Associations between alcohol use disorders and polymorphisms of genes influencing ethanol metabolism have been widely reported, but gene-gene and gene-sex interaction studies have rarely been examined. Using a set of samples collected during an epidemiological study of alcohol use disorders AUDs in a Tibetan population in China, we performed a case-control study to investigate the relationship between the functional polymorphisms of genes influencing ethanol metabolism and AUDs. The sample included 383 individuals with an AUDIT score >or=10 and 350 control subjects with the AUDIT score <or=5. All participants were genotyped for ALDH2*1/*2, ADH1B*1/*2, and CYP2E1*c1/c2*. Data were analyzed employing an integrated strategy using MDR, SPSS, and UNPHASED software. The MDR analysis showed that the four-factor model including ADH1B*1/*2, ALDH2*1/*2, and CYP2E1*c1/*c2 polymorphisms, and sex was the most accurate model associated with AUDs with the highest OR 3.299. It also revealed that CYP2E1 *c1/*c2 polymorphism interacted significantly with sex. Independent analysis confirmed that both ADH2*2 and ALDH2*2 allele were significantly associated with AUDs (OR: 0.441 for ADH2*2 and 0.137 for ALDH2*2). CYP2E1*c2 was positively associated with AUDs only in males homozygotic for ALDH2*1 and ADH1B*1 (OR: 2.585). Cumulative association analysis showed the number of protective alleles and genotypes were negatively associated with AUDs. In conclusion, ALDH2*2 and ADH1B*2 alleles were not only independently associated with AUDs but also demonstrated cumulative dosage effects. However the positive association between CYP2E1*c2 allele and AUDs might only exist in males homozygotic for ALDH2*1 and ADH1B*1.
(c) 2009 Wiley-Liss, Inc.