Complex activities of daily living in mild cognitive impairment: conceptual and diagnostic issues

Background: the impact of cognitive impairment on activities of daily living (ADL) is being used as a major criterion for differentiating between mild cognitive impairment (MCI) and dementia. The concept of an ADL threshold that separates MCI from dementia, however, appears to be improbable for several reasons. Objectives: to determine whether complex ADL are impaired in patients with MCI; to examine the usefulness of the assessment of ADL impairment for the diagnosis of MCI; to explore whether both cognitive testing and assessment of impaired ADL are significant predictors of the diagnosis according to the diagnostic gold standard of MCI. Design: cross-sectional study. Setting: university-based outpatient clinic. Subjects: a total of 45 elderly MCI patients diagnosed according to research diagnostic criteria and 30 age-matched cognitively unimpaired controls. Methods: clinical assessment - Alzheimer's disease Assessment scale, cognitive subscale (ADAS-cog) for the assessment of cognitive functions, Alzheimer's disease Cooperative Study scale for ADL in MCI (ADCS-MCI-ADL) for the assessment of impairments of complex ADL. Statistical evaluation - Mann-Whitney U tests for significant differences on measures of cognition and everyday functioning. Non-parametric correlations for associations between ADL and cognitive ability. Receiver operator curve (ROC) analyses to identify optimal cut-off scores on the ADCS-MCI-ADL and ADAS-cog scales to differentiate between MCI patients and controls. Binary logistic regression analyses to predict the diagnosis of MCI on the basis of the above-mentioned instruments. Results: patients scored significantly higher than controls on the ADAS-cog scale and significantly lower on the ADCS-MCI-ADL scale. There was a significant negative correlation of the above-mentioned scales in MCI patients (r = -0.46, P < 0.01). Both instruments discriminated well between patients and controls (ADCS-MCI-ADL: optimal cut-off 52 points, sensitivity 0.89, specificity 0.97; ADAS-cog: optimal cut-off 10 points, sensitivity 0.78, specificity 1.0). With regard to the linear predictor in the logistic regression built, both instruments were strong predictors of the diagnosis according to the diagnostic gold standard (ADCS-MCI-ADL: P = 0.002; ADAS-cog: P = 0.041). Conclusion: impairment of ADL is already present in MCI. Therefore, intact ADL cannot be used as a criterion to define the syndrome of MCI and to distinguish it from mild dementia. The assessment of complex ADL is probably useful for the diagnosis of MCI.