Case Report: non-monotonic longitudinal RSWA with tonic–phasic dissociation across serial attended video-polysomnography in isolated REM sleep behaviour disorder

Background Quantitative REM sleep without atonia (RSWA) metrics are increasingly used to characterize isolated REM sleep behavior disorder (iRBD) and as candidate biomarkers. Cohort follow-up often reports rising RSWA, but within-individual trajectories, particularly when tonic and phasic components diverge, remain sparsely documented. Methods A man in his mid-seventies with iRBD underwent four attended video-polysomnograms over −6.5 years. Submental EMG was quantified using a REM sleep atonia index (RAI) with pre-specified exclusion of REM mini-epochs contaminated by artifact or physiological activation (arousal-linked activation and respiratory-event adjacency). Tonic and phasic submental RSWA components were extracted concurrently. Diary-anchored interviews around each visit captured enactment frequency, ordinal ratings of dream vividness/recall and sleep-related accidental harm. Results Corrected RAI varied non-monotonically (0.268, 0.401, 0.313, 0.463). Tonic and phasic RSWA dissociated: tonic activity declined overall, whereas phasic activity rose early and remained elevated. PSG-night enactment frequency tracked REM sleep opportunity (Spearman ρ = 0.89; n = 4) more closely than corrected RAI or tonic/phasic indices. Sleep-related harm was absent at the final visit despite persistent phasic activity. Conclusion In this case, longitudinal RSWA did not behave as a single monotone scalar. Component-wise reporting alongside explicit REM sleep opportunity and exclusion accounting may be essential for interpreting apparent longitudinal change and relating electrophysiology to behavioral expression.