Complexity in neuropsychological assessments of cognitive impairment: A network analysis approach

In a neuropsychological assessment, each test aims at measuring a single cognitive function. However, test performance depends on an interconnected system of cognitive functions and individual characteristics. For a better understanding of cognitive deficits, it is fundamental to recognize this complexity and study the relationships between test performances. This study aims to evaluate complexity in neuropsychological assessment through network analysis (NA) in 165 healthy older adults, 191 patients with Alzheimer's disease (AD), and 129 patients with vascular encephalopathy (VaE). NA is a flexible method to explore different domains where many variables are correlated with each other, and the relationships are key for understanding the domains. We included general aspects of individual differences (i.e., age, sex, years of education) and the raw scores of a clinically used neuropsychological battery in the network. Healthy subjects showed a segregated pattern, suggesting a good specificity of each test in measuring a specific cognitive function. Moreover, the scores were related to age and education. In the patient groups, the identified patterns changed in a consistent manner, showing less specificity and new relationships between the various tests, thereby reducing the impact of age and education on the performance. In particular, AD patients showed worse performance on the tests but also a different balance between different tasks, suggesting a reorganization of the cognitive system and not a mere decline. These results provide a new perspective in looking at the complexity of cognitive function assessment.