Neuropsychological predictors of conversion from amnestic mild cognitive impairment (aMCI) to dementia: a 4-year clinic-based longitudinal study

Abstract

BACKGROUND: Elderly people who demonstrate memory impairment that falls short of dementia, are referred to as having amnestic Mild Cognitive Impairment (aMCI). AMCI patients have an elevated risk of developing dementia, although not all will do so. Clinical criteria for Alzheimer’s Disease (AD) and aMCI do not specify how impairment of a cognitive nature should be defined. The process of differentially diagnosing these conditions can be improved, if knowledge of neuropsychological measures that best discriminate between neurodegenerative and non-neurodegenerative cognitive impairment is used to implement diagnostic criteria for aMCI and AD.

AIMS: We sought to 1) determine the frequency of aMCI referrals to our specialist memory clinic, 2) characterise the detailed neuropsychology of a group of patients with aMCI, 3) determine the utility in differential diagnosis and test-retest reliability of these neuropsychological measures, and 4) establish a subset of neuropsychological measures that were of prognostic utility in aMCI.

METHODS: The case notes of 187 consecutive referrals received by our Royal Edinburgh Hospital memory assessment service across an 18-month period were reviewed retrospectively and numbers of patients fulfilling aMCI criteria were recorded. The baseline neuropsychological performances of 46 patients with aMCI, 20 patients with very early stage AD, 20 elderly patients with depressive symptoms and 24 healthy elderly participants were compared in order to determine their usefulness in differential diagnosis. AMCI participants were followed-up across an average of 4 years. Baseline neuropsychological performances of the aMCI dementia converters and aMCI non-converters were compared. Logistic regression analysis was applied to ascertain the predictive accuracy of a combination of these.

RESULTS: One quarter of referrals received by our memory assessment service met criteria for aMCI, most of whom displayed additional neuropsychological impairments of a non-memory nature, all the while performing above the highest cut off points on even the most comprehensive dementia screening measures. A number of neuropsychological measures were highly sensitive and specific to early AD however, similar combinations of both high sensitivity and specificity to aMCI were not achieved. Forty one percent of patients presenting to our service who fulfilled criteria for aMCI, received a clinical diagnosis of dementia across an average 4-year period. Performances on a comprehensive cognitive screening measure and a measure of delayed word recognition accuracy at baseline, classified 74% of aMCI patients comprising our clinic sample in accordance with their prognostic fate.

CONCLUSION: A significant proportion of patients presenting to specialist memory clinics display episodic and semantic memory/ or executive impairment that falls short of dementia and that is not detectable using traditional bedside screening measures. The vast majority of such patients (i.e. 72%) experience persisting or progressive cognitive impairment, and a significant proportion (41%) go on to receive a clinical diagnosis of dementia. The baseline neuropsychological performance of aMCI patients who do and do not develop dementia differs, and contributes over and above clinical information to the prediction of long-term diagnostic outcome. The high frequency with which aMCI is encountered in clinical practice, coupled with the minority of aMCI patients who experience resolution of their cognitive impairment, and the sensitivity and prognostic utility of several neuropsychological tasks, has implications for the clinical management of patients with aMCI.