To determine whether sleep quality is influenced by the mode of mechanical ventilation, we performed polysomnography on 11 critically ill patients. Because pressure support predisposes to central apneas in healthy subjects, we examined whether the presence of a backup rate on assist-control ventilation would decrease apnea-related arousals and improve sleep quality. Sleep fragmentation, measured as the number of arousals and awakenings, was greater during pressure support than during assist-control ventilation: 79 ± 7 versus 54 ± 7 events per hour (p = 0.02). Central apneas occurred during pressure support in six patients; heart failure was more common in these six patients than in the five patients without apneas: 83 versus 20% (p = 0.04). Among patients with central apneas, adding dead space decreased sleep fragmentation: 44 ± 6 versus 83 ± 12 arousals and awakenings per hour (p = 0.02). Changes in sleep–wakefulness state caused greater changes in breath components and end-tidal CO2 during pressure support than during assist-control ventilation. In conclusion, inspiratory assistance from pressure support causes hypocapnia, which combined with the lack of a backup rate and wakefulness drive can lead to central apneas and sleep fragmentation, especially in patients with heart failure.
Mechanical ventilation is used primarily to improve gas exchange and achieve respiratory muscle rest (1). To achieve this goal, it is important that a patient does not make respiratory efforts out of synchrony with the cycling of the ventilator (2, 3). Because behavioral stimuli are decreased during sleep, respiratory muscle rest might be greater during sleep as compared with wakefulness. The operation of a ventilator, however, including its alarms, may disrupt sleep (4). Although disruptions of sleep may adversely affect critically ill patients, little information is available about the interplay between patient–ventilator synchrony and sleep. Studies of healthy volunteers and animals suggest that sleep disruption could result in negative energy balance (5, 6), reduced host immunity (7), and decreased ventilatory responses to hypoxemia and hypercapnia (8, 9). These deleterious effects might prolong the duration of mechanical ventilation in critically ill patients.
Cooper and coworkers (10) have shown that mechanically ventilated patients arouse repeatedly from sleep. In patients free of critical illness, arousals can result from a derangement of arterial blood gases or an increase in respiratory effort (11–13), and both are common occurrences in critically ill patients. Patient–ventilator dysynchrony or a particular ventilator mode might also lead to sleep disruption (10). Pressure support predisposes to an abnormal breathing pattern, specifically central apneas with consequent hyperpnea (14). Meza and coworkers (14) demonstrated that healthy subjects develop central apneas during pressure support when their carbon dioxide tension (Pco2) decreases by a few torr below the apnea threshold. Whether pressure support causes apneas in critically ill patients and whether the apneas lead to disruption of sleep are unknown. Disruption of sleep secondary to central apneas, however, can cause cardiopulmonary abnormalities in ambulatory patients (8, 9, 15–20), and these effects may be magnified in critically ill patients.
Unlike pressure support, assist-control ventilation delivers a fixed tidal volume on every breath, and it can be set to deliver breaths when a patient fails to make an effort. The backup rate will prevent the development of apneas, and perhaps decrease arousals. If pressure support causes central apneas by lowering arterial Pco2 below the apnea threshold, then adding dead space to the ventilator circuit might also decrease both the apneas and arousals.
The purpose of our study was to determine whether the mode of mechanical ventilation influences sleep consolidation in critically ill patients. We hypothesized that sleep would be more fragmented during pressure support than during assist-control ventilation, primarily because of the development of central apneas. A second goal was to determine whether adding dead space during pressure support would decrease sleep fragmentation as a result of decreasing the frequency of central apneas. Some of the results of these studies have been previously reported in the form of abstracts (21, 22).
Additional details are provided in the online data supplement.
Eleven mechanically ventilated patients were recruited (Table 1)
Patient No. | Diagnosis | Age (yr) | Days of Ventilator Support |
---|---|---|---|
1 | ARDS, CHF | 90 | 26 |
2 | ARDS, CHF | 65 | 7 |
3 | ARDS | 75 | 16 |
4 | ARDS | 54 | 31 |
5 | CHF | 67 | 3 |
6 | CHF | 49 | 4 |
7 | CHF | 72 | 50 |
8 | Pneumonia | 66 | 7 |
9 | Pneumonia | 59 | 2 |
10 | Pneumonia, CHF | 85 | 34 |
11 | Pulmonary hemorrhage | 58 | 4 |
Patients were ventilated through an endotracheal tube or tracheostomy connected to a Puritan-Bennett 7200 ventilator (Mallinckrodt, Hazelwood, MO). End-tidal CO2 was measured at the distal end of the endotracheal or tracheostomy tube. Airflow and airway pressure were measured in the inspiratory limb of the ventilator circuit, close to the Y-piece. The signals were measured between the end-tidal CO2 connector and the Y-piece when dead space (100 ml) was introduced. Rib-cage and abdominal motion, pulse oximetry, electroencephalography, right and left electrooculography, submental electromyography, and electrocardiography were recorded (see online data supplement).
The ventilator was initially set in the assist-control mode with a backup rate of 4 breaths per minute and tidal volume (Vt) of 8 ml/kg. Over 5 to 10 minutes of quiet wakefulness, the patient's respiratory rate on the ventilator was measured. The backup rate on assist-control ventilation was then set at 4 breaths below the patient's respiratory rate and kept at that setting for the rest of the study. Pressure support was adjusted to achieve a Vt equivalent to that during assist-control ventilation, namely 8 ml/kg. Patients were randomized to receive at least 2 hours each of the following three modes: assist-control ventilation, pressure support alone, and pressure support with dead space.
All studies were performed between 10:00 p.m. and 06:00 a.m. Apneas, and electroencephalogram (EEG) arousals and awakenings were manually scored (23). EEG arousals were defined as an abrupt shift in EEG frequency consisting of theta, α, and/or frequencies greater than 12 Hz lasting 3 seconds or longer and suggestive of a return toward wakefulness (24). Awakenings were defined as EEG features compatible with wakefulness and lasting more than half of a 30-second epoch (23). Arousals and awakenings were classified as apnea related if they occurred within 3 breaths of the end of an apnea (25, 26).
Elastance and resistance of the respiratory system were measured as described previously (27). Mechanical inspiratory time (Ti), expiratory time (Te), total respiratory cycle time (Ttot), end-tidal CO2, and Vt were measured on a breath-by-breath basis. Apnea threshold was determined from the end-tidal CO2 of the breath immediately before the onset of an apnea (28) (see online data supplement).
Number of apneas, arousals, and awakenings that occurred per hour of sleep and the coefficient of variation of end-tidal CO2 were calculated for each mode of ventilation and compared by one-way analysis of variance (ANOVA). For a given mode, breath components during wakefulness were quantified as percentage change from sleep to wakefulness, and vice versa. This approach was used to assess the effect of a state (sleep or wakefulness) on breathing pattern while still accounting for differences in breath components among patients. During a given state (sleep or wakefulness), the differences in magnitude of breath components between ventilator modes are expressed in absolute numbers. Two-way ANOVA was used to compare breath components during different ventilator modes and sleep–wakefulness states. For ANOVA results, Newman–Keuls multiple range tests were applied.
All 11 patients achieved sleep while receiving each mode of ventilation. Total durations of sleep were as follows: 90 ± 6 minutes during assist-control ventilation, 75 ± 6 minutes during pressure support alone, and 82 ± 7 minutes during pressure support with added dead space. Rapid eye movement sleep was observed in 4 of the 11 patients, but only 1 patient achieved rapid eye movement sleep with all three modes. Slow wave sleep was not seen in any patient. Five patients had features of Stage 2 sleep (spindles and K complexes), and the other six patients had features of only Stage 1 sleep. The efficiency of maintaining sleep (time asleep divided by duration of the study arm, 2 hours) was 63 ± 5% during pressure support alone; efficiency was higher with the addition of dead space, 81 ± 7 (p < 0.05), or with assist-control ventilation, 75 ± 5% (p < 0.05).
Six of the 11 patients developed apneas during pressure support but not during assist-control ventilation by virtue of the backup rate (p < 0.05; Figure 1

Figure 1. Polysomnographic tracings during assist-control ventilation and pressure support in a representative patient. Electroencephalogram (C4-A1, O3-A2), electrooculogram (ROC, LOC), electromyograms (Chin and Leg), integrated tidal volume (VT), rib-cage (RC), and abdominal (AB) excursions on respiratory inductive plethysmography are shown. Arousals and awakenings, indicated by horizontal bars, were more numerous during pressure support than during assist-control ventilation.
[More] [Minimize]
Figure 2. Number of apneas per hour during assist-control ventilation (AC; open triangles), pressure support (PS; closed circles), and pressure support with added dead space (open circles). Six of 11 patients developed apneas during pressure support as compared with none during assist-control ventilation (top). In the patients who developed apneas, the number of apneas decreased with the addition of dead space (bottom). Individual and group mean values are shown. Bars represent the SE.
[More] [Minimize]In the 11 patients, the frequency of arousals was equivalent during pressure support and assist-control ventilation: 39 ± 6 and 35 ± 7 arousals/hour, respectively (p = 0.8). Awakenings were more frequent during pressure support than during assist-control ventilation: 39 ± 7 versus 19 ± 3 awakenings/hour (p < 0.01). Total sleep fragmentation, measured as the sum of arousals and awakenings, was greater during pressure support than during assist-control ventilation: 79 ± 7 versus 54 ± 7 arousals and awakenings/hour (p < 0.05). During pressure support, apnea-related sleep fragmentation consisted of 10 ± 4 arousals/hour, 13 ± 6 awakenings/hour, and 23 ± 7 arousals and awakenings/hour as contrasted with rates of zero during assist-control ventilation (p < 0.05).
In the six patients who developed apneas during pressure support, arousals and awakenings were more frequent during pressure support than during assist-control ventilation: 83 ± 12 versus 43 ± 7 events/hour (p < 0.05). In the five patients who did not develop apneas during pressure support, the frequency of arousals and awakenings was not different during pressure support and assist-control ventilation: 74 ± 6 and 68 ± 12 events/hour, respectively (p = 0.9). In the six patients who developed apneas during pressure support, the addition of dead space decreased the frequency of arousals and awakenings from 83 ± 12 to 44 ± 6 events/hour (p < 0.01).
Compared with wakefulness, respiratory rate was 14.9 ± 3.2% lower (p = 0.0001; Figure 3

Figure 3. Mechanical expiratory time (TE; top left), mechanical inspiratory time (TI; top middle), respiratory rate (top right), minute ventilation (VI; bottom left), tidal volume (VT; bottom middle), and end-tidal CO2 (CO2; bottom right) during assist-control ventilation (AC) and pressure support (PS). For each mode, the lines connect the mean value of each patient during wakefulness (left) and sleep (right). The group mean values are shown for wakefulness (closed symbols) and sleep (open symbols). Compared with wakefulness, TE, VT, and PCO2 were greater and respiratory rate and VI were less during sleep. The differences between sleep and wakefulness for respiratory rate, TI, and TE were greater for pressure support than for assist control.
[More] [Minimize]Minute ventilation (Vi) was lower during sleep as compared with wakefulness: 17.3 ± 3.9% during assist-control ventilation (p < 0.05) and 24.0 ± 8.0% (p = 0.003) during pressure support (Figure 3, bottom left). By design, Vt during wakefulness was equivalent during assist-control ventilation and pressure support: 0.57 ± 0.03 versus 0.64 ± 0.05 L (p = 0.3; Figure 3, bottom middle). During assist-control ventilation, Vt did not change between wakefulness and sleep (p = 0.4). During pressure support, however, Vt was greater during sleep as compared with wakefulness (p = 0.001; Figure 3, bottom middle). The differences in respiratory rate, Ti, and Te between sleep and wakefulness were greater during pressure support than during assist-control ventilation (p < 0.05; Figure 3).
During wakefulness while receiving pressure support, the five patients without apneas had a greater Vt, 0.75 ± 0.08 versus 0.54 ± 0.02 L (p = 0.02), and a greater Vi, 17.2 ± 2.5 versus 10.0 ± 1.3 L/minute (p = 0.03; Figure 4)

Figure 4. Minute ventilation during pressure support while awake (left) and asleep (right) in six patients with apneas (closed symbols) and five patients without apneas (open symbols). Minute ventilation during sleep was greater in patients who did not develop apneas than in patients who developed apneas (p = 0.02), suggesting that increased respiratory drive protects against the development of central apneas.
[More] [Minimize]Mean inspiratory flow, a measure of respiratory drive, was higher in patients without apneas than in patients with apneas: 642 ± 104 versus 359 ± 48 ml/second (p < 0.05) during sleep; and 690 ± 128 versus 402 ± 55 ml/second (p < 0.05) during wakefulness.
The patients with apneas and without apneas had equivalent respective levels of pressure support, 16.8 ± 1.5 and 19.6 ± 2.6 cm H2O (p = 0.6), static elastance, 25.9 ± 4.1 and 29.7 ± 5.7 cm H2O/L (p = 0.3), and respiratory resistance, 13.5 ± 2.0 and 14.1 ± 1.6 cm H2O/L per second (p = 0.8; unpaired t tests). The doses of sedatives did not differ between the patients with and without apneas.
End-tidal CO2 was greater during sleep than during wakefulness: by 4.6 ± 1.8% during assist-control ventilation (p = 0.01) and by 11.0 ± 2.3% (p = 0.0001) during pressure support (Figure 3, bottom right). The extent of change in end-tidal CO2 was greater during pressure support than during assist-control ventilation (p = 0.02; Figure 3, bottom right). The coefficient of variation of end-tidal CO2 levels was greater during pressure support than during assist-control ventilation: 8.7 ± 1.4 versus 4.7 ± 0.7% (p = 0.03). In the six patients who developed apneas, the addition of dead space resulted in a 4.3 ± 1.4 mm Hg increase in end-tidal CO2 (p = 0.01) and it tended to decrease the coefficient of variation in end-tidal CO2 to 5.6 ± 0.6% (p = 0.06). In all patients, oxygen saturation was never below 93%, and it was not affected by differences in sleep–wakefulness state or the mode of ventilation.
The number of central apneas was significantly related to the difference between end-tidal CO2 and the apnea threshold point during wakefulness (r = −0.6; p = 0.02; Figure 5)

Figure 5. The difference between the average end-tidal CO2 and the apnea threshold plotted against the number of central apneas per hour of pressure support alone (closed symbols) and pressure support with added dead space (open symbols) in six patients. The average end-tidal CO2 was measured during sleep (left), wakefulness (middle), and during both sleep and wakefulness (right). The mean number of central apneas per hour was strongly correlated with the end-tidal CO2 during a mixture of both sleep and wakefulness (including the transitions between sleep and wakefulness) (r = −0.83, p < 0.001), but not during sleep alone (r = −0.52, p = 0.08).
[More] [Minimize]The study provides several insights into the effect of ventilator mode on sleep quality of critically ill patients: (1) patients receiving mechanical ventilation have severely fragmented sleep, but pressure support may further aggravate the fragmentation in susceptible patients; (2) at commonly used ventilator settings, central apneas occur during pressure support but not during assist-control ventilation. These apneas can worsen sleep fragmentation; (3) adding dead space decreases the frequency of central apneas and sleep fragmentation during pressure support; and (4) in critically ill patients, the patterns of breathing and gas exchange during sleep differ from those during wakefulness, and these differences are greater with pressure support as compared with assist-control ventilation.
Cooper and coworkers (10) studied sleep quality in mechanically ventilated patients. In eight patients with electrophysiological features of sleep, 42 arousals and awakenings occurred per hour of sleep. This level of sleep fragmentation is similar to that seen in patients with obstructive sleep apnea who have excessive day-time sleepiness and impaired cognition (29). It is also comparable to the fragmentation we observed during assist-control ventilation. Unlike the current study, Cooper and coworkers (10) did not investigate the effect of ventilator mode on sleep quality.
Central apneas can cause hypoxia and hypercapnia with consequent increase in respiratory effort, and all three factors can cause arousal from sleep (25, 30, 31). The patients who developed central apneas did so more frequently during pressure support than during pressure support with added dead space, and, in turn, more frequently than during assist-control ventilation (Figure 2). Sleep fragmentation followed a similar trend. Compared with assist-control ventilation, the increase in apneas during pressure support, with and without dead space, was associated with a proportional increase in the number of awakenings (r = 0.66; p = 0.01). These data suggest that central apneas during pressure support may worsen sleep quality.
Sleep fragmentation during pressure support resulted from an increased frequency of awakenings rather than arousals; indeed, arousals did not differ between the two modes. These data suggest that not only the frequency, but also the degree, of sleep fragmentation was greater during pressure support. Changes in sleep depth are thought to be more important than the mere frequency of disruptions in causing daytime sleepiness in patients with obstructive sleep apnea (32). Although daytime sleepiness may not be a concern in critically ill patients, sleep fragmentation can adversely affect the cardiorespiratory and autonomic nervous systems (16, 17, 19, 20).
To delineate the influence of central apneas on sleep fragmentation, we divided arousals and awakenings into those that did or did not occur within 3 breaths of the end of an apnea. A substantial number of apnea-related arousals and awakenings occurred during pressure support. These events were decreased by adding dead space and completely eradicated with assist-control ventilation. At first glance, it might seem that categorizing arousals and awakenings as apnea or nonapnea related would lead to predictable results (because the backup rate on assist-control ventilation should prohibit apnea-related arousals). The classification, however, highlights that all other (non-apnea-related) arousals and awakenings were equally frequent during assist-control ventilation and pressure support. This observation suggests that critically ill patients have a background level of sleep disturbance, probably secondary to nonrespiratory factors such as pain, medications, staff interruptions, noise, and light (10, 33, 34). Nevertheless, a change in the ventilator mode from assist-control ventilation to pressure support can further aggravate sleep disruption.
To minimize the effect of variations in environmental factors (33–35), pain, acuity of illness, and medications (36) on sleep fragmentation, the patients were randomized to three arms on a single night. Visits from staff were minimized and data during such visits were excluded from analyses.
This is the first study to assess whether critically ill patients develop central apneas while receiving mechanical ventilation (Figure 1). During pressure support, six patients developed apneas and five patients did not (Figure 2). Heart failure was more common among the patients with apneas than among the patients without apneas: 83 versus 20% (p = 0.04). This finding suggests that mechanisms similar to those seen in patients with Cheyne–Stokes respiration could be involved (37). Occurrence of central apneas during pressure support is not unique to patients with heart failure. Central apneas can occur in healthy subjects receiving pressure support (14, 38). Moreover, pressure support has inherent oscillatory behavior that predisposes to substantial variations in tidal volume even when patient effort is constant (39, 40). This instability combined with higher levels of pressure support (14) predisposes to central apneas and consequent sleep fragmentation in patients without heart failure, and may explain the occurrence of apneas in our patient who did not have heart failure.
If pressure support predisposes to apneas in healthy subjects, it might appear puzzling that five of our patients did not develop apneas during pressure support. The lack of apneas may have resulted from dampening of the chemical control system secondary to abnormal respiratory mechanics and elevated dead space (41); a low functional residual capacity and low levels of pressure support may have also contributed. The level of pressure support and the elastance of the respiratory system were not different in the six patients with apneas and the five patients without apneas (p > 0.3). During both sleep and wakefulness, minute ventilation was greater in patients without apneas than in patients with apneas (Figure 4). The higher ventilation in the patients without apneas suggests that a greater respiratory drive, intrinsic dead space, or both may have stabilized ventilation and prevented central apneas. Mean inspiratory flow, a measure of respiratory drive (42), was indeed higher in patients without apneas than in the patients with apneas during both sleep and wakefulness (p < 0.05). The elevated respiratory drive in these five patients may have stabilized ventilation and prevented apneas from occurring (43). An elevated drive may have also prevented the occurrence of apneas during pressure support in one of our patients with heart failure.
The addition of dead space produced a mean increase in end-tidal CO2 of 4.3 mm Hg, which resulted in a decrease in the frequency of central apneas (Figure 2, bottom) and sleep disruptions. Hypocapnia can trigger central apneas in patients with heart failure (44), and the correction of hypocapnia can abolish central apneas in patients with idiopathic central sleep apnea (44) and central sleep apnea due to heart failure (45). Adding dead space has not been previously shown to prevent central apneas during pressure support.
Inspiratory assistance, pain, discomfort, and sepsis can cause chronic hyperventilation (46, 47), which, in turn, may lower the Pco2 below the apnea threshold and cause central apneas. Therefore, as expected, the level of pressure support and elastance of the respiratory system correlated with the frequency of central apneas (14). Similarly, acute hyperventilation due to awakenings from sleep can also lower Pco2, and thereby cause central apneas (48). As expected, the number of awakenings correlated with the number of central apneas. In line with the preceding conceptual framework, a smaller difference between the end-tidal CO2 and the apnea threshold point would result in more numerous central apneas. In the current study, the frequency of central apneas correlated most strongly with the difference between the end-tidal CO2 (during both wakefulness and sleep) and the apnea threshold point (r = −0.83, p = 0.01; Figure 5). In a multiple regression model, however, the only factor that significantly predicted the frequency of central apneas was the difference between the level of end-tidal CO2 (during both wakefulness and sleep) and the apnea threshold point.
The relative importance of hyperpnea during sleep, wakefulness, and the transition between wakefulness and sleep in causing central apneas is controversial. To address this issue, we measured end-tidal CO2 during unequivocal sleep (Figure 5), during unequivocal wakefulness (Figure 5), and during a mixture of sleep and wakefulness (to include the transitions between the two states) (Figure 5). The frequency of central apneas was strongly correlated with values of end-tidal CO2 that included a mixture of wakefulness and sleep (r = −0.83; p < 0.001; Figure 5) but was not significantly correlated with values of end-tidal CO2 during sleep alone (r = −0.52; p = 0.08). This observation stresses that the level of CO2 during sleep is not the sole determinant of central apneas. If a patient is chronically hyperventilating during wakefulness and then rapidly switches from wakefulness to sleep, not enough time will be available at the point of sleep onset for Pco2 to increase. Consequently, central apneas will ensue.
When physicians set the ventilator during the daytime they do not base it on whether a patient is asleep or awake because they usually cannot tell what state a patient is in. In the current study, both the pattern of breathing and gas exchange differed between sleep and wakefulness during both assist-control ventilation and pressure support (Figure 3) and the state-dependent changes were greater during pressure support than during assist-control ventilation. In clinical practice, the level of pressure support is adjusted in accordance with a patient's respiratory rate, which provides reasonable guidance as to a patient's inspiratory effort (49). Respiratory rate was lower during sleep as compared with wakefulness (Figure 3, top right). If a physician adjusts pressure support according to patient's sleeping respiratory rate, patient effort will likely increase on awakening.
Changes in ventilator settings are commonly based on arterial blood gas measurements. With pressure support, end-tidal CO2 was up to 7 mm Hg higher during sleep as compared with wakefulness. The coefficient of variation for end-tidal CO2 during pressure support was 8.1% (range, 1.5 to 10.3%). Differences of this magnitude between sleep and wakefulness, arising from variations in CO2, may cause physicians to change ventilator settings when a change is not necessary. Consequently, underventilation or overventilation may result (50).
The changes in breathing pattern and gas exchange between sleep and wakefulness carry important implications for research. During pressure support, our patients experienced an increase in mechanical inspiratory time on moving from wakefulness to sleep (Figure 3, top middle). This change was associated with an increase in tidal volume, thereby inducing hypocapnia and apnea during pressure support. Research into patient–ventilator interaction also needs to carefully control for the sleep–wakefulness state because of its effect on breath components and gas exchange (51, 52).
It would be imprudent to disregard central apneas and the associated fragmentation of sleep as aberrations of little consequence. Central apneas can cause repetitive arousals that can, in turn, produce elevations in the levels of catecholamines (16) and blood pressure (15), and contribute to the progression of cardiac failure (17, 18) and arrhythmias (19). These deleterious consequences of central apneas might increase morbidity and mortality in critically ill patients. In summary, critically ill patients experience greater fragmentation of sleep during pressure support than during assist-control ventilation because of the development of central apneas, and this effect is especially prominent in patients with heart failure. In conclusion, selecting a ventilator mode has a marked influence on the quality of sleep in a critically ill patient, and a patient's response to ventilator settings can differ significantly between sleep and wakefulness.
The authors thank Dr. Franco Laghi for his careful review of the manuscript.
1. | Tobin MJ. Advances in mechanical ventilation. N Engl J Med 2001;344:1986–1996. |
2. | Tobin MJ, Jubran A, Laghi F. Patient–ventilator interaction. Am J Respir Crit Care Med 2001;163:1059–1063. |
3. | Sinderby C, Navalesi P, Beck J, Skrobik Y, Comtois N, Friberg S, Gottfried SB, Lindstrom L. Neural control of mechanical ventilation in respiratory failure. Nat Med 1999;5:1433–1436. |
4. | Krachman SL, D'Alonzo GE, Criner GJ. Sleep in the intensive care unit. Chest 1995;107:1713–1720. |
5. | Scrimshaw NS, Habicht JP, Pellet P, Piche ML, Cholakos B. Effects of sleep deprivation and reversal of diurnal activity on protein metabolism of young men. Am J Clin Nutr 1966;19:313–319. |
6. | Rechtschaffen A, Gilliland MA, Bergmann BM, Winter JB. Physiological correlates of prolonged sleep deprivation in rats. Science 1983;221:182–184. |
7. | Benca RM, Quintas J. Sleep and host defenses: a review. Sleep 1997;20:1027–1037. |
8. | Cooper KR, Phillips BA. Effect of short-term sleep loss on breathing. J Appl Physiol 1982;53:855–858. |
9. | White DP, Douglas NJ, Pickett CK, Zwillich CW, Weil JV. Sleep deprivation and the control of ventilation. Am Rev Respir Dis 1983;128:984–986. |
10. | Cooper AB, Thornley KS, Young GB, Slutsky AS, Stewart TE, Hanly PJ. Sleep in critically ill patients requiring mechanical ventilation. Chest 2000;117:809–818. |
11. | Issa FG, Sullivan CE. Arousal and breathing responses to airway occlusion in healthy sleeping adults. J Appl Physiol 1983;55:1113–1119. |
12. | Gleeson K, Zwillich CW, White DP. The influence of increasing ventilatory effort on arousal from sleep. Am Rev Respir Dis 1990;142:295–300. |
13. | Berry RB, Mahutte CK, Light RW. Effect of hypercapnia on the arousal response to airway occlusion during sleep in normal subjects. J Appl Physiol 1993;74:2269–2275. |
14. | Meza S, Mendez M, Ostrowski M, Younes M. Susceptibility to periodic breathing with assisted ventilation during sleep in normal subjects. J Appl Physiol 1998;85:1929–1940. |
15. | Davies RJO, Belt PJ, Robert SJ, Ali NJ, Stradling JR. Arterial blood pressure responses to graded transient arousal from sleep in normal humans. J Appl Physiol 1993;74:1123–1130. |
16. | Naughton MC, Bernard DC, Liu PP. Effects of CPAP on sympathetic activity in patients with heart failure and central sleep apnea. Am J Respir Care Med 1995;152:472–479. |
17. | Mann DL, Kent RL, Parsons B. Adrenergic effects on the biology of the adult mammalian cardiocyte. Circulation 1992;85:790–804. |
18. | Leung RST, Bradley TD. Sleep apnea and cardiovascular disease. Am J Respir Crit Care Med 2001;164:2147–2165. |
19. | Meredith IT, Broughton A, Jennings GL. Evidence of a selective increase in cardiac sympathetic activity in patients with sustained ventricular arrhythmias. N Engl J Med 1991;325:618–624. |
20. | Bennett LS, Barbour C, Langford B, Stradling JR, Davies RJ. Health status in obstructive sleep apnea: relationship with sleep fragmentation and daytime sleepiness, and effects of continuous positive airway pressure treatment. Am J Respir Crit Care Med 1999;159:1884–1890. |
21. | Parthasarathy S, Malik I, Jubran A, Tobin MJ. Does ventilatory mode influence sleep quality in critically ill patients? [abstract]. Am J Respir Crit Care Med 2001;163:A483. |
22. | Parthasarathy S, Jubran A, Tobin MJ. Does addition of dead space alter sleep quality during pressure support ventilation? [abstract]. Am J Respir Crit Care Med 2001;163:A180. |
23. | Retchshaffen A, Kales A. A manual of standardized terminology, techniques, and scoring system for sleep stages of human subjects. Los Angeles: UCLA BIS/BRI Publications; 1968. |
24. | Bonnet M, Carley D, Carskadon M, Easton P, Guilleminault C, Harper R. A preliminary report from the sleep disorders atlas task force of the American Sleep Disorders Association. Sleep 1992;15:173–184. |
25. | Xie A, Wong B, Phillipson EA, Slutsky AS, Bradley TD. Interaction of hyperventilation and arousals in the pathogenesis of idiopathic central sleep apnea. Am J Respir Crit Care Med 1994;250:489–495. |
26. | Trinder J, Merson R, Rosenberg JI, Fitzgerald F, Kleiman J, Douglas Bradley T. Pathophysiological interactions of ventilation, arousals, and blood pressure oscillations during Cheyne–Stokes respiration in patients with heart failure. Am J Respir Crit Care Med 2000;162:808–813. |
27. | Jubran A, Tobin MJ. Pathophysiologic basis of acute respiratory distress in patients who fail a trial of weaning from mechanical ventilation. Am J Respir Crit Care Med 1997;155:906–915. |
28. | Xie A, Skatrud JB, Dempsey JA. Effect of hypoxia on the hypopnoeic and apnoeic threshold for CO2 in sleeping humans. J Physiol. 2001;535:269–278. |
29. | Guilleminault C, Partinen M, Quera-Salva MA, Hayes B, Dement WC, Nino-Murcia G. Determinants of daytime sleepiness in obstructive sleep apnea. Chest 1988;94:32–37. |
30. | Berthon-Jones M, Sullivan CE. Ventilatory and arousal responses to hypoxia in sleeping humans. Am Rev Respir Dis 1982;125:632–639. |
31. | Berthon-Jones M, Sullivan CE. Ventilatory and arousal responses to hypercapnia in sleeping humans. J Appl Physiol 1984;57:59–67. |
32. | Bennett LS, Langford BA, Stradling JR, Davies RJ. Sleep fragmentation indices as predictors of daytime sleepiness and nCPAP response in obstructive sleep apnea. Am J Respir Crit Care Med 1998;158:778–786. |
33. | Freedman NS, Gazendam J, Levan L, Pack AI, Schwab RJ. Abnormal sleep/wake cycles and the effect of environmental noise on sleep disruption in the intensive care unit. Am J Respir Crit Care Med 2001;163:451–457. |
34. | Walder B, Francioli D, Meyer JJ, Lancon M, Romand JA. Effects of guidelines implementation in a surgical intensive care unit to control nighttime light and noise levels. Crit Care Med 2000;28:2242–2247. |
35. | Gabor JY, Cooper AB, Lee B, Hanly PJ. Impact of the ICU environment on sleep in healthy subjects [abstract]. Am J Respir Crit Care Med 2001;163:A689. |
36. | Kress JP, Pohlman AS, O'Connor MF, Hall JB. Daily interruptions of sedative medications in critically ill patients undergoing mechanical ventilation. N Engl J Med 2000;342:1471–1475. |
37. | Hanly PJ, Zuberi-Khokhar NS. Increased mortality associated with Cheyne–Stokes respiration in patients with congestive heart failure. Am J Respir Crit Care Med 1996;153:272–276. |
38. | Morrell M, Shea SA, Adams L, Guz A. Effects of inspiratory support upon breathing during wakefulness and sleep. Respir Physiol 1993;93:57–70. |
39. | Hotchkiss JR, Adams AB, Stone MK, Dries DJ, Marini JJ, Crooke PS. Oscillations and noise: inherent instability of pressure support ventilation? Am J Respir Crit Care Med 2002;165:47–53. |
40. | Hotchkiss JR, Adams AB, Dries DJ, Marini JJ, Crooke PS. Dynamic behavior during noninvasive ventilation: chaotic support? Am J Respir Crit Care Med 2001;163:374–378. |
41. | Younes M, Ostrowski M, Thompson W, Leslie C, Shewchuk W. Chemical control stability in patients with obstructive sleep apnea. Am J Respir Crit Care Med 2001;163:1181–1190. |
42. | Grunstein MM, Younes M, Milic-Emili J. Control of tidal volume and respiratory frequency in anesthetized cats. J Appl Physiol 1973;35:463–476. |
43. | Khoo MC, Kronauer RE, Strohl KP, Slutsky AS. Factors inducing periodic breathing in humans: a general model. J Appl Physiol 1982;53:644–659. |
44. | Xie A, Rankin F, Rutherford R, Bradley TD. Effects of inhaled CO2 and added dead space on idiopathic central sleep apnea. J Appl Physiol 1997;82:918–926. |
45. | Khayat RN, Xie A, Patel AK, Skatrud JB. Ventilatory and cardiovascular effects of added dead space inpatients with congestive heart failure and central sleep apnea [abstract]. Am J Respir Crit Care Med 2001;163:A756. |
46. | Preas HL II, Jubran A, Vandivier RW, Reda D, Godin PJ, Banks SM, Tobin MJ, Suffredini AF. Effect of endotoxin on ventilation and breath variability: role of cyclooxygenase pathway. Am J Respir Crit Care Med 2001;164:620–626. |
47. | Meah MS, Gardner WN. Post-hyperventilation apnoea in conscious humans. J Physiol 1994;477:527–538. |
48. | Hall MJ, Xie A, Rutherford R, Ando S, Floras JS, Bradley TD. Cycle length of periodic breathing in patients with and without heart failure. Am J Respir Crit Care Med 1996;154:376–381. |
49. | Jubran A, Van De Graaff WB, Tobin MJ. Variability of patient–ventilator interaction with pressure support ventilation in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1995;152:129–136. |
50. | Younes M. Mechanisms of respiratory load compensation. In: Dempsey JA, Pack AI, editors. Lung biology in health and disease. Vol. 79: Regulation of breathing. New York: Marcel Dekker; 1994. p. 867–922. |
51. | Parthasarathy S, Jubran A, Tobin MJ. Assessment of neural inspiratory time in ventilator-supported patients. Am J Respir Crit Care Med 2000;162:546–552. |
52. | Stell IM, Paul G, Lee KC, Ponte J, Moxham J. Noninvasive ventilator triggering in chronic obstructive pulmonary disease: a test lung comparison. Am J Respir Crit Care Med 2001;164:2092–2097. |