Lipid profile dysregulation predicts alcohol withdrawal symptom severity in individuals with alcohol use disorder.

Alcohol withdrawal syndrome (AWS) is a serious medical condition of high variability in alcohol use disorder (AUD) after drinking cessation. Identification of clinical biomarkers capable of detecting severe AWS is needed. While alcohol consumption and withdrawal are linked with lipid profile dysregulation, the relationship between lipid levels (high-density lipoprotein cholesterol [HDL-C], low-density lipoprotein cholesterol [LDL-C], and triglycerides) and AWS is unknown. Therefore, this study investigated whether HDL-C, LDL-C, and triglycerides conferred risk for moderate-to-severe AWS symptoms in treatment-seeking individuals (n = 732) admitted to the National Institute on Alcohol Abuse and Alcoholism (NIAAA) alcohol treatment program. Lipid levels were measured upon admission, and the Clinical Institute Withdrawal Assessment of Alcohol Scale, Revised (CIWA-Ar) assessed AWS severity for generating a three-level AWS typology (none-to-mild, moderate, and severe). Multivariable multinomial logistic regression examined whether lipid levels were associated with risk for moderate-to-severe AWS. We found significant predictive relationships between AWS and HDL-C, LDL-C, and triglycerides. While extremely high HDL-C (≥100 mg/dL) conferred the highest odds for moderate (4.405, 95% CI, 2.572-7.546, p < 0.001) and severe AWS (5.494, 95% CI, 3.541-8.523, p < 0.001), the lowest odds ratios for moderate AWS (0.493, 95% CI, 0.248-0.981, p = 0.044) and severe AWS (0.303, 95% CI, 0.223-0.411, p < 0.001) were associated with high LDL-C (≥160 mg/dL). The present study demonstrates that altered lipid levels, measured upon admission for inpatient AUD treatment, may help to predict which individuals are at risk for medically relevant moderate-to-severe AWS. This suggests that further research into the role of lipid biomarkers in AWS may be beneficial for identifying biologically determined risk profiles in AUD. Published by Elsevier Inc.