Background and purpose: Previous research examining mild cognitive impairment (MCI) has highlighted the heterogeneity of outcome in MCI sufferers. MCI is associated with greater risk of progression to dementia; however, a substantial proportion of those identified with MCI have alternative outcomes including recovery to unimpaired status. This heterogeneity may in part reflect insufficient sensitivity and specificity in identifying subclinical memory impairment.

Method: The present study examined learning in a sample of 109 adults aged 61–91 years with persistent amnestic MCI, persistent non-amnestic MCI, recovered MCI and healthy controls. At the final assessment point, learning for words recalled across each trial of the Rey Auditory Verbal Learning Test was examined for each group.

Results: It was found that persistent amnestic MCI participants displayed significantly lower learning compared with recovered MCI and healthy control groups.

Discussion: The results of this study indicated that poor learning across trials may be a defining feature of persistent amnestic MCI. Further research is required to establish the predictive utility of within trial list learning performance to identify individuals with persistent and progressive variants of MCI.