MMSE versus MoCA for cognitive screening in Brazilian clinical practice: a comparative analysis

Abstract

Background: Identifying cognitive decline—from mild cognitive impairment (MCI) to dementia—remains a major challenge in clinical practice. The Mini-Mental State Examination (MMSE) is the most widely used screening instrument, whereas the Montreal Cognitive Assessment (MoCA) was designed to detect subtle executive deficits. In Brazil, the diagnostic accuracy of both tools is strongly influenced by educational level: the MMSE often shows a ceiling effect in highly educated individuals, potentially masking early decline, while the more sensitive MoCA demonstrates educational bias, increasing false-positive rates among individuals with low formal education. Objective: To compare the performance of the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA) in identifying cognitive decline, from mild cognitive impairment (MCI) to dementia, while considering the impact of educational heterogeneity in the Brazilian population. Methods: A systematic review was conducted in accordance with PRISMA principles. A paired analysis of studies identified in the PubMed, Embase, and SciELO databases was performed using the following search strategy: (“MMSE”) AND (“MoCA”) AND (“comparison” OR “screening”) AND (“Brazil”). A total of 58 studies were initially identified. Articles that did not address the clinical predictive performance of cognitive tests were excluded. After screening and full-text assessment, six studies met the eligibility criteria and were included in the final analysis. Results: The diagnostic accuracy of cognitive screening instruments varied according to the stage of cognitive decline and cognitive reserve. The MoCA demonstrated higher sensitivity for early detection of cognitive impairment, with AUC values ranging from 0.71 to 0.99 and sensitivity exceeding 87% in some populations, identifying executive and attentional deficits not captured by the MMSE. Conversely, the MMSE exhibited a ceiling effect among highly educated individuals, with sensitivity as low as 18% for mild cognitive impairment. However, the MoCA also showed educational bias in low-education populations when the conventional cutoff (≥26) was applied, increasing false-positive rates. Neuroimaging evidence suggests that lower MMSE scores are more frequently associated with medial temporal lobe atrophy, whereas MoCA scores correlate with microstructural alterations and early executive dysfunction. Furthermore, combining the MMSE with the Clock Drawing Test increased diagnostic sensitivity from 43.9% to 97.5% in specific clinical contexts. Conclusion: The MoCA demonstrated greater sensitivity for the early detection of cognitive decline, particularly in mild cognitive impairment, whereas the MMSE showed greater diagnostic stability in populations with lower educational levels. These findings highlight the influence of educational background on test accuracy and support the use of context-adapted screening strategies in clinical practice.