Test the best: Classification accuracies of four cognitive rating scales for parkinson⇔s disease mild cognitive impairment

Abstract

Objective: A progressive cognitive impairment is one of the frequent non-motor symptoms during Parkinson’s disease (PD) course. A short and valid screening tool is needed to detect an incipient cognitive deficit at the mild cognitive impairment stage in Parkinson’s disease (PD–MCI). Method: The present study aims to evaluate the classification accuracies of four cognitive screenings: Montreal Cognitive Assessment (MoCA), Mattis Dementia Rating Scale second edition (DRS–2), Mini-Mental State Examination (MMSE) and Frontal Assessment Battery (FAB) in a cohort of PD patients (PD–MCI, n = 46; and Parkinson’s disease with normal cognition, PD-NC, n = 95) and Controls (n = 66). All subjects underwent a standard neuropsychological battery as recommended by the International Parkinson and Movement Disorder Society and underwent all four screening tools. Results: In the detection of PD-MCI versus PD-NC, the MoCA showed a sensitivity of 84% and a specificity of 66% with a screening cutoff score at ≤25 points. The MoCA’s AUC was 86% (95% CI 78.7–93.1). In the detection of PD-MCI versus Controls, the FAB displayed 84% sensitivity and 79% specificity with a cutoff ≤16 points, to screen. The FAB’s AUC was 87% (79.0–95.0). Conclusions: Our results show that the MoCA is the most discriminative tool for screening MCI in the PD population.