Our findings also suggest that the subtle differences in substance vulnerability definitions may not be of major concern in genetic analyses. All three alternative measures of substance vulnerability showed broadly consistent estimates, and the high SNP-correlations among them suggest the same genetic influences affect the three phenotypes. Thus, it is likely that results of genetic analyses across these phenotypes can be meaningfully compared, and meta- or mega- analyses of datasets using alternative phenotypic definitions of substance vulnerability will be informative.
