Polysomnographic versus video scoring of large muscle group movements during sleep in children with restless sleep

Background: Restless sleep is a common complaint in children with sleep and other disorders, when it occurs as an isolated symptom it can depict a separate condition called restless sleep disorder (RSD). A key diagnostic feature of RSD is the polysomnographic (PSG) scoring of five or more large muscle movements (LMM) per hour of sleep. Initial publications on RSD used video-PSG (vPSG) analysis to count LMM. However, recently, scoring criteria for LMM on PSG signals alone have been published, but their correlation with vPSG scoring has not been done. In this study we compare vPSG to PSG scoring of LMM in a group of children with restless sleep. Methods: Video-PSG studies from 20 children (14 boys and six girls, mean age 11.3 years, SD 4.17) were reviewed by a blinded scorer. First, scorer used a video analysis (video-LMM), then on a different day the scorer used PSG criteria for LMM without video (PSG-LMM). Results: PSG-LMM were found to be significantly more numerous than video-LMM (mean LMM index 8.6 ± 2.95 SD vs. 6.9 ± 2.40 SD, respectively) while their duration was not statistically different (mean LMM duration 11.6 ± 2.33 s vs. 11.6 ± 2.99 s, respectively). 95.4% of video-LMM were also detected by PSG. Both LMM indexes and durations obtained by the two methods correlated significantly between them. A cut off of 6.24 LMM/hour by PSG-LMM was calculated to be equivalent to 5 LMM by video-LMM. Conclusions: PSG-LMM scoring criteria detects >95% of LLM scored by video-PSG. PSG-LMM lead to the detection of a higher number of movements, indicating their higher sensitivity in picking up LMM in PSG recording. This further supports that these criteria need adequate consideration for their eventual inclusion into the current sleep scoring manual. A higher index may be needed to diagnose RSD.