In the logistic regression analyses for the effects of CTQ scores and (a) haplotypes and (b) rs11503014 on substance dependence, the only tests that were independent were the tests where the outcomes were single diagnoses of alcohol, cocaine and heroin dependence, resulting in a Bonferroni corrected significant p value of p < 0.008 for the whole model tests. A more stringent correction for all 14, non-independent logistic regression analyses would result in a Bonferroni corrected significant p value of p < 0.004 for each of the logistic regression whole model tests.
