Chapter 22 Respiratory Physiology

Understanding the Control of Ventilation

Abstract

During sleep, the relative regularity of breathing in resting wakefulness is replaced by marked respiratory variability due to changes in the drive to both the respiratory pump muscles and the upper airway dilating muscles. Breathing during wakefulness is controlled by several factors including voluntary and behavioral elements; chemical factors, including low oxygen levels; high carbon dioxide (CO₂) levels and acidosis; and mechanical signals from the lung and chest wall (see Chapter 21 for more information). During sleep, there is loss of voluntary control and a decrease in the usual ventilatory responses to both low oxygen (hypoxemic) and high carbon dioxide (hypercapnic) levels. Both the hypoxemic and hypercapnic responses are most depressed in rapid eye movement (REM) sleep.

These blunted ventilatory responses during sleep are clinically important. They permit the marked hypoxemia that occurs during REM sleep in patients with lung or chest wall disease and may also be important in the pathogenesis of upper airway obstruction during sleep.

Sleep alters both breathing pattern and respiratory responses to many external stimuli. These changes allow the development of sleep-related hypoxemia in patients with respiratory disease and may contribute to the pathogenesis of apneas in patients with the sleep apnea syndrome. This chapter reviews the control of ventilation in adults during sleep and its relevance to these clinical problems (see Section 13 of this volume).

Introduction

Many respiratory problems during sleep are related to sleep-induced changes in the control of ventilation. Much is known about the control of breathing during wakefulness, but less about how breathing is controlled in sleeping normal subjects or in patients with sleep disorders. The major goal of the respiratory control system is homeostasis, to keep blood gases tensions in a tight range to allow the body to maintain normal metabolic functions.

Unlike the heart, the respiratory muscles do not have a built-in pacemaker. These muscles receive impulses from a region in the medulla that has been called the respiratory center, respiratory oscillator, respiratory signal generator, and respiratory pacemaker. For breathing to change as physiological conditions change, the respiratory center receives and responds to three general types of information: chemical information (from chemoreceptors responding to PaO₂, PaCO₂, and pH), mechanical information (from receptors in the lung and chest wall), and behavioral information (from higher cortical centers) to allow breathing to be altered during speech, swallowing, and other activities. The aspects of awake control that may have impact on control of breathing during sleep are reviewed later.

Chemical Information

The carotid body senses PaO₂ (arterial concentration of O₂) which sends impulses to the medulla via the ninth cranial nerve. Ventilation usually increases only when PaO₂ is less than 60 mm Hg. When PaO₂ drops acutely below 30 to 40 mm Hg, the medulla (see Chapter 21 for more information of the brainstem network associated to the control of breathing) may be depressed by the hypoxemia; thus, ventilation may decrease. Carbon dioxide is sensed in both the carotid body and a region of the medulla called the central chemoreceptor. As PaCO₂ (arterial concentration of CO₂) increases, there is a brisk linear increase in ventilation. Drugs that depress central nervous system function (e.g., some benzodiazepines, opioids) may profoundly depress chemical drives to breathe in normal altitude (see Chapter 24 for the paradoxical effects of benzodiazepines on respiration during sleep in high altitude).

Mechanical Information

In the presence of lung diseases or changes in the load of the respiratory system, receptors in the lung and chest wall send impulses to the medulla. Those in the lung are the best understood: The receptors respond to irritation, inflation (stretch), deflation, and congestion of blood vessels. The information from these sensors travels centrally via the vagus nerve. The major result of stimulating these sensors is to shorten inspiration and reduce tidal volume to cause a rapid, shallow breathing pattern.

Information from Higher Central Nervous System Centers

The respiratory system is used for many nonbreathing functions (e.g., singing, laughing, crying, and speaking) that are directed by higher brain centers. The efferent pathways involved in these activities may actually bypass the respiratory center in the medulla; thus, these nonrespiratory activities can override the metabolic homeostatic function of the respiratory control system. It appears that just as abnormal chemistry may increase the drive to breathe, there is a respiratory drive linked to wakefulness. As one sleeps, arouses, and dreams, there may be dramatic changes in the information from higher central nervous system (CNS) centers.

Physiology of Ventilatory Control During Sleep

This chapter largely focuses on the effects of sleep on the various mechanisms controlling breathing in adult humans.

Hypoxic Ventilatory Response during Sleep

Adult Human Beings

The ventilatory response to hypoxia falls during sleep in healthy adults (Fig. 22-1).¹^–⁴ The hypoxic ventilatory response was lower during non–rapid eye movement (NREM) sleep than during wakefulness in the studies in which the subjects were exclusively or mostly men¹^,² but the responses were similar in wakefulness and NREM sleep in the studies in which women predominated.³^–⁵ It is not clear why there is such a difference between the genders in the effect of sleep on the hypoxic ventilatory response and whether hormonal factors might be important. Comparison of the results in men and women (Fig. 22-2) shows that the major gender difference is in the levels of ventilatory response during wakefulness, which is much higher in men than in women,⁴ whereas the responses in NREM sleep were not different between the genders. During NREM sleep both the ventilatory responses to hypocapnic hypoxia and the posthypoxic ventilatory decline were similar between the genders.⁶ It was also found that the posthypoxic ventilatory decline in NREM sleep was not associated with a decline in respiratory frequency nor with any change in airway mechanics.⁷

Figure 22-1 Mean relationship between expired ventilation and decreasing O₂ saturation in the different sleep stages in 12 subjects, 6 male ¹ and 6 female.⁴

(From Douglas NJ. Control of ventilation during sleep. Clin Chest Med 1985;6:563.)

Figure 22-2 Comparison of hypoxic ventilatory responses in different sleep stages in four studies ^1,^3,⁵ in healthy subjects indicating the unchanged responses in NREM sleep in women, and the similarity of responses in REM sleep in all studies.

The hypoxic ventilatory response during rapid eye movement (REM) sleep is lower than in NREM sleep in both men and women.¹^–⁴ The hypoxic ventilatory response during REM sleep was remarkably similar in all three directly comparable studies:^1,^2,⁴ about 0.4 L/min/percentage SaO₂ (see Fig. 22-2).

Adult Animals

The isocapnic hypoxic ventilatory response has been found to be unchanged ⁸ or decreased⁹ during sleep in dogs with tracheostomies and decreased during sleep in goats.¹⁰

Hypercapnic Ventilatory Response during Sleep

Adult Human Beings

The hypercapnic ventilatory response is depressed during sleep in adult human beings (Fig. 22-3). Studies performed in either electroencephalogram-documented¹¹^–¹⁴ or presumed¹⁵^–¹⁷ NREM sleep showed that the slope of the ventilation-CO₂ response line falls during NREM sleep, compared with wakefulness. Two large studies¹²^,¹³ suggested that the decrease in response from wakefulness to NREM sleep is approximately 50%. Although Bülow¹² reported lower responses in stages 3 and 4 sleep than in stage 2 sleep, later researchers¹³^,¹⁸ found no such difference. In fact, Bülow did not apply statistics, and his data indicate considerable overlap (see Figs. 21, 26, and 27 in reference 12). Indeed, no separation is evident until high CO₂ levels are reached when the number of data points is small.

Figure 22-3 Mean relationship between expired ventilation and increasing end-tidal PCO₂ in 12 subjects, 6 male and 6 female, indicating the mean + SE resting ventilation-CO₂ set-point during wakefulness.

(From Douglas NJ. Control of ventilation during sleep. Clin Chest Med 1985;6:563. [Redrawn from Douglas NJ, White DP, Weil JV, et al. Hypercapnic ventilatory response in sleeping adults. Am Rev Respir Dis 1982;126:758-762.])

NREM sleep does not significantly change the position of the ventilation-CO₂ response line, as assessed by the rebreathing technique with extrapolation of the CO₂ response line to the intercept at zero ventilation.¹³^,¹⁸ However, as the slope of the response line decreases during NREM sleep, the CO₂ increment between the awake resting ventilatory set-point and the ventilatory response line is increased significantly during NREM sleep (see Fig. 22-3).¹²^,¹³

Berthon-Jones and Sullivan ¹⁸ found that healthy women did not change the hypercapnic ventilatory response from wakefulness to NREM sleep, a result consistent with the finding by Davis and colleagues¹⁹ of higher ventilatory responses during sleep in women than in men. In contrast, neither Douglas and colleagues¹³ nor Gothe and colleagues²⁰ identified any difference between the genders in any stage. Three groups have measured the hypercapnic ventilatory response during REM sleep in adults. Douglas and colleagues¹³ found that in 10 of 12 subjects, the hypercapnic response was lowest in REM sleep, compared with either wakefulness or NREM sleep. The mean hypercapnic ventilatory response during REM sleep was 28% of the level during wakefulness (see Fig. 22-3). These authors were unable to accurately define the position of the hypercapnic ventilatory response line because some responses during REM sleep had negative slopes. Berthon-Jones and Sullivan¹⁸ and White²¹ found a tendency for the lowest responses to occur during REM sleep.

Adult Animals

As in adult human beings, the hypercapnic ventilatory response is lower in NREM sleep than in wakefulness in dogs ²²^,²³ and cats,²⁴ with a further drop in response from NREM to REM sleep.²²^–²⁵ Sullivan and colleagues²³ found that the hypercapnic ventilatory response was lower in phasic than in tonic REM sleep in dogs.

Ventilatory Response to Chemical Stimuli: Conclusions

In summary, there appear to be genuine gender and species differences in the effect of sleep on the hypoxic ventilatory response. Although there are no major species differences in the effect of sleep on the hypercapnic ventilatory response, there may be gender differences. In human adults without disordered breathing, both ventilatory responses are reduced during sleep and are at their lowest during REM sleep.

Added Resistance and Airway Occlusion During Sleep

Adult Human Beings

Sleep modifies the ventilatory response to added inspiratory resistance.²⁶^–²⁹ All studies agree that NREM sleep blunts the respiratory timing response to added resistance,²⁶^–³⁰ with some reporting a net effect on ventilation as well.^27,^29,³⁰ In adults with asthma, the ventilatory response to induced bronchoconstriction is not modified by sleep.³¹ Issa and Sullivan³² reported that, in response to airway occlusion, there was a progressive increase in respiratory effort during NREM sleep and rapid, shallow breathing during REM sleep.

Animals

In both dogs ³³^,³⁴ and cats,²⁴ ventilatory compensation for added loads is maintained during NREM sleep at the level found in wakefulness. However, when the respiratory system is further stressed by combining resistive loading with hypercapnia, the ventilatory and airway occlusion responses to loading are found to be impaired during both NREM and REM sleep.²⁴

In summary, it is not yet clear precisely what effect increased airflow resistance has on ventilation during sleep, particularly during REM sleep, in adults.

Arousal Responses

Isocapnic Hypoxia

In normal subjects, isocapnic hypoxia is a poor stimulus to trigger arousal from sleep (Fig. 22-4), with many subjects remaining asleep with SaO₂ as low as 70%^1,^2,⁵ and no difference between NREM and REM sleep in arousal threshold. Conversely, the arousal sensitivity to hypoxia is decreased in REM sleep in patients having obstructive sleep apnea with asphyxial hypoxia,³⁵ in dogs with isocapnic hypoxia,⁸^,⁹ and in cats with hypocapnic hypoxia.³⁶ Carotid body denervation reduces arousal sensitivity in dogs⁷ but does not alter the arousal threshold in cats.³⁶

Figure 22-4 Arousal responses to hypoxia and hypercapnia ¹³ in different sleep stages, indicating whether the subjects remained asleep (red circles) or aroused (blue circles). The hypoxic responses are total data for 9 subjects; the hypercapnic plots are for each of 12 individuals, 6 females (1-6) and 6 males (7-12), indicating with the dotted line mean PETCO₂ in wakefulness in each subject.

(From Douglas NJ. Control of ventilation during sleep. Clin Chest Med 1985;6:563).

Hypercapnia

Hypercapnia also produces arousal from sleep at variable levels ^3,^11,¹³ but awakens most subjects before the end-tidal CO₂ has risen by 15 mm Hg above the level in wakefulness.^3,¹¹^–¹³ Berthon-Jones and Sullivan¹⁶ reported arousal thresholds up to 6 mm Hg higher in slow-wave sleep than in either stage 2 or REM sleep in male but not in female subjects, whereas Douglas and associates¹³ found no gender or sleep stage-related differences. Hypoxia increases the sensitivity to CO₂ arousal.³⁵

Added Resistance

Human beings tend to arouse from sleep following either the addition of an inspiratory resistance ²⁸ or the occlusion of inspiration.³⁷ Added inspiratory resistance was found to produce a similar percentage increase in arousal frequency in stages 2 and 3/4 and REM sleep.²⁶ However, the arousal frequency during control sleep periods was lowest in stages 3/4 sleep and remained significantly lower during slow-wave sleep than in REM sleep after the addition of inspiratory resistance (Fig. 22-5).²⁴^,²⁵

Figure 22-5 Timing of arousal after application of inspiratory resistance of 4, 7, or 10 cm H₂O/L/sec in stages 2 and 3/4 and REM sleep. Those occasions in which arousal occurred within 2 min of addition of resistance are indicated by red circles, and those in which arousal did not occur within 2 min by blue circles.

(From Gugger M, Molloy J, Gould GA, et al. Ventilatory and arousal responses to added inspiratory resistance during sleep. Am Rev Respir Dis 1989;140:1301-1307.)

Arousal from REM sleep after airway occlusion ³²^,³⁸ is far more rapid than arousal from NREM sleep (Fig. 22-6; nearly 3 times quicker in REM), whereas patients with obstructive apneas have longer apneas during REM sleep. Upper airway receptors would be exposed to respiratory pressure changes in normal people with airway occlusion but in patients with sleep apnea receptors above the level of occlusion would not experience pressure swings. However, it is unclear why these upper airway receptors should be more effective in REM sleep than in NREM sleep in normal subjects.