Submental and diaphragmatic muscle activity during and at resolution of mixed and obstructive apneas and cardiorespiratory arousal in preterm infants.

Abstract

Pathomechanisms involved in obstructive apneas remain obscure. Apnea arousal failure has been proposed as a cause for sudden death during sleep. The present study hypothesizes an interdependency between upper airway dilating submental muscle electromyogram (EMG) activity (EMGsub), diaphragmatic muscle activity (EMGdia), incidence of bradycardia, and transcutaneous measured PO2 (tcpO2) upon termination of apnea. Polygraphic recordings, including surface EMG (EMGsub, EMGdia), EEG, ECG, and transcutaneous PO2/PCO2 (tcpO2/tcpCO2) were performed on 10 preterm infants at 36, 40, 44, and 52 wk of conceptional age. EMGsub increased initially, then decreased in 28 of 33 non-rapid eye movement (N-REM) sleep apneas (REM: 35 of 69 events). This correlated with a decrease of tcpO2 during N-REM sleep (p < 0.05). A parallel decrease of EMGsub and EMGdia was correlated with the occurrence of bradycardia (REM and N-REM: p < 0.01). Concomitant termination of apnea and bradycardia (n = 22), occurred in the presence of a phasic, simultaneous activation of EMGsub and EMGdia in 64% of REM sleep and in 79% of N-REM sleep-related event, was characterized by a deep inspiration preceded by a short expiration, and correlated with the extent of tcpO2-decline during REM sleep apneas (p < 0.05). In one apnea with bradycardia that progressed to asystolia, this mechanism was missing, but was evoked by a slight tactile stimulation, where-upon cardiorespiratory functions were immediately reestablished whereas N-REM sleep continued uninterrupted. Our data demonstrate an interdependency between changes of EMGsub and EMGdia activity, tcpO2 decline, and occurrence of bradycardia. A "cardiorespiratory arousal" terminated apneas and bradycardia without a change in sleep phase.