Sudden drop in muscle tone during REM sleep in human beings [Caída abrupta del tono muscular al entrar a sueño MOR en el ser humano]

Relatively low tonic electromyographic activity of the mentalis or submentalis muscles constitutes one of the three electrophysiological signs for identifying rapid eye movement sleep (REM), described in the standardized manual for scoring sleep stages in human subjects. The other two signs, low voltage mixed frequency EEG activity and episodic rapid eye movements are inadequate for delimiting the start of REM sleep, because EEG activity resembles that of stage 1 and rapid eye movements are not constantly present. The term "relatively low" tonic EMG and not "low tonic EMG" is used according to the standardized manual because tonic EMG shows considerable variation from subject to subject and from session to session, and more important because low EMG values may be reached during other sleep stages. Therefore, REM sleep scoring is based on "relatively low" tonic EMG. Despite the relevance of the loss of muscular tone for scoring the start of REM sleep and for sleep disorders-such as narcolepsy and REM sleep behavioral disorder, where loss of muscle tone or the lack of it is implicated-, very few quantitative studies of EMG activity during REM sleep in humans have been performed. Amplitude analysis of mentalis and orbicularis oris muscles and spectral power analysis of suprahyoid, masseter and temporalis muscles have demonstrated that EMG activity is lower during REM than during NREM sleep. The mentalis muscle maintains tonically the lowest values during REM sleep with very low variability during the same REM sleep episode and across REM episodes, except for very brief phasic activations, whereas during NREM sleep muscle tone shows large variations within the same sleep stage and along the night. Only one study exists which analyzes the time course of the loss of tone during the transition from NREM to REM sleep integrating the EMG amplitude. However, it was done for long time windows of 20 seconds that does not allow identifying the precise moment of EMG activity drop. Given that the fall in EMG activity is one of the main keys for REM sleep scoring, the objective of the present investigation is to describe the EMG activity of the mentalis muscle during the NREMREM sleep transition by analyzing short time windows of two seconds. Ten healthy, young adult, right-handed subjects (5 men and 5 women) participated in the study after giving informed consent. All had regular sleeping habits, were in good health and were free of drugs, medication or caffeine intake as assessed by interviews and questionnaires on sleeping habits and health. Polysomnography (PSG) was recorded using a Grass model 8-20E polygraph with filters set at .03 and 70 Hz. Additionally to EEG (C3-A2 and C4-A1), electroculogram (EOG) and EMG of the mentalis muscle, nasaloral air flow and EMG of anterior tibialis muscles were recorded to remove those subjects showing signs of sleep apnea or periodic limb movement disorder. EEG, EMG and EOG were digitized at 1024 Hz through an analog-to-digital converter of 12 bits resolution using the acquisition program Gamma (version 4.4). The initiation of the first three REM sleep episodes of one night for each subject was indicated in the PSG recordings, following the standardized rules of the manual for scoring sleep stages of human subjects. The fourth REM sleep episode was not considered for analysis because not all subjects had a fourth REM episode. EMG activity of the mentalis muscle of three 30-second epochs around the start of REM sleep (the previous one, the REM entrance and the posterior one) was analyzed. EMG activity was submitted to Fast Fourier Transform and absolute power for every 250 msec (256 points) was obtained for two broad bands: one from 24 to 28 Hz and the other from 28 to 32 Hz, as these have demonstrated significant differences between REM and NREM sleep, in previous studies. Absolute power values were log-transformed previous to statistical analysis to approximate them toward normal distribution. The time course of the drop in muscle tone was established in the case of each individual NREM-REM sleep transition for two second time windows, both visually on the EMG signal and also by statistically comparing consecutive 2-second averages of EMG absolute power (8 means of 250 msec). When there was no clear visual or statistical evidence of decreased EMG activity, the 30-second epoch was divided in half. Additionally, the first rapid eye movement was visually identified. EMG signals were visually inspected and absolute power values of two-second epochs containing eye movement or phasic EMG artifacts were substituted by the average of the preceding and following twosecond means. This procedure was chosen instead of rejection in order to maintain the time sequence. The average of substituted epochs was lower than 1 for the NREM-REM sleep transitions. Once the significant differences were established for the individual NREM-REM sleep transitions, the absolute power for the 20 seconds prior and the 20 seconds after the turning point was averaged for the group and compared using the Student t test. A level of p <0.05 was required for significance for both individual and group analyses. EMG drop was statistically identified in 15 out of the 30 NREMREM sleep transitions (p < 0.05). In 14 cases more than one significant difference was found due to phasic increases shorter than two seconds. Thus, EMG drop was established where both visual inspection of EMG signal and statistical differences were matched. It was necessary to divide the 30-second epoch in half just in one individual case. The comparison of EMG power after averaging for the group the 20 seconds before and the twenty seconds after the individual turning point showed that EMG absolute power was significantly different for the two bands (p < 0.0001 for both bands). The first eye movement occurred after the EMG drop in 28 out of the 30 NREM-REM sleep transitions within a range of 2 and 52 seconds. EMG fall was simultaneous to the first eye movement in one case and eye movement preceded EMG drop in just one NREM-REM sleep transition. Present results indicate that the loss of muscle tone of the mentalis muscle during the transition from NREM to REM sleep occurs suddenly rather than gradually, within a time window lasting no longer than 2 sec. This could be appreciated in individual as well as in group analysis. It still remains a matter of debate if REM sleep is under the control of a single generator that simultaneously commands the start of all of its physiological changes, or if each of the physiological systems involved in REM sleep is under its own command starting at its own time and are only orchestrated by a common mechanism. The loss of muscle tone occurred before the first rapid eye movement in 29 out of 30 of the REM sleep onset episodes analyzed, upholding the proposition that physiological systems involved in REM sleep follow different time courses in agreement with non-simultaneous onset of the different physiological mechanisms as it happens with pontogeniculate-occipital waves in cats that begin long before EEG desynchronization and EMG fall and with results observed in two studies in man which report that EMG amplitude decreases before eye movements. The sudden drop in muscle tone during NREM-REM sleep transition may help to understand the physiological mechanisms involved in sleep disorders where loss of muscle tone or the lack of it is implicated, such as narcolepsy and REM sleep behavioral disorder. It can also be used as an objective sign to establish the onset of REM sleep in research where the precise moment of REM sleep onset is needed. The time relationship among muscle tone fall and other physiological signs of REM sleep remains to be investigated