Genioglossus reflex responses to negative upper airway pressure are altered in people with tetraplegia and obstructive sleep apnoea

Abstract

Key points: Protective reflexes in the throat area (upper airway) are crucial for breathing. Impairment of these reflexes can cause breathing problems during sleep such as obstructive sleep apnoea (OSA). OSA is very common in people with spinal cord injury for unknown reasons. This study shows major changes in protective reflexes that serve to keep the upper airway open in response to suction pressures in people with tetraplegia and OSA. These results help us understand why OSA is so common in people with tetraplegia and provide new insight into how protective upper airway reflexes work more broadly.

Abstract: More than 60% of people with tetraplegia have obstructive sleep apnoea (OSA). However, the specific causes are unknown. Genioglossus, the largest upper-airway dilator muscle, is important in maintaining upper-airway patency. Impaired genioglossus muscle function following spinal cord injury may contribute to OSA. This study aimed to determine if genioglossus reflex responses to negative upper-airway pressure are altered in people with OSA and tetraplegia compared to non-neurologically impaired able-bodied individuals with OSA. Genioglossus reflex responses measured via intramuscular electrodes to ∼60 brief (250 ms) pulses of negative upper-airway pressure (∼-15 cmH₂ O at the mask) were compared between 13 participants (2 females) with tetraplegia plus OSA and 9 able-bodied controls (2 females) matched for age and OSA severity. The initial short-latency excitatory reflex response was absent in 6/13 people with tetraplegia and 1/9 controls. Genioglossus reflex inhibition in the absence of excitation was observed in three people with tetraplegia and none of the controls. When the excitatory response was present, it was significantly delayed in the tetraplegia group compared to able-bodied controls: excitation onset latency (mean ± SD) was 32 ± 16 vs. 18 ± 9 ms, P = 0.045; peak excitation latency was 48 ± 17 vs. 33 ± 8 ms, P = 0.038. However, when present, amplitude of the excitation response was not different between groups, 195 ± 26 vs. 219 ± 98% at baseline, P = 0.55. There are major differences in genioglossus reflex morphology and timing in response to rapid changes in airway pressure in people with tetraplegia and OSA. Altered genioglossus function may contribute to the increased risk of OSA in people with tetraplegia. The precise mechanisms mediating these differences are unknown.

Keywords: sleep-disordered breathing; spinal cord injury; upper airway physiology.

Figure 1. Schematic of the experimental set‐up

Participants lay supine connected to a solenoid valve system on the other side of a soundproof wall while breathing through a nasal mask attached to a pneumotachograph and an expiratory valve. A differential pressure sensor (mask pressure) and two nasal pressure catheters (choanal pressure and epiglottic pressure sensors) were used to measure stimulus characteristics and airway collapsibility. One arm of the solenoid was connected to room air (atmospheric pressure) and the other was connected to a negative pressure reservoir evacuated to approximately −100 cmH₂O. Brief pulses (250 ms) were delivered every 2–10 breaths during early inspiration when airflow reached 2 l min⁻¹. An adjustable valve limited the pressure delivered to the mask to approximately −15 cmH₂O.

Figure 2. Example of a genioglossus reflex response in an able‐bodied control

The lower trace indicates an ensemble average of the rectified genioglossus EMG of 75 negative pressure pules in a 31‐year‐old male with obstructive sleep apnoea (apnoea/hypopnoea index = 15 events per hour of sleep). The top trace shows the ensemble averaged choanal pressure tracing.

Figure 3. Examples of phasic genioglossus EMG activity during quiet breathing in two people with tetraplegia (top two tracings) and two able‐bodied controls (lower two tracings)

Each example shows 15 s of data (four breaths) including the mask pressure tracing, airflow signal and raw genioglossus electromyogram (EMG_raw). An example of a transient negative pressure pulse delivered during early inspiration in each example is also displayed. The horizontal line indicates zero flow.

Figure 4. Scatter plots of excitation peak latency (A) and amplitude (B) in people with spinal cord injury and able‐bodied participants

Horizontal lines indicate mean ± SD.

Figure 5. Example of a suppression‐only genioglossus reflex response in an individual with tetraplegia and obstructive sleep apnoea

The lower trace shows the ensemble average of the rectified genioglossus EMG in response to 44 negative pressure pulses in a 69‐year‐old male with an apnoea/hypopnoea index of 47 events per hour of sleep. The top trace shows the ensemble average of the choanal pressure tracing.