Working memory in mild cognitive impairment and Alzheimer's disease: Contribution of forgetting and predictive value of complex span tasks.

OBJECTIVE: This study examines working memory (WM) in mild cognitive impairment (MCI) and Alzheimer's disease (AD). METHOD: Performances on sentence span and operation span were measured in individuals meeting criteria for MCI (n = 20) and AD (n = 16) as well as in healthy older adults (n = 20). In addition, the effect of retention interval was assessed by manipulating the length of first and last items of trials (long-short vs. short-long), as forgetting might contribute to impaired performance in AD and MCI. RESULTS: Results show a group effect (p < .001, η² = .47): In both conditions and for both material types, WM span is lower in AD than in MCI (p < .001), which in turn is lower than in healthy aging (p < .05). An effect of retention interval on complex span was found for all groups (p < .001, η² = .57), supporting a role for forgetting within WM. When computing a proportional interval effect (p < .05, η² = .12), it was found that persons with AD were more sensitive to retention interval than were healthy older adults (p < .05). Among persons with MCI, those who later showed significant clinical deterioration or progression to AD were more affected by retention interval (p < .05, η² = .28) than were those who remained stable. Furthermore, deficits in AD are associated with a higher proportion of intrusion errors, particularly those from the current trial (p < .05, η² = .15), which could reflect inhibitory processes. CONCLUSIONS: Overall, these results indicate impaired WM in age-related disorders with a gradient between MCI and AD. Retention interval increases deficit in persons with AD. It also shows potential in predicting a negative prognosis in those with MCI.