RESUMO
OBJECTIVE: To evaluate the applicability and the psychometric properties of Montreal Cognitive Assessment Brazilian Version (MoCA-BR) in the elderly, as well as comparing its accuracy as a tracking test for mild cognitive impairment (MCI) and mild Alzheimer's disease (AD) with the accuracy of Mini-Mental State Examination (MMSE). METHOD: A transversal study was performed in 4 reference medical centers that care for the elderly. In all, 229 elderly participated in the study. To select the sample, the clinical history of the elderly, Pfeffer Functional Activities Questionnaire, and neuropsychological battery, apart from MMSE and MoCA-BR cognitive tests, were selected. The elderly were classified into control, MCI, and mild AD groups. RESULTS: There was a significant statistical difference between the MoCA-BR scores of the elderly and the control group, MCI, and mild AD (p < 0.001). The Cronbach alpha for MoCA-BR was 0.77, indicating a good internal consistency. The test-retest reliability was elevated, with intraclass correlation coefficient (ICC) 0.91. The inter-examiner reliability was excellent (ICC 0.96). The area under curve of the receiver operating characteristics curve was 0.95, when evaluating the ability of MoCA-BR to discriminate between the elderly with cognitive impairment and cognitively healthy elderly. CONCLUSIONS: The results of the study show that the Brazilian version of MoCA is a reliable cognitive tracking tool and is accurate for the detection of MCI and early stage AD, with good applicability on the elderly with education equal to or more than 4 years and adequate to discriminate between cognitively healthy elderly, and those with MCI and mild, proving to be superior to MMSE in tracking MCI and similar to this test when tracking mild AD.