A haplotype analysis of CYP2E1 polymorphisms in relation to alcoholic phenotypes in mexican americans

Background Studies regarding the association between the 4 polymorphisms of CYP2E1 (CYP2E1*1D, *5B, *6, and *1B) and alcoholism are inconsistent and inconclusive. The purpose of the present study was to clarify previously discordant studies by haplotype analysis in the Mexican American population. Methods The 4 polymorphisms of CYP2E1 were studied in 334 alcoholics and 365 controls. Genotype, allele, and haplotype frequency comparisons between alcoholics and controls were assessed. Patterns of linkage disequilibrium (LD) at CYP2E1 were determined. Reconstructed haplotypes were tested for associations with clinical phenotypes (age onset of drinking, Maxdrinks, and smoking status). Results No significant associations between the 4 polymorphisms of CYP2E1 and alcoholism were revealed by single allele tests. High LD was found between the CYP2E1 c2 and C alleles in Mexican Americans. Eleven haplotypes were present in the 699 participants. The 6 main haplotypes with frequencies higher than 1% made up 97% of the total halpotypes. The frequency of subjects carrying H6 (1C-c2-C-A2) was significantly higher in alcoholics than in controls (p = 0.0001). In contrast, the frequencies of H7 (1C-c2-C-A1) and H10 (1C-c2-D-A1) were significantly lower in alcoholics than in controls (p = 0.0072 for H7 and p = 0.0407 for H10). The frequency of H6 was significantly higher in alcoholics who had late onset of drinking than in nonalcoholic controls. Furthermore, the frequencies of H6 haplotype were also consistently higher in groups who had high number of maximum drinks (9 to 32 drinks) than in controls. When smokers are excluded, the frequencies of H6, H7, and H9 (1C-c2-D-A2) showed statistically significant differences between alcoholics and controls (p < 0.05). Moreover, the association between H6 and alcoholism become more robust when smokers are excluded. Furthermore, the frequency of H1 (1C-c1-D-A2) in alcoholic-smokers was much higher than in alcoholic-nonsmokers (p = 0.0028). In contrast, alcoholic-smokers carried less H2 (1C-c1-D-A1) in comparison with alcoholic-nonsmokers (p = 0.0417). The H3 (1D-c2-C-A2) frequency in alcoholic-smokers was much lower than in alcoholic-nonsmokers (p = 0.0042) and control-smokers (p = 0.0363). Conclusion Our data demonstrate that carrying haplotype H6 might enhance susceptibility to developing alcoholism, but possessing the H7 or H10 haplotype appears to decrease this susceptibility. The H6, H7, and H9 haplotypes may play certain roles in different clinical phenotypes in Mexican American alcoholics. In addition, our data suggest that the H1, H2, and H3 haplotypes are associated with alcohol drinking and smoking. These results support that haplotype analysis is much more informative than single allele analysis. Our findings clearly indicate the importance of H6 haplotype in alcohol drinking in Mexican Americans.