Effect of genetic polymorphisms and smoking histories of participants on nicotine dependence in current smokers: Regression analysis of the effect of combinations of genetic polymorphisms on nicotine dependence
Each analysis was performed for the whole cohort (the left side) and the male subgroup only (the right side). Forward-selection regression began with the effect of the SLC6A3 polymorphism alone. Variables were added one at a time to the model until no remaining variable produced a significant result. SLC6A3: input 1 or 0 for the 10r/10r or other genotype, respectively; ANKK1/DRD2: input 1 for the A2/A2 genotype, 0 for the A1/A2 or A1/A1 genotypes; CYP2A6: input 1 for the *1/*1 genotype, 0 for genotypes including the *4 allele. R2, squared multiple correlation coefficient adjusted for degrees of freedom; AIC, Akaike’s information criterion. *The appropriate model was selected on the basis of minimising AIC.