Quantitative EMG criteria for diagnosing idiopathic REM sleep behavior disorder.

Abstract

The purpose of this study is to determine the diagnostic cutoff for the proportion of rapid eye movement (REM) sleep with tonic and phasic activities of the submentalis muscle activity that can be used to diagnose REM sleep behavior disorder (RBD). Seventeen patients clinically diagnosed as idiopathic RBD and 15 age- and gender-matched controls were studied. Surface electromyography was recorded from the submentalis muscle, and two sleep technologists manually identified epochs with tonic and phasic activities during REM sleep. Receiver operating characteristic (ROC) curves were constructed to find the optimal cutoff values for diagnosing RBD using the proportion of REM sleep with tonic and phasic activities of the submentalis muscle. Cohen's kappa coefficient was calculated to evaluate interrater reliability. The cutoff value with the optimal sensitivity and specificity was 6.5% for the proportion of REM sleep with tonic activity (sensitivity, 94.1%; specificity, 93.3%; area under the ROC curve, 0.976) and 9.5% for the proportion of REM sleep with phasic activity (sensitivity, 94.1%; specificity, 93.3%; area under the ROC curve, 0.992). The cutoff value required to achieve a specificity of 100% was 8.9% for tonic activity and 11.1% for phasic activity. Cohen's kappa coefficient between two scorers was 0.96 (95% confidence interval, 0.95-0.97) and 0.95 (95% confidence interval, 0.94-0.95) for tonic and phasic activities, respectively (both p < 0.001). Identifying periods of tonic and phasic activities of the submentalis muscle during REM sleep is useful to discriminate patients with idiopathic RBD from controls.