Sleep-related breathing disorders in COPD

INTRODUCTION — Chronic obstructive pulmonary disease (COPD) is frequently associated with sleep-related breathing disorders (SRBD), including sleep-related hypoxemia, obstructive sleep apnea, central sleep apnea, respiratory effort-related arousals (RERAs), and sleep-related hypoventilation. These SRBDs may be associated with nonrestorative sleep and daytime sleepiness and fatigue [1-10].

The various forms of SRBD in COPD will be reviewed here, with special emphasis on (1) the diagnostic procedures required to detect the COPD-specific SRBD and (2) specific treatment options for SRBD in patients with COPD. A description of SRBD and the evaluation, diagnosis, and treatment of COPD and obstructive and central sleep apnea occurring independently are discussed separately. (See "Chronic obstructive pulmonary disease: Diagnosis and staging" and "Stable COPD: Initial pharmacologic management" and "Polysomnography in the evaluation of sleep-disordered breathing in adults" and "Central sleep apnea: Risk factors, clinical presentation, and diagnosis" and "Central sleep apnea: Treatment" and "Clinical presentation and diagnosis of obstructive sleep apnea in adults".)

EPIDEMIOLOGY — Sleep-related breathing disorders (SRBD) are common in patients with COPD [1], occurring in approximately 30 to 40 percent of patients [11-13].

There are four major domains of sleep-disordered breathing in COPD patients.

●Sleep-related hypoxemia – The prevalence of sleep-related hypoxemia (also called nocturnal hypoxemia) increases along with the severity of COPD [2]. Sleep-related hypoxemia is often associated with daytime hypoxemia, but not always. Isolated sleep-related hypoxemia (fall in partial pressure of arterial oxygen [PaO₂] of >10 mmHg or pulse oxygen saturation [SpO₂] below 88 percent for more than five minutes) has been reported in up to 70 percent of COPD patients, and can occur in patients with daytime oxygen saturation of 90 to 95 percent [3,14].

●Coexisting obstructive sleep apnea – Moderate to severe obstructive sleep apnea (OSA) may be present in 10 to 30 percent of patients with COPD, a prevalence that is similar to that of the general population [1-3,15].

The coexistence of OSA and COPD has been referred to as the "overlap syndrome" [16], although it is unclear whether it is truly an overlap syndrome or just two common diseases presenting together.

●Hypoventilation during sleep – Sleep-related hypoventilation is defined as a greater than normal increase in arterial tension of carbon dioxide (PaCO₂) during sleep (PaCO₂ of >50 mmHg or rise in PaCO₂ of >10 mmHg for more than 10 minutes compared to wakefulness). The prevalence of sleep-related hypoventilation is approximately 43 percent in hypercapnic patients with COPD but is lower (approximately 6 to 10 percent) and often limited to rapid eye movement (REM) sleep in patients with mild COPD and normal daytime arterial blood gases [1,17].

●Respiratory effort related arousals – Respiratory effort related arousals [RERAs], also described as sleep fragmentation due to worsening of breathing mechanics, can be seen in COPD (see 'Pathophysiology' below). The prevalence, however, remains unclear [1,18,19], and the clinical significance may be limited to patients with advanced stages of COPD with greater degrees of hyperinflation and muscle weakness [20]. RERAs are accounted for in summary measures of OSA severity and are generally considered to be on the pathophysiologic spectrum of OSA. However, COPD-specific alterations in breathing due to lower airway abnormalities may also contribute to RERAs in patients with COPD. (See "Polysomnography in the evaluation of sleep-disordered breathing in adults", section on 'Respiratory effort-related arousals'.)

●Others – A meta-analysis of 60 studies in patients with COPD and sleep disorders reported a prevalence of 22 percent for restless leg syndrome and 29 percent for insomnia [13].

PATHOPHYSIOLOGY — While the exact pathogenesis of nocturnal hypoxemia in COPD is unclear, normal sleep-related decline in ventilatory motor output exacerbates ventilation-perfusion mismatch and hypoventilation in individuals with COPD and can cause a marked deterioration in gas transfer, particularly during rapid eye movement (REM) sleep [17,21,22]. In addition, COPD can lead to elevations in closing volume, which means that the small airways in the lung are closed during tidal volume breathing and not participating fully in gas exchange. The increase in closing volume in combination with a reduction in functional residual capacity during sleep can exacerbate ventilation-perfusion mismatch and hypoxemia [23].

Sleep is associated with decreased hypoxic and hypercapnic ventilatory responses [24]. This sleep-related ventilatory depression leads to modest elevations in arterial tension of carbon dioxide (PaCO₂) in normal individuals compared with awake values [24]. In addition, sleep is associated with a loss of drive to upper airway dilator muscles [25,26], leading to upper airway narrowing and subsequent inspiratory airflow limitation in normal non-snoring and snoring individuals without sleep apnea [27]. Reductions in inspiratory airflow can further compromise ventilation and increase PaCO₂. (See "Pathophysiology of upper airway obstruction in obstructive sleep apnea in adults".)

COPD is associated with alterations in respiratory mechanics and gas exchange that can lead to dynamic hyperinflation and/or hypoventilation [28]. During wakefulness, patients adopt specific mechanisms to minimize hyperinflation (by prolonging expiratory time) and to preserve alveolar ventilation (with pursed lip breathing) [29-31]. Sleep is associated with several changes in respiratory control that interfere with the compensatory mechanisms used during wakefulness. In COPD, reductions in ventilation during sleep can produce marked elevations in PaCO₂ because of concomitant elevations in dead space ventilation [32]. Mechanisms to minimize hypoventilation will exacerbate dynamic hyperinflation, whereas lengthening expiratory time could compromise ventilation during sleep.

●A loss of ventilatory drive leads to significant reductions in tidal volume [33], and compensatory mechanisms to prevent hypoventilation are left to changes in respiratory pattern. An increase in inspiratory time would increase tidal volume but would reduce expiratory time, potentially exacerbating air trapping. Similarly, an increase in respiratory rate would maintain minute ventilation but also reduce expiratory time. In either case, respiratory patterns during sleep in COPD can worsen dynamic hyperinflation by shortening expiratory time [34].

●Inspiratory flow limitation (IFL) is commonly observed during sleep and even mild degrees of IFL lengthen inspiratory time, further comprising expiratory time [35].

Thus, sleep induces specific alterations in breathing pattern that compromise ventilation and increase ventilatory loads in COPD. It has been shown that hyperinflation is associated with reductions in sleep efficiency [20]. Moreover, the mechanism by which these changes contribute to sleep fragmentation and morning fatigue in COPD is likely due to increased arousal frequency (RERAs) and is under clinical investigation [36].

Obstructive sleep apnea (OSA) is characterized by increased collapsibility of the upper airway leading to inspiratory flow limitation, manifest by snoring, hypopneas, and apneas. The interrelationship of COPD and OSA is complicated, as the manifestations of COPD can vary among individual patients. As an example, some patients have a more emphysematous presentation and others have more airways disease (chronic bronchitis) with less air trapping. Increasing lung volumes due to air trapping are associated with improvements in upper airway collapsibility [34], which may explain why COPD patients with emphysema (pink puffer) typically do not present with concomitant OSA while obese COPD patients (blue bloater) exhibit significant OSA more often. Regardless, the coexistence of OSA and COPD has been associated with more pronounced hypoxemia and hypercapnia compared with COPD or OSA individually.

CONTRIBUTING FACTORS — The majority of patients with COPD have multiple comorbid conditions that are associated with an increased risk of sleep-related breathing disorders, such as metabolic syndrome (approximately 25 to 50 percent), smoking, cardiovascular diseases, and opioid use [37,38].

Obesity — The prevalence of obesity (body mass index [BMI] ≥30 kg/m²) in COPD is increasing, ranging from 25 to 42 percent [38]. In patients with COPD and obesity, respiratory disturbances and blood gas abnormalities (hypoxemia and hypercapnia) first occur during sleep, when ventilation responsiveness to chemical stimuli is blunted and the wakefulness drive is reduced. In addition, decreases in accessory muscle activity, particularly during rapid eye movement (REM) sleep, place an additional burden on the diaphragm, which is already under mechanical disadvantage.

Moreover, obesity is the strongest predictor for having coexisting obstructive sleep apnea (OSA). A BMI >30 kg/m² increases the probability of having OSA to more than 50 percent in males and approximately 20 to 30 percent in females.

In patients with COPD and obesity, positive airway pressure may not only mitigate OSA but also offset the mechanical disadvantage in obese COPD patients [37].

Smoking — Current smokers have a greater likelihood of snoring and SRBD relative to never smokers [39]. Inhalation of nicotine in cigarette smoke and withdrawal from nicotine during smoking cessation can also cause sleep disruption [40,41]. Conversely, oxyhemoglobin desaturation may be obscured by the presence of carboxyhemoglobin, which shifts the oxyhemoglobin dissociation curve to the left [42].

Heart failure — Central sleep apnea is not commonly seen in patients with COPD but can develop in the presence of heart failure or other cardiovascular comorbidities. The primary predisposing condition for central sleep apnea is heart failure with a reduction in cardiac output, leading to Cheyne Stokes type of central apnea. The presence of heart failure with reduced ejection fraction (below 40 percent) increases likelihood of having central sleep apnea to more than 50 percent. (See "Sleep-disordered breathing in heart failure".)

Opioid use — Chronic opioid use is an increasingly recognized risk factor for SRBD. Central sleep apnea is the most common form, although obstructive and mixed patterns also occur. Central sleep apnea related to opioid use is often marked by irregular respiratory patterns, including ataxic breathing. (See "Sleep-disordered breathing in patients chronically using opioids".)

CLINICAL FEATURES — The symptoms and signs of SRBD in patients with COPD are essentially the same as for patients without COPD (table 1). However, patients may be asymptomatic or their symptoms may overlap with the features of COPD itself and are thus not reported.

A large study demonstrated that nighttime symptoms in COPD patients frequently go unnoticed by physicians and may not be reported by patients themselves [6,43,44]. Thus, a specific sleep history should be sought from both the patient and the bed partner because, in many cases, it is only the bed partner who is aware of the abnormal ventilatory pattern.

The symptoms and signs of SRBD in patients with COPD include snoring, insomnia, awakening with a sensation of gasping or choking, morning headaches, daytime sleepiness or fatigue, and poor concentration or memory impairment (table 1). Additional features that may suggest SRBD include a large neck circumference, shallow or "crowded airway," and obesity.

Patients with severe COPD typically have disrupted sleep and often have desaturations during non-rapid eye movement (NREM) sleep that are more pronounced during rapid eye movement (REM) sleep [45]. In a population-based study, nocturnal oxygen desaturation was more common among those whose forced expiratory volume in one second/forced vital capacity (FEV₁/FVC) ratio was <0.65 [2].

While patients with acute hypercapnia may develop symptoms of hypercapnia (eg, anxiety, dyspnea, daytime sluggishness, morning headaches, hypersomnolence, confusion) when the arterial tension of carbon dioxide (PaCO₂) is above 75 mmHg, patients with chronic hypercapnia may not develop symptoms until the PaCO₂ rises above 90 mmHg. (See "Mechanisms, causes, and effects of hypercapnia", section on 'Effects of hypercapnia'.)

DIAGNOSIS — The diagnosis of SRBD in patients with COPD requires a high index of suspicion and formal sleep testing with polysomnography, since clinical features and questionnaires lack sensitivity and often overlap with symptoms of COPD.

Detecting SRBD is challenging in patients with COPD due to the following:

●The primary sleep problems reported by patients are difficulties initiating and maintaining sleep and reduced sleep time, and these often supersede other symptoms of SRBD such as unrefreshed sleep, waking up gasping and choking, morning headaches, and increased daytime sleepiness [1].

●Nocturnal hypoxemia, while commonly observed in COPD patients with borderline daytime oxygen levels of 90 to 95 percent, does not produce specific symptoms [3] and often remains undetected.

●Standardized questionnaires for screening and determining probability of obstructive sleep apnea (OSA), such as the Epworth Sleepiness Scale and the Sleep Apnea Clinical Score, have limited utility and are not validated for use in COPD patients [46].

When should SRBD be suspected in COPD? — The following criteria have been recommended as indications for evaluation for possible sleep-related breathing disorders in COPD [47].

●Patients with COPD who report typical symptoms suggestive of OSA such as snoring (especially expiratory), gasping and choking, morning headaches, or increased daytime sleepiness [48]

●Obesity (body mass index [BMI] >30 in males, 40 in females) with or without specific symptoms of OSA

●Reduced daytime pulse oxygen saturation (below 93 percent) at rest or during exercise

●Daytime hypercapnia

●Neck circumference >43 cm (17 inches) in males and >41 cm (16 inches) in females

●Signs of pulmonary hypertension or right heart failure

●Polycythemia

●Morning headaches, especially in response to oxygen therapy

●Patients who use drugs that are known to affect breathing such as opioids and hypnotic medications

●Comorbidities known to be associated with OSA (eg, atrial fibrillation, diabetes type 2, end-stage kidney disease, heart failure, stroke, systemic arterial hypertension)

Overnight pulse oximetry — For asymptomatic patients with a resting awake pulse oxygen saturation (SpO₂) of 90 to 95 percent, overnight oximetry at home may be an appropriate initial step as a screening device for transient or sustained nocturnal hypoxemia and certain types of SRBD [3,49,50].

Among patients with sleep-related desaturation detected by overnight oximetry, approximately half have OSA, and an in-laboratory polysomnogram (PSG) or home sleep apnea test (HSAT) is needed to identify these patients (algorithm 1). However, overnight oximetry when measured alone is not recommended as a screening test for OSA. (See "Home sleep apnea testing for obstructive sleep apnea in adults", section on 'Pulse oximetry'.)

The reliability of overnight pulse oximetry in COPD was found to be consistent in identifying nocturnal hypoxemia in one study [50] that found transient or sustained hypoxemia without evidence of sleep apnea in a significant proportion (38 percent) of patients with moderate to severe COPD (mean forced expiratory volume in one second [FEV₁] 37 percent) who did not qualify for home oxygen therapy based on their daytime partial pressure of arterial oxygen (PaO₂) of 56 to 69 mmHg [50]. In contrast, another study found that significant variability between two overnight oximetry studies, such that 35 percent of patients had a different result on the second night [51]. Thus, more than one night of observation may be appropriate to adequately evaluate nocturnal hypoxemia.

Home sleep apnea test — HSAT is widely used for the diagnosis of OSA in otherwise healthy individuals and has the potential advantage of more rapid implementation of treatment with positive airway pressure. Improvements in HSAT technology, including devices that monitor sleep stages, respiratory abnormalities during sleep, and oximetry, and supervision of HSAT by a board-certified sleep physician have led to greater acceptability of HSAT in patients with mild-to-moderate COPD and suspected OSA (algorithm 1) [52]. Patient selection, features of acceptable devices, and use of HSAT for the diagnosis of OSA are discussed separately. (See "Home sleep apnea testing for obstructive sleep apnea in adults".)

The 2017 guidelines of the American Academy of Sleep Medicine recommend a step-wise approach that includes in-lab studies when a single overnight screening has not detected (ruled in) either significant sleep apnea or nocturnal hypoxemia. Rather than repeating HSAT, they recommend moving one step up to a full polysomnography study in the sleep lab [52].

One difficulty is determining the severity of COPD that would make HSAT inappropriate, and the American Academy of Sleep Medicine Guidelines (AASM) do not provide a specific threshold re COPD severity. One study of 72 patients with stable COPD Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage 2 or 3 and symptoms of OSA found a high rate of recording failures with HSAT and variable correlation with in-laboratory PSG [53]. These results suggest that HSAT is less likely to be helpful in patients with COPD GOLD stage 2 or higher. In addition, HSAT does not include sufficient monitoring to detect the full range of COPD-specific SRBD, such as hypoventilation without severe hypoxemia (requires CO₂ monitoring devices) or periods of inspiratory and expiratory airflow limitation leading to frequent arousal [47].

If HSAT is nondiagnostic, guidelines suggest an in-laboratory PSG, rather than a repeat HSAT. In-laboratory PSG is still preferred over HSAT for patients with moderate-to-severe cardiorespiratory disease [52].

In-laboratory polysomnography — In-laboratory PSG is the gold standard for the diagnosis of SRBD and is able to detect the full range of COPD-specific SRBD, such as hypoventilation without severe hypoxemia (using CO₂ monitoring devices), central sleep apnea, and periods of inspiratory and expiratory airflow limitation leading to frequent arousal [47]. An approach to choosing between HSAT and in-laboratory PSG is provided in the algorithm (algorithm 1). (See "Polysomnography in the evaluation of sleep-disordered breathing in adults".)

Indications for an in-laboratory study rather than HSAT include the following [52]:

●Suspicion of nonrespiratory sleep disorders (eg, parasomnia, severe insomnia, sleep-related movement disorder)

●Comorbid conditions that increase the likelihood of nonobstructive sleep-disordered breathing (eg, significant cardiorespiratory disease, potential respiratory muscle weakness, severe insomnia, and history of stroke or chronic opiate use). As noted above, significant COPD would generally include those with GOLD stage 2 or higher and COPD associated with hypercapnia or resting SpO₂ <89 percent, although these thresholds have not been clearly defined. (See 'Home sleep apnea test' above.)

Full-night study (preferred) — For patients with symptoms or signs of a SRBD, the preferred test is a standard in-laboratory PSG with continuous monitoring of carbon dioxide via an end-tidal (etCO₂) or transcutaneous (tcCO₂) system, particularly in those patients with daytime hypercapnia or daytime hypoxemia.

Full-night PSG is preferred over a split-night study for patients with COPD, even if severe sleep apnea is apparent in the first third of the night in a range that would qualify for continuous positive airway pressure (CPAP), because patients with COPD may have more than one type of SRBD. Patients with severe emphysema often use accessory intercostals muscles to support ventilation. These patients are more vulnerable to decompensation during rapid eye movement (REM) sleep due to muscle atonia/hypotonia, and REM sleep usually predominates in the second half of the night. A full-night PSG will allow determination of whether REM sleep-related hypotonia/atonia worsens OSA, gas transfer, and breathing mechanics. (See 'Split-night protocol' below.)

Additional considerations relevant to patients with COPD include the following:

●Sleep-related hypercapnia – In patients with COPD, tcCO₂ monitoring is the preferred method to detect sleep-related hypercapnia or worsening of hypercapnia because etCO₂ underestimates CO₂ due to abnormalities in breathing mechanics and shortened expiratory time during inspiratory flow limitation. In addition, etCO₂ cannot be performed during titration of PAP. However, etCO₂ monitoring is more widely available than tcCO₂ monitoring. (See "Polysomnography in the evaluation of sleep-disordered breathing in adults", section on 'Ventilation'.)

●Periodic leg movements – Special emphasis should also be placed on monitoring periodic leg movements, as periodic leg movements of sleep (PLMS) are common during non-REM (NREM) sleep in COPD patients and they can be mistaken for respiratory events, when associated with either arousal or inspiratory flow limitation [7]. (See "Polysomnography in the evaluation of abnormal movements during sleep".)

●Supplemental oxygen – In patients with daytime hypoxemia who desaturate readily during sleep, supplemental oxygen will markedly alter the frequency and severity of nocturnal oxyhemoglobin desaturations. Most COPD sleep centers initiate a sleep study omitting supplemental oxygen in patients who have pulse oxygen saturation (SpO₂) levels >88 percent and then add oxygen if SpO₂ decreases to below 88 percent in NREM sleep for more than 10 minutes.

Split-night protocol — For patients with a high likelihood of OSA and without baseline hypercapnia, it may be practical to order a split-night study with initial documentation of OSA and then initiation of CPAP, following the usual acceptability criteria that are described separately. (See "Overview of polysomnography in adults", section on 'Split-night protocol'.)

Although split-night studies are now well established for sleep apnea patients, their value in COPD patients remains unclear for the following reasons: (1) REM sleep represents a particular vulnerability for patients with COPD and documentation of the REM effect often cannot be obtained from the first two to three hours of sleep; (2) Initiation of CPAP can worsen hyperinflation and may cause patients to reject CPAP in the latter part of the night. A careful adaptation to CPAP and its modalities are warranted in these cases.

A split-night study must document an apnea hypopnea index (AHI) ≥40 during a minimum of two hours and then titrate CPAP for more than three hours, as respiratory events can worsen during the latter part of the night. CPAP must eliminate respiratory events during REM and NREM sleep, including REM sleep when supine.

Diagnostic criteria — Diagnostic criteria for the various forms of SRBD have been established by the AASM [54].

●Sleep-related hypoxemia – Significant sleep-related hypoxemia is defined as a SpO₂ of ≤88 percent (≤90 percent in children) for ≥5 minutes that is not fully explained by sleep-related hypoventilation, OSA, or other SRBD.

●Obstructive sleep apnea – OSA is defined by the presence of ≥5 predominantly obstructive respiratory events (ie, obstructive and mixed apneas, hypopneas, or respiratory effort-related arousals [RERAs]) per hour of sleep (for in-laboratory PSG) or recording time (for out-of-center sleep testing) and symptoms (eg, sleepiness, fatigue, insomnia, awakening with gasping or choking, snoring); or ≥15 predominantly obstructive respiratory events, independent of symptoms [54]. (See "Clinical presentation and diagnosis of obstructive sleep apnea in adults", section on 'Diagnosis'.)

●Central sleep apnea – Central sleep apnea (CSA) is defined by the presence of ≥5 central apneas and/or hypopneas per hour of sleep (or recording time if a home study) plus a requirement that more than 50 percent of the total number of apneas and hypopneas are central [54]. In addition, patients must have one or more symptoms related to the disorder (eg, sleepiness, insomnia, frequent awakenings, nonrestorative sleep, awakening short of breath, witnessed apneas). (See "Central sleep apnea: Risk factors, clinical presentation, and diagnosis", section on 'Diagnostic criteria'.)

●Sleep-related hypoventilation – Sleep-related hypoventilation is defined as an arterial tension of CO₂ (PaCO₂) >55 mmHg for ≥10 minutes or an increase in PaCO₂ ≥10 mmHg during sleep (compared with an awake supine value) to a value >50 mmHg for ≥10 minutes [55]. The primary cause of hypoventilation is related to COPD (or another pulmonary disorder), but not obesity hypoventilation syndrome, medication use, or a known congenital central alveolar hypoventilation syndrome [54]. Sleep-related hypoventilation is generally most severe during REM sleep. Hypoxemia may or may not be present. (See "Polysomnography in the evaluation of sleep-disordered breathing in adults", section on 'Hypoventilation'.)

For patients with COPD, hypercapnia can be due to dead space ventilation and ventilation-perfusion mismatching from COPD or concomitant obesity hypoventilation syndrome, OSA, or neuromuscular weakness (eg, diaphragmatic paralysis). (See "Mechanisms, causes, and effects of hypercapnia", section on 'Mechanisms and etiologies of hypercapnia'.)

DIFFERENTIAL DIAGNOSIS — The differential diagnosis of SRBD includes a variety of non-respiratory causes of sleep disturbances and daytime sleepiness. Sleep fragmentation may be due to coughing or shortness of breath and other processes that can cause SRBD that are unassociated with COPD (eg, heart failure) [38]. Depression and anxiety are common among patients with COPD and can contribute to poor sleep quality and excessive daytime sleepiness [56,57]. Polysomnography (PSG) is often necessary to exclude SRBD in patients with depression or anxiety. (See "Clinical manifestations and diagnosis of obesity hypoventilation syndrome" and "Sleep-disordered breathing in heart failure" and "Clinical presentation and diagnosis of obstructive sleep apnea in adults", section on 'Differential diagnosis'.)

TREATMENT — Management of SRBD in patients with COPD should be individualized according to the type and severity of SRBD. Treatment options range from nocturnal oxygen in patients with isolated sleep-related hypoxemia to positive airway pressure therapy in patients with obstructive or central sleep apnea. The goals of therapy are to alleviate hypoxemia during sleep, improve sleep quality, and reduce COPD-related morbidity and possibly mortality [38,43,58].

Pharmacotherapy including anticholinergic and beta-agonist drugs and theophylline are well-established treatments to improve bronchial airway properties and ventilation; thus they can also improve SRBD in COPD and poor sleep quality [43]. The use of these treatments in COPD is discussed separately. Similarly, although lung volume reduction therapy has been shown to improve breathing mechanics and sleep, the primary indication for this intervention is not sleep-disordered breathing, and thus, it is also discussed separately. (See "Stable COPD: Initial pharmacologic management" and "Lung volume reduction surgery in COPD".)

Sleep-related hypoxemia

Indications for treatment — Patients with severe awake hypoxemia (arterial oxygen tension [PaO₂] ≤55 mmHg [7.32 kPa] or pulse oxygen saturation [SpO₂] ≤88 percent) meet criteria for long-term oxygen therapy (LTOT) and clearly benefit from nocturnal oxygen. However, the benefit of nocturnal oxygen in patients with mild-to-moderate hypoxemia and nocturnal desaturation has not been clearly demonstrated [59,60]. In the absence of severe daytime hypoxemia, we prescribe nocturnal oxygen for patients with nocturnal desaturation (SpO₂ ≤88 percent) who have associated signs or symptoms attributable to hypoxemia (table 2). (See 'Efficacy' below.)

The treatment of sleep-related hypoxemia, after exclusion of other SRBDs, is generally guided by the severity and duration of hypoxemia and associated clinical features. Medicare criteria for nocturnal oxygen supplementation include [61]:

●A PaO₂ ≤55 mmHg (7.32 kPa) or SpO₂ ≤88 percent.

●A decrease in SpO₂ >5 percent for at least five minutes during sleep.

●Associated symptoms or signs reasonably attributed to hypoxemia (eg, impaired cognition, nocturnal restlessness, morning headaches, pulmonary hypertension, erythrocytosis).

●Absence of another cause of sleep-related hypoxemia (eg, OSA). If determination of nocturnal desaturation is made by overnight oximetry, concomitant OSA should be excluded by careful history and/or polysomnography. (See 'Diagnostic criteria' above.)

When indicated, supplemental oxygen is generally supplied by nasal cannula at a flow rate sufficient to maintain the SpO₂ in the range of 90 to 95 percent [43]. (See "Long-term supplemental oxygen therapy".)

Importantly, nocturnal oxygen therapy alone is not recommended for patients with comorbid COPD and OSA (algorithm 1). (See 'Obstructive sleep apnea' below.)

Efficacy — The strength of the evidence in favor of administering supplemental oxygen to prevent long-term consequences of sleep-related hypoxemia depends on the degree of daytime hypoxemia.

●Severe daytime hypoxemia – Continuous home oxygen therapy is recommended in patients with severe resting awake hypoxemia (partial pressure of arterial oxygen [PaO₂] less than 55 mmHg [8 kPa]) due to COPD because it improves survival and quality of life [62,63]. The use of LTOT in COPD is described separately. (See "Long-term supplemental oxygen therapy", section on 'Benefits'.)

The exact flow rate of supplemental oxygen needed to prevent sleep-related hypoxemia may be higher than the amount needed when the patient is awake. In the Nocturnal Oxygen Therapy Trial, the oxygen flow was routinely increased by 1 L/minute during sleep [62]. We typically obtain an overnight oximetry study to guide oxygen supplementation in patients with pulmonary hypertension or significant hyperinflation due to COPD. These patients often have significant perfusion/ventilation mismatch or a right-left shunt with severe hypoxemia that is unresponsive to supplemental oxygen.

●Mild to moderate daytime hypoxemia – While LTOT improves survival in COPD patients with severe hypoxemia, oxygen therapy during sleep has not been shown to improve survival in patients with mild to moderate daytime hypoxemia (eg, SpO₂ 89 to 95 percent) or arterial desaturation only at night [60,64,65].

The hypothesis that supplemental oxygen might improve survival in patients with mild-to-moderate daytime hypoxemia (SpO₂ 89 to 93 percent) was challenged by the long-term oxygen treatment trial (LOTT), which examined the effect of continuous supplemental oxygen in 738 such patients [65]. In this trial, supplemental oxygen did not result in any difference in time to death or first hospitalization, rate of COPD exacerbations, or rate of COPD-related hospitalizations, nor was there a benefit in quality of life, lung function, or distance walked in six minutes.

As the LOTT trial did not specifically assess nocturnal hypoxemia, the International Nocturnal Oxygen (INOX) trial was designed to answer whether nocturnal oxygen is of benefit in COPD patients with nocturnal hypoxemia who did not meet criteria for LTOT (table 2) [59]. The multicenter randomized INOX trial compared nocturnal supplemental oxygen versus air from a sham concentrator in 243 patients with an SpO₂ <90 percent for ≥30 percent of the nocturnal home oximetry recording time [59,60]. In the active group, oxygen was titrated to maintain SpO₂ >90 percent for >90 percent of the recording time. At three years of follow-up, similar numbers of patients in the two groups met the composite outcome of death or resting SpO₂ ≤88 percent. Exacerbation rates, hospitalizations and quality of life were similar in the two groups. Interpretation of the trial is hampered by low recruitment (243 participants instead of planned 600) and incomplete retention.

We continue to prescribe nocturnal oxygen for patients with COPD who have mild to moderate daytime hypoxemia and comorbidities such as pulmonary hypertension, right heart failure, cardiac arrhythmia, and coronary artery disease. The use of supplemental oxygen in patients with mild to moderate daytime hypoxemia is discussed in greater detail separately. (See "Stable COPD: Overview of management", section on 'Supplemental oxygen'.)

Risk of hypoventilation with supplemental oxygen — In approximately 20 percent of COPD patients requiring oxygen, oxygen therapy at night can worsen hypoventilation during sleep, but this rise in CO₂ rarely worsens hypercapnia and acidosis in the morning [66]. Even in patients with severe COPD (mean arterial carbon dioxide tension [PaCO₂] 53 mmHg), the administration of supplemental oxygen at night was associated with only small (<6 mmHg) increases in PaCO₂ throughout sleep [67]. Nevertheless, if the rise in PaCO₂ offsets the beneficial effect of oxygenation, institution of positive airway pressure (PAP) therapy can mitigate the worsening of hypoventilation.

Obstructive sleep apnea — Treatment of OSA includes a combination of patient education, lifestyle modifications (eg, weight loss for patients who are overweight or obese, alcohol avoidance), and PAP. An overview of OSA management is provided separately along with guidance regarding the initiation and titration of PAP. (See "Obstructive sleep apnea: Overview of management in adults" and "Titration of positive airway pressure therapy for adults with obstructive sleep apnea" and "Mode selection for titration of positive airway pressure in adults with obstructive sleep apnea".)

Initiating positive airway pressure — PAP is the mainstay of therapy for OSA, almost always starting with continuous positive airway pressure (CPAP). Many patients with mild-to-moderate COPD and OSA are candidates for initiation of PAP at home following diagnosis by HSAT (algorithm 1) (see 'Home sleep apnea test' above). By contrast, patients with moderate-to-severe COPD and OSA generally require initiation and manual titration of PAP in a sleep laboratory; moderate-to-severe COPD and oxygen-dependent COPD are considered contraindications to initiating CPAP with auto-titrating devices at home. (See "Titration of positive airway pressure therapy for adults with obstructive sleep apnea", section on 'Modes of positive airway pressure'.)

●Initiation of PAP – The initiation of PAP in patients with COPD and OSA, generally follows the protocol outlined for routine treatment of OSA and starts with CPAP at 4 cm H₂O. Heated humidification is generally recommended for patients receiving CPAP. However, some patients with COPD are sensitive to changes in humidity and may report increased cough or dyspnea when the level of humidity is too high (or too low) for them. (See "Titration of positive airway pressure therapy for adults with obstructive sleep apnea".)

●Efficacy – For patients with coexisting obstructive sleep apnea and COPD, use of PAP has been associated with lower mortality, morbidity, and exacerbation rates [68,69]. In a prospective observational study, 651 COPD patients (228 with OSA and treated with CPAP, 213 with OSA but not treated with CPAP, and 210 patients without OSA) were followed for a median duration of 9.4 years [68]. Patients with OSA not treated with CPAP had a higher mortality (relative risk 1.79, 95% CI 1.16-2.77) and were more likely to experience a severe COPD exacerbation leading to hospitalization compared with the COPD-only group. In addition, patients with OSA treated with CPAP had no increased risk for mortality or exacerbations compared with patients with COPD alone.

Larger observational studies have also shown an association between PAP adherence and improved outcomes in this population [69,70]. As an example, health insurance claims data in the United States were used to identify 6810 patients with COPD and OSA who were prescribed PAP; among these, 4482 patients (66 percent) were adherent based on PAP user data [69]. During two years of PAP therapy compared with the year before therapy in 1412 propensity score-matched patients, PAP-adherent patients showed greater reductions in emergency room visits, inpatient hospitalizations, and severe acute exacerbations compared with nonadherent patients.

CPAP is also effective in mitigating the excess risk of mortality in hypercapnic patients [71] and in hypoxemic COPD patients [72].

CPAP failure or intolerance — While CPAP is sufficient in most patients with concomitant OSA and COPD, some patients require high levels of PAP to achieve control of OSA. The American Academy of Sleep Medicine (AASM) suggests changing to bilevel positive airway pressure (BPAP) in the spontaneous mode, if residual obstructive events or snoring are observed on CPAP therapy at a pressure of ≥15 cm H₂O [73].

The application of PAP in patients with severe emphysema can also worsen dynamic hyperinflation and increase work of breathing, particularly at pressures exceeding 10 cm H₂O [17]. In these patients, newer PAP devices that allow reduction in expiratory pressures or even the application of bilevel PAP may help to get patients adjusted to PAP and lower the side effects.

Other modalities of inspiratory pressure support (adaptive server ventilation [ASV] or auto-adapting bi-positive pressure ventilation [AVAP]) may further improve nocturnal and diurnal blood gases, although it remains unclear whether these modalities provide additional clinical benefit in these patients compared to CPAP alone [74]. (See "Modes of mechanical ventilation", section on 'Adaptive support ventilation'.)

Central sleep apnea — Data regarding the consequences of central sleep-disordered breathing events in patients with COPD are lacking. Nevertheless, as for central sleep apnea (CSA) in the absence of COPD, treatment should be considered if CSA and the underlying cause (ie, heart failure) impose additional health risks to the patient [1,75]. (See "Central sleep apnea: Treatment".)

Some patients develop central apneas during the initiation of CPAP for OSA, a phenomenon known as treatment emergent CSA. The management of treatment-emergent CSA generally involves continuing CPAP, as many patients will spontaneously improve. Patients who do not improve (and do not have heart failure) can be switched to ASV; BPAP with a backup rate is a reasonable alternative. Treatment-emergent CSA is described separately. (See "Mode selection for positive airway pressure titration in adult patients with central sleep apnea syndromes" and "Treatment-emergent central sleep apnea".)

Sleep-related hypoventilation — Patients with severe COPD and intermittent or persistent daytime hypercapnia often experience worsening hypoventilation during sleep even in the absence of OSA.

Noninvasive ventilation — The optimal role for noninvasive ventilation (NIV; eg, CPAP or BPAP) in these patients is unclear and data are limited [76]. Patients with poor sleep quality and/or daytime sleepiness may be more likely to derive benefit.

Nocturnal NIV has been used in combination with supplemental oxygen in patients with daytime hypercapnia and may improve survival, but not necessarily quality of life [76]. Nocturnal NIV with supplemental oxygen improved sleep quality and daytime blood gases better than oxygen alone in some studies [77,78], but not in others [79]. The benefits on sleep and daytime blood gases have been attributed to improvements in breathing mechanics such as reductions in micro-atelectasis with positive pressure, prevention of collapse in the intrapulmonary airways, and reduction in work of breathing all of which contribute to resting of chronically fatigued respiratory muscles [43]. The rationale for nocturnal NIV in COPD, its implementation, and effects on survival and quality of life in patients with COPD are discussed separately. (See "Nocturnal ventilatory support in COPD".)

For patients with sleep-related hypoventilation due to COPD, PAP is typically initiated in a sleep laboratory, if it has not already been titrated during a hospitalization. The details of initiation of nocturnal NIV in hypercapnic patients with COPD are discussed separately. (See "Nocturnal ventilatory support in COPD", section on 'Practical aspects'.)

Pharmacotherapy — Respiratory stimulants such as progestational agents, theophylline, acetazolamide, and protriptyline were once used to treat hypercapnia, but the benefits were modest and not maintained. They are rarely utilized for this indication. (See "Central sleep apnea: Treatment", section on 'Pharmacologic therapy' and "Disorders of ventilatory control", section on 'Drugs affecting ventilatory drive'.)

Other sleep disorders — Treatment of restless leg syndrome and insomnia are discussed separately. (See "Management of restless legs syndrome and periodic limb movement disorder in adults" and "Overview of the treatment of insomnia in adults".)

FUTURE DIRECTIONS — Ongoing research is aimed at determining whether nocturnal oxygen therapy is beneficial in mild hypoxemia (see 'Sleep-related hypoxemia' above), whether adaptive server ventilation (ASV) or auto-adapting bi-positive pressure ventilation (AVAP) have a role in the treatment for central sleep apnea (CSA), and whether more comfortable and effective interfaces and modes for treating SRBD in patients with COPD will improve sleep quality and adherence.

As an example, nasal high flow therapy with oxygen or room air is being studied for a potential role in SRBD. Positive airway pressure (continuous positive airway pressure [CPAP], bilevel positive airway pressure [BPAP]), noninvasive ventilation (NIV), and long-term oxygen therapy are the mainstream treatments for sleep-disordered breathing in COPD, but adherence rates are low [80-82]. Preliminary data suggest that nasal high flow of warm and humidified air (NHF) may represent an alternative means to improve sleep-disordered breathing in these patients.

NHF of oxygen or room air through an open nasal cannula was first introduced to improve oxygenation in infants and children with hypoxic respiratory failure [83]. It has subsequently been extended to adult pulmonary care [84-86]. Moreover, several studies demonstrate that it may improve arterial blood gases and exacerbation rate in COPD patients with hypercapnic respiratory failure [87,88]. The mechanisms have been attributed to reductions in dead space ventilation [89], slight increases in positive expiratory pressure [90], and reductions in work of breathing [89], compared with supplemental oxygen alone [91].

Although several nasal high flow medical devices exist for use in various hospital settings using pressurized air outlets (eg, TNI Soft Flow, Optiflow, Precision Flow) there are only a few devices that provide adequate comfort and minimal noise required for use during sleep. Whether nocturnal use of NHF air can prevent nocturnal hypercapnia and improve daytime outcomes similar to CPAP or long-term oxygen therapy is being examined in several clinical trials [92,93].

PROGNOSTIC IMPLICATIONS — SRBD may adversely affect quality of life, morbidity, and mortality in patients with COPD.

●Sleep-related hypoxemia – Several lines of evidence suggest that sleep-related hypoxemia may increase morbidity and mortality in COPD. First, in the Nocturnal Oxygen Treatment Trial [14,62], a post hoc analysis revealed increased mortality in patients with nocturnal hypoxemia. Second, sleep-related oxygen desaturation has been associated with worsening pulmonary hypertension and cor pulmonale [94,95]. Third, sleep-related hypoxemia is associated with neurocognitive dysfunction [96,97], which can improve with treatment for sleep-disordered breathing [98]. Nevertheless, studies examining the effects of nocturnal oxygen administration on outcomes in COPD have been inconclusive due to small sample sizes and high attrition rates [14,18,99].

●Obstructive sleep apnea – Several large studies now show that the coexistence of obstructive sleep apnea (OSA) is a negative prognostic predictor for both cardiovascular disease and mortality in COPD patients [68,100,101]. In a retrospective case control study consisting of 10,981 males with OSA, comorbid COPD was associated with a seven-fold increased risk for all-cause mortality [100]. Decreased survival has been attributed to a greater frequency of exacerbations, more severe hypoxemia and hypercapnia, pulmonary hypertension, and cor pulmonale with potentially lethal arrhythmias.

●Pulmonary hypertension – While pulmonary hypertension (PH) is not common in patients with even severe OSA, the risk for having pulmonary arterial hypertension with OSA is increased in patients with coexisting COPD, chronic hypoventilation, and obesity [64]. The prevalence of pulmonary hypertension in patients with COPD Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage 2 to 3 is approximately 25 percent (22 to 27 percent) and increases to 33 to 53 percent in patients with GOLD stage 4 COPD [102,103]. The combination of OSA and COPD increased right ventricular remodeling compared to COPD alone and may explain the increased mortality compared to COPD or OSA alone. Nevertheless, compared with daytime hypoxia, hypercapnia and reduced forced expiratory volume in one second (FEV₁), indices that reflect OSA severity (such as apnea hypopnea index [AHI] or oxygen desaturation index), play a minor role for developing pulmonary hypertension and right heart failure [43,58]. Moreover, while daytime hypoxemia is the strongest predictor for the development of PH, there is no evidence that nocturnal hypoxia alone increases the risk for PH.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sleep-related breathing disorders in adults".)

SUMMARY AND RECOMMENDATIONS

●Definition – Chronic obstructive pulmonary disease (COPD) is associated with a spectrum of sleep-related breathing disorders (SRBD; formerly called sleep disordered breathing) that include sleep-related hypoxemia, obstructive sleep apnea (OSA), central sleep apnea, respiratory effort-related arousals, sleep-related hypoventilation, restless leg syndrome, and insomnia. (See 'Introduction' above.)

●Contributing factors – The majority of patients with COPD have at least one comorbid condition and more than half have at least four coexisting conditions (eg, obesity, heart failure, smoking, opioid use) that are associated with a higher prevalence of SRBDs. (See 'Contributing factors' above.)

●Clinical features – The symptoms and signs of SRBD in patients with COPD are essentially the same as for those without COPD (eg, snoring, insomnia, awakening with gasping or choking, morning headaches, daytime sleepiness or fatigue, and poor concentration or memory impairment). However, patients may be asymptomatic or may have symptoms that overlap with the features of COPD itself and thus not report them. (See 'Clinical features' above.)

●Diagnostic evaluation – The diagnosis of SRBD in patients with COPD requires a high index of suspicion and formal sleep testing with polysomnography. Indications include symptoms of SRBD, obesity (body mass index [BMI] >35), neck circumference >43 cm (17 inches) in males or >41 cm (16 inches) in females, pulse oxygen saturation [SpO₂] <93 percent, daytime hypercapnia, pulmonary hypertension, or cor pulmonale. (See 'When should SRBD be suspected in COPD?' above.)

•For patients with a high likelihood of OSA and without baseline hypercapnia, it may be practical to obtain a split-night study due to issues with accessibility; otherwise a full night polysomnogram is preferred. When possible, continuous monitoring of carbon dioxide via an end-tidal (etCO₂) or transcutaneous (tcCO₂) system, is helpful, particularly in patients with daytime hypercapnia or daytime hypoxemia (algorithm 1). (See 'In-laboratory polysomnography' above.)

•Diagnostic criteria for the various SRBD have been established by the American Academy of Sleep Medicine. (See 'Diagnostic criteria' above.)

●Management

•Continuous supplemental oxygen is indicated for patients with severe awake hypoxemia (ie, PaO₂ ≤55 mmHg [7.32 kPa] or SpO₂ ≤88 percent). A clinical benefit to nocturnal oxygen has not been demonstrated for most patients with COPD and sleep-related desaturation, but without severe daytime hypoxemia or other SRBD. For patients with symptoms or signs reasonably attributable to nocturnal desaturation (eg, impaired cognition, nocturnal restlessness, morning headaches, pulmonary hypertension, erythrocytosis), a trial of nocturnal oxygen supplementation may be reasonable. Indications for oxygen supplementation in patients who have mild-to-moderate hypoxemia and certain comorbidities are listed in the table (table 2). (See 'Sleep-related hypoxemia' above and "Stable COPD: Overview of management", section on 'Supplemental oxygen'.)

•Treatment of OSA in patients with COPD includes a combination of patient education, lifestyle modifications, and positive airway pressure (PAP). Initiation and titration of PAP should be performed manually in a sleep laboratory. (See 'Obstructive sleep apnea' above.)

•For patients with hypercapnia due to severe COPD, nocturnal noninvasive ventilation may improve survival but not necessarily quality of life. The role of nocturnal NIV in patients with severe COPD is discussed separately. (See 'Noninvasive ventilation' above.)

ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Peter Gay, MD, and Hartmut Schneider, MD, PhD, who contributed to earlier versions of this topic review.