Introduction Idiopathic rapid eye movement behaviour disorder (iRBD) is a strong predictor for alpha-synucleinopathies including Parkinson’s disease. In this study, we examined whether the percentage of REM sleep without atonia (RSWA) during polysomnography (PSG) can predict phenoconversion to neurodegenerative disease in idiopathic REM sleep behaviour disorder (iRBD).

Methods Patients with PSG-confirmed iRBD, including those that phenoconverted during follow-up after diagnosis and those that remained free of neurodegeneration, were identified from an existing database. Tonic, phasic, mixed and ‘’any’’ RSWA activity from the mentalis, tibialis anterior and flexor digitorum superficialis muscles was analysed. Demographic, clinical, PSG and RSWA variables were compared between converters and non-converters. RSWA cut-offs predicting phenoconversion were established using receiver operating characteristic analysis.

Results Six (46%) patients developed parkinsonism (n = 4) or Lewy Body dementia (n = 2). Phenoconverters had significantly higher percentages of RSWA at iRBD diagnosis than non-converters (p = 0.04). Optimal cut-off values to predict phenoconversion were 59.2% for ‘’any’’ (mentalis) RSWA (67% sensitivity; 100% specificity) and 0.26% for tonic without mixed RSWA (83% sensitivity; 100% specificity), with respective area under the curve values of 0.857 and 0.905.

Discussion Patients with an increased percentage of RSWA at iRBD diagnosis were shown to have an increased risk of subsequent neurodegenerative disease. Recent changes to the AASM Manual for Scoring Sleep and Associated Events (v2.6) state that reporting a RSWA index using SINBAR criteria is optional; this study provides further evidence for the value and clinical relevance in producing and reporting a RSWA index.