Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2000 Jul 1;526(Pt 1):195–202. doi: 10.1111/j.1469-7793.2000.00195.x

The respiratory response to inspiratory resistive loading during rapid eye movement sleep in humans

Mary J Morrell *, Helen A K Browne *, Lewis Adams *
PMCID: PMC2270002  PMID: 10878111

Abstract

  1. We investigated the respiratory response to an added inspiratory resistive load (IRL) during rapid eye movement (REM) sleep in humans and compared this with those in non-REM (NREM) sleep and wakefulness.

  2. Results were obtained from 7 out of 15 healthy subjects (n = 7; 32 ± 9 years, mean ± s.d.). Linearised IRLs (4 and 12 cmH2O l−1 s−1) were applied for five breaths during NREM sleep (4-10 trials per subject; total 101), REM sleep (2-5 trials; total 46) and wakefulness (2-3 trials; total 40). Respiratory variables were compared, between unloaded breathing (UL: mean of 5 breaths preceding IRL) and the 1st (B1) and 5th (B5) loaded breaths in each state.

  3. During wakefulness, 12 cmH2O l−1 s−1 IRL produced an immediate respiratory compensation with prolongation of inspiratory time (TI; UL: 2.0 ± 0.6; B1: 2.6 ± 0.7 s) and an increase in tidal volume (VT; UL: 0.49 ± 0.12; B1: 0.52 ± 0.12 l). During REM sleep, TI was prolonged (UL: 2.0 ± 0.3; B1: 2.2 ± 0.5 s), although VT fell (UL: 0.27 ± 0.15; B1: 0.22 ± 0.10 l). For both wakefulness and REM sleep the TI response was significantly greater than seen in NREM sleep (UL: 1.9 ± 0.3; B1: 1.9 ± 0.3 s.). For VT, only the wakefulness response was significantly different from NREM sleep (UL: 0.31 ± 0.14; B1: 0.21 ± 0.10 l). The B5 responses were not significantly different between states for any of the variables.

  4. REM sleep is associated with partial respiratory load compensation suggesting that exacerbation of sleep disordered breathing in REM (compared to NREM) sleep is unlikely to be secondary to an inability to overcome increases in upper airway resistance.

During sleep, changes in respiratory mechanics place additional demands on respiratory control mechanisms if adequate gas exchange is to be maintained. Loss of consciousness is accompanied by a reduction in efferent neural activity to both upper airway and respiratory pump muscles leading to potential decreases in inspiratory muscle force and increases in resistance to airflow (Orem et al. 1977; Remmers et al. 1978; Sauerland & Harper, 1981; Hudgel et al. 1984). When awake, an experimentally induced increase in the resistance to airflow typically leads to a prolongation of inspiratory duration of the first breath, resulting in a maintenance of tidal volume and ventilation, although some variation between individual responses can occur (Iber et al. 1982; Wilson et al. 1984; Hudgel et al. 1987; Wiegand et al. 1988; Badr et al. 1990). This immediate respiratory compensatory response appears to be almost lost during sleep such that imposition of a load is associated with an acute hypoventilation (Iber et al. 1982; Wilson et al. 1984; Hudgel et al. 1987; Wiegand et al. 1988; Gugger et al. 1989; Badr et al. 1990). However, the majority of studies supporting this observation have been carried out during non-rapid eye movement (NREM) sleep. The aim of the present study was to test the immediate respiratory response to an added inspiratory resistive load during REM sleep.

The inability to compensate for a respiratory load is likely to be a particular problem during REM sleep because the respiratory mechanics are more compromised secondary to the occurrence of atonicity in some respiratory-related skeletal muscles (Pack, 1995). In healthy adults, observations of breathing during REM sleep have shown that it is often irregular (Bulow, 1963). In patients with sleep apnoea, REM sleep is associated with prolonged apnoeas (Findley et al. 1980) and in some cases apnoeas occur only during this sleep stage (Kass et al. 1996). In patients with chronic lung disease, sleep-disordered breathing is seen predominantly in REM sleep and is often associated with severe hypoxaemia (Douglas et al. 1979). One explanation for these observations is that the ability to mount a respiratory compensation to fluctuations in resistance is impaired during REM sleep.

In the present study we wished to measure ventilation in response to an added inspiratory resistive load during REM sleep. In designing this study, we needed to consider the difficulties of making experimental interventions during REM sleep. This phase of sleep is characterised by skeletal muscle atonia and frequent episodes of rapid eye movement (phasic REM) interspersed with periods of no eye movements (tonic REM). Hypoventilation is known to be more pronounced during the relatively short (typically <30 s) periods of phasic REM sleep (Gould et al. 1988). In addition, arousal occurs more often when an intervention is applied during REM compared to NREM sleep (Gugger et al. 1989). For these reasons, we elected to focus on the 1st breath respiratory response to an added load during REM sleep; we also chose to analyse the 5th breath post intervention in order to document any longer term compensatory responses. We aimed to compare these responses during REM sleep to those occurring during wakefulness and NREM sleep, in healthy young people. Our hypothesis was that the immediate respiratory compensation to an added resistive load would be less in REM sleep compared to NREM sleep and wakefulness.

METHODS

Subjects

Fifteen subjects (age 23-46 years; 9 male) were studied. Each subject was a regular nocturnal sleeper, not obese, with no history of respiratory, cardiovascular, or neurological disease (see Table 1). The hospital research ethics committee approved the study and subjects gave written informed consent prior to participation.

Table 1.

Subject characteristics

Subject Age(years) Sex Height(m) Weight(kg) BMI(kg m−2) Smoker (pack/years)
1 36 M 1.83 80.7 24.1 0.0
2 23 F 1.64 64.0 23.8 0.0
3 46 M 1.78 74.0 23.4 0.0
4 34 F 1.68 60.0 21.3 15.0
5 23 M 1.80 67.6 20.9 1.0
6 25 M 1.67 60.3 21.6 0.1
7 36 M 1.77 76.0 24.3 0.0
Mean ± s.d. 32 ± 9 1.74 ± 0.07 68.9 ± 7.5 22.8 ± 1.4 2.3 ± 5.2
Open in a new tab

Measurements

The sleep state was measured using two electroencephalograms (EEG; C3-A2, C4-A1), two electroenoculograms (EOG) and an electromyogram (EMG) of the submental muscle. The electrodes were positioned according to the International 10-20 system of electrode placement (Harner & Sannit, 1974).

Airflow was measured using a Fleisch No. 2 pneumotachograph with a differential pressure transducer (MP45, ±2 cmH2O; Validyne, CA, USA) attached to a nasal mask. Expired air was sampled through a flexible probe placed just within the nostril. From this, end-tidal PCO2 (PET,CO2) was measured using a quadrupole mass spectrometer (QP 9000; PK Morgan Ltd, Kent, UK); this was taken as an estimate of PCO2 in arterial blood.

Oesophageal pressure (Poes), reflecting intrathoracic pressure, was measured via a catheter tipped with a pressure transducer (CTC/6F; Gaeltec, Isle of Skye, Scotland); this was taken as an index of respiratory effort. Prior to passing the catheter, both nasal passages and the back of the throat were sprayed with a small amount of topical anaesthetic (4 % lignocaine hydrochloride solution, Astra Pharmaceuticals Ltd, Hertfordshire, UK). The catheter was then passed through the nasal passage until the tip of the catheter was 30 cm from the nostril. Poes was measured with respect to atmospheric pressure.

All signals were recorded on a digital computer via an analogue to digital interface (1401 Plus, Cambridge Electronic Design Ltd, Cambridge, UK). Digital signals were then analysed using commercially written software (Spike 2, Cambridge Electronic Design) to provide breath by breath measurements of inspiratory time (TI), expiratory time (TE), total breath time (Ttot), tidal volume (VT), inspired minute ventilation (V̇I), PET,CO2 and maximal change in Poes during inspiration (ΔPoes).

Protocol

All subjects were studied overnight after restricting their sleep (max. 4 h) on the previous night. They were asked to refrain from drinking alcohol or coffee for 4 h before the study.

Following the application of the EEG, EOG and EMG electrodes, and the passing of the oesophageal catheter, subjects lay supine on the bed. The PET,CO2 probe was secured and a nasal mask and headgear fitted. To check that the mask was airtight the subject attempted to breathe with its opening occluded. When any leaks had been eliminated, the pneumotachometer was attached to the nasal mask. Distal to this, a one-way valve (T-shaped non-rebreathe valve – model 2600; Hans Rudolph, Kansas City, MO, USA) was connected to allow separation of inspired and expired air. The deadspace of the pneumotachometer/respiratory valve was approximately 60 ml. The subject was then allowed to fall asleep. All subjects slept in the supine position. Visual observation of the EEG, EOG and EMG was used to determine when the subject had reached stable NREM or REM sleep. Once stable sleep had been achieved a linearised resistive load was applied to the inspiratory limb of the respiratory valve; two levels of load (4 and 12 cmH2O l−1 s−1) were used. Each inspiratory resistive load (IRL) was maintained for five breaths unless an arousal occurred in which case it was immediately removed. For each subject as many IRLs as possible were applied during NREM and REM sleep. Each application was separated by at least a 3 min period. At the end of each study three IRLs at each level were applied during wakefulness. In three subjects IRLs could not be applied at the end of the study due to discomfort; these subjects returned to the laboratory on a separate occasion when the IRLs were applied during wakefulness.

Analysis

The sleep stage was determined from computer recordings of the EEG, EOG and EMG according to standard criteria (Rechtschaffen & Kales, 1968). Trials were not analysed if (1) an arousal occurred during either the five breaths preceding or five breaths during application of the IRL (criteria of the American Sleep Disorders Association (1992)), (2) the sleep stage was ambiguous, or (3) a respiratory artefact such as swallowing occurred. For five breaths preceding and five breaths during each IRL, TI, TE, Ttot, VT, V̇I, PET,CO2 and ΔPoes were analysed. For each of the respiratory variables, a mean value for the five unloaded (UL) breaths was calculated for each subject for each IRL trial. All trials in each of the wakefulness, NREM and REM sleep conditions were then averaged for each subject. Similar within subject means were calculated separately for the 1st (B1) and 5th (B5) breath during the IRL application. These group mean data for the UL, B1 and B5 were then compared across within and between conditions (see below).

Trials carried out during REM sleep were further classified as either phasic or tonic based on the presence or absence of eye movements during the UL breaths and/or the 1st breath following the application of IRL. Assessment of the respiratory response to each 12 cmH2O IRL during phasic vs. tonic REM sleep was carried out by comparing the TI of the 1st breath immediately preceding the application of IRL to the 1st breath following load application.

Statistical analysis

For each respiratory variable, group mean data were compared using analysis of variance (ANOVA) with repeated measures. Breathing during the different states (wakefulness, NREM and REM sleep) without an added IRL was compared using ‘state’ as a single within factor. Any variables showing statistically significant state changes were further examined by comparing only two conditions (wakefulness vs. NREM sleep, wakefulness vs. REM sleep and NREM vs. REM sleep) using paired t statistics. The effect of IRL on breathing was compared using state and breathing response (UL, B1 IRL, B5 IRL) as two within factors. Since we were primarily interested in the effect of changes in state on the response to IRL we focused on the ANOVA interaction statistic. Any variables showing statistically significant breath responses were further examined by comparing UL vs. B1 IRL or UL vs. B5 IRL for pairs of states. Comparisons of TI during phasic and tonic REM sleep were made using Student's paired t test. Statistical significance was defined as P = 0.05.

RESULTS

Fifteen healthy subjects were studied (mean ±s.d.: age 30.1 ± 8.0 years; BMI 23.5 ± 2.7 kg m−2; 9 males). REM sleep data were collected in seven representative subjects (Table 1). Those in whom REM sleep data were not recorded were either unable to tolerate the recording apparatus and did not sleep at all, or failed to achieve REM sleep.

In the seven subjects studied, each of the 4 and 12 cmH2O l−1 s−1 IRLs was applied during wakefulness (2-3 trials: 4 cmH2O l−1 s−1 total n = 20; 12 cmH2O l−1 s−1 total n = 20), NREM sleep (4-10 trials: 4 cmH2O l−1 s−1 total n = 48; 12 cmH2O l−1 s−1 total n = 53) and REM sleep (2-5 trials: 4 cmH2O l−1 s−1 total n = 24; 12 cmH2O l−1 s−1 total n = 22). The 12 cmH2O l−1 s−1 IRLs applied during REM sleep were subdivided into those occurring during phasic (n = 13) and tonic (n = 9) REM sleep.

The addition of 4 cmH2O l−1 s−1 IRL resulted in no change in the group mean level of any of the respiratory variables during wakefulness, NREM or REM sleep. These data are thus not presented here. The effects of a change in state and of adding 12 cmH2O l−1 s−1 IRL on breathing are shown in Fig. 1.

Figure 1. State-related changes in respiratory load compensation.

Figure 1

Open in a new tab

Group mean (±s.e.m.) results for unloaded breathing (UL, mean of 5 breaths) and for the first (B1 IRL) and the fifth (B5 IRL) breaths following application of a 12 cmH2O inspiratory resistive load during wakefulness (•), NREM sleep (□) and REM sleep (▴).

Changes in breathing from wakefulness to sleep without an added IRL

During UL breathing, VT was significantly reduced during NREM (P = 0.04) and REM (P = 0.001) sleep compared to wakefulness. TE was significantly shorter during REM sleep compared to wakefulness (P = 0.04) and NREM sleep. The changes resulted in a significantly lower V̇I during REM sleep (P = 0.01) compared to wakefulness. TI, Ttot and ΔPoes were not different between states.

Measurements of PET,CO2 were made in five subjects during wakefulness and NREM sleep and in three subjects during REM sleep. During unloaded breathing, the PET,CO2 for this group (n = 3) was 43.8 mmHg (range, 41.8 to 46.7) during wakefulness, 45.7 mmHg (range, 43.6 to 47.5) during NREM sleep and 46.0 mmHg (range, 44.0 to 50.6) during REM sleep.

Changes in breathing following 12 cmH2O l−1 s−1 IRL

Figure 2 shows a typical respiratory response to a 12 cmH2O l−1 s−1 IRL applied during wakefulness, NREM and REM sleep in one subject. Note the greater reduction in airflow following application of the IRL during NREM sleep compared to wakefulness and REM sleep.

Figure 2. An example of application of IRL in one subject.

Figure 2

Open in a new tab

Recording of chin electromyogram (EMGchin), central electroencephalogram (EEG, C4-A1), left and right eye electrooculogram (EOG), airflow (V), oesophageal pressure (Poes), mask pressure (Pmask) and end tidal PCO2 (PET,CO2), in one subject during wakefulness, NREM and REM sleep before and during application of an inspiratory resistive load (IRL).

First breath IRL response

The first breath response to the addition of an IRL in the different states is depicted in Fig. 1. The prolongation of TI was significantly less in NREM (P = 0.01) and REM (P = 0.01) sleep compared to wakefulness; and the prolongation during NREM was significantly less compared to REM sleep (P = 0.01). Associated with these effects on TI, the changes in Ttot were significantly different between NREM sleep and wakefulness (P = 0.01). The application of an IRL produced a reduction in VT during both sleep states; the change was significantly different for both NREM (P = 0.02) and REM (P = 0.03) sleep compared to the increase in VT observed during wakefulness. The first breath responses for TE, V̇I or ΔPoes were not significantly different between states.

Fifth breath IRL response

The fifth breath response to the addition of an IRL in the different states is depicted in Fig. 1. Compared to unloaded breathing, this response was not significantly different between states for any of the variables shown.

Phasic vs. tonic REM sleep IRL response

Figure 3 shows a typical respiratory response to a 12 cmH2O l−1 s−1 IRL applied during tonic and phasic REM sleep in one subject. Note the suggestion of a reduction in airflow during the burst of eye movements pre IRL. For each of the phasic (n = 13) and tonic (n = 9) REM sleep trials the effect of 12 cmH2O l−1 s−1 IRL on the respiratory response in TI (breath immediately preceding IRL (UL) vs. B1 IRL) is shown in Fig. 4. During phasic REM sleep TI was significantly prolonged (UL: 1.94 ± 0.35 s; B1: 2.14 ± 0.41 s; P = 0.008); it was also prolonged during tonic REM sleep, although this did not reach statistical significance (UL: 1.91 ± 0.30 s; B1: 2.09 ± 0.45 s; P = 0.08).

Figure 3. An example of application of IRL during tonic vs. phasic REM sleep in one subject.

Figure 3

Open in a new tab

Recording of chin electromyogram (EMGchin), central electroencephalogram (EEG, C4-A1), left and right eye electrooculogram (EOG), diagraphamatic electromyogram (EMGdia) and airflow (V), in one subject during tonic (left panel) and phasic (right panel) REM sleep. The vertical line shows where the inspiratory resistive load was applied.

Figure 4. Respiratory load compensation during tonic and phasic REM sleep.

Figure 4

Open in a new tab

Individual (○) and group mean (•) responses in inspiratory time (TI) following application of a 12 cmH2O inspiratory resistive load (UL: 1st breath immediately preceding 12 cmH2O IRL vs. B1 IRL: 1st breath immediately following IRL) during tonic and phasic REM sleep.

DISCUSSION

The aim of the present study was to compare the respiratory response to an inspiratory resistive load (IRL) applied during wakefulness, NREM and REM sleep. Specifically, we wanted to determine whether respiratory compensation was reduced during REM sleep. Consistent with previous reports we found that during wakefulness, but not during NREM sleep, significant compensation occurred within the first breath following application of an IRL. The key finding in our study was that some immediate respiratory load compensation occurred during REM sleep, although it was insufficient to prevent a fall in ventilation.

Immediate respiratory compensation to resistive loading during NREM vs. REM sleep

Immediate respiratory load compensation, assessed by chest wall inspiratory EMG activity, mouth occlusion pressure or the prolongation of TI has been shown to be reduced (Wiegand et al. 1988, Gugger et al. 1989; Badr et al. 1990) or absent (Iber et al. 1982; Wilson et al. 1984; Hudgel et al. 1987) during NREM sleep compared to wakefulness. In the present study, using prolongation of TI as an index of respiratory compensation, we confirmed an absence of compensation during NREM sleep, but found evidence of a small compensation during REM sleep with a 12 cmH2O resistive load. Our failure to see any respiratory compensation, during wakefulness or sleep, using a 4 cmH2O resistive load is consistent with all (Wiegand et al. 1988; Daubenspeck, 1995; Brack et al. 1998) but one (Gugger et al. 1989) of the studies in which a comparably small load has been tested.

There are few human studies investigating the effect of REM sleep on respiratory load compensation. Wiegand et al. (1988) reported that the ventilatory response in REM was similar to that in NREM sleep, although they found considerable within and between subject variability. Gugger et al. (1989) reported frequent arousals following the application of a load and they were unable to quantify respiratory responses. In the present study, using coefficient of variation in ΔTI as a measure of variability (calculated on the same number of trials per subject; 17 trials; 2-3 per subject) we found no evidence that the response to an IRL was more variable during REM sleep compared to either NREM sleep or wakefulness (UL vs. B1 IRL: REM sleep 117 %; NREM sleep 284 %; and wakefulness 152 %). Therefore we are confident that the observed respiratory compensation that occurred was not an artefact of a relatively small number of trials undertaken when breathing was variable.

The mechanisms responsible for immediate respiratory load compensation are uncertain. The lack of a prolongation in TI (Iber et al. 1982), an increase in chest wall EMG activity (Hudgel et al. 1987) or any augmentation in mouth occlusion pressure (Wilson et al. 1984) on the first breath following the application of an IRL during NREM sleep suggests the compensation requires the excitatory influences of wakefulness. REM sleep is similar to wakefulness in that there is a desynchronisation of the EEG and thus a potential excitatory influence on respiratory motor output.

Sleep is associated with a modulation of respiratory reflexes. For example, diaphragmatic EMG activity in dogs is reduced following occlusion of the upper airway both in REM and NREM sleep (Smith et al. 1997). In addition, the reflex response to negative pressure in the upper airway is diminished during NREM sleep in humans (Horner et al. 1994) and absent during REM sleep in dogs (Harms et al. 1996). Taken together, these observations do not argue for a relative augmentation of respiratory reflexes during REM compared to NREM sleep and our findings are probably not explained by such a mechanism.

The loss of wakefulness is accompanied by a loss of conscious perception of a load; such cortical inputs are likely to be important in respiratory load compensation. It has previously been suggested that during REM sleep the pattern of breathing is determined by the sleep-related neural activity per se rather than by modulation of chemo- or mechanoreflex control (Orem, 1980a). Thus our demonstration of a REM-related prolongation of TI may be explained if the central neural processing of the sensory information related to application of an inspiratory respiratory load was different in this state compared to NREM sleep. Furthermore, the transient nature of the respiratory response observed may be related to the inherent variability or neuronal activity during REM sleep (see below).

Respiratory compensation in response to sustained resistive loading

Sustained application (approximately 2-3 min) of an IRL has been shown to produce an augmentation in respiratory effort, measured using mouth occlusion pressure and surface EMG activity (Wilson et al. 1984; Wiegand et al. 1988; Gugger et al. 1989; Badr et al. 1990). This respiratory compensation to a sustained respiratory load is likely to be mediated by chemostimulation. In humans, Lopata et al. (1980) showed that breathing CO2 augmented ventilatory load compensation, and Wilson et al. (1984) showed that breathing a hyperoxic gas mixture delayed it. However the strongest evidence for a chemically mediated response to a sustained respiratory load comes from the studies of Bruce et al. (1974). These workers found that the gradual recovery of tidal volume and ventilation observed following application of a 25 cmH2O load in anaesthetised cats was abolished if circulation to the head was isolated and pH/blood gas levels of the blood perfusing the carotid bodies and medullary chemoreceptors were kept constant. This intervention did not abolish the initial immediate fall in ventilation following application of the load.

Taking account of previous studies, we elected to examine the time course of REM-related respiratory compensation by also analysing the 5th breath response, which would be minimally influenced by chemoreceptor-mediated respiratory augmentation (Iber et al. 1982). Our analysis revealed that the 1st breath prolongation of TI was not sustained by the 5th breath but that a fall in TE resulted in a maintenance of VT and V̇I relative to the 1st breath. Our inability to measure a response to a sustained IRL may have been due to a variability in the pattern of breathing which has previously been shown during wakefulness (Shea et al. 1987; Brack et al. 1998). However, as stated above, calculation of the coefficient of variation indicated that this was probably not a problem in the present study. One explanation of our findings is that the TI response for the 1st breath during REM sleep was related to a ‘detection’ of a change in resistance rather than the magnitude of the resistance per se. Indeed for TI a similar pattern of response was seen during wakefulness.

Respiratory compensation during phasic and tonic REM

In humans, phasic REM sleep is associated with marked between subject variability in breathing pattern (Neilly et al. 1991). Recordings of neural activity in animals have shown that REM sleep exerts both an excitatory and inhibitory influence on the respiratory system, e.g. thoraco-abdominal atonia vs. increase in respiratory rate (Orem 1996). These different influences combine to produce specific REM-related changes in breathing. Some of the changes are sustained throughout both tonic and phasic REM sleep, e.g. thoraco-abdominal atonia, whereas others appear to be more transient in nature.

The close association between respiratory-related neuronal activity recorded in the medulla and pontogeniculoccipital waves indicates the specific effect of phasic REM sleep on breathing (Orem 1980b). Inhibition of ventilatory responses to respiratory stimuli (such as lung inflation) is more pronounced during phasic compared with tonic REM sleep (Sullivan et al. 1979). No evidence of ventilatory compensation was observed in sleeping dogs during phasic REM sleep (Smith et al. 1997). In an English bulldog model of sleep apnoea the activity of the diaphragm and upper airway dilator muscles was found to be reduced during phasic REM sleep (Hendricks et al. 1991). Finally, in anaesthetised cats the magnitude of the REM-related respiratory inhibition was related to the frequency of phasic eye movements (Kline et al. 1986). Taken together these studies would suggest that phasic REM sleep, compared to tonic REM sleep, may be associated with a reduced or more variable respiratory load compensation. However, in the present study, using a relatively small data set, we found a significant prolongation during phasic but not tonic REM sleep. This observation may reflect the difficulty in arbitrarily deciding when phasic and tonic REM starts and finishes. Put another way, if neural inhibition is present during phasic REM sleep, how long does the inhibition last? We chose to define our interventions as occurring during phasic REM if eye movements were present during any of the five UL breaths and/or the 1st breath following the application of IRL. If a phasic REM-related inhibition is short lived and there were no eye movements immediately preceding the IRL then some of our trials classified as occurring during phasic REM sleep might have been better identified as tonic REM sleep.

Limitations of the study

The main limitation of the present study was our ability to obtain data during REM sleep. Our subjects were heavily instrumented (i.e. nose mask plus oesophageal pressure catheter) because our aim was to obtain comprehensive data to determine the mechanisms responsible for any changes in breathing following application of IRL. In some subjects the instrumentation led to inability to sleep long enough to allow REM sleep to develop. In addition, the short duration and intermittent occurrence of REM sleep also resulted in a smaller number of applications of IRLs compared to NREM sleep. This is a problem which has been noted by others (Wiegand et al. 1988). Therefore to ensure that our mean responses were representative, we took particular care to include only subjects with more than one IRL applied during REM sleep in our analysis. We also took care to ensure that the REM periods analysed were free from any overt EEG arousals (>3 s, American Sleep Disorders Association). However, it is possible that the addition of IRLs led to more subtle arousals, undetectable by the EEG/EMG criteria used, and that these were associated with the observed respiratory response.

The use of a nasal mask and breathing circuit allowed us to make accurate measurements of breathing during sleep. However, the instrumentation did produce a small respiratory load and dead space. We do not believe that this would have confounded our findings since the use of the one-way breathing valve minimises the dead space and the circuit resistance was a constant load across all sleep states.

Finally it should be acknowledged that our experimentally mediated responses to added IRL may differ from the endogenous sleep-related increments in respiratory load produced by increases in upper airway resistance, despite the fact that we chose levels of resistance similar to those which occur in normal healthy subjects during sleep.

Implications of the study

Our finding of an immediate prolongation of TI in response to an added IRL during REM sleep suggests that the irregular breathing in normal subjects and sleep-disordered breathing in patients, which occur in this state, do not result from inadequate respiratory compensation in response to changes in upper airway resistance. Evidence from other studies shows REM-related impairment of respiratory reflexes that are at least as pronounced as for NREM sleep. The present data thus provide support for an influence of sleep state-related central neural processing in determining the respiratory response to changes in mechanical events.

Acknowledgments

We wish to thank the Breathlessness Research Charitable Trust for providing essential equipment. M.J.M. was supported by a Wellcome Trust Research Career Development Fellowship.

References

  1. American Sleep Disorders Association. EEG arousals: Scoring rules and examples. A preliminary report from the Sleep Disorders Atlas Task Force of the American Sleep Disorders Association. Sleep. 1992;15:173–184. [PubMed] [Google Scholar]
  2. Badr MS, Skatrud JB, Dempsey JA, Begle RL. Effect of mechanical loading on inspiratory and expiratory muscle activity during NREM sleep. Journal of Applied Physiology. 1990;68:1195–1201. doi: 10.1152/jappl.1990.68.3.1195. [DOI] [PubMed] [Google Scholar]
  3. Brack T, Jubran A, Tobin JM. Effect of resistive loading on variational activity of breathing. American Journal of Respiratory and Critical Care Medicine. 1998;157:1756–1763. doi: 10.1164/ajrccm.157.6.9704114. [DOI] [PubMed] [Google Scholar]
  4. Bruce EN, Smith JD, Grodins FS. Chemical and reflex drives to breathing during resistance loading in cats. Journal of Applied Physiology. 1974;37:176–182. doi: 10.1152/jappl.1974.37.2.176. [DOI] [PubMed] [Google Scholar]
  5. Bulow K. Respiration and wakefulness in man. Acta Physiologica Scandinavica. 1963;59:7–9. [PubMed] [Google Scholar]
  6. Daubenspeck JA. Mechanical aspects of loaded breathing. In: Roussos C, editor. The Thorax. 2. New York: Marcel Decker; 1995. pp. 953–985. [Google Scholar]
  7. Douglas NJ, Calverley PMA, Leggett RJE, Bush HM, Flenley DC, Brezinova V. Transient hypoxemia during sleep in chronic bronchitis and emphysema. Lancet. 1979;1:1–4. doi: 10.1016/s0140-6736(79)90451-3. [DOI] [PubMed] [Google Scholar]
  8. Findley LJ, Wilhoit SC, Suratt PM. Apnea duration and hypoxemia during REM sleep in patients with obstructive sleep apnea. Chest. 1980;78:432–436. doi: 10.1378/chest.87.4.432. [DOI] [PubMed] [Google Scholar]
  9. Gould GA, Gugger M, Molloy J, Tsara V, Shapiro CM, Douglas NJ. Breathing pattern and eye movement density during REM sleep in humans. American Review of Respiratory Diseases. 1988;138:874–877. doi: 10.1164/ajrccm/138.4.874. [DOI] [PubMed] [Google Scholar]
  10. Gugger M, Molly J, Gould GA, Whyte KF, Raab GM, Shapiro CM, Douglas NJ. Ventilatory and arousal responses to added inspiratory resistance during sleep. American Review of Respiratory Diseases. 1989;140:1301–1307. doi: 10.1164/ajrccm/140.5.1301. [DOI] [PubMed] [Google Scholar]
  11. Harms CA, Zeng YJ, Smith CA, Vidruk EH, Dempsey JA. Negative pressure-induced deformation of the upper airway causes central apnea in awake and sleeping dogs. Journal of Applied Physiology. 1996;80:1528–1539. doi: 10.1152/jappl.1996.80.5.1528. [DOI] [PubMed] [Google Scholar]
  12. Harner PF, Sannit T. A Review of the International Ten-Twenty System of Electrode Placement. Quincy, MA, USA: Grass Instrument Company; 1974. [Google Scholar]
  13. Hendricks JC, Kovalski RJ, Kline LR. Phasic respiratory muscle patterns in sleep disordered breathing during rapid eye movement sleep in the English bulldog. American Review of Respiratory Diseases. 1991;144:1112–1120. doi: 10.1164/ajrccm/144.5.1112. [DOI] [PubMed] [Google Scholar]
  14. Horner RL, Innes JA, Morrell MJ, Shea SA, Guz A. The effect of sleep on reflex genioglossus muscle activation by stimuli of negative pressure in humans. The Journal of Physiology. 1994;476:141–151. [PMC free article] [PubMed] [Google Scholar]
  15. Hudgel DW, Martin RJ, Johnson B, Hill P. Mechanics of the respiratory system and breathing pattern during sleep in normal humans. Journal of Applied Physiology. 1984;56:133–137. doi: 10.1152/jappl.1984.56.1.133. [DOI] [PubMed] [Google Scholar]
  16. Hudgel DW, Mulholland M, Hendricks C. Neuromuscular and mechanical responses to inspiratory resistive loading during sleep. Journal of Applied Physiology. 1987;63:603–608. doi: 10.1152/jappl.1987.63.2.603. [DOI] [PubMed] [Google Scholar]
  17. Iber C, Berssenbrugge A, Skatrud JB, Dempsey JA. Ventilatory adaptations of resistive loading during wakefulness and non-REM sleep. Journal of Applied Physiology. 1982;52:607–614. doi: 10.1152/jappl.1982.52.3.607. [DOI] [PubMed] [Google Scholar]
  18. Kass JE, Akers SM, Bartter TC, Pratter MR. Rapid-eye-movement-specific sleep-disordered breathing: a possible cause of excessive daytime sleepiness. American Journal of Respiratory and Critical Care Medicine. 1996;154:167–169. doi: 10.1164/ajrccm.154.1.8680674. [DOI] [PubMed] [Google Scholar]
  19. Kline LR, Hendricks JC, Davies RO, Pack AI. Control of activity of the diaphragm in rapid eye movement sleep. Journal of Applied Physiology. 1986;61:1293–1300. doi: 10.1152/jappl.1986.61.4.1293. [DOI] [PubMed] [Google Scholar]
  20. Lopata M, Pearle JL. Diaphragmatic EMG and occlusion pressure response to elastic loading during CO2 rebreathing in humans. Journal of Applied Physiology. 1980;49:669–675. doi: 10.1152/jappl.1980.49.4.669. [DOI] [PubMed] [Google Scholar]
  21. Neilly JB, Gaipa EA, Maislin G, Pack AI. Ventilation during early and late rapid-eye-movement sleep in normal humans. Journal of Applied Physiology. 1991;71:1201–1215. doi: 10.1152/jappl.1991.71.4.1201. [DOI] [PubMed] [Google Scholar]
  22. Orem J. Neuronal mechanisms of respiration in REM sleep. Sleep. 1980a;3:251–267. doi: 10.1093/sleep/3.3-4.251. [DOI] [PubMed] [Google Scholar]
  23. Orem J. Medullary respiratory neuron activity: relationship to tonic and phasic REM sleep. Journal of Applied Physiology. 1980b;48:54–65. doi: 10.1152/jappl.1980.48.1.54. [DOI] [PubMed] [Google Scholar]
  24. Orem J. Excitatory drive to respiratory system in REM sleep. Sleep. 1996;19:S154–S156. doi: 10.1093/sleep/19.suppl_10.154. [DOI] [PubMed] [Google Scholar]
  25. Orem J, Netick A, Dement WC. Increased upper airway resistance to breathing during sleep in the cat. Electroencephalography and Clinical Neurophysiology. 1977;43:14–22. doi: 10.1016/0013-4694(77)90190-0. [DOI] [PubMed] [Google Scholar]
  26. Pack A. Changes in respiratory motor activity during REM sleep. In: Dempsey JA, Pack AI, editors. Regulation of Breathing. New York: Marcel Dekker, Inc.; 1995. pp. 983–1010. [Google Scholar]
  27. Rechtschaffen A, Kales A. A Manual of Standardized Terminology, Techniques and Scoring Systems for Sleep Stages of Human Subjects. Bethesda, MD, USA: US Government Printing Office; 1968. [Google Scholar]
  28. Remmers JE, DeGroot WJ, Sauerland EK, Anch AM. Pathogenesis of upper airway occlusion during sleep. Journal of Applied Physiology. 1978;44:931–938. doi: 10.1152/jappl.1978.44.6.931. [DOI] [PubMed] [Google Scholar]
  29. Sauerland EK, Harper RM. The human tongue during sleep: electromyographic activity of the genioglossus muscle. Experimental Neurology. 1981;51:160–170. doi: 10.1016/0014-4886(76)90061-3. [DOI] [PubMed] [Google Scholar]
  30. Shea SA, Walter J, Murphy K, Guz A. Evidence for individuality of breathing patterns in resting healthy man. Respiration Physiology. 1987;68:331–344. doi: 10.1016/s0034-5687(87)80018-x. [DOI] [PubMed] [Google Scholar]
  31. Smith CA, Henderson KS, Xi L, Chow C-M, Eastwood PR, Dempsey JA. Neural-mechanical coupling of breathing in REM sleep. Journal of Applied Physiology. 1997;83:1923–1932. doi: 10.1152/jappl.1997.83.6.1923. [DOI] [PubMed] [Google Scholar]
  32. Sullivan CE, Murphy E, Kozar LF, Phillipson EA. Ventilatory responses to CO2 and lung inflation in tonic versus phasic REM sleep. Journal of Applied Physiology. 1979;47:1304–1310. doi: 10.1152/jappl.1979.47.6.1304. [DOI] [PubMed] [Google Scholar]
  33. Wiegand L, Zwillich C, White D. Sleep and the ventilatory response to resistive loading in normal man. Journal of Applied Physiology. 1988;64:1186–1195. doi: 10.1152/jappl.1988.64.3.1186. [DOI] [PubMed] [Google Scholar]
  34. Wilson PA, Skatrud JB, Dempsey JA. Effects of slow wave sleep on ventilatory compensation to inspiratory elastic loading. Respiration Physiology. 1984;55:103–120. doi: 10.1016/0034-5687(84)90120-8. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES