ABSTRACT. The preclinical stages of dementia include subtle neurocognitive changes that are not easily detected in standard clinical evaluations. Neuropsychological evaluation is important for the classification and prediction of deterioration in all the phases of dementia. Objective: Compare the neuropsychological performance in healthy older adults with subjective cognitive decline (SCD) and with mild cognitive impairment (MCI) using principal components analysis. Methods: We evaluated 94 older adults with a clinical protocol which included general measures of mental, emotional and functional state. The neuropsychological protocol included tasks of memory, executive function, attention, verbal fluency and visuoconstructional abilities. We used principal component analysis (PCA) to reduce variables´ dimensionality on neuropsychological evaluation. Results: 33(35%) participants had a normal cognitive function, 35(37%) had subjective cognitive decline and 26(28%) had a mild cognitive impairment. The PCA showed seven factors: processing speed, memory, visuoconstruction, verbal fluency and executive components of cognitive flexibility, inhibitory control and working memory. ANOVA had shown significant differences between the groups in the memory (F=4.383, p=0.016, η2p=0.087) and visuoconstructional components (F=5.395, p=0.006, η2p=0.105). Post hoc analysis revealed lower memory scores in MCI than SCD participants and in visuospatial abilities between MCI and SCD and MCI and Normal participants. Conclusions: We observed differentiated cognitive profiles among the participants in memory and visuoconstruction components. The use of PCA in the neuropsychological evaluation could help to make a differentiation of cognitive abilities in preclinical stages of dementia.
P242: Prediction of falls risk with six different biomarkers in mild cognitive impairment using principal component analysis
