© 2006 American Thoracic Society doi: 10.1164/rccm.2601005
Update in Sleep and Control of Ventilation 2005Departments of Medicine and Physiology, and Centre for Sleep Medicine and Circadian Biology at the University of Toronto; Sleep Research Laboratory of the Toronto Rehabilitation Institute; and Department of Medicine, Toronto General Hospital of the University Health Network, Toronto, Ontario, Canada Correspondence and requests for reprints should be addressed to Richard L. Horner, Ph.D., Room 6368, Medical Sciences Building, 1 Kings College Circle, Toronto, ON, Canada M5S 1A8. E-mail: richard.horner{at}utoronto.ca CONSEQUENCES OF HYPOXIA Some infants with obstructive sleep apnea (OSA) demonstrate delayed arousal from sleep after compromised ventilation. Accordingly, Waters and Tinworth (1) tested the hypothesis that exposure to intermittent asphyxia during development can impair arousal responses. In 10-d-old piglets, hypercapnic and hypoxic stimuli (8% O2, 7% CO2) and air (recovery) were each applied at 6-min intervals for a total of 48 min. Stimuli were applied for 4 d and arousal responses were compared with control animals receiving room air. Even on the first day of exposure there were marked delays in time to arousal from the first to the fourth asphyxic episode (17 ± 7 to 42 ± 29 s), with these arousal deficits exacerbated on the fourth day of interventions. The authors conclude that the experience of intermittent asphyxia, even on a background of normal development, can lead to acute and chronic arousal deficits in infants. Eckert and coworkers (2) used respiratory-related evoked potentials (RREPs) to determine the sensory processes mediating hypoxia-induced suppression of respiratory load sensation in normal healthy subjects. RREPs were measured over the central and parietal cortex after application of short-duration inspiratory resistive loads; loads were applied during and after isocapnic normoxia or hypoxia (arterial O2 saturation of about 80%). The subjects also rated the perceived magnitude of externally applied inspiratory resistive loads under the same conditions. After the brief inspiratory loads, the amplitude of the first positive peak (P1) of the RREP (at 40–110 ms) was significantly reduced by about 36% during isocapnic hypoxia compared with normoxia; P1 is thought to reflect the arrival of ascending respiratory signals at the somatosensory cortex. Hypoxia also reduced the perceived magnitude of the externally applied resistive loads. The data confirm that isocapnic hypoxia suppresses respiratory load sensation. The authors also conclude that the changes in P1 suggest that this is mediated, at least in part, by suppression of respiratory afferent information before arrival at the sensory cortex. Some patients with OSA exhibit residual hypersomnolence despite treatment, suggesting long-lasting neural changes, perhaps mediated by hypoxia. Using a mouse model of long-term intermittent hypoxia (IH), Zhan and coworkers (3) determined whether inducible nitric oxide synthase (iNOS) was activated in brain regions involved in sleep–wake regulation, and also compared the effects of chronic IH on hypersomnolence, oxidative injury, and proinflammatory responses in iNOS-deficient mice with those in wild-type control mice. For delivery of IH, ambient O2 was reduced from 21 to 10% for 5 s at 90-s intervals (causing arterial O2 saturations of about 85%), with the stimuli applied for 10 h/d for 6 wk. When measured 2 wk after cessation of IH, the wild-type mice showed increased sleep times and shortened sleep latencies consistent with persistent hypersomnolence, and this was associated with increased iNOS activity in wake-active regions of the lateral basal forebrain and posterior lateral (orexin-containing) hypothalamus. There were no effects of IH on sleep times or latencies in the iNOS-deficient mice, and these mice were also resistant to IH-induced lipid peroxidation and proinflammatory gene responses. Importantly, inhibition of iNOS in the wild-type mice was effective in reducing the proinflammatory gene responses to IH. The authors conclude that the data support an important role for iNOS in the development of persistent hypersomnolence, lipid peroxidation, and proinflammatory responses in wake-active brain regions after IH, and suggest a potential role for inhibition of inducible NO to protect against these effects in patients with OSA. Zhan and coworkers (4) then hypothesized that activation of NADPH oxidase played a major role in oxidation injury after long-term IH. Eight weeks of IH led to increased NADPH oxidase gene and protein responses in most wake-active brain regions tested, including the lateral basal forebrain, dorsal raphe, and locus coeruleus neurons. Importantly, in mice without NADPH oxidase activity, or in those in which NADPH oxidase was pharmacologically inhibited with apocynin, there was resistance to development of persistent hypersomnolence after IH and also resistance to carbonylation, lipid peroxidation injury, and proinflammatory responses in the wake-active brain regions. Overall, these findings significantly increase our understanding of the potentially important role of NADPH oxidase in mediating the neurobehavioral, redox, and proinflammatory responses to IH, and suggest that inhibiting NADPH oxidase may help prevent such oxidation-mediated morbidities in patients with OSA. Because patients with OSA often exhibit impaired neuropsychological function, O'Donoghue and coworkers (5) hypothesized that high-resolution magnetic resonance imaging may reveal structural brain changes consistent with such impairments. Studies were performed on 27 patients with severe OSA (mean apnea–hypopnea index [AHI], 72 ± 17 incidents per hour of sleep) and 24 control subjects. In contrast to another study (6), the results showed no areas of gray matter volume change in patients with OSA when using an optimized voxel-based morphometry technique. Twenty-three patients with OSA had repeat imaging after 6 mo of treatment with nasal continuous positive airway pressure (CPAP); no changes in gray matter density or regional volumes occurred. The authors conclude there is no marked structural brain damage even in severe OSA, but are also careful to stress that this does not indicate there are no changes in neural or psychologic function in patients with OSA, but rather that the basis for those deficits is not gray matter or brain volume changes. In addition to the adverse impact of oxidative stress on central neuronal function and sleepiness (3, 4), Chen and coworkers (7) hypothesized that chronic IH in rats would lead to myocardial dysfunction, oxidative stress, and decreased antioxidant capacity. The IH for this study consisted of minimum O2 levels of about 5%, with the stimuli applied every 60 s for 8 h/d, 5 d/wk for 5 wk. The chronic IH led to left ventricular (LV) hypertrophy, dilation, and decreased fractional shortening. Left ventricular end-diastolic pressures were also increased, and cardiac output decreased, compared with control subjects. These changes in heart function were also associated with markers of increased oxidative stress as LV myocardial lipid peroxides were almost doubled in rats exposed to IH compared with control rats (1,258 ± 703 vs. 715 ± 240 µm/mg protein). There was also evidence of decreased antioxidant capacity after the IH as LV myocardial Cu-Zn superoxide dismutase levels were lowered compared with control rats (10.3 ± 4.9 vs. 18.6 ± 8.2 U/mg protein). The authors conclude that chronic IH leads to oxidative stress over the same time course as development of LV myocardial dysfunction. However, a causal relationship between oxidative stress and myocardial alterations in this setting remains to be determined. UPPER AIRWAY Characterizing the activity of single genioglossus motor units is important because a change in discharge frequency recorded peripherally in muscle reflects changes in excitability of single hypoglossal motoneurons centrally in the medulla. Thirty genioglossus motor units were recorded in anesthetized rats by John and coworkers (8). All units were active during inspiration, with most (77%) also showing increased discharge with hypercapnia, thereby indicating CO2-dependent excitatory inputs to hypoglossal motoneurons. Interestingly, 23% of the motor units did not respond to hypercapnia, showing an effective lack of CO2-sensitive inputs to a significant proportion of hypoglossal motoneurons. Of relevance, motor units were selected for recording on the basis of the presence of clear respiratory-related activity whereas tonically discharging units were not studied. Because tonic activity constitutes a significant component of genioglossus activation in conscious animals (9), it is important in future animal or human experiments to determine the respiratory and nonrespiratory excitatory drives to hypoglossal motoneurons and their sleep state dependence. Studies examining the respiratory control of the tongue have focused almost exclusively on the extrinsic muscles, especially genioglossus. Until recently, however, control of the intrinsic muscles, which comprise the bulk of the tongue, had not been studied. Accordingly, Bailey and coworkers (10) recorded the activity of intrinsic (superior longitudinal) and extrinsic (hyoglossus) retractor muscles in anesthetized rats. These muscles were coactivated during inspiration both before and during mild hypoxia and airway occlusion. The authors suggest these results support the concept that coactivation of both intrinsic and extrinsic protrudor and retractor muscles contributes to the maintenance of airway patency.
Although suppression of pharyngeal muscle activity in sleep is important in predisposing to OSA, the central neural mechanisms underlying the effects of sleep on pharyngeal motoneurons are not well understood. Fenik and coworkers (11) hypothesized that suppression of hypoglossal nerve activity in a pharmacologic model of "rapid eye movement (REM) sleep" is mediated by withdrawal of excitatory serotonergic and noradrenergic inputs to hypoglossal motoneurons. The pharmacologically induced REM sleep–like state was elicited by microinjection of the cholinergic agonist carbachol into the pontine reticular formation of anesthetized and vagotomized rats. During anesthesia before carbachol administration, serotonergic and noradrenergic ( Although the above study implicated withdrawal of serotonin as an important contributor to decreased genioglossus muscle activity in sleep, those experiments were performed in anesthetized and vagotomized rats in a pharmacologically induced REM sleep–like state (11). Accordingly, Sood and coworkers (9) determined the role of endogenous serotonin at the hypoglossal motor nucleus in modulating genioglossus activity in conscious animals. Rats were implanted with electrodes to record genioglossus and diaphragm activities and sleep–wake states. Microdialysis probes in the hypoglossal motor nucleus were used to perfuse artificial cerebrospinal fluid (control), or mianserin or MDL100907 (serotonin and 5-HT2 receptor antagonists). The results showed a normally weak endogenous serotonergic drive modulating genioglossus activity across sleep–wake states, although a role for serotonergic mechanisms in genioglossus motor control was increased during CO2-stimulated breathing. The data also showed a significant endogenous serotonergic drive modulating genioglossus activity in anesthetized and vagotomized rats but not vagus nerve–intact rats. This additional result suggested that previous observations in reduced preparations may have been influenced by vagotomy such that the potential importance of serotonin in modulating genioglossus activity may have been overemphasized. Overall, these results have implications for pharmacologic interventions aiming to increase genioglossus activity by manipulating endogenous serotonin in patients with OSA, a strategy that has met with limited clinical benefit. These two studies (9, 11) were the subject of an editorial in which the need to consider data from multiple preparations was highlighted (12). There is major interest in developing strategies to increase motor output to genioglossus muscle and prevent OSA. Bellemare and coworkers (13) determined whether inspiratory flow limitation occurs in anesthetized and spontaneously breathing supine rabbits, and whether this can be reversed by electrical stimulation of the hypoglossal nerves. All the anesthetized rabbits snored, and 80% also had inspiratory flow limitation that was reversed by hypoglossal nerve stimulation. Interestingly, the recruitment characteristics of hypoglossal nerve fibers appeared steep, with significant mechanical effects obtained with stimulus intensities at 36 to 60% of maximum. The authors conclude that these results support the potential feasibility of hypoglossal nerve stimulation as a treatment option for OSA. However, the achievable mechanical effects on the airspace within the tolerable window of stimulation intensities may be a limiting factor in some sleeping humans. The upper airway of children with OSA is more collapsible than the upper airway of children without OSA. Accordingly, Arens and coworkers (14) used respiratory-gated magnetic resonance imaging to quantify changes in upper airway size and shape during tidal breathing. Several differences in upper airway anatomy were observed in the sedated children with OSA (4.3 ± 2.3 yr; AHI, 5.7 ± 4.7 episodes per hour of sleep) compared with similarly sedated age-matched control subjects. The children with OSA had smaller upper airways especially during inspiration, with airway size increasing during expiration. Airway cross-sectional areas spanning the palatal to glosso-pharynx varied more in the patients with OSA compared with control subjects (317–922 vs. 15–24%, respectively). The authors conclude that upper airway size changes markedly during breathing in children with OSA compared with control subjects, although the contributions of altered resistive loading or compliance to those changes need to be determined. CHEMORECEPTION Individuals exhibit inherent differences in the level of alveolar ventilation and PaCO2, but the mechanisms underlying these differences are unknown. The firing pattern of brainstem chemosensitive neurons is, at least in part, stimulated by CO2-mediated increases in intracellular free protons, which are then extruded via Na+/H+ exchangers (NHEs). Of the eight NHEs currently identified, NHE3 is expressed in respiratory-related regions of the brainstem. In rabbits, Wiemann and coworkers (15) determined levels of NHE3 in the obex region of the medulla and related these to the levels of ventilation and PaCO2 in awake animals. Interestingly, alveolar ventilation was about 20% lower, and PaCO2 correspondingly higher, in the animals with higher levels of brainstem NHE3 compared with those with lower levels of NHE3. Potential confounding factors such as metabolic CO2 production were not different between groups. The authors conclude that levels of brainstem NHE3, likely via intracellular pH modulation, contribute to the individual level of alveolar ventilation and PaCO2 by adjusting the set point and loop gain of the respiratory system. Two further articles debate topical and important issues regarding sites of central chemoreception (widespread vs. localized and specialized chemoreceptor cells), the relative roles of serotonergic neurons of the medullary raphe nuclei versus glutamatergic neurons of the retrotrapezoid nucleus in mediating respiratory responses to CO2, and whether serotonergic neurons are intrinsically chemosensitive and mediate respiratory responses in vivo (16, 17). A further review highlights breakthroughs in the control of breathing in the last 100 yr (18) and another focuses on advances in the pathogenesis of obstructive and central sleep apnea (19). ANIMAL MODELS OF SLEEP-DISORDERED BREATHING Central congenital hypoventilation syndrome (CCHS) is characterized by autonomic nervous system disturbances and sleep-disordered breathing. A high proportion of patients with CCHS are heterozygous for the Phox2b polyalanine expansion mutation. Accordingly, Durand and coworkers (20) determined whether newborn mice heterozygous for the targeted deletion of the transcription factor Phox2b would display sleep-disordered breathing. Breathing pattern was measured by whole-body plethysmography in 5-d-old mutant mice (Phox2b+/–) and wild-type control mice. Despite normal amounts of sleep–wake states, total apnea time in the mutant pups was markedly increased compared with control mice, due specifically to disordered breathing events in sleep. In contrast to typical findings in patients with CCHS, however, the disordered breathing largely occurred in active rather than quiet sleep. Interestingly, an adult with late-onset central hypoventilation syndrome was also found to have a mutation in the Phox2b gene (21). The pre-Bötzinger complex in the ventrolateral medulla is essential for respiratory rhythm generation in rodents. Accordingly, McKay and coworkers (22) hypothesized that ablation of pre-Bötzinger complex neurons would disrupt respiratory pattern, especially during sleep. In adult rats instrumented for recording sleep–wake states and breathing, substance P conjugated to the toxin saporin was injected bilaterally into the pre-Bötzinger complex to selectively ablate neurokinin-1 receptor–expressing neurons. Ablation of these neurons disrupted respiratory pattern, initially in REM sleep (about 4 d postinjection), then in non-REM sleep (about 6 d), and finally in wakefulness (about 9 d). Because sleep-disordered breathing (SDB) is prevalent in elderly humans and in some patients with neurodegenerative disease, the authors suggest that loss of pre-Bötzinger complex neurons could be a contributing factor to SDB in those conditions. SDB AND METABOLISM Neuropeptide Y (NPY) and leptin may contribute to the development of atherosclerosis. Barceló and colleagues (23) investigated the independent effects of OSA and obesity on plasma NPY and leptin levels in patients with OSA. NPY levels were increased (p < 0.01) in patients with OSA, independent of obesity. Leptin levels were also increased in OSA, but mainly in association with obesity. Among patients with OSA, CPAP applied for at least 4 h per night in a nonrandomized, uncontrolled fashion was accompanied by reductions in NPY levels in all patients, and by leptin levels only in nonobese patients. The authors concluded that NPY and leptin levels are increased in patients with OSA, but whereas the former appears independent of obesity, the latter is closely linked to obesity. Using self-reported questionnaire data, Shin and coworkers (24) examined the association of habitual snoring with glucose and insulin metabolism in a cross-sectional analysis of 2,719 nonobese Korean men free of diabetes. Whereas there were no significant differences in fasting blood glucose and insulin levels between habitual snorers and nonsnorers, habitual snoring was independently associated with elevated glucose and insulin levels during glucose tolerance tests. The authors concluded that habitual snoring may affect glucose–insulin metabolism, independently of diabetes in nonobese, middle-aged Korean men. Because polysomnography was not performed, this study leaves unclear whether the relationship between habitual snoring and abnormal glucose metabolism was related to snoring per se, or to associated SDB.
Reichmuth and associates (25) performed cross-sectional and longitudinal analyses of 1,387 participants in the Wisconsin Sleep Cohort to determine whether SDB is independently related to the prevalence and incidence of type II diabetes. The prevalence of diabetes increased with increasing AHI. The odds ratio (OR) for having diabetes with an AHI SDB AND CARDIOVASCULAR DISEASES Whereas C-reactive protein (CRP) levels have been reported to be elevated in adults with SDB, and may be associated with development of cardiovascular diseases, it remained unknown whether the same was true in children. Kaditis and colleagues (26) found no significant relationship between the presence of SDB and CRP levels in children. Drager and coworkers (27) tested the hypothesis that OSA could contribute to the risk for atherosclerosis by altering functional and structural properties of large arteries. In healthy volunteers and patients with OSA they measured arterial compliance, and carotid intima–media thickness and diameter. They found more evidence of early carotid artery atherosclerosis in subjects with OSA, characterized by lower arterial compliance, and greater carotid intima–media thickness and diameter, than in subjects without OSA. Minoguchi and coworkers (28) also found greater carotid intima–media thickness in association with higher serum levels of CRP, interleukin 6 (IL-6), and IL-8 in patients with OSA than in control subjects. These findings suggest that SDB could predispose to stroke.
To determine whether SDB is associated with stroke, Arzt and colleagues (29) performed cross-sectional and longitudinal analyses of the 1,475 and 1,189 participants, respectively, of the Wisconsin Sleep Cohort. In the cross-sectional analysis, subjects with an AHI With respect to the pulmonary circulation, although local tissue activation of the endothelin system contributes to the development of pulmonary hypertension, the impact of isolated chronic plasma hyperendothelinemia on the pulmonary circulation is unknown. Migneault and associates (30) infused endothelin-1 in rats over 7 or 28 d, and found that plasma endothelin-1 levels doubled (p < 0.05), and lung tissue endothelin-1 levels increased fourfold after 7 d (p < 0.001) but was no longer significantly elevated after 28 d. Right ventricular systolic pressure was unaffected. Maximum dilatory responses to both acetylcholine (p < 0.01) and sodium nitroprusside (p < 0.001) were reduced after 28 d and were normalized by addition of the nitric oxide synthase inhibitor L-NG-nitro-L-arginine and the antioxidant N-acetyl-L-cysteine. Therefore chronic hyperendothelinemia reduces pulmonary vasodilator reserve in response to nitric oxide. Correction by an antioxidant and L-NG-nitro-L-arginine suggests that this relates to increased production of reactive oxygen species. However, endothelin does not appear to be elevated in patients with OSA. In a previous study involving patients with bradyarrhythmias, atrial overdrive pacing improved both central sleep apnea (CSA) and OSA. To determine whether these findings could be reproduced, Lüthje and colleagues (31) studied 20 patients who had implanted dual-chamber pacemakers or cardioverter defibrillators for therapy of various cardiac arrhythmias and who also had predominantly OSA. Patients underwent polysomnography on three consecutive nights in a randomized, single-blind, cross-over study in which pacemakers were programmed for nonpacing, or for overdrive pacing at 7 or 15 beats/min faster than the mean nocturnal heart rate. Overdrive pacing at either of the two pacing rates had no effect on AHI. Pepin and coworkers (32) confirmed these findings in patients with OSA who had no other indication for a pacemaker. These data do not support the use of atrial overdrive pacing for therapy of OSA. CONTROL OF BREATHING AND SLEEP APNEA
Nopmaneejumruslers and associates (33) tested the hypotheses that CSA in patients with stroke is associated with nocturnal hypocapnia and LV systolic dysfunction (LV ejection fraction [LVEF] During sleep, maintenance of rhythmic breathing is dependent on PCO2. If PCO2 falls below the apneic threshold, CSA ensues. Xie and coworkers (34) tested the hypothesis that one factor that might contribute to hypocapnia and respiratory control system instability in patients with HF and CSA is attenuated cerebrovascular response to CO2. Cerebral blood flow velocity in the middle cerebral artery was measured in patients with HF with and without CSA. The overall cerebrovascular reactivity to CO2 was lower in patients with CSA than in the control group (p < 0.05). These investigators concluded that patients with HF and CSA have a diminished cerebrovascular response to PCO2. The reduced cerebrovascular reactivity to CO2 might compromise the ability of the brain to wash CO2 from the brainstem, thus predisposing to CSA by facilitating ventilatory overshooting during hypercapnia and undershooting during hypocapnia. Low PCO2 predisposes to CSA in patients with HF. Javaheri and colleagues (35) sought to determine whether the same was true in patients with liver cirrhosis and normal LV systolic function. In 13 hypocapnic (PCO2 < 36 mm Hg; mean, 33 mm Hg) patients with HF, the mean AHI was 28/h and most events were central. In contrast, among 10 hypocapnic (PaCO2 < 36 mm Hg; mean, 32 mm Hg) patients with cirrhosis and normal LVEF, the mean AHI was only 2/h. The authors concluded that, in contrast to patients with HF, the presence of hypocapnia does not predict CSA in patients with cirrhosis. The reason for this discrepancy was not determined. Brack and associates (36) examined end-expiratory lung volume in 12 patients with Cheyne-Stokes respiration. Compared with regular breathing, end-expiratory lung volume during hyperpnea increased in all 12 primarily because of alterations in tidal volume and expiratory time rather than postinspiratory inspiratory muscle activity. The implications of these findings for the pathogenesis of CSA in patients with HF were not discussed. CLINICAL FEATURES, MONITORING, AND THERAPY OF SDB The AHI, the severity of nocturnal desaturation, and the frequency of movement arousals are poor predictors of the degree of daytime sleepiness in patients with OSA. Chervin and coworkers (37) found that the tendency for sigma (13–15 Hz) electroencephalographic power to vary with the respiratory cycle during periods of regular breathing predicted next-day sleepiness measured by the multiple sleep latency test. The predictive value was greater in patients with OSA than in those without this problem. In contrast, the AHI and minimal oxygen saturation did not predict sleepiness as well. These findings suggested that altered electroencephalographic activity during nonapneic breathing may be linked to daytime sleepiness. Because previous studies have demonstrated that upper airway size and resistance are influenced by lung volume during wakefulness, Heinzer and colleagues (38) tested the effects of increasing and decreasing lung volume during non-REM sleep on the level of CPAP required to prevent flow limitation in subjects with OSA. Using a rigid head-out shell equipped with a positive/negative pressure attachment to manipulate extrathoracic pressure, they demonstrated that increasing lung volume decreased the CPAP level required to prevent flow limitation (p < 0.001). Conversely, reducing lung volume increased the CPAP (p < 0.001) required to prevent flow limitation. Thus changes in lung volume have important effects on upper airway patency in subjects with OSA during sleep. It has been suggested that rising asthma rates may be partly due to increasing obesity, but the causal mechanisms are unclear. In 788 children, Sulit and coworkers (39) examined the potential relationships of obesity and SDB with wheezing and asthma. They found that children with wheeze were significantly more likely to be male (OR, 1.62), black (OR, 1.90), and obese (OR, 1.57), and to have a maternal history of asthma (OR, 1.93). Further adjustment for SDB attenuated the association between obesity and wheeze (OR, 1.45), but did not substantially alter the association between obesity and asthma. Although SDB and obesity are each associated with asthma and wheeze, potential mechanisms for these relationships were not identified. Children with mild SDB, not meeting criteria for adenotonsillectomy, frequently exhibit neurocognitive and behavioral problems. Goldbart and associates (40) tested the hypothesis that antiinflammatory therapy by leukotriene inhibition might attenuate SDB. In a nonrandomized, open-label trial, montelukast administration for 16 wk in 24 children with SDB induced significant reductions in adenoid size and SDB. They concluded that leukotriene inhibition may attenuate SDB in children, but that it cannot be recommended for this purpose until randomized controlled trials confirm their findings. Whitelaw and associates (41) measured the accuracy with which clinicians predict clinical improvement of OSA on CPAP, using polysomnography compared with oximeter-based home monitoring. Patients with a clinical suspicion of OSA were randomized to have polysomnography or home monitoring. The likelihood of clinical success of CPAP, that is, greater or less than 50%, was estimated on the basis of clinical and test data. All patients were treated for 4 wk with autoadjusting CPAP irrespective of the presence of OSA or symptoms. Among 288 patients, the correct prediction rate was similar with polysomnography (0.61) and with home monitoring (0.64). These investigators concluded that the ability of physicians to predict the outcome of CPAP treatment in individual patients is not significantly better with polysomnography than with home oximeter–based monitoring. However, these results are confounded by the fact that several patients receiving CPAP did not have OSA, and that optimal levels of CPAP were not necessarily achieved. The clinical relevance of these findings therefore remains unclear. Constrained resources have resulted in long waiting times for polysomnography. Hukins (42) therefore studied 91 patients with OSA who were randomized to either arbitrary pressure CPAP based on body mass index commenced at home before CPAP titration during polysomnnography, or to CPAP at settings determined by polysomnography. Both interventions resulted in similar improvements in clinical outcomes and CPAP compliance after 3 mo. This approach to initiating treatment with CPAP may therefore be feasible when there are long waiting lists for polysomnography. Patients with neuromuscular disease normally have their ventilator set by empirical means. In a randomized trial, Fanfulla and coworkers (43) tested the efficacy of two noninvasive pressure support ventilation (nPSV) settings in nine patients with neuromuscular disease: (1) nPSV setting titrated on simple clinical parameters and (2) physiologic setting, tailored to the patient's respiratory effort. During wakefulness, the two settings caused similar improvements in gas exchange and minute ventilation, and the degree of diaphragm unloading. However, during sleep, compared with clinical nPSV, physiological nPSV caused significantly better gas exchange, sleep efficiency (p < 0.01) and percentage of REM sleep (p < 0.01). These improvements correlated significantly with the reduction in ineffective efforts. They concluded that physiological nPSV causes greater improvement in sleep architecture and nighttime gas exchange than clinical nPSV. Several articles addressed the relationship between SDB, and its treatment, on cardiovascular morbidity and mortality. Gami and coworkers (44) reported a significantly higher sudden death rate in patients with OSA during sleep time (12 midnight to 6:00 A.M.) than in patients without OSA in whom the peak sudden death rate occurred later in the morning after sleep. This suggests that OSA was playing a role in triggering sudden death during sleep, most likely through ventricular arrhythmias. Support for this possibility was provided by Ryan and coworkers (45), who found, in a randomized trial, that treatment of OSA by CPAP in patients with HF significantly reduced the frequency of ventricular ectopic beats during sleep. In a long-term nonrandomized observational study lasting 12 yr, Marin and coworkers (46) reported that among men with OSA not treated with CPAP, those with severe untreated OSA (AHI > 30/h) had a higher rate of fatal and nonfatal cardiovascular events than did men without OSA. In contrast, among patients whose severe OSA was treated with CPAP, the cardiovascular morbidity and mortality rate did not differ from those of subjects without OSA. In a long-term randomized multicenter trial, Bradley and associates (47) tested the effects of CPAP on cardiovascular function, morbidity, and mortality in 258 patients with HF and CSA over mean and maximum follow-up periods of 24 and 54 mo, respectively. CPAP reduced the central AHI by 50%, increased nocturnal O2 saturation, increased LVEF and 6-min walking distance, and lowered daytime plasma norepinephrine levels. However, it did not affect the combined rate of mortality and cardiac transplantation (32 events in both groups, p = 0.54), nor the frequency of hospitalizations. However, the trial lacked sufficient statistical power to conclude that CPAP either does or does not reduce mortality and heart transplantation rates in such patients. The authors concluded that their data did not support the routine use of CPAP to extend life in HF patients with CSA. The American Thoracic Society published a document outlining its official position on core training requirements for pulmonary residency training programs to obtain proficiency in sleep disorders medicine (48). It includes recommendations on the types of didactic and clinical training required to acquire basic scientific and practical knowledge on the pathophysiology and pathogenesis of SDB and other nonrespiratory sleep disorders, as well as on the principles of their diagnosis and therapy. FOOTNOTES Conflict of Interest Statement: R.L.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. T.D.B. received peer-reviewed research funding from Respironics, ResMed, and Tyco in partnership with the Canadian Institutes of Health Research from 1998 to 2006: $2,000,000. Received in original form January 6, 2006; accepted in final form January 6, 2006 REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1) receptor antagonism at the hypoglossal motor nucleus decreased respiratory-related hypoglossal nerve activity, indicating that hypoglossal motoneurons are under tonic excitatory serotonergic and noradrenergic drives. Importantly, after microinjection of these antagonists, pontine carbachol no longer caused depression of hypoglossal nerve activity. The authors conclude that depression of hypoglossal nerve activity after pontine carbachol can be fully explained by the combined withdrawal of excitatory noradrenergic and serotonergic inputs to hypoglossal motoneurons, and speculated that similar mechanisms may be operating in natural REM sleep. 
15/h versus an AHI < 5/h was 2.30 (p = 0.005) after adjustment for age, sex, and body habitus. However, the odds for developing diabetes within 4 yr with a baseline AHI 
40%). In 93 patients with stroke who underwent polysomnography and echocardiography, CSA was found in 19% who had lower nocturnal PCO2 (p = 0.015) and a higher prevalence of LVEF