Section of Respiratory and Critical Care Medicine, Department of Medicine, University of Illinois at Chicago and VA Chicago Health Care System, Chicago, Ill., USA
Obstructive sleep apnea syndrome (OSAS) afflicts about 5% of adults in Western countries and is thought to play an important role in the pathogenesis of cardiovascular disorders and diabetes mellitus. Although the etiology of OSAS is uncertain, intense local and systemic inflammation are present in these patients. In the upper airway, this process may promote oropharyngeal inspiratory muscle dysfunction and amplify upper airway narrowing and collapsibility thereby worsening the frequency and duration of apneas during sleep. The presence of systemic inflammation, characterized by elevated levels of certain potent pro-inflammatory mediators, such as C-reactive protein, leptin, TNF-α, IL-1β, IL-6, reactive oxygen species and adhesion molecules, may predispose to the development of cardiovascular complications observed in patients with OSAS. Treatment with nasal CPAP abrogates, in part, local and systemic inflammation in these patients. Whether therapeutic interventions aimed at abating inflammation could be a useful adjunct in the treatment of OSAS merits further investigation.
Copyright © 2003 S. Karger AG, Basel
Obstructive sleep apnea syndrome (OSAS) is defined as repeated episodes of upper airway occlusion during sleep with consequent excessive daytime sleepiness and abnormalities in cardiopulmonary function. It is estimated that up to 5% of adults in Western countries are likely to have undiagnosed OSAS . Anatomic narrowing of the airway, increased collapsibility of the airway tissues, disturbances in reflexes that affect upper airway caliber and pharyngeal inspiratory muscle function contribute to upper airway occlusion during sleep . Viewed as predominantly a mechanical problem, the present treatment of OSAS is also mechanical, splinting of the airway by external application of positive pressure.
Despite the attractiveness of such a simplistic model, the past decade has seen an explosion of data implicating a role for inflammation in the pathogenesis of OSAS. In this perspective, we will review some of these data and argue that upper airway and systemic inflammation may, in part, underlie the process of upper airway occlusion and contribute to the clinical manifestations of OSAS and its cardiovascular complications. We have based our review on a Medline search of articles published in the English language since 1966. The search phrases obstructive sleep apnea (OSA) and inflammation, OSA and cytokines, OSA and leptin were used in the process.
Half of patients with OSAS obstruct at the level of the uvulopalatal arch while the other half at the base of the tongue . Magnetic resonance imaging has suggested that upper airway narrowing in patients with obstructive sleep apnea is due to increased thickness of the lateral pharyngeal walls . Interestingly, the increase in thickness is not due to an increase in the size of parapharyngeal fat pads or abnormality in the bony structures but to an increase in soft tissue. It is conceivable that such increase could be partly due to inflammation-induced swelling. Several investigators have shown presence of inflammation in upper airway soft tissue. Woodson et al.  studied histopathologic features of pharyngeal tissue from apneic patients, snorers and normal patients. Extensive edema with vascular congestion and dilatation, acanthosis of epidermis and mucous gland hypertrophy were observed in apneic patients and snorers, albeit to a lesser extent in the latter group.
Detailed histopathological analysis of tissues obtained from 48 patients undergoing uvulopalatopharyngoplasty (UVPP) for OSAS, revealed marked subepithelial edema in 79% of the patients . Sekosan et al.  found extensive plasma cell infiltration and interstitial edema in the lamina propria of the uvula mucosa of patients with moderate OSAS. Recently, Paulsen et al.  examined the epithelial-connective tissue interface in patients with OSAS, habitual snorers and cadavers. A significant reduction in surface area of connective tissue papillae that are specialized structures providing anchorage for epithelium and surface for nutrient transfer, acanthosis and lymphocytic infiltration (predominantly T cells) were seen in patients and snorers. The authors speculated that vibration trauma might have incited connective tissue inducers that are initiated by T cells and plasma cells.
It should be noted that nasal inflammation is also present, which is remote from the site of obstruction and snoring-induced trauma, in patients with OSAS. Previous studies in our laboratory indicate that the number of polymorphonuclear leukocytes and concentrations of bradykinin and vasoactive intestinal peptide are increased in the nasal lavage of patients with OSAS [9, 10]. Conceivably, nasal inflammation in the absence of clinical evidence of rhinitis and sinusitis could contribute to upper airway obstruction .
Increased Airway Collapsibility
Neutral endopeptidase, an ectoenzyme that cleaves and inactivates inflammatory peptides, such as bradykinin and substance P, which elicit interstitial edema, was significantly decreased in the uvula epithelium of patients with OSAS . Down-regulation of neutral endopeptidase would lead to increased concentration of bradykinin and substance P, which may amplify vasodilation and interstitial edema in the uvula mucosa. In turn, local vasodilation in the pharynx increases upper airway collapsibility probably due to edema-induced reduction in pharyngeal cross-sectional area .
Impairment of Upper Airway Reflexes
Upper airway anesthesia evokes obstructive apneas in healthy individuals  and prolongs the apnea duration in patients with OSAS . These findings underline the importance of mechanoreceptors and sensory nerves in the maintenance of upper airway patency. Friberg et al.  demonstrated abnormal varicose nerve endings invading the soft palatal epithelium in patients with sleep apnea and habitual snorers. Such sprouting of nerve endings appears to be a sign of afferent neuropathy also seen during wound healing. In this study, calcitonin gene-related peptide (CGRP) and substance P (SP), neurotransmitters assumed to be present in afferent nerve fibers of the C and A-δ type, were abundant by immunohistochemistry. Interestingly, substance P is also implicated in neurogenic plasma extravasation and inflammation of the skin .
In addition to the morphologic evidence, several investigators showed functional disturbances in sensory function. Kimoff et al.  studied two-point discrimination and vibration thresholds in the oropharynx of 37 patients with OSAS, 12 nonapneic snorers and 15 control subjects. Both sensory modalities were impaired in patients with OSAS and nonapneic snorers when compared to controls. Following nasal CPAP therapy for six months, significant improvement in both two-point discrimination and vibratory sensation thresholds were noted. In an earlier study, Larsson et al.  demonstrated impaired thermal sensitivity in the oropharynx of patients with OSAS. It has been argued that vibration injury from snoring may account for injury to sensory nerves. To this end, snoring results in a vibration frequency of 70 Hz, which is in the range associated with soft tissue damage .
Pro-inflammatory cytokines, such as interleukin-1 (IL-1) and tumor necrosis factor (TNF)-α, have been shown to modulate physiological sleep. Plasma IL-1β concentration is highest at the onset of sleep in humans suggesting a sleep wake cycle variation of this cytokine . Cloned IL-1 receptor antagonist transiently inhibits NREM sleep in rabbits . Compounds that promote IL-1β production such as muramyl peptides, endotoxin and TNF-α, enhance NREM sleep, while substances that inhibit its production, such as prostaglandin E2 and corticosteroids, inhibit sleep . Elevated plasma TNF-α levels are associated with enhanced slow wave amplitudes during sleep . In rabbits given a fragment of TNF-α soluble receptor, spontaneous NREM sleep was suppressed by 40% .
Vgontzas et al.  determined plasma levels of IL-1β, TNF-α and IL-6 in patients with OSAS, narcolepsy, and hypersomnia in comparison to control subjects. The concentration of TNF-α was significantly elevated in patients with OSAS and narcolepsy and correlated with the intensity of sleepiness measured as mean nap sleep latency. IL-6 levels were elevated only in OSAS and correlated with body mass index. Given that IL-6 and TNF-α have been associated with increased sleepiness and fatigue in other conditions, these findings suggest that the cardinal symptoms of OSAS may be mediated by both cytokines.
The same investigators also found that the concentrations of IL-6 and TNF-α were higher in obese men with OSAS compared to non-apneic obese men . A significant independent correlation was found between OSAS, high cytokine levels, amount of visceral fat and insulin resistance. They concluded that OSAS is associated with elevated plasma concentrations of IL-6 and TNF-α independent of obesity. In the aforementioned studies, blood was drawn either only in the morning  or in the morning and evening .
Postulating a circadian rhythm of cytokine release as a regulator of sleep, Entzian et al.  determined TNF-α, IL-1β, IL-6 and interferon-=γ release in ex-vivo short term culture of blood samples every four hours during the day and at 2-hour intervals during sleep in 10 patients with OSAS and 10 healthy volunteers. While the circadian rhythms of IL-1β, IL-6 and interferon-=γ were unaltered, there was a significant difference in TNF-α secretion in patients with OSAS; the normal nocturnal peak was absent and an additional daytime peak had developed. Following treatment with nasal CPAP for three months, repeat study of seven patients revealed persistence of the disturbance in the circadian rhythm of TNF-α release. The reason(s) underlying altered TNF-α release in these patients is unclear. Given that TNF-α has been shown to modulate somnolence and fatigue, it may play a role in the pathogenesis of OSAS.
Circulating TNF-α levels are elevated in inflammatory diseases that are associated with muscle weakness in laboratory animals [29, 30]. For instance, diaphragmatic contractility is significantly reduced in hamsters after exposure to TNF-α . When murine diaphragm and limb muscle preparations were incubated with TNF-α, maximum tetanic force diminished by approximately 20% in both types of muscle . Transgenic mice with cardiac restrictive overexpression of TNF-α develop heart failure and elevated plasma levels of TNF-α from cardiac spillover. Diaphragm muscle excised from these mice displayed greater intracellular oxidant levels and almost 50% less force generation compared with control mice . Muscle weakness was partially reversed with N-acetylcysteine suggesting oxidative stress as a mediator of contractile dysfunction. Taken together, these data suggest that increased circulating levels of TNF-α in OSAS may promote pharyngeal inspiratory muscle dysfunction thereby worsening apneic episodes during sleep.
Reactive oxygen and nitrogen species (ROS and RNS) are atoms or molecules with unpaired electron(s) at their outer orbits, which render them highly prone to chemical reactions. ROS such as superoxide anion (O2–), hydroxyl radical (OH–) and peroxynitrite (ONOO–) are injurious to cells mainly through a process called lipid peroxidation, which damages cell membranes and generates more ROS and RNS in a chain reaction. These compounds are also implicated in ischemia-reperfusion injury that occurs following repletion of oxygen or restoration of blood flow to tissues.
In this scheme of events, during reperfusion or reoxygenation, hypoxia-activated enzyme xanthine oxidase utilizes oxygen and hypoxanthine, a degradation product of ATP, to create oxidizing agents such as superoxide anion, hydrogen peroxide and uric acid . Reperfusion injury is operant in many disease states including myocardial infarction and stroke. The intermittent nature of nocturnal hypoxia in OSAS may predispose patients to such ischemia-reperfusion injury and its consequences, such as acute cardiac and cerebral ischemia .
Previous work in our laboratory showed that although granulocyte count in nasal lavage fluid of patients with OSAS was elevated in comparison to controls, phagocytosis of opsonized bacteria and oxidative burst were not different . In this study, nasal lavage fluid was obtained at various times during the day. We also measured levels of exhaled pentane, a product of lipid peroxidation due to ROS, in patients with OSAS . Exhaled nasal and oral pentane levels after sleep were significantly higher than pre-sleep values in patients with OSAS. Although there was no significant difference in pentane levels before sleep between control subjects and apneic patients, exhaled pentane levels were significantly higher in apneic patients upon awakening. Sahebjani et al.  showed that excretion of uric acid, a marker for ROS production, was increased and correlated significantly with apnea-hypopnea index in patients with OSAS. In this study, nasal CPAP therapy resulted in a reduction of uric acid excretion to normal levels.
Schultz et al.  measured ex vivo release of superoxide from stimulated polymorphonuclear neutrophils in patients with severe OSAS, normal volunteers and patients with lung cancer. Superoxide release was markedly increased in OSAS when compared to normals and lung cancer patients. Interestingly, application of nasal CPAP for two nights completely reversed this phenomenon implying intermittent hypoxia as the priming factor for the increased neutrophil burst. However, the investigators found no relationship between superoxide levels and degree of nocturnal arterial oxygen desaturation.
Reactive oxygen species are associated with activation of inflammatory cells in OSAS. Dyugovskaya et al.  found increased monocyte adhesion molecule expression and oxidative metabolism in patients with OSAS as compared with controls. After 6 months of nasal CPAP therapy, monocyte adhesion molecules CD15 and CD11c and oxidative metabolism were reduced to normal levels. A rebound in adhesion molecule expression and oxidative metabolism were observed when same subjects were studied after one night off nasal CPAP. When monocytes from control patients were incubated in hypoxic medium, their expression of adhesion molecules increased to the level of OSAS patients. Collectively, these findings suggest hypoxia as the stimulus for monocyte activation in OSAS.
A recent study by Vassilakopoulos et al.  may offer a link between increased levels of ROS and increased cytokine levels in patients with OSAS. They reported the effects of resistive breathing on plasma cytokine levels and monocyte function in healthy subjects. Resistive breathing induced early (during loading) increases in levels of IL-6 and IL-1β as well as late (30 min after loading) increase in TNF-α. After receiving high-dose oral antioxidants, the resistive breathing-induced TNF-α response was abolished, IL-1β levels were undetectable and IL-6 response was blunted. Modification of cytokine levels with antioxidants during resistive breathing suggested a role for ROS as the mediator for cytokine release. Intracellular flow cytometry did not show an increase in cytokine positive monocytes suggesting that resistive breathing-induced cytokine production was not monocyte-dependent. The authors proposed that respiratory muscles or pulmonary epithelial cells exposed to increased transpulmonary pressure during resistive breathing could be the source for hypercytokinemia. Given that patients with OSAS go through shorter but more frequent resistive breathing runs nocturnally, it is tempting to speculate that such intermittent loading may lead to ROS-mediated hypercytokinemia in addition to hypoxia-induced monocyte priming in these patients. Clearly, additional studies are indicated to support or refute this hypothesis.
C-reactive protein (CRP) is a sensitive marker for systemic inflammation. An elevated plasma level of this acute phase reactant indicates heightened activity of inflammation in humans. Presence of elevated CRP levels was also associated with increased risk of cardiovascular and cerebrovascular mortality. In a study of over 16,000 subjects, high body mass index was associated with an elevated CRP level . Post-hoc analysis excluding smokers and persons with inflammatory diseases, cardiovascular disease, and diabetes mellitus did not change this association. In a recent study which aimed to single out the confounding variable of obesity, Shamsuzzaman et al.  showed that CRP levels were higher in 22 patients with newly diagnosed OSAS when compared with 20 control subjects. The groups were matched for age, and importantly, body mass index (BMI). Furthermore, CRP levels were independently associated with severity of sleep-disordered breathing. Yokoe et al.  compared levels of CRP and IL-6 in 30 men with newly diagnosed OSAS against a control group of 14 obese men and also before and after nasal CPAP therapy. Monocyte production of IL-6, serum levels of IL-6 and CRP were elevated in patients with OSAS compared to obese control subjects. One month of nasal CPAP therapy was associated with marked reduction in serum levels and monocyte production of IL-6 and CRP. Adipose tissue produces IL-6 and BMI has been shown to positively correlate with IL-6 levels. In this study, BMI did not change during nasal CPAP therapy. Furthermore, IL-6 levels were independently associated with nocturnal hypoxia suggesting hypoxia-primed monocytes as the source of the rise. Elevation of CRP in OSAS may be secondary to IL-6 since the latter is a known inducer of the former . The aforementioned studies [42, 43] strongly suggest that systemic inflammatory state seen in patients with OSAS exists, at least in part, independent of obesity.
Leptin is an adipocyte-derived hormone that reduces appetite, increases energy expenditure and decreases body weight. Interestingly, most obese subjects have high levels of this hormone suggesting that obesity is a leptin-resistant state. Recently, Ip et al.  showed that patients with OSAS had significantly higher serum levels of leptin compared with BMI-matched controls. Notably, patients with OSAS also had more adverse vascular risk factor profiles such as dyslipidemia, higher diastolic blood pressure, insulin resistance and greater adiposity. Following treatment with nasal CPAP for six months, there was a significant reduction in leptin levels in the absence of any change in BMI. As stress and increased muscle sympathetic nerve activity are associated with elevated leptin levels, the authors speculated that the increased sympathetic nerve discharge associated with nocturnal arousals might account for the increase in leptin expression. TNF-α administration increases plasma leptin levels . Thus, elevated levels of TNF-α in OSAS may be another putative stimulus for hyperleptinemia. Leptin is a pleiotropic hormone that has immunomodulatory properties including enhancement of phagocytic function, stimulation of memory T-cell maturation and certain anti-inflammatory properties .
An expert panel of the World Health Organization  defined metabolic syndrome as glucose intolerance and/or insulin resistance plus two or more of the following: hypertension, elevated triglycerides and/or reduced HDL cholesterol, central obesity and microalbuminuria. Metabolic syndrome is strongly associated with increased cardiovascular mortality and morbidity. Centerpiece of the syndrome appears to be visceral fat accumulation in the mesentery and the omentum (central obesity), which is a better predictor of coronary artery disease than body mass index. Chin et al.  studied the relationship between intra-abdominal visceral fat and serum leptin levels in 22 patients with OSAS before and after treatment with nasal CPAP. They showed that nasal CPAP reduced leptin levels in a few days and the amount of visceral fat after six months of therapy. Interestingly, reduction of visceral fat during nasal CPAP therapy occurred without weight loss as well as with weight loss. Rapid reduction in leptin levels corroborates Ip et al.’s  findings and suggests that increased sympathetic nerve discharge during nocturnal arousals could account for the increase in leptin expression in OSAS. However, the mechanism of visceral fat reduction after nasal CPAP therapy is unknown. Nevertheless, reduction in visceral fat may explain, in part, the salutary cardiovascular effects of nasal CPAP therapy.
Summary and Future Directions
There is a growing body of clinical evidence that local and systemic inflammation is present in patients with OSAS. Inflammation may contribute to anatomic upper airway narrowing, abnormalities in upper airway reflexes, upper airway collapsibility and inspiratory pharyngeal muscle dysfunction. These unfavorable processes may increase the severity of OSAS setting up a vicious cycle. Activation of neutrophils and monocytes and elaboration of pro-inflammatory cytokines, TNF-α and IL-6 in particular, could play a role in the pathogenesis of OSAS. Symptoms of lethargy and excessive daytime hypersomnolence may be mediated, in part, by these processes. Although the sources of cytokine and chemokine production are not yet established, ROS and perhaps RNS priming of immune effector cells and subsequent elaboration of leukocyte and endothelial adhesion molecules may be the mediators of systemic inflammation. Local production of cytokines in the inflamed upper airway, respiratory muscles and pulmonary epithelial cells exposed to stress with ‘spillage’ into systemic circulation are other putative sources. Given the recent realization of the significance of inflammation in ischemic heart disease pathogenesis, it is possible that cardiovascular consequences of OSAS may, in part, be mediated by heightened systemic inflammatory state in OSAS (fig. 1).
Fig. 1. Interactions between inflammation, obstructive sleep apnea syndrome and cardiovascular complications.
Clearly, further research is indicated to elucidate the putative role of inflammation in the genesis and natural history of OSAS. Research efforts should target the possible link between upper airway and systemic inflammation and consequent vascular wall inflammation. Sources of cytokine production other than monocytes and inflammatory cellular trafficking need to be better defined in OSAS. Clinical trials investigating the effects of antioxidants on systemic inflammation and cardiovascular morbidity in OSAS are desirable. What, if any, role does the systemic inflammatory state of obesity play in triggering OSAS? Although inflammatory state appears to be ameliorated with nasal CPAP therapy, is the treatment of obesity (with its attendant inflammatory state) necessary to prevent recurrence of OSAS? Finally, the intriguing observations of reduction in visceral fat deposition and reversal of leptin resistance with nasal CPAP therapy deserve further study to delineate mechanisms.
Prof. Israel Rubinstein, MD
Department of Medicine (M/C 719), University of Illinois at Chicago
Building 911, Room 173, 840 South Wood Street
Chicago, IL 60616-7323 (USA)
Tel. +1 312 996 8039, Fax +1 312 996 4665, E-Mail IRubinst@uic.edu
Received: June 25, 2003
Accepted: September 7, 2003
Number of Print Pages : 7
Number of Figures : 1, Number of Tables : 0, Number of References : 48
Respiration (International Review of Thoracic Diseases)
Founded 1944 as ‘Schweizerische Zeitschrift für Tuberkulose und Pneumonologie’ by E. Bachmann, M. Gilbert, F. Häberlin, W. Löffler, P. Steiner and E. Uehlinger, continued 1962–1967 as ‘Medicina Thoracalis’ as of 1968 as ‘Respiration’, H. Herzog (1962–1997)
Official Journal of the European Association for Bronchology and Interventional Pulmonology
Vol. 70, No. 6, Year 2003 (Cover Date: November-December 2003)
Journal Editor: C.T. Bolliger, Cape Town
ISSN: 0025–7931 (print), 1423–0356 (Online)
For additional information: http://www.karger.com/journals/res