Objective: It has been suggested that neuropsychological functioning and cognitive factors influence substance use and treatment outcomes in youth. This study examined a model in which language skills moderate the extent to which expectancies about the positive effects of alcohol predict the persistence of alcohol involvement in youth over an 8-year period.
Method: Participants were substance use disordered adolescents recruited from inpatient alcohol and drug treatment centers (N = 139). Exclusion criteria included major head trauma, neurological illness and psychiatric disorders. Participants were administered neuropsychological tests, expectancy questionnaires and substance involvement interviews that spanned an 8-year period from ages 16 to 24 on average. Substance involvement was assessed by self-report, collateral reports and urine toxicology screens.
Results: Using latent class growth analysis of alcohol use over 8 years, participants were classified as abstainers, infrequent users, worse with time or frequent users. Language x Expectancy interactions were significant at all time points (p range .05 to .0001, effect size eta2 range 0.03 to 0.20). This interaction significantly predicted 8-year alcohol dependence symptoms over and above effects accounted for by covariates or main effects (F = 2.98, 5/100 df, p < .05; R2delta = 4%, beta = 0.21, p <.05).
Conclusions: For youths with above average language skills, positivealcohol expectancies predicted alcohol use frequency and dependence symptoms in the 8 years following treatment; expectancies were less related to outcomes for youths with poorer language scores. Results suggest that verbal skills may magnify the relationship between alcohol expectancies and drinking behavior.