Statistically Defined Parkinson’s Disease Executive and Memory Cognitive Phenotypes: Demographic, Behavioral, and Structural Neuroimaging Comparisons

Article type: Research Article

Authors: Crowley, Samuel J.^a | Banan, Guita^{b; c} | Amin, Manish^{b; c} | Tanner, Jared J.^{a; c} | Hizel, Loren^a | Nguyen, Peter^a | Brumback, Babette^d | Rodriguez, Katie^a | McFarland, Nikolaus^{e; f} | Bowers, Dawn^{a; f} | Ding, Mingzhou^{c; g} | Mareci, Thomas A.^{b; c} | Price, Catherine C.^{a; f; *}

Affiliations: [a] Department of Clinical and Health Psychology, University of Florida, Gainesville, FL, USA | [b] Department of Biochemistry and Molecular Biology, Gainesville, FL, USA | [c] McKnight Brain Institute, University of Florida, Gainesville, FL, USA | [d] Department of Biostatistics, University of Florida, Gainesville, FL, USA | [e] Department of Neurology, Gainesville, FL, USA | [f] Norman Fixel Institute for Neurological Diseases, University of Florida, Gainesville, FL, USA | [g] Department of Biomedical Engineering, University of Florida, Gainesville, FL, USA

Correspondence: [*] Correspondence to: Catherine C. Price, PhD, Associate Professor, Clinical and Health Psychology, 1225 Center Drive, Gainesville, FL, 32611, USA. Tel.: +1 352 273 5929; Fax: +1 352 265 0096; E-mail: [email protected].

Abstract: Background:Some individuals with Parkinson’s disease (PD) experience working memory and inhibitory difficulties, others learning and memory difficulties, while some only minimal to no cognitive deficits for many years. Objective:To statistically derive PD executive and memory phenotypes, and compare PD phenotypes on disease and demographic variables, vascular risk factors, and specific neuroimaging variables with known associations to executive and memory function relative to non-PD peers. Methods:Non-demented individuals with PD (n = 116) and non-PD peers (n = 62) were recruited to complete neuropsychology measures, blood draw, and structural magnetic resonance imaging. Tests representing the cognitive domains of interest (4 executive function, 3 memory) were included in a k-means cluster analysis comprised of the PD participants. Resulting clusters were compared demographic and disease-related variables, vascular risk markers, gray/white regions of interest, and white matter connectivity between known regions involved in executive and memory functions (dorsolateral prefrontal cortices to caudate nuclei; entorhinal cortices to hippocampi). Results:Clusters showed: 1) PD Executive, n = 25; 2) PD Memory, n = 35; 3) PD Cognitively Well; n = 56. Even after disease variable corrections, PD Executive had less subcortical gray matter, white matter, and fewer bilateral dorsolateral-prefrontal cortex to caudate nucleus connections; PD Memory showed bilaterally reduced entorhinal-hippocampal connections. PD Cognitively Well showed only reduced putamen volume and right entorhinal cortex to hippocampi connections relative to non-PD peers. Groups did not statistically differ on cortical integrity measures or cerebrovascular disease markers. Conclusion:PD cognitive phenotypes showed different structural gray and white matter patterns. We discuss data relative to phenotype demographics, cognitive patterns, and structural brain profiles.

Keywords: Entorhinal cortex, white matter, hippocampus, prefrontal cortex, dementia, executive function, memory

DOI: 10.3233/JPD-202166

Journal: Journal of Parkinson's Disease, vol. 11, no. 1, pp. 283-297, 2021