Conventional polysomnographic (PSG) sleep staging to sleep staging based on a static-charge-sensitive bed (SCSB) recording in infants and young children was compared. The study consisted of whole-night clinical sleep studies in 22 children at 24 weeks (SD 24, range 1–79 weeks) of age. Most of the children presented with respiratory disturbances during sleep. From the SCSB record, sleep stages were differentiated according to regularity of breathing, presence of body movements, and most important, presence of high-frequency components of breathing (SCSB spikes). With both methods, three sleep/wake stages were distinguished: rapid eye movement (REM) sleep, non-rapid eye movement (NREM) sleep and wakefulness. The average inter-scorer reliability of the PSG sleep staging controlled in nine subjects was 88%. The average concordance between the two methods ranged from 82 to 85%, depending on the criteria used for scoring the SCSB. The mean sensitivity of the SCSB to detect NREM sleep ranged from 77 to 90% and the mean sensitivity to detect REM sleep ranged from 61 to 86%. The mean positive predictive value was 89–96% for NREM sleep and 54–67% for REM sleep. In conclusion, REM sleep is characterized by irregular breathing with superimposed fast respiratory movements. These changes are specific enough to allow distinction between episodes of NREM sleep, REM sleep and wakefulness with the non-invasive SCSB method in infants and young children. Incomplete concordance between PSG and SCSB score was most frequently observed during sleep stage transition periods, where the behavioural state and electrophysiological criteria disagreed. When combined with the PSG, the SCSB provides complementary information about the behavioural state of child.
Department of Physiology, University of Turku, Finland 2:
Department of Paediatrics, John Hunter Hospital, Newcastle, Australia 3:
Department of The Diseases of the Chest, University of Turku, Finland 4:
Department of Medicine, University of Sydney, Australia