An Evaluation of Brief Cognitive Tests for the Identification of Mild Cognitive Impairment

Mild cognitive impairment (MCI), a term used to describe the transitional state between normal aging and established dementia, has been identified as a potentially effective
time point at which to target interventions to prevent or slow the decline into dementia. An efficient way of identifying people with MCI is required as a first step in conducting large scale studies to develop and evaluate interventions targeted at this high risk population.

The aim of the thesis was to investigate the effectiveness of a range of brief cognitive tests to be used for the purpose of identifying people with MCI in an efficient and
accurate manner. A systematic review of the literature found that over 40 brief cognitive tests have been developed and tested to identify amnestic MCI (aMCI). However, the majority of these previous studies were conducted in secondary care settings and were at high risk of unblinding the assessment process, which may have exaggerated the diagnostic accuracy of the assessed tests.

The Memory Alteration Test (M@T) and Test Your Memory test (TYM) were selected as potentially useful brief cognitive tests for aMCI and their validity was assessed within a cohort of older people recruited from the community, without prior knowledge of their cognitive status. A total of 472 older people were assessed for MCI according to
the Petersen criteria using a standardised battery of neuropsychological tests. A prevalence of MCI of 16.5% was found within the assessed cohort and the M@T was found to be more accurate than the TYM at detecting aMCI, performing with a sensitivity of 85% and specificity of 84%.

The ability of reaction time task derived measures to identify cognitive impairment within the cohort was also assessed. These were not as accurate at predicting cognitive status as the M@T. However, they did demonstrate some promising discriminative abilities and should be further explored as potentially useful alternatives to the
traditionally used memory tests, which may be influenced by administrator bias, education level and language ability.