The Authors: Jean-Christian Borel, PhD, is the Head of the Research and Development Department at ‘AGIR à dom’, a non-profit home care provider. He has developed an expertise in the field of non-invasive ventilation, home-based training programmes and integrated care. He carried out his PhD thesis at the Hypoxia Pathophysiology Laboratory (HP2) on the clinical and cardiovascular consequences of the obesity hypoventilation syndrome, followed by a postdoctoral fellowship on the pathophysiology of the sleep apnoea syndrome in Dr Frederic Series' team at the Quebec Heart and Lung Institute.
Anne-Laure Borel, MD, PhD from Grenoble University Hospital, France, has completed a postdoctoral fellowship in Quebec, Canada in Dr Jean-Pierre Despres' team of the Quebec Heart and Lung Institute. She is an endocrinologist with a clinical activity focused on obesity and type 2 diabetes. Her research interests are in the field of the cardiometabolic complications of an excess visceral adiposity and more specifically, the association between sleep-disordered breathing, excess visceral fat and metabolic abnormalities.
Denis Monneret, MSc, PhD, is a pharmacist with a diploma of medical biology from Lille Faculty of Pharmacy, France. His MSc and PhD degrees specialized in biology of obstructive sleep apnoea and obesity hypoventilation syndrome. After working as Assistant Professor in the Biochemistry and Hormonology Unit at Grenoble University Hospital, he joined the HP2 Laboratory where he works on oxidative stress and metabolic biomarkers of obstructive sleep apnoea, and obesity hypoventilation syndrome.
Renaud Tamisier, MD, PhD, is an Associate Professor of Physiology at Grenoble University School of Medicine, France. He works in the HP2 Laboratory and has been involved in both research and clinical studies in sleep-disordered breathing for the last 15 years. His research interests are mainly on the cardiovascular consequences of intermittent hypoxia and obstructive sleep apnoea. In the last 15 years, he has published over 40 papers in the best international journals of the respiratory and sleep fields.
Jean-Louis Pépin, MD, PhD is Professor of Clinical Physiology and Head of the Clinical Physiology, Sleep and Exercise Department at Grenoble University Hospital. He is the President of the French Sleep Research Society. His clinical and research interests are mainly on the cardiovascular consequences associated with chronic and intermittent hypoxia, chronic respiratory failure and non-invasive ventilation.
Patrick Lévy MD, PhD is Professor of Physiology at Grenoble University, France. He is the Head of the HP2 Laboratory (Inserm Unit 1042) and is the Clinical Vice-President of the European Sleep Research Society. He has been involved in sleep-disordered breathing for the last 25 years, and his clinical and research interests are mainly on obstructive sleep apnoea. He leads one of the most active research teams on obstructive sleep apnoea and intermittent hypoxia, which published over 200 manuscripts in the best international journals of the respiratory, cardiovascular and sleep fields.
SERIES EDITOR: AMANDA J PIPER
Jean-Christian Borel, AGIR à dom, Département Recherche et Développement, 29-31 Bd Des Alpes, 38244Meylan Cedex, France. Email: email@example.com
Obesity hypoventilation syndrome (OHS) is defined as a combination of obesity (body mass index ≥ 30 kg/m2), daytime hypercapnia (partial arterial carbon dioxide concentration ≥45 mm Hg) and sleep-disordered breathing after ruling out other disorders that may cause alveolar hypoventilation. Through the prism of the International Classification of Functioning, OHS is a chronic condition associated with respiratory, metabolic, hormonal and cardiovascular impairments, leading to a decrease in daily life activities, a lack of social participation and high risk of hospitalization and death. Despite its severity, OHS is largely underdiagnosed and the health-related costs are higher than those of apnoeic or obese eucapnic patients. The present review discusses the definition, epidemiology, physiopathology and treatment modalities of OHS. Although nocturnal positive airway pressure therapies represent first-line treatment and are effective in improving patient outcomes, there is a need to offer combined treatment strategies and to assess the effect of multimodal therapeutic strategies on morbidity and mortality.
Although obesity, defined by a body mass index (BMI) ≥30 kg/m2, is associated with an increased rate of death from cardiovascular diseases and certain cancers,1 the mechanisms involved in these relationships are incompletely understood. Obstructive sleep apnoea syndrome is commonly associated with obesity2 and is a risk factor for cardiovascular morbidity.3 Beyond obstructive sleep apnoea syndrome, a particular subgroup of obese patients is affected by chronic respiratory failure, the so-called obesity hypoventilation syndrome (OHS). These patients are characterized by a greater morbi-mortality than obese apnoeic patients, although this condition is often underdiagnosed. The present work reports definition, epidemiology, physiopathology and treatment modalities of OHS.
OHS is defined as a combination of obesity (BMI ≥ 30 kg/m2), daytime hypercapnia (partial arterial carbon dioxide concentration ≥45 mm Hg) and various types of sleep-disordered breathing after ruling out other disorders that may cause alveolar hypoventilation (severe obstructive or restrictive pulmonary diseases, chest wall disorders, neuromuscular diseases, severe hypothyroidism, and congenital central hypoventilation syndrome).4 Seventy to 90% of patients with OHS also exhibit obstructive sleep apnoea syndrome,5,6 while 10–15% of sleep apnoea patients referred through sleep laboratories have diurnal hypercapnia and can be classified as OHS.7
Although the overall prevalence of OHS has never been directly assessed in the general population, it is currently estimated at 3.7/1000 persons in the US population, which is one of the most overweight in the world.8 OHS prevalence has been more frequently assessed in patients referred to sleep clinics with a potential diagnosis of sleep-disordered breathing,6,9 in patients already diagnosed with obstructive sleep apnoea (OSA),7,10–13 as well as in a cohort of obese hospitalized patients.14 The prevalence is estimated between 10% and 20% in patients referred to sleep laboratories and rises to 20–30% or more in patients already diagnosed with OSA, as well as in hospitalized obese adult patients. There is a dose–response relationship between obesity as expressed by BMI and OHS prevalence.6,7,14 The different estimates of prevalence observed in these studies could be partly explained by the inclusion of patients with concomitant chronic obstructive pulmonary disease12 or patients with unstable medical status.14 Over the past decade, the progression of obesity seems to have stabilized in the United States.15 However, extreme obesity among adults is still increasing.16,17 Thus, it could be anticipated that both prevalence and incidence of OHS will rise in the near future, and we can anticipate OHS to be a considerable health burden.
GENERAL HEALTH STATUS, SOCIOECONOMICAL CONSEQUENCES AND MORTALITY OF OHS
Among OHS patients admitted to hospital, only 20% had previously received the diagnosis of OHS.14 A large proportion of patients are diagnosed with OHS only when presenting with acute respiratory failure.11,18–20 This highlights that OHS is underdiagnosed or that the diagnosis is dramatically delayed. Yet, OHS patients are more likely to suffer from congestive heart failure,9 pulmonary hypertension5 and diabetes mellitus9,21 than obese eucapnic OSA patients. Moreover, the number of hospitalizations in the past few years,22 as well as the number of admissions in intensive care unit,14 are higher in newly diagnosed OHS patients than in eucapnic obese patients. As a consequence, the direct health costs (i.e. general practice services, hospital services, medications) per year of an OHS patient can reach twice that of an apnoeic patient23 or an obese eucapnic patient,22 and about six times that of an age, gender and socioeconomic status-matched control subject.23 Furthermore, labour income and employment rate are lower in OHS patients than in apnoeic patients or control subjects.23 The lack of social participation in OHS patients reflected by high unemployment rates is also in line with a poor social functioning score reported through questionnaire SF-36.24 OHS patients experience excessive daytime sleepiness9,24,25 that could be related to the severity of nocturnal rapid eye movement (REM) sleep hypoventilation (i.e. the higher the percentage of REM sleep spent in hypoventilation, the higher the excessive daytime sleepiness).25
Beyond the additional burden of comorbidities compared with eucapnic obese individuals, OHS patients have a higher risk of death, particularly when they are left untreated.14,19 In the study of Nowbar et al.,14 the mortality rate was 23%, 18 months after hospital discharge, compared with 9% in eupnoeic obese patients. OHS patients had a significantly higher hazard ratio for mortality (hazard ratio = 4.0) after adjustment for age, gender, BMI, electrolyte abnormalities, renal insufficiency, history of thromboembolism and history of hypothyroidism. In observational cohorts, mortality rate was reduced when OHS patients were treated with non-invasive ventilation (NIV),18,26–28 but it may remain higher than long-term mortality rates observed in large cohorts of obese patients.29
Because most OHS patients in a stable state are diagnosed when they present to sleep clinics, most studies looking for determinants of hypoventilation in obesity have consisted of cohorts of sleep apnoeic patients.7,9,10,13,30,31 In a recent meta-analysis, Kaw et al.31 have shown that daytime hypoventilation in obese sleep apnoeic patients was associated with severity of sleep apnoea, along with impaired chest wall mechanics and severity of obesity. These findings highlight that daytime hypoventilation in obese patients results from the conjunction of several factors that interact to overwhelm the compensatory mechanisms of CO2 homeostasis (Fig. 1).32,33
A decrease in pulmonary volumes (vital capacity, total lung capacity, residual functional capacity) has been reported as a determinant of daytime hypercapnia in obese patients.6,7,12–14,30 This respiratory restriction, more pronounced in hypercapnic than in eucapnic obese patients,6,21,31,34,35 is associated with decreases in total respiratory system compliance and lung compliance;36 moreover, expiratory flow limitation in awake subjects can occur at low lung volume.37,38 Taken together, these alterations increase the work of breathing.34,39,40
Although mechanisms underlying this pulmonary restriction are incompletely understood, the decrease in lung volume is usually related to fat mass41 and to intrathoracic and abdominal fat distribution.41,42 These fat deposits could have direct mechanical effects on respiratory function by impeding diaphragm motion, changing the balance of elastic recoil between chest and lung. Another possible mechanism may be related to the metabolic syndrome and the low-grade inflammation associated with excess visceral and intrathoracic adiposity. Lin et al.43 have shown, in a large cohort study of more than 46 000 subjects without previously known respiratory disease, that metabolic syndrome was associated with a higher risk of restrictive lung impairment after adjustment for confounders (age, gender, BMI, physical activity, alcohol consumption). Additionally, other studies have reported an association between central adiposity and pulmonary restriction,41,44 and lower muscle strength in extra-respiratory muscles, as demonstrated by handgrip.45 Similarly, we have observed that OHS patients exhibit lower lung function, higher low-grade inflammation21 and higher waist/hip ratio (an anthropometric marker of central adiposity) than obese eucapnic patients. The low-grade inflammation associated with metabolic syndrome and the higher work of breathing46 could induce specific muscle impairment in OHS leading to a deterioration in pulmonary function.
Respiratory muscle function
Although studies have shown that weight loss improves strength/endurance of respiratory muscles47,48 and sometimes hypoventilation,49 the question of inspiratory muscles weakness in OHS remains open. OHS patients are able to temporarily correct hypoventilation with voluntary breathing (i.e. hyperventilation).12 Additionally, during spontaneous breathing34 as well as during hypercapnic-induced hyperventilation,50 transdiaphragmatic pressure has been demonstrated higher or at least equivalent in OHS patients compared with obese eucapnic patients.34,50 Moreover, we recently observed that OHS patients exhibited somatotropic axis impairment compared with eucapnic obese patients. Insulin-like growth factor-I levels were inversely related to partial arterial carbon dioxide concentration levels and to vital capacity, which is an indirect marker of diaphragmatic strength.51 Taken together, these results suggest that muscle function is slightly impaired and does not allow compensation for the overload imposed on the respiratory system by OHS. The underlying causes for respiratory muscle dysfunction that may coexist in OHS include inflammation and somatotopic axis impairment. Further studies are needed to explore in detail respiratory muscle structure and function in OHS.
Sleep-disordered breathing encountered in OHS patients
Obstructive sleep apnoea syndrome
It is largely recognized that most OHS patients suffer from an obstructive sleep apnoea syndrome52 (Fig. 2) that could explain or at least contribute to daytime hypoventilation.31 The pathophysiology of sleep apnoea is multifactorial and has been discussed in detail by Isono53 in a previous paper in this series. In brief, upper airway closure occurs when pharyngeal dilating forces cannot overcome the collapsing effect of the negative transmural inspiratory pressure gradient and the tissue weight.54 Excessive fat deposition induces an enlargement of soft tissue surrounding the upper airway, compromising the pharyngeal airspace55 and therefore predisposing the airway to closure during sleep.54 Moreover, the reduced lung volume, particularly present in OHS patients,31 reduces the inspiratory-related caudal tracheal traction56 that stabilizes upper airway structures.54 Beyond these anatomical predispositions to sleep apnoea, it has recently been proposed that fluid shifts from the legs to the neck during sleep (because of the recumbent position), contributing to the pathogenesis of sleep apnoea.57,58 Although this hypothesis remains debated,59 Redolfi et al. have shown that prevention in daytime fluid accumulation in the legs reduces apnoea/hypopnoea index.60,61 As OHS patients are likely to exhibit cor pulmonale22 and consequently fluid overload and peripheral oedema, their upper airway collapsibility might be increased by nocturnal rostral fluid shift. To test the importance of this phenomenon, the severity of sleep apnoea and the ratio of central versus obstructive events from acute exacerbation to a chronic stable condition remain to be studied.
The pattern of sleep apnoea is a key in the development of hypercapnia in OHS patients. The mean duration of apnoea and hypopnoea is increased with a reduced duration and amplitude of interapnoea ventilation.62 This insufficient post-event ventilatory compensation63 may contribute to the interaction between sleep apnoea syndrome and diurnal hypoventilation. Accordingly, the severity of sleep apnoea syndrome—appreciated by overnight apnoea/hypopnoea index—contributes to diurnal hypoventilation in obese apnoeic patients.31
Another type of sleep respiratory abnormality encountered in patients with OHS is central hypoventilation (Fig. 3) that is characterized by a sustained reduction in ventilation associated with a constant or reduced respiratory drive.25,64 Hypoventilation is more pronounced during REM sleep64 and the proportion of REM sleep hypoventilation is associated to awake ventilatory response to CO2. Indeed the lower awake CO2 ventilatory response, the higher the percentage of REM sleep spent in hypoventilation.25
There is increasing evidence that sleep-breathing disorders contribute to the pathogenesis of diurnal hypoventilation via alteration of ventilatory drive.65
Norman et al.65 have recently proposed a computational model that unifies acute hypercapnia during sleep-breathing events with chronic sustained hypoventilation during wakefulness. In this model, the persistence of elevated bicarbonate concentration after the sleep period (characterized by repetitive acute hypercapnia events), possibly due to either alteration of renal excretion or reduction in ventilatory drive (or both), further blunts ventilatory response to CO2, which in turn increases partial arterial carbon dioxide concentration. In a recent study, we observed that the improvement in bicarbonate concentration after 1 month of nocturnal NIV was related to NIV-induced nocturnal oxygen blood saturation improvement.66 This result supports the hypothesis of a link between nocturnal breathing disorders and daytime hypoventilation (Fig. 2b). Nevertheless, although diurnal partial arterial carbon dioxide concentration and bicarbonate concentration were improved after 1-month NIV, the ventilatory response to CO2 remained unchanged, suggesting that other factors may play a role in respiratory drive impairment in OHS.
The hypothesis of familial impairment of respiratory drive has been evoked, but Jokic et al. have shown that ventilatory chemoresponsiveness was similar between first-degree relatives of OHS patients compared with age- and BMI-matched control subjects.67 Another hypothesis, supported by animal studies, is the participation of neurohumoral agents such as adipokine leptin that acts as a powerful ventilatory stimulant.68,69 In humans, congenital leptin deficiency exists but is very rare;70 on the contrary, in obese subjects circulating leptin levels are often higher than those in lean subjects.71–75 These results suggest that a central resistance to leptin may occur in obesity. A possible mechanism for this central resistance could be deficient leptin transport through the blood–brain barrier.76,77
Inflammation, insulin resistance and somatotropic impairment in OHS
Obesity is a disease state characterized by chronic systemic low-grade inflammation and associated inflammatory changes in the adipose tissue.78 Inflammatory status in OHS patients has not been extensively studied, but Budweiser et al. found that C-reactive protein, a systemic biomarker of inflammation, was associated with poor survival.26 We recently observed that moderate OHS patients had a higher level of high-sensitivity C-reactive protein, a higher level of the proatherogenic chemokine regulated on activation, normal T cell expressed and secreted and a lower adiponectin level (an antiatherogenic and insulin-sensitizing adipokine) compared with age- and BMI-matched eucapnic control subjects.21 Accordingly, OHS patients exhibited higher insulin resistance, higher glycated Hb and were more likely to be treated by glucose-lowering medications. Endothelial dysfunction, an early key event of atherosclerosis and a strong predictor of incident cardiovascular events,79 was also more impaired in OHS patients compared with eucapnic obese patients.21 Additionally, the impairment in somatotropic axis, as demonstrated by low insulin-like growth factor-I level,51 may contribute to this endothelial dysfunction.80 Taken together, these results support a particular cardiovascular risk associated with OHS and confirm the results of observational cohorts data, demonstrating a higher prevalence of cardiovascular diseases.9,22,26
Through the prism of the International Classification of Functioning81 (Fig. 1), OHS is a chronic condition associated with impairments of body structures (upper airway, respiratory muscles) or body functions (control of breathing, sleep, cardiovascular, metabolic), decrease in activities and a lack of social participation. Thus, in the absence of guidelines on treatment modalities being clearly established in OHS, clinicians should adapt treatment modalities aimed at improving specific impairments, dysfunctions and handicaps of each patient. From this perspective, the three main tools available for clinicians are positive airway pressure therapies (continuous positive airway pressure (CPAP) and/or NIV), body weight loss strategies and rehabilitation.
Positive airway pressure therapies to abolish sleep-breathing disorders
Positive airway pressure therapies (CPAP and/or NIV) represent the first-line therapy for sleep-breathing disorders. The choice of nocturnal CPAP or NIV is often based on the underlying sleep-related respiratory abnormality encountered.82 CPAP is efficient in reversing diurnal hypoventilation82–85 in OHS patients presenting mainly OSA-related hypoventilation. Piper et al.85 have shown that more than 60% of OHS patients who exhibited an incomplete initial response to CPAP have a good response after 3 months of CPAP, suggesting that incomplete immediate response to CPAP does not preclude midterm efficacy. However, some OHS patients continue to exhibit sleep hypoventilation,82,83,85,86 particularly in REM,25,35 even though upper airway obstruction is abolished with CPAP, and require NIV to overcome this central hypoventilation. In contrast to CPAP, NIV is more expensive and more complex to set up a large range of ventilatory parameters (ventilatory mode, inspiratory and expiratory pressures, back-up respiratory rate, triggering sensitivity, pressurization slope).87 Moreover, residual respiratory events occur frequently in OHS patients treated with nocturnal NIV. Additional attention is required to ensure treatment efficacy and avoid machine-related undesirable respiratory events.88,89 Thus, in the absence of long-term randomized trials comparing NIV with CPAP in patients with predominant OSA-related hypoventilation,85 NIV should be reserved for OHS patients presenting with primarily central hypoventilation during sleep. Manufacturers have proposed new ventilatory modes (example: average volume-assured pressure support, volume targeting by bi-level positive pressure ventilation90,91 and auto-titrating NIV92), but the efficacy of these modes needs to be assessed in larger cohort studies and on a long-term follow-up basis.
Long-term uncontrolled studies have consistently reported a lower mortality rate in OHS patients treated with NIV26–28 compared with those in a cohort principally composed of untreated patients.14 Additionally, these studies have shown an improvement in gas exchange.26,28 Beyond hypoventilation, basal atelectasis or coexisting lung disease may lead to obesity-related hypoxia. Therefore, additional oxygen therapy may be necessary to correct persistent hypoxaemia even under efficient positive airway pressure therapy. Of note, the need for supplemental oxygen therapy and baseline hypoxaemia have been shown as independent predictors of mortality.26–28 In a recent randomized controlled trial including moderate OHS patient, partial arterial oxygen concentration was not significantly improved after 1 month of nocturnal NIV.66 Even though baseline hypoxaemia was very mild, this result supports the concept that obesity-related hypoxaemia might be more difficult to correct than hypercapnia. Additionally, this randomized controlled trial failed to demonstrate that short-term nocturnal NIV could impact favourably on metabolic, cardiovascular or inflammatory markers in spite of the improvements in pulmonary function and sleep parameters. This suggests that positive airway pressure therapies should not be the only form of treatment for these patients.66
Effect of body weight loss, changes in lifestyle habits and rehabilitation programmes
Previous studies have largely documented the benefits of body weight loss to correct or attenuate sleep apnoea related to obesity. Indeed, Peppard et al.93 have studied a cohort of 690 subjects and demonstrated that 10% weight gain or weight loss were respectively associated with a 32% increase or a 26% decrease in the apnoea/hypopnoea index. Intervention studies designed to allow weight loss in obese patients have largely shown that weight loss achieved either by bariatric surgery (Roux-en-Y gastric bypass, vertical-banded gastroplasty)94,95 or by lifestyle intervention96–99 was associated with an improvement in sleep-related breathing disorders.
Body weight loss obviously seems to be the etiological treatment in OHS. The correction of OHS related to weight loss following bariatric surgery has been addressed in some studies that have reported an improvement in diurnal hypoventilation, as well as an improvement in spirometric parameters.49,100–102 Whereas bariatric surgery seems to be an interesting option for treatment of OHS, two points have to be considered: first, a multicentre study identified sleep apnoea syndrome as an independent factor associated with an increased risk of major adverse outcome during the 30 days following bariatric surgery in a large cohort of patients;103 and second, recurrent sleep-disordered breathing, as well as deterioration of blood gas and spirometric parameters,102 were observed in a follow-up of bariatric surgery patients, even though these patients did not regain weight.104
The level of physical activity is another characteristic of lifestyle habits that is associated with sleep-disordered breathing and may be corrected by lifestyle intervention programmes. Indeed, the level of physical activity is inversely correlated to the apnoea/hypopnoea index.93 In addition, an increase in physical activity is also associated with a decrease in visceral fat accumulation,105 and with an improvement in low-grade inflammation. Physical activity also improves metabolic profile and muscle function. Therefore, OHS patients may benefit from an increase in physical activity, as well as from weight loss, acting synergistically to improve the respiratory disturbances and comorbidities of OHS.
Compared with obese eucapnic OSA patients, individuals with OHS exhibit a lower exercise tolerance.106 In contrast, in a study of patients with a range of disorders causing chronic respiratory failure and treated with long-term NIV, Budweiser et al.107 have shown that OHS patients had a relatively preserved functional capacity (i.e. 6MWD test) compared with patients with chronic obstructive pulmonary disease or thoracic restriction. Moreover, among these different causes of respiratory failure, exercise tolerance was improved after 2 months of nocturnal NIV.108 However, in OHS patients, in spite of correction of partial arterial carbon dioxide concentration after 3 months of NIV, these patients continued to exhibit exercise-induced hypoventilation.106 In a recent study aimed at comparing a low-calorie diet and a physical activity programme plus respiratory muscles training versus the same programme without respiratory muscle training, the former strategy induced greater improvement in dyspnoea and exercise capacity than the latter one.109 On the other hand, NIV during exercise training might also be an aid to enhance exercise capacity in rehabilitation programmes.110 Taken together, these results suggest that specific exercise-based rehabilitation programmes including NIV and respiratory muscle training are likely to benefit these patients, and the onset of NIV could be the appropriate time to start such programmes. Nevertheless, the best modalities to include in these programmes have to be determined in order to improve motivation,111 adherence and long-term benefits.
CONCLUSION AND PERSPECTIVES
Although largely underdiagnosed, OHS is characterized by sleep, respiratory, metabolic and cardiovascular impairments leading to a decrease in activities, a lack of social participation and a higher mortality rate. Although CPAP/NIV represent first-line therapy to treat sleep-related breathing disorders, they cannot be considered as the only therapeutic strategy. Weight loss strategies (surgical and non-surgical), as well as changes in lifestyle habits and rehabilitation programmes, should be proposed. Finally, long-term, large-scale multicentre studies are required to assess the effect of multimodal therapeutic strategies on morbidity and mortality.
The authors thank Suzanne Lamothe for language revision.