To examine whether the extent of substance use moderated the relationship between FA and neuropsychological performance, hierarchical regressions examined mean FA and substance use (i.e., lifetime marijuana use, marijuana use days per month, lifetime alcohol use, drinks per month), and their interaction as predictors of neuropsychological test score. Analyses were conducted after outliers (>2.5 SD) on substance use were removed. Regression analyses examining the significance of FA in the right anterior limb of the internal capsule and lifetime marijuana use in predicting performance on WMS-III Logical Memory delayed recall (n = 32) indicated that increased use was associated with decreased performance on delayed recall (β = -.53; p <.01). The main effect of FA and the FA × use interaction were nonsignificant. Extent of marijuana use did not significantly moderate the relationship between FA and test performance in other regions. In addition, there were no significant main effects of alcohol use (lifetime or drinks per month) or FA × alcohol use interactions in predicting test performance.
