A domain-based framework for cognitive profile identification in Parkinson's disease across diverse samples

Abstract

Introduction: Parkinson's disease (PD) is a progressive neurodegenerative disorder characterized by motor symptoms and heterogeneous cognitive impairments influenced by factors such as age, disease duration, and severity. Traditional neuropsychological assessments often fall short in capturing the multifaceted nature of PD-related cognitive dysfunction due to their reliance on single test metrics. This study provides empirical support for the implementation of domain-based cognitive assessments, structured in line with Movement Disorder Society recommendations, to provide a multidimensional evaluation of cognitive profiles in PD patients.

Methods: Neuropsychological and clinical data were analyzed from 316 PD patients recruited from three Spanish hospitals-Hospital Clínico San Carlos (Madrid), the University Complejo Universitario de Santiago de Compostela (Galicia), and Hospital Virgen del Rocío (Sevilla)- and a control group of 96 older individuals, whose age difference from the PD group was statistically significant. Five cognitive domains were constructed, addressing attention/working memory, executive functions, memory, visuospatial abilities, and language, using composite z-scores derived from standardized neuropsychological tests.

Results: Latent Cluster Analysis identified three distinct cognitive profiles: (1) a fronto-striatal profile characterized by mild deficits in executive and attention functions and intact visuospatial abilities, (2) a posterior cortical profile marked by severe memory and visuospatial impairments but strong language performance, and (3) a preserved profile displaying mild deficits across multiple domains. Comparisons between PD clusters and controls revealed significant differences in cognitive trajectories, emphasizing the value of a domain-based approach for differentiating neurodegenerative patterns from normal aging.

Discussion: The findings highlight the potential of domain-based assessments to unify data across diverse samples, fostering standardized cross-cohort comparisons and facilitating large-scale research initiatives. By enabling methodological consistency, this approach provides a robust framework for advancing the understanding of cognitive dysfunctions in PD and improving clinical decision-making.

Keywords: Parkinson’s disease; aging; cognitive domains; deep brain stimulation; latent cluster analysis; neuropsychological assessment.

FIGURE 1

(A) Shows a representation of the neuropsychological profiles of the three clusters identified in the control group. (B) Represents the profiles of the three neuropsychological clusters identified in the PD group. (C) Shows the overall cognitive profile comparison between the control group and the entire PD group. (D) Compares the three PD clusters with the control group. The figures display scores across neuropsychological domains: ATT/WM, attention/working memory; MEM, memory; EF, executive functions; VOSP, visuospatial function, and LANG, language. Dots closer to the center of the pentagon indicate higher (i.e., better) cognitive scores, while dots farther from the center represent poorer (more negative) scores.