Periodic oscillatory slowing and aperiodic broadband shifts in isolated REM sleep behavior disorder are associated with phenoconversion.

STUDY

Objectives

Conventional EEG slowing in isolated REM sleep behavior disorder (iRBD) may be confounded by disease-related shifts in the aperiodic background. We separated periodic and aperiodic EEG components to clarify oscillatory slowing and examine associations with phenoconversion and clinical measures.

Methods

Resting-state EEG was recorded from 134 patients with iRBD and 51 healthy controls (HC). During follow-up, 26 converted to Parkinson's disease or dementia with Lewy bodies. Power spectra (2-50 Hz) were parameterized using FOOOF to extract aperiodic offset, low-frequency slope (2-30 Hz), and high-frequency slope (30-50 Hz), and to derive the periodic spectrum after aperiodic removal. Group differences, spatial distributions, survival associations, and clinical correlations were analyzed.

Results

After removing the aperiodic component, iRBD showed higher band-averaged periodic 4-8 Hz power than HC (p=0.042), consistent with slowing of the dominant alpha rhythm. In aperiodic analyses, iRBD exhibited higher offset than HC (p<0.001). Converters showed more pronounced oscillatory slowing (p=0.008). Topographically, converters showed higher offset in the occipital region. Survival analyses showed that higher posterior periodic 4-8 Hz power (hazard ratio = 1.342, 95% confidence interval = 1.1-1.6) and offset (hazard ratio = 7.457, 95% confidence interval = 2.6-21.3) were associated with phenoconversion risk. Among converters, MoCA-K correlated with low-frequency slope (r=-0.613), whereas MDS-UPDRS-III correlated with offset (r=0.605).

Conclusions

Separating periodic from aperiodic EEG activity revealed distinct RBD- and phenoconversion-related oscillatory slowing and aperiodic alterations. These findings suggest that aperiodic components may provide additional prognostic and clinical information beyond conventional band power.