Finally, because this is a population sample that was ascertained for adoption status, not for illness status, it is possible for us to provide an estimate of the odds ratios for the effects of these haplotypes on risk for SUD in the population. Therefore, we calculated odds ratios for nicotine dependence using Model 2 for ND (those who have not ever smoked were excluded), specifying haplotypes 2, 3 and 4 as the risk haplotypes and categorizing those with 3 or more symptoms of ND as ND and those with 2 or less as non-nicotine dependent. Using this classification system, the odds ratio for ND for those with 2 risk haplotypes as compared to those with no risk haplotypes was 1.82 (95% CI 1.06 to 3.16, n=221). Consistent with the regression analyses, the odds ratio was considerably greater for females (3.73, 95% CI 1.71 to 8.39, n=113) than in males (0.89, 95% CI 0.41 to 1.93, n=108).
