Heart Rate Variability in evaluation of autonomic dysfunction in idiopathic REM-sleep behaviour disorder

Nearly 80% of people diagnosed with idiopathic REM sleep behaviour disorder (iRBD) via video-polysomnography (v-PSG) are expected to be in the prodromal stage of an alpha-synucleinopathy. Signs of autonomic dysfunction can appear earlier than motor or cognitive alpha-synucleinopathy symptoms. Heart Rate Variability (HRV) can potentially be an objective measurement of autonomic dysfunction, and furthermore can be obtained directly from v-PSG.

The aim of this study was to evaluate dysautonomia in iRBD subjects using HRV obtained during different sleep stages and wakefulness from v-PSG.

Subjects positively screened by an RBD screening questionnaire (RBD-SQ) underwent v-PSG to diagnose RBD. HRV obtained from v-PSG recordings was correlated to dysautonomia evaluated from a Non-Motor Symptoms Scale (NMSS) questionnaire. Optimal cut-off values of HRV parameters to predict dysautonomia were calculated using receiver operating characteristics (ROC) - area under the curve (AUC) analysis. The effect of confounder variables was predicted with binomial logistic regression and multiple regression analyses.

Out of 72 positively screened subjects, 29 subjects were diagnosed as iRBD (mean age 66 ± 7.7 years) by v-PSG. Eighty-three per cent of the iRBD subjects in our cohort were at the time of diagnosis classified as having possible or probable prodromal Parkinson's Disease (pPD) compared to zero subjects being positively screened in the control group. The iRBD-positive subjects showed significant inverse correlations of NMSS score, particularly to log low-frequency (LF) component of HRV during wakefulness: r = -0.59 (p = 0.001). Based on ROC analysis and correlation between NMSS score, log LF during wakefulness (AUC 0.74, cut-off 4.69, sensitivity 91.7%, specificity 64.7%, p = 0.028) was considered as the most accurate predictor of dysautonomia in the iRBD group. Apnoea-hypopnoea index (AHI) negatively predicted dysautonomia in the iRBD group. None of the HRV components was able to predict the presence of iRBD in the full cohort. Age, gender, and PSG variables were significant confounders of HRV prediction.

The presented study did not confirm the possibility of using HRV from v-PSG records of patients with iRBD to predict dysautonomia expressed by questionnaire methods. This is probably due to several confounding factors capable of influencing HRV in such a cohort.