Objective sleep data as predictors of cognitive decline in dementia with Lewy Bodies and Parkinson's disease

Introduction: Parkinson's disease (PD) and Dementia with Lewy Bodies (DLB) prognosis depends on cognitive function evolution. Sleep disorders, as objectivated by polysomnography (PSG), are intimately connected with PD and DLB pathophysiology, but have seldomly been used to predict cognitive decline. Methods: 20 DLB and 49 PD patients underwent one-night in-lab video-PSG. Sleep variables were defined, including REM sleep motor events, Tonic and phasic REM sleep muscular tone and RBD diagnosis. Cognitive state (assessed with the Global Deterioration Scale (GDS) was collected from case files for 6 months intervals, for a maximum period of 3.5 years or until death/drop-out.). The relation between PSG data at baseline and variation of GDS scores over time was tested with mixed linear regression analysis. Results: GDS scores were higher in DLB, than in PD. We confirmed significant cognitive decline in both disorders, but no significant differences in progression between them. There were no significant interactions between PSG data and GDS variation for the entire group and DLB separately. In PD patients, there was a significant interaction between RBD diagnosis and tonic excessive muscular tone and GDS increase. Conclusion: Our data suggests that PSG data can be useful in predicting cognitive decline in PD but not in DLB patients. In PD patients, an RBD diagnosis is predictive of cognitive deterioration, confirming the notion that this non-motor symptom relates to a malignant sub-type. Tonic excessive muscular activity, but not other RBD features, had predictive value in this group, pointing to a specific relation with the disease pathophysiology.