The Impact of Removing Former Drinkers from Genome-wide Association Studies of AUDIT-C.

C. Dao,Hang Zhou,Aeron M. Small,K. Gordon,Boyang Li,R. Kember,Yixuan Ye,J. Gelernter,Ke Xu,H. Kranzler,Hongyu Zhao,A. Justice
DOI: https://doi.org/10.1111/add.15511
2021-04-16
Addiction
Abstract:BACKGROUND AND AIMS The Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) questionnaire screens for harmful drinking using a 12-month timeframe. A score of 0 is assigned to individuals who report abstaining from alcohol in the past year. However, many middle-aged individuals reporting current abstinence are former drinkers (FDs). Because FDs may be more genetically prone to harmful alcohol use than life-long abstainers (LAs) and are often combined with LAs, we evaluated the impact of differentiating them on the identification of genetic association. DESIGN AND SETTING The UK Biobank (UKBB) includes AUDIT-C and alcohol drinker status. PARTICIPANTS 131,510 Europeans, including 5,135 FDs. MEASUREMENTS We compared three genome-wide association (GWAS) analyses to explore the effects of removing FDs: the full AUDIT-C data, AUDIT-C data without FDs, and data from a random sample numerically matched to the data without FDs. Because prior studies show a consistent association of the ADH1B polymorphism rs1229984 with both alcohol consumption and alcohol use disorder, we compared allele frequencies for rs1229984 stratified by AUDIT-C value and FD versus LA status. Additionally, we calculated polygenic risk scores (PRS) of related diseases. FINDINGS The rs1229984 allele frequencies among FDs were numerically comparable to those with high AUDIT-C scores and very different from those of LAs. Removing FDs from GWAS yielded a stronger association with rs1229984 (p-value after removal: 1.9 x10-70 versus 1.7x10-65 and 2.5 x10-62 ), more statistically significant single nucleotide polymorphisms (SNPs) (after removal: 11 versus 9 and 8), and genomic loci (after removal: 11 versus 9 and 7). Additional independent SNPs were identified after removal of FDs: rs2817866 (PTGER3), rs7105867 (ANO3), and rs17601612 (DRD2). For PRS of alcohol use disorder and major depressive disorder, there are statistically significant differences between FDs and LAs. CONCLUSIONS Differentiating between former drinkers and life-long abstainers can improve Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) genome-wide association results.
Medicine
What problem does this paper attempt to address?