ABSTRACT: Lysophosphatidic acid (LPA) is an endogenous lysophospholipid and a bioactive lipid that is synthesized by the enzyme autotaxin (ATX). The ATX-LPA axis has been associated with cognitive dysfunction and inflammatory diseases, mainly in a range of nonalcoholic liver diseases. Recently, preclinical and clinical evidence has suggested a role of LPA signaling in alcohol use disorder (AUD) and AUD-related cognitive function. However, the ATX-LPA axis has not been sufficiently investigated in alcoholic liver diseases. An exploratory study was conducted in 136 participants, 66 abstinent patients with AUD seeking treatment for alcohol (alcohol group), and 70 healthy control subjects (control group). The alcohol group was divided according to the presence of comorbid liver diseases (i.e., fatty liver/steatosis, alcoholic steatohepatitis, or cirrhosis). All participants were clinically evaluated, and plasma concentrations of total LPA and ATX were measured using enzyme-linked immunosorbent assays. Data were primarily analyzed using analysis of covariance (ANCOVA) while controlling for age, body mass index, and sex. Logistic regression models were created to assess the association of the ATX-LPA axis and AUD or liver disease. LPA and ATX were log10-transformed to fit the assumptions of parametric testing.The main results were as follows: total LPA and ATX concentrations were dysregulated in the alcohol group, and patients with AUD had significantly lower LPA (F(1,131) = 10.677, p = 0.001) and higher ATX (F(1,131) = 8.327, p = 0.005) concentrations than control subjects; patients with AUD and liver disease had significantly higher ATX concentrations (post hoc test, p < 0.05) than patients with AUD but not liver disease; significant correlations between AUD-related variables and concentrations of LPA and ATX were only found in the non-liver disease subgroup (the duration of alcohol abstinence with LPA and ATX (r = +0.33, p < 0.05); and the severity of AUD with ATX (rho = -0.33, p < 0.05)); and a logistic regression model with LPA, ATX, and AUD-related variables showed an excellent discriminative power (area under the curve (AUC) = 0.915, p < 0.001) for distinguishing patients with AUD and comorbid liver disease. In conclusion, our data show that the ATX-LPA axis is dysregulated in AUD and suggest this lipid signaling, in combination with relevant AUD-related variables, as a reliable biomarker of alcoholic liver diseases.