Models of functional organization as a method for detecting cognitive deficits: data from a sample of social drinkers.

Literature on the cognitive deficits associated with social drinkers' chronic use of alcohol at moderate to heavy levels is equivocal. As an alternative to detecting impairment through measures of mean performance levels, the functional organization of cognitive skills in infrequent and heavy alcohol users was compared. Subjects (N = 364) were adolescent and young adult participants in a longitudinal study of health status and psychoactive substance use. LISREL was used to identify group invariance in the number and nature of cognitive components underlying performance. Results showed that a model with three cognitive components (general intelligence/abstraction, spatial relations/visual-motor speed, and immediate memory) best represented performance in both infrequent use and heavy use groups. There were some group differences in the role of unspecified processing components, but no clear evidence for alcohol-related shifts in functional organization was found. The hypothesis of cognitive compensation, which highlights methodological problems in deficit-detection research, is evaluated with respect to the potential value of using changes in functional organization, that is, the latent structure of performance, to uncover the neurotoxic effects of alcohol or other drug use. More definitive tests of the compensation hypothesis will require prospective, within-subject comparisons of functional organization in clinical as well as nonclinical samples.