high rates of EA participants in the SDPS made it difficult to evaluate how results might differ across racial or ethnic subgroups. Third, care must be taken in comparing male proband results with those from the combined group of male and female offspring. Fourth, while the key measure relating to the R2 changes in Table 6 remained significant for most analyses after Bonferroni correction, the significance or the R2 change for offspring prospective data did not meet that standard. Fifth, only 1 measure was used for most characteristics, including sensation seeking and impulsivity, and it is possible that different results might be observed if different measures had been included. Sixth, only a limited number of covariates were used in the analyses presented in Tables 5 and 6 and it is possible additional covariates might have affected the results differently. Seventh, reflecting the high cost for alcohol challenges, this measure was not used in the offspring and could only be included in the proband analyses.
