The profiles and practice of cognitive function in Parkinson's disease

The overall aim of this thesis was to gain a better understanding of the clinical and neurobiological characteristics of cognitive impairment in Parkinson’s disease (PD) and the efficacy of cognitive training for its treatment. We used neuropsychological and neuroimaging data from a large outpatient cohort of PD patients to study cognitive heterogeneity and the underlying dopaminergic and serotonergic deficiencies. We further systematically reviewed the literature on the neural correlates of cognitive training (CT) and performed a randomized controlled trial in 140 PD patients on the neuropsychological and brain network effects of an eight-week online cognitive training program. We performed a cluster analysis in a large sample of 226 mostly early-stage PD patients to assess the heterogeneity in the clinical symptom profile of PD, with a focus on cognitive heterogeneity. We distinguished four symptom profiles, including two symptom profiles with similar mean age and disease stage, but significant differences in their clinical cognitive symptom profile. Briefly, the results suggest the existence of parallel disease pathways that may have predictive value for the rate of cognitive decline, for the profile and course of neuropsychiatric symptoms and for treatment efficacy. To evaluate the dopaminergic and serotonergic deficiencies associated with cognitive impairment in PD, we performed a dopamine transporter SPECT study in the same patient cohort. Binding of 123I-FP-CIT in the caudate head and anterior putamen was associated with processing speed tasks, while extra-striatal binding did not show any significant associations with cognitive function. This study showed that deterioration of the striatal dopamine system may be involved in impairment of mental processing speed. We additionally performed a randomized controlled trial, the COGnitive Training In Parkinson Study (COGTIPS), to study the effects of home-based, multi-domain CT in PD patients with subjective cognitive complaints. In this thesis, we describe the study protocol, primary results and effects on functional connectivity and network topology. 140 PD patients were randomized into an experimental adaptive and challenging cognitive training or an active control condition. The eight-week cognitive training did not improve planning accuracy, but did produce small significant differences between the groups on processing speed during executive function tasks. Improvement on the experimental training games correlated with the improvement on the Tower of London and Stroop Color Word Test. Interestingly, post-hoc sub-group analyses suggested that CT was more efficacious in patients that were cognitively more impaired and fulfilled the diagnostic criteria for PDD, but this sub-sample was small. At six-months follow-up, the initial positive effects were no longer present. To improve our understanding of the neural processes underlying the effects of CT on cognitive function, we performed a systematic review of studies that used functional MRI (fMRI) before and after CT in healthy elderly and subjects with a neurodegenerative disease. This literature study provided convincing evidence that CT can alter brain activation and connectivity. In healthy populations, CT induced lower regional brain activity potentially as a result of enhanced neural efficiency. Conversely, in individuals with neurodegenerative diseases brain activity increased after CT, presumably showing that CT increased the neural effort that can be delivered to successfully fulfill task demands. Resting-state fMRI studies showed that CT increased connectivity and enhanced segregation between task-positive and task-negative resting-state networks. We aimed to replicate these findings in the COGTIPS sample, where we studied the effects of CT on resting-state network connectivity and the brain network topology. This study showed that our eight-week intervention did not affect global brain network connectivity and topology. Post-hoc analyses at the nodal level showed that CT enhanced connectivity of the dorsal anterior cingulate cortex and dorsolateral prefrontal cortex.