Cognitive Impairment Prediction in MCI and Alzheimers Disease: The Role of MMSE and Face Recognition Deficits

OpenAlex İngilizce

Objective: Face recognition deficits may serve as early indicators of cognitive decline in neurodegenerative disorders such as Mild Cognitive Impairment (MCI) and Alzheimer's disease (AD). This study investigates facial recognition performance in individuals with MCI and AD compared to cognitively healthy controls, exploring its association with global cognitive functioning (MMSE), demographic factors, and depressive symptoms. Methods: Neuropsychological records of individuals aged 65–80 were retrospectively reviewed. Participants were grouped as cognitively healthy controls, MCI, or AD. All had completed the Mini-Mental State Examination (MMSE), Benton Face Recognition Test (BFRT), and Geriatric Depression Scale (GDS). Statistical comparisons were conducted using ANOVA, Kruskal-Wallis, and chi-square tests, followed by correlation and regression analyses. Results: Significant group differences were observed in MMSE scores, BFRT long form (age- and education-adjusted), short form, and Part A scores. No differences were found in GDS scores or BFRT Part B. The control group outperformed both MCI and AD groups in face recognition tasks, particularly in basic face matching (BFRT Part A), with more pronounced deficits in AD. Regression analysis identified MMSE as the strongest predictor of cognitive impairment. While facial recognition impairments were observed in both MCI and AD groups, the MMSE was more effective in identifying overall cognitive decline. Conclusion: Face recognition deficits, especially in basic perceptual matching, are evident in early AD and may provide supplementary diagnostic information. However, MMSE remains the most robust tool for detecting general cognitive impairment. Combining global and domain-specific assessments may enhance early detection strategies. Limitations include retrospective design, group differences in age and education, and absence of neuroimaging or biomarker data.