Cross-sectional studies suggest a correlation between alterations in dream content reports and executive dysfunction tests in Parkinson's disease (PD), but this has not been assessed in longitudinal studies. Our objective was to assess the predictive value of dream content for progression of cognitive dysfunction in PD. We prospectively addressed all consecutive, non-demented patients with PD attending an outpatient clinic during a 1-year period. Dream reports were collected at baseline by means of a dream diary and analysed according to the Hall and Van de Castle system. Patients were assessed at baseline for rapid eye movement sleep behaviour disorder, motor stage, mood disorder and psychosis. The Montreal Cognitive Assessment (MoCA) was applied at baseline and 4 years later. Linear regression analysis was used to the test the relation between each dream index (predictors), demographic and other motor and non-motor variables (covariates), and change in MoCA scores (dependent variable). In all, 58 patients were assessed at both time points and 23 reported at least one dream (range 1-27, total 148). Aggression, physical activities, and negatively toned content predominated in dream reports. The MoCA scores decreased significantly from baseline to follow-up. In the multivariate model, negative emotion index was the strongest predictor of cognitive decline. We found a significant positive association between negative emotions in dreams at baseline and subsequent reduction in MoCA scores. These findings suggest that some dream content in patients with PD could be considered a predictor of cognitive decline, independent of other factors known to influence either dream content or cognitive deterioration.
Keywords: Parkinson’s disease; cognitive dysfunction; dream content; longitudinal; predictors.
© 2020 European Sleep Research Society.