Subjects

The present study was carried out on 10 male volunteers with a mean age of 27.1 ± 1.66 years (mean ± standard deviation [SD]) and a mean body mass index of 22.0 ± 2.27 kg/m² . Subjects with ongoing respiratory disorders or infections, subjects taking medication known to affect muscle activity, and subjects with severe orofacial skeletal disharmonies were excluded from the study. Informed consent was obtained from each subject before the study.

Simultaneous recordings of GG EMG activity and respiratory movement

EMG signals from the left GG muscle were recorded in a monopolar fashion using a stainless steel fine-wire electrode (diameter, 0.03 mm, MT Giken, Tokyo, Japan) via the perioral approach.10 The tip of the urethane-insulated wire was bared approximately 2 mm. The wire was inserted using a 27-gauge needle, which was removed immediately after the electrodes were placed. A reference electrode was placed on the left earlobe. EMG activity of the GG muscle was identified by (1) tonic burst activity during tongue protrusion, and (2) rhythmic phasic activity coinciding with inspiration. Respiratory movement of the chest wall was recorded simultaneously by an inductance band (TR-751T, Nihon-Kohden, Tokyo, Japan).

Protocol

Recordings were made in 8 body and head positions: (1) upright body with head straight, (2) upright body with head rotated to the right, (3) upright body with head rotated to the left, (4) supine body with head straight, (5) supine body with head rotated to the right, (6) supine body with head rotated to the left, (7) lateral recumbent body position to the right, and (8) lateral recumbent body position to the left.

In the upright-body position, the subject sat erect in a reclining chair without a headrest and kept the Frankfort horizontal parallel to the floor. In the supine body position, the chair was reclined flat and the subject lay down, using a headrest. The Frankfort horizontal was kept perpendicular to the floor, and the longitudinal head axis coincided with the trunk axis through the use of a pillow. In the lateral recumbent body position, the subject lay on his right or left side and, using a pillow, held his head in line with his trunk. In positions with the head rotated, the subject was instructed to rotate his head as much as possible without overstraining or producing discomfort.

In each position and at least 5 minutes after setting up the electrodes and monitoring respiratory movement and body and head position, GG EMG activity was recorded for 3 minutes. During these recordings, subjects were instructed to breathe 20 times per minute through the nose with the mouth closed and to keep their teeth in contact in the intercuspal position.

After each recording session with the head rotated, subjects were instructed to return their heads to the straight position. Similarly, after each recording session in the lateral recumbent position, subjects returned to a supine position, and we confirmed recovery of GG EMG activity. EMG signals were amplified, band-pass filtered at 50 Hz to 3 kHz, full-wave rectified, and integrated at a time-constant of 0.3 seconds. After the signals were passed through an A/D converter (Maclab/8S, AD Instruments, Castle Hill, Australia), they were stored on a personal computer (Macintosh Performa 5270, Apple Computer, Cupertino, Calif) for data analysis. We also examined the reproducibility and temporal changes in GG EMG activity in 2 subjects, respectively. All EMG signals were recorded and stored in the same manner.

Data analysis

Integrated GG EMG signals were averaged with reference to respiratory movement. Tonic GG EMG activity was defined as the minimum GG EMG activity during expiration. Phasic GG EMG activity was defined as the difference between the maximum EMG activity during inspiration and the minimum EMG activity during expiration. EMG activity was normalized to the maximum value during tongue protrusion in the upright position. In studies of reproducibility and temporal changes in GG EMG activity, EMG activity was standardized to the maximum value during inspiration in the upright position. The paired t-test was used to evaluate the dependence of GG EMG activity on laterality, eg, upright body position with the head rotated to the right and left. A 1-way, repeated analysis of variance (ANOVA) was used to investigate significant differences. For the purpose of multiple comparisons, contrasts were used to compare GG EMG activities among different body and head positions in both phasic and tonic activities. Statistical significance was established at P < .05.

GG EMG activity and changes in body positions

Many studies have examined changes in GG EMG activity when subjects move from an upright position to a supine position. Sauerland and Mitchell2 observed activation of the GG muscle with a change from the head-up to the supine-body position. In addition, Pae et al14 reported that GG EMG activity was approximately 34% greater in the supine position than in the upright position. In the present study, both phasic and tonic GG EMG activity increased significantly in association with a change from the upright to the supine position, which was in accordance with previous studies.2,14 Takahashi et al29 suggested that gravity would pull the tongue posteriorly with a change from the upright to the supine position and that augmented GG EMG activity in the supine position could counteract the increased effect of gravity to maintain adequate upper-airway patency. In contrast, Wasicko et al30 reported that there was less GG EMG activity in the supine position than in the head-up position. They suggested that the increment in GG EMG activity reported by Sauerland and Mitchell2 was short-term and may have been related to movement of the head per se, which could possibly activate GG EMG activity via vestibular mechanisms. It is known that vestibular afferents are stimulated during dynamic head movements.31–33 In the present study, because each subject kept his head and jaw in a given position during recording, it is unlikely that vestibular mechanisms were involved in inducing changes in GG EMG activity. Although we examined temporal changes in GG EMG activity in only 2 subjects for up to 20 minutes, the above assumption is supported by the consistent GG EMG activity we observed.

Head rotation and GG EMG activity

We were interested in examining the effect of the direction of head rotation on GG EMG activity, because it is known that head rotation activates the tonic neck reflex pathway and that this reflex affects extreme muscles in a laterally asymmetric manner. These reflexes have been characterized in the orofacial region. Funakoshi and Amano26 studied the effect of the tonic neck reflex on jaw muscles in the decerebrated rat. They demonstrated that when the head was rotated, the temporal, masseter, and digastric muscles were activated with ipsilateral dominance. In our study, however, there was no laterality in the effects of head rotation on GG EMG activity. This may be because of specific anatomical features of the GG muscle, which originates from the mental spine of the mandible and runs midsagittally toward the tongue body. Therefore, it may not be necessary for the GG muscles on either side to behave differently when the head is rotated. However, further studies are needed to explain why head rotation causes GG EMG activity to decrease rather than increase.

Phasic GG EMG activity decreased during head rotation in both the upright and supine positions in the present study. Ohmae et al34 and Karaho et al35 examined morphological changes in the pharynx during head rotation. They reported that bony structures, such as the mandible, hyoid bone, and thyroid cartilage, rotated with the cervical spine but that less rotation was observed for more caudal structures. They also showed that upper-esophageal resting pressure decreased significantly during head rotation. Their studies suggest that morphological changes in the upper airway are attributable to changes in the position of the hyoid bone and the larynx. The hyoid bone is suspended by supra- and infrahyoid muscles and surrounding soft tissue that could be deformed by head rotation. Because it has been shown that forces acting on the hyoid bone can change upper-airway resistance during respiration,1,36 it is likely that displacement of the hyoid bone by head rotation can maintain adequate upper-airway patency. If this is the case, the GG muscle would not necessarily be activated during head rotation. Therefore, we speculate that the change in position of the hyoid bone may account for the decreased GG EMG activity during head rotation.

Lateral recumbent position and GG EMG activity

Little is known about changes in upper-airway physiology in the lateral recumbent position. In our study, we observed significant decreases in both phasic and tonic GG EMG activity when subjects moved from the supine to the lateral recumbent position. Using the acoustic reflection technique, Jan et al4 found that there were no significant differences between upper-airway cross-sectional areas in the supine and lateral recumbent positions in normal awake subjects. A factor that may explain the our discrepant findings is the difference in the relative direction of the gravitational pull on the tongue: there is less gravitational effect in the lateral recumbent position than in the supine position. Pevernagie et al16 studied the influence of changes in the direction of the gravitational pull on upper-airway structures in patients with OSA. They reported that in both the supine and lateral recumbent positions, subjects with positional OSA, ie, patients in whom the rate of apnea and hypopnea during sleep in the supine position was 2 or more times greater than the rate in the lateral position, had a larger upper-airway cross-sectional area during wakefulness than subjects with nonpositional OSA.21,22 They also demonstrated that subjects with positional OSA had a greater transverse upper-airway diameter in the lateral recumbent position than did subjects with nonpositional OSA. Accordingly, they assumed that gravity predisposed patients with nonpositional OSA to upper-airway collapse when they slept on their sides and that gravity had a smaller effect on patients with positional OSA. Although we did not examine morphological changes in the upper airway in response to movement from the supine to the lateral recumbent position in the current study, it is possible that the decreased gravitational force on the tongue in the lateral recumbent position diminishes encroachment of the tongue on the upper airway,37 which in turn requires less GG EMG activity.