Respiratory muscle activity during REM sleep in patients with diaphragm paralysis

Methods.

Five patients with isolated BDP participated (table 1). Diaphragm paralysis was diagnosed as an unpotentiated bilateral twitch transdiaphragmatic pressure < 2 cm H₂O.4 Four normal subjects matched for sex and body mass index (BMI) were also studied. All participants were nonsmokers and regular nocturnal sleepers. The patients with BDP had not received nocturnal ventilatory support. Our ethics committee approved the study. Participants gave written informed consent.

Sleep state was measured using two EEG (C3-A2, C4-A1), left and right electro-oculograms (EOG), and submental EMG. We also measured chest wall and abdominal movement, died airflow (thermistor, Jaeger) and oxygen saturation (Nellcor-N-200E). EMG of five respiratory muscles (sternocleidomastoid, pectoralis minor, parasternal intercostals, diaphragm, and rectus abdominis) were recorded using surface electrodes (Neuroline, Medicotest).

Results.

Patient details are shown in table 1. The four normal subjects had mean (SD) age 36.3 (2.6) years and BMI 23.5 (1.3) Kg/m². Epworth sleepiness scores were normal in both groups (mean [SD] patients: 6.0 [2.5]; normal subjects: 4.5 [2.6]).

All participants had both NREM and REM sleep (table 2); the quantity of REM sleep was within the normal range in patients with BDP and normal subjects (18.6% of TST for patients and 28.9% for normal subjects; see table 2). However, the amount of phasic REM in the patients with BDP was almost double that of normal subjects (see table 2). There were no differences in the distribution of NREM sleep stages (stage II–IV) between the two groups. In addition, the AI across the whole night (NREM plus REM) was similar between the patients with diaphragm paralysis and normal subjects, but greater when REM only was analyzed.

All patients with BDP maintained activity of the accessory respiratory muscles during REM. Specifically, activity in the sternocleidomastoids EMG (EMG_SCM) was present in all patients with BDP, during both tonic and phasic REM sleep (figure). Parasternal intercostal EMG (EMG_PS) was present in three of the five patients, in both phasic and tonic REM. Of note, the patient with the highest REM apnea/hypopnea index (AHI) (no. 5) also failed to generate EMG_PS activity. Grade 1 pectoralis minor activity was observed during REM in one patient with BDP only. Grade 1 rectus abdominis activity was obtained from two patients with BDP during REM sleep. Inspiratory diaphragm EMG (EMG_DIA) was never observed in the patients with BDP, whereas clear EMG_DIA was always present in the normal subjects in all stages of sleep. Expiratory (abdominal) muscle activity was observed in the EMG_DIA in four of the five patients with BDP during NREM sleep and in two during REM. No expiratory activity from any site was observed in the normal subjects.

• Download figure
• Open in new tab
• Download powerpoint

Figure. Original data from a patient with isolated bilateral diaphragm paralysis during wakefulness (top panel), non-REM (NREM) sleep (middle panel), and REM sleep (bottom panel). Inspiratory muscle activity was observed in wake, NREM (stage II), and REM sleep. Note the preservation of activity in both tonic and phasic REM, and the expiratory activity in the diaphragm during NREM sleep. Data from a normal subject are shown in the right panel.

The AHI was greater in the patients with BDP compared to the normal subjects; one of these (no. 4) had mild obstructive sleep apnea on AHI criteria. The apneas/hypopneas in patients with BDP were associated with a lower basal SaO₂, minimum SaO₂, and lower mean SaO₂ across all desaturation events (see table 2).

Discussion.

Patients with isolated BDP have a normal quantity of REM sleep albeit with slightly increased arousals. These patients maintain ventilation during both tonic and phasic REM sleep by recruitment of the accessory muscles of respiration, especially the sternocleidomastoids, and in a proportion the parasternal intercostals and the abdominal muscles. In contrast, normal subjects did not recruit these muscle groups during REM sleep. These data support our hypothesis that CNS reorganization, in this case at the level of the brainstem, compensates for the impairment of the respiratory muscle pump that results from diaphragm paralysis.

Our data support the hypothesis that change in brainstem neuronal activity occurs in patients with diaphragm paralysis and serves to minimize the hypoventilation that would otherwise occur. Abdominal muscle recruitment is also a compensatory strategy during wakefulness.3 During both sleep and wakefulness abdominal muscle activity in expiration serves to displace the abdominal contents in a cephalad direction; their subsequent caudal movement during abdominal muscle relaxation augments inspiration.

Few studies have investigated breathing during REM sleep in humans with BDP. Four patients with isolated diaphragm dysfunction have been previously reported.6 All four subjects entered REM, but the diagnosis was not confirmed by phrenic nerve stimulation and, unexpectedly, two of these subjects exhibited diaphragm activity during NREM sleep and wakefulness. Data also exist from patients with neuromuscular disease although only two studies (in patients with amyotrophic lateral sclerosis [ALS]) evaluated diaphragm function using phrenic nerve stimulation. In one, 35 patients with ALS were studied.7 REM sleep was present even in patients with established respiratory failure, but the quantity was reduced compared with patients with ALS without CO₂ retention (8 versus 13%). In the other, 13 patients with diaphragm weakness due to ALS were evaluated8; as a group, the percentage of TST spent in REM was reduced (7%), but the patients with the longest periods of REM were observed to have inspiratory activity of the sternocleidomastoid muscle during REM, consistent with our data. Indeed, the fact that only a minority of the latter series (as opposed to all in ours) demonstrated this compensation suggests that it may be the upper motor neuron element of ALS that prevents compensatory adaptation.

Although symptoms are clearly dependent on the nature and severity of the insult to the respiratory system, they may be modified by neuronal compensation. This process may be relevant to non-neurologic disease such as chronic obstructive pulmonary disease (COPD) (for example, patients with COPD maintain accessory inspiratory muscle during REM).9,10⇓ We speculate that manipulation of compensatory neuronal reorganization may offer approaches to symptom relief in patients whose primary disease cannot be effectively treated.