days (AUC = 0.90, 95% CI, 0.86-0.92), with a PEth cutoff of 0.145 μmol/L (101.9 ng/ml), yielding 95% sensitivity (95% CI, 86-98) and 80% specificity (95% CI, 61-92). CONCLUSIONS: Our findings support the use of PEth as an objective measure of heavy drinking days in a clinical setting. Further research is needed to validate these findings in larger cohorts.