Regression analyses used past year drinking days to predict composite NP change scores. For females (n=29), more drinking days in the year prior to follow-up predicted greater reduction in performance in Visuospatial Functioning from baseline to follow-up, above and beyond baseline Visuospatial Functioning scores (F(2,24)= 11.42, p<.001; R2Δ=10%, p<.05; β=−.33; see and ). For males, follow-up drinking did not predict NP domain change scores.
Standardized betas showing extent to which adolescent drinking days in the past year predicted change in neuropsychological performances
For female adolescents, more past year drinking days predicted a greater reduction in performance on the Visuospatial Functioning composite score (R2Δ =10%, β = −.33, p < .05).
Exploratory hierarchical regressions used task-specific NP change scores as dependent variables (i.e., the 12 non-premorbid functioning measures in ), the corresponding baseline NP score and any covariates on Block 1, and substance involvement variables on Block 2. Variables that correlated (p<.05) with outcome change scores were used as covariates in analyses. For females, age was used as a covariate for Coding, Block Design, and CVLT Long Delay Free Recall change scores. For males, age was a covariate for Coding and Color Word Interference Inhibition/Switching; CDQ total for Complex Figure copy; family history density for Coding and Block Design; and SES for CVLT Long Delay Free Recall change scores. For females, past year drinking days predicted Complex Figure change scores, similar as above (F(2,20)= 19.04, p<.001; R2Δ=8%, p<.05; β=−.32). For males, higher follow-up HSS scores predicted slowed DVT completion times from baseline to follow-up, above and beyond baseline DVT time (F(3,39)=8.65, p<.001; R2Δ=7%, p<.05; β=.27). DVT omission and commission errors were not significantly related to HSS scores, suggesting that the decrements in time completion were due to processing speed rather than accuracy.
Follow-up regressions using lifetime drinking, average past 3-month drinking, and HSS scores were used to predict individual NP test change scores within the Visuospatial Functioning domain for females. More past 3-month drinking predicted greater reduction in Complex Figure delay performance from baseline to follow-up, above and beyond baseline Complex Figure delay scores (F(2,20)= 19.17, p<.001; R2Δ=8%, p<.05; β=−.32; see ). Lifetime drinking days and HSS scores did not predict reductions on Block Design or Complex Figure copy.
For female adolescents, more drinks per month predicted a greater reduction in performance on complex figure delay (R2Δ = 8%, β = −.32, p < .05).
Results above remained unchanged when using non-transformed alcohol involvement variables. After controlling for alcohol consumption, lifetime and past month tobacco and marijuana use reported at follow-up did not significantly predict any follow-up NP change score. Recency of drinking did not predict follow-up NP change measures, and results above remained unchanged when controlling for days since last alcohol use.