Working memory impairments in alcohol-dependent participants without clinical amnesia

Background: The delayed alternation (DA) task is highly sensitive to the deficits of nonhuman animals with alcohol-related brain damage. DA is thought to measure working memory which serves as a temporary store for processing of information. However, performance on this type of task has only been investigated in alcohol-dependent humans with severe cognitive deficits. The aim of the current study was to explore the validity of DA as a test sensitive to alcohol-related brain damage by manipulating storage and processing components in three versions of the task. It was hypothesized that alcohol-dependent people would perform worse than control participants and that their deficits would be more pronounced in DA versions with maximal working memory demands. Methods: A sample of 12 alcohol-dependent participants without Wemicke-Korsakoff syndrome was compared with a sample of 12 nonalcohol-dependent controls on three versions of DA. These versions, in order of increasing working memory demand, were single alternation (LR), double alternation (LLRR), and asymmetric alternation (LRRR). DA was administered on a personal computer and performance measured by the number of trials taken to reach criterion. Results: Alcohol-dependent participants, compared with the control participants, took more trials to reach learning criterion on DA on all versions when analyzed together (p = 0.002). Performance on DA was also found to deteriorate with increased working memory demands in both groups of participants (p < 0.001). However, the deficits of alcohol-dependent participants were most pronounced on the DA task with moderate (LLRR) as opposed to extreme (LRRR) working memory demands. Conclusions: The results indicate that both storage and processing demands are necessary for task performance and demonstrate sensitivity of DA to alcohol-related brain injury.