Opioid Use Disorder, Sleep Deficiency, and Ventilatory Control: Bidirectional Mechanisms and Therapeutic Targets

Abstract

Opioid use continues to rise globally. So too do the associated adverse consequences. Opioid use disorder (OUD) is a chronic and relapsing brain disease characterized by loss of control over opioid use and impairments in cognitive function, mood, pain perception, and autonomic activity. Sleep deficiency, a term that encompasses insufficient or disrupted sleep due to multiple potential causes, including sleep disorders, circadian disruption, and poor sleep quality or structure due to other medical conditions and pain, is present in 75% of patients with OUD. Sleep deficiency accompanies OUD across the spectrum of this addiction. The focus of this concise clinical review is to highlight the bidirectional mechanisms between OUD and sleep deficiency and the potential to target sleep deficiency with therapeutic interventions to promote long-term, healthy recovery among patients in OUD treatment. In addition, current knowledge on the effects of opioids on sleep quality, sleep architecture, sleep-disordered breathing, sleep apnea endotypes, ventilatory control, and implications for therapy and clinical practice are highlighted. Finally, an actionable research agenda is provided to evaluate the basic mechanisms of the relationship between sleep deficiency and OUD and the potential for behavioral, pharmacologic, and positive airway pressure treatments targeting sleep deficiency to improve OUD treatment outcomes.

The opioid crisis continues to devastate lives and communities worldwide. Opioids are commonly used for acute and chronic pain management (1). Low-dose morphine is also the only approved pharmacotherapy for chronic refractory breathlessness (2). In the 1990s, healthcare providers increased opioid prescribing in response to the “pain as fifth vital sign” campaign, a downplay of the abuse potential of opioids, and aggressive marketing of drugs such as Oxycontin. However, opioids are indeed highly vulnerable to abuse, and owing to their central nervous depressant effects, accidental deaths. Opioid use disorder (OUD) is a chronic and relapsing brain disease characterized by loss of control over opioid use and deficits in cognitive function, mood, pain perception, and autonomic activity. OUD affects over 3 million U.S. citizens and 16 million individuals worldwide and causes one overdose death every 20 minutes (3). Most opioid-related deaths occur during sleep owing to respiratory failure, where opioid-induced respiratory depression, sleep-related loss of respiratory drive, and loss of behavioral protective mechanisms that control breathing cooccur.

Because the endogenous opioid system plays a central role in mood and well-being modulation, the endogenous opioid system is believed to importantly contribute to the development of OUD. The endogenous opioid system consists of three G protein coupled receptors, mu, delta, and kappa, which are stimulated by a family of endogenous opioid peptides (4). Opioid receptors can also be activated exogenously by alkaloid opiates, the prototype of which is morphine. The finding that morphine’s analgesic and addictive properties are abolished in mice lacking the mu receptor (5) points to the mu opioid receptor as a key molecular player in the therapeutic effects of opioids and in OUD.

Effective evidence-based frontline treatments exist in the form of FDA-approved medications for OUD (MOUD). These include methadone (full mu-opioid receptor agonist), buprenorphine (partial mu-opioid agonist and/or antagonist), and naltrexone (long-acting mu-opioid antagonist) (6). These treatments help decrease illicit drug use, treatment attrition, and disease transmission and improve social functioning. However, there is significant variability in treatment responses. Relapse rates are high and are associated with lack of retention in treatment and a continued cycle of relapse to illicit opioid use, risk for injection-related infectious complications, overdose, and death (7, 8). Better outcomes for individuals who are maintained on MOUD highlight the need to identify effective strategies to improve MOUD retention. New approaches to complement or enhance MOUD, improve retention in treatment, and foster the learning of skills necessary for long-term recovery could transform OUD care.

One strategy is to identify and target a neurobiological system that may be linked to OUD relapse, namely the sleep and circadian system. Sleep deficiency is an important correlate of OUD. It is associated with overlapping cognitive deficits in executive function and reward processing (9, 10). Sleep is increasingly being recognized as a key determinant of brain health, as it is an important factor in several physical and mental health disorders (11, 12). A growing scientific consensus has identified sleep as a critical component of OUD, both during the active disease state and during recovery. A 2018 FDA public meeting that included patients with OUD identified sleep disturbance as a primary contributor to relapse and treatment attrition (13). More recently, NIH committed $25 million to fund a research program entitled “Sleep dysfunction as a core feature of opioid use disorder and recovery” as part of the Helping End Addiction Long-term initiative (14). Sleep disturbance is common and often severe in persons with OUD (15, 16). A challenge in identifying the mechanistic role of sleep deficiency in OUD is that the mechanisms differ over the trajectory of OUD during active disease, opioid withdrawal, and recovery (Figure 1).

Figure 1.

Schematic of opioid use disorder trajectory and mechanisms of sleep deficiency leading to opioid use disorder. This is an illustration of the trajectory of opioid use disorder (OUD). Sleep deficiency accompanies OUD throughout this trajectory. The lower portion of this figure describes potential mechanisms whereby sleep deficiency may contribute to OUD. MOUD = medication for opioid use disorder.

Although opioid mortality prevalence is higher in people who are middle-aged, with a history of substance abuse and psychiatric comorbidity (17), at an individual level, our ability to identify who is most at risk of harm is poor. This is owing to incomplete knowledge of individual differences in vulnerability to opioid-induced respiratory depression and likely complex bidirectional interactions among sleep, opioid use disorder, and breathing control (16).

Mechanisms of Opioid Use Disorder Leading to Sleep Deficiency

Effects of Opioids on Sleep

Morpheus was the Greek god of dreams and the son of Hypnos, the god of sleep. Despite the historical inference, few studies have systematically investigated the effects of opioids on sleep. Paradoxically, rather than dream and sleep promotion, findings from early studies indicate that acute opioid use increases wake time and reduces slow wave sleep and REM sleep in a dose-dependent manner in addictive disorders (18, 19). More recent findings from studies in healthy participants show similar adverse consequences on sleep architecture with acute opioid use (20, 21). Mechanistically, opioid-induced sleep deficiency effects are likely mediated, at least in part, via inhibition of sleep-promoting neurons in the ventrolateral preoptic area of the anterior hypothalamus (22).

Nonetheless, improvements in sleep architecture may occur with chronic opioid use (23). Indeed, opioids can improve sleep quality and increase sleep time in people with chronic nonmalignant pain (24, 25). Given the evidence for a bidirectional relationship between sleep and pain whereby poor sleep worsens pain perception and vice versa (Figure 2) (26, 27), opioids may improve sleep, at least in part, via reductions in pain. Similar sleep quality–dependent mechanisms may also contribute to the therapeutic improvements in breathlessness with low-dose morphine in people with chronic breathlessness (28). However, recent data suggest persistence of prolonged sleep disturbance as measured via the Pittsburgh Sleep Quality Index in chronic, noncancer pain opioid addiction, even in those who have been opioid free for more than 6 months (29). Thus, whether opioid-related improvements in sleep quality is a consistent effect with chronic use across cohorts remains uncertain. Given that strategies to improve sleep can reduce pain (30), evaluation of the effects of opioids on sleep may be an important consideration in the clinical management of pain in people taking opioids. Although there is some evidence that opioid use is associated with perceived daytime sleepiness (15, 31), data on objective alertness is scarce. Further information on the differential effects of opioid use on key sleep parameters in different populations, including potential acute versus chronic differences, is covered in detail elsewhere (31) and is summarized in Table 1. Ultimately, effects likely depend on factors such as differences in patient characteristics, opioid type, dose, and concurrent use of other central nervous system depressants (24).

Figure 2.

Cycle of opioid dose escalation and the potential to impair sleep. This figure illustrates a vicious cycle leading to opioid dose escalation, whereby opioids impair sleep, which lowers pain threshold, and leads to greater opioid use and worsened pain.

Table 1.

Summary of Current Evidence on the Effects of Opioids on Key Sleep Parameters

Variable	Acute	Chronic
Sleep disturbance	↑	↕ ↔
Sleep onset latency	↑ ↔	↔
REM sleep latency	↑	↔
REM sleep	↓ ↔	↓ ↔
Slow wave sleep	↓	↑ ↔ ↓
Lighter stages of N1 or N2 sleep	↑	↔
Wake after sleep onset	↑ ↔	↓ ↔
Arousal index	↑ ↔	↔
Next-day perceived sleepiness	↑	↑

As highlighted in Table 1, although there are some consistent findings on the effects of acute opioid use on certain sleep parameters such as reductions in slow-wave sleep and sleep disturbance, findings on acute and chronic effects of opioids for most sleep parameters vary between studies and between different populations, doses, and opioid classes. Very few studies have been appropriately designed to investigate these parameters definitively, especially for chronic use. ↓ = a decrease, ↑ = an increase, ↔ = no reported change, and ↕ = conflicting findings of report that indicate either an increase or a decrease. Refer to the text for further detail.

Effects of Opioids on Sleep-disordered Breathing, Sleep Apnea Endotypes and Ventilatory Control

Like high altitude exposure, whereby everyone will eventually develop SDB if they go high enough (32), the same is likely true for SDB and opioids. Indeed, breathing can slow and become irregular, leading to hypercapnia and hypoxia, with high doses of opioids (33). Accordingly, central sleep apnea is very common with high-dose opioid use and is associated with impaired psychomotor vigilance performance (34, 35). Many people prescribed opioids also have comorbidities. This includes preexisting respiratory diseases such as SDB. Thus, defining the prevalence per se of SBD with opioid use and the different manifestations that are not always clearly defined in the literature is challenging but may be as high as 80% in certain clinical referral populations (36, 37). Given the loss or downregulation of many protective respiratory control mechanisms that occurs during sleep, including loss of stimulatory wakefulness drive and decreased chemoresponsiveness to CO₂ (32, 38), people who also have preexisting SDB may be at increased risk of harm and death with opioid use, especially at high doses.

However, identifying which individuals with SDB are most at risk of harm with opioid use remains a major clinical challenge. There is no single clinical predictor. Rather, there are likely multiple contributing risk factors (Figure 3). Knowledge of the heterogeneous causes of SDB and how opioids affect each of the underlying causes is crucial to understanding individual risk profiles.

Figure 3.

Spectrum of potential variable effects of opioids on breathing stability during sleep. Whether opioids worsen or improve breathing stability during sleep is influenced by a range of factors. (A) Examples of influential physiological factors are highlighted within the box on the top of the seesaw. (B) The balance of these various factors and their effects on sleep and breathing may vary within (e.g., with different sleep stages) and between nights (e.g., with concurrent central nervous system depressant use) and may result in different manifestations of sleep-disordered breathing. This can range from obstructive, mixed, to different forms of central sleep apnea from quite typical central apnea patterns to unique atypical patterns where it appears that central breathing pattern generator effectively misses beats and seemingly random intervals. In some cases, the highlighted physiological factors may be protective whereby the addition of opioids may reduce sleep-disordered breathing (e.g., via reductions in loop gain/breathing instability in those with high baseline loop gain and via A/G OPRM1 genotype protective effects) such that the balance of the seesaw tips to the left/blue (stable breathing). Conversely, opioids may perpetuate sleep-disordered breathing and increase the risk of respiratory failure in other cases (e.g., worse hypoxemia in those with already low loop gain at baseline and via A/A OPRM1 genotype deleterious effects) such that the balance of the seesaw tips to the right/orange (unstable breathing toward respiratory failure). Variability in degrees of pain, opioid dose and type, and previous history of opioid use may also influence an individual’s propensity toward breathing stability or respiratory failure with opioid use. Refer to the text for further details. OSA = obstructive sleep apnea.

Central sleep apnea

Physiologically, opioids cause central respiratory depression and central sleep apnea in a dose-dependent manner (34, 35). Loss of the wakefulness drive to breathe that occurs at sleep onset can cause central apnea (32, 38). This mechanism may be magnified when combined with opioid-induced central respiratory depression via activation of μ-opioid receptors (39). Recent animal data highlight the importance of the medullary pre-Bötzinger complex and the pontine Kölliker–Fuse nucleus mediating opioid-related respiratory disturbances (40). Opioid-induced sleep disturbance, such as reduced deep sleep, can also induce sleep state breathing instability to further drive central apnea. Indeed, transient arousals alone can cause centrally mediated reductions in breathing and apnea upon the return to sleep owing to the ventilatory response to arousal and accompanying reductions in CO₂, the main driver of breathing during sleep (32, 38). Chemosensitivity decreases from wakefulness to non-REM sleep with further reductions in REM sleep (41, 42). Interactions between sleep stage–related reductions in chemosensitivity combined with opioid-induced changes in chemosensitivity may further contribute to breathing instability during sleep whereby central breathing instability is common in lighter non-REM sleep but rare in REM sleep where hypoventilation may predominate (32). Increased hypoxic and decreased hypercapnic ventilatory responses (chemoresponsiveness) that have been reported in people on methadone maintenance treatment (43) can further drive cyclical breathing patterns and central apnea (44). Thus, the presence, severity, and specific manifestations of opioid-induced central sleep apnea (e.g., Figure 3) varies owing to multiple factors, including sleep stage, the type, duration, and dose of opioid, and other pathophysiological and clinical characteristics unique to the individual (e.g., body habitus and blood morphine concentration) (Figure 3). Finally, given the multiple breathing instability mechanisms that opioids can adversely influence, sleep apnea is common. So too is an increased risk of respiratory failure and death in people on opioids with sleep apnea (45).

Obstructive sleep apnea

The potential link between obstructive sleep apnea (OSA) and opioid use and the risk of harm is less clear than the link between opioid use and central apnea (46, 47). There is overlap in the pathophysiology between central and OSA. For example, centrally mediated loss of respiratory drive causes not only central apnea but also causes upper airway narrowing and closure owing to concurrent central reductions in the neural drive to the pharyngeal dilator muscles (48). Unstable control of breathing (high loop gain) is not only a contributor to central apnea but is also a feature of OSA for at least 30% of patients (49). Thus, opioids have been postulated to worsen OSA. However, randomized controlled trial data to support this concern are lacking (46). Rather, the existing data indicate no systematic change in OSA severity with opioids (46), including acute oral administration of 30–40 mg of MS Contin versus placebo (50, 51). However, there is considerable interindividual variability. As discussed in detail below, variability in responses is likely explained by interindividual differences in 1) blood morphine concentration (50–53), 2) opioid polymorphisms (50), and 3) the underlying pathophysiology of OSA (52, 54). Understanding these and other clinically revenant contributors on a per-patient basis is essential to identify people with OSA most at risk of harm and, conversely, those in whom opioids can be prescribed safely. This is important as people with OSA are approximately two times more likely to be prescribed opioids than people without OSA (55). This may be at least in part because of the bidirectional relationship between sleep disruption and increased pain intensity (Figures 1 and 2) (26, 27).

There are at least four key endotypes that contribute to OSA pathophysiology (Figure 4) (49, 54, 56, 57). Impaired pharyngeal anatomy, or a collapsible upper airway, is the pathophysiological trait common to all patients with OSA. However, the magnitude of anatomical impairment varies markedly between individuals (49, 56). More than 70% of people with OSA also have one or more nonanatomical endotypes that contribute to their OSA (49, 56). These include 1) inadequate upper-airway dilator muscle activity during sleep, 2) unstable control of breathing/excessive sensitivity to minor changes in CO₂ (high loop gain), and 3) a low respiratory arousal threshold (waking up too easily to minor airway narrowing), which prevents deeper, more stable sleep (49, 56).

Figure 4.

Schematic of the four key endotypes that contribute to obstructive sleep apnea (OSA) pathophysiology and the currently known effects of opioids on OSA endotypes and severity. Key endotypes include impaired pharyngeal anatomy or increased upper airway collapsibility, impaired pharyngeal dilator muscle function, low respiratory arousal threshold and unstable respiratory control (examples of desired nonanatomical endotypes highlighted in blue, gray tracings indicate an abnormal or impaired response). The lower schematic text highlights currently known effects of opioids on OSA endotypes (outside arrows/light gray shading), and OSA severity and the interaction with OSA endotypes (inside arrows in dark gray shading) based on human studies whereby ↑ = increase, ↔ = no change, and ↓ = decrease. Refer to the text for further detail. *Expected change based on the findings from Meurice and colleagues (58).

An earlier study indicated that the opioid receptor antagonist naloxone reduces upper airway collapsibility during sleep in healthy individuals (58). The synthetic opioid fentanyl suppresses hypoglossal motoneuron output in rats to the largest upper-airway dilator muscle, genioglossus (59). These changes, combined with respiratory depression, would tend to increase the propensity for OSA. However, during wakefulness, opioids reduce chemosensitivity (60). Thus, opioids may paradoxically stabilize breathing by reducing overly sensitive ventilatory responses to CO₂ in certain people with OSA (i.e., the >30% of patients with high loop gain). Indeed, overnight oxygenation improves with an acute morphine administration in some patients with OSA, which is related to chemosensitivity, and blood morphine concentration, which can vary 20-fold between individuals for a single morphine dose (50–53). Differences in opioid polymorphisms are also likely important. For example, people with the A118G OPRM1 polymorphism of A/A tend to have worse hypoxemia with an acute 40 mg dose of morphine, whereas those with A/G tend to have improvements in overnight hypoxemia (50). Given that OSA improves during deep sleep (61), whether opioids promote or disrupt sleep will also have potentially variable effects on OSA severity, including potential REM suppression effects (62), where OSA tends to be most severe (61).

To date, only one study has systematically investigated the effects of morphine on the key endotypes that contribute to OSA (52). An acute 40 mg dose of morphine before sleep does not systematically change upper-airway collapsibility, the respiratory arousal threshold, or genioglossus muscle responsiveness to airway narrowing during sleep (Figure 4) (52). However, consistent with wakefulness reports, loop gain and ventilatory responses to hypercapnia during sleep are reduced with morphine versus placebo. Together, assuming opioid-respiratory depression is minimal, these physiological changes would tend to stabilize breathing during sleep in people with high loop gain. Similarly, the addition of a hypnotic with an opioid, which is typically cautioned against, may actually promote sleep (via an increase in the arousal threshold) and reduce sleep apnea risk in patients with chronic pain (63), many of whom are not obese, in whom a low respiratory arousal threshold endotype is common (64). However, compensatory increases in genioglossus muscle activity occur with hypercapnia with placebo but not with morphine (52). Thus, people with high baseline respiratory arousal thresholds (hard to wake up) and already low loop gain (e.g., obesity hypoventilation phenotype) may be particularly vulnerable to respiratory depression during sleep and harm with opioids. Accordingly, knowledge of OSA endotypes for which simplified quantification techniques are being developed (57), and knowledge of individual opioid polymorphisms, may be useful in managing the potential risk of opioid use in the postoperative setting and beyond in people with OSA (54). Nonetheless, there are clear gaps in knowledge in this area that require further careful investigation of the potential risk of harm before this targeted approach can be considered clinically. This includes the effects of different doses, types of opioids, and acute versus chronic use (Table 2).

Table 2.

Knowledge Gaps and Research Agenda Priorities

Priority #1: Investigate the mechanisms by which sleep deficiency may contribute to OUD Priority #2: Investigate the impact of how treatments that improve sleep deficiency, including pharmacological and non-pharmacological strategies, may improve OUD outcomes Priority #3: Advance knowledge on the potentially disparate effects of opioids on sleep and breathing and the underlying mechanisms responsible Priority #4: Develop clinically applicable biomarkers to identify who is most at risk of SDB and harm with opioids versus those in whom opioids may be beneficial and safe To do this we need to conduct studies to advance knowledge on: How sleep deficiency may impact key neurocognitive addiction brain circuits (connectomes) involved in executive function, incentive salience and negative emotionality (i.e., via functional magnetic resonance imaging) How sleep deficiency may act through intermediate mechanisms such as stress, negative affect, pain, and other substance use (e.g., nicotine, stimulants, alcohol) to influence important clinical outcomes such as retention in medication for OUD programs and percentage of days with illicit drug use How more proximal ecological factors (e.g., psychosocial, family, and sleep environmental) that lead to sleep deficiency influence treatment and prevention efforts Whether cognitive behavioral therapy for insomnia consolidates sleep in people with OUD and sleep deficiency and its impact on key outcomes Whether chronobiological therapies such as melatonin and bright light therapy improve delayed sleep phase and social jet lag and improve outcomes in people with OUD The dose–response effects of opioids, including different durations and types of opioids, on: Sleep architecture and sleep quality (noting current evidence is mixed) SBD prevalence including the specific SDB manifestations and severity (noting current evidence suggests non-linear disparate effects with reductions in SDB at certain doses vs. worsening at others) Pathophysiological endotypes to provide mechanistic insight Acute vs. chronic including longitudinal studies and accompanying effects on cognition, perceived sleepiness, and objective alertness The effects on SDB when combined with other central nervous system depressants including sleep promotion aids The effects of these factors between different groups (i.e., those with vs. without comorbidities, with vs. without SDB, obese vs. nonobese, specific endotype target groups, etc.) The role of existing and emerging therapies, including novel pharmacotherapies with respiratory stimulatory properties in different settings (e.g., postoperative) and populations on opioid induced SBD and the influence on key outcomes The role of orexin antagonists for use in OUD, including the dosing regimen, side effects, safety profiles either alone or in combination with conventional opioid substation therapies, particularly around the risk of SDB

Definition of abbreviations: OUD = opioid use disorder; SDB = sleep-disordered breathing.

Sleep and Sleep Deficiency across the OUD Trajectory

Sleep and Sleep Deficiency

The two-process conceptual model of sleep regulation posits that two constituent processes, 1) a sleep-wake dependent homeostatic Process S (sleep drive) and 2) Process C (circadian rhythmicity), generates the timing of sleep and wakefulness. The American Academy of Sleep Medicine recommends at least 7 hours of sleep daily for adults, based on evidence linking sleep duration to health outcomes (65). In contrast to a focus of individual sleep disorders (e.g., insomnia), sleep deficiency, as defined by the NIH, is a broader construct that includes insufficient sleep duration (sleep deprivation), sleep out of sync with the body’s circadian rhythm (noncircadian sleep), not getting all the different types of sleep that the body needs (impaired sleep architecture), and poor sleep quality (e.g., owing to untreated sleep apnea, insomnia, pain, mental health issues, or stress).

Sleep Deficiency Accompanies OUD across the Trajectory of Disease

In addition to the direct effects of opioids on sleep (Table 1), sleep deficiency accompanies OUD across the trajectory of this addiction from initial medical or recreational use through misuse, addiction, withdrawal, recovery, relapse, overdose, and death.

Data have also emerged that disrupted circadian rhythms, such as those observed in shift workers and adolescents, increase susceptibility to addiction. There appears to be correlation between delayed sleep phase chronotype and addiction vulnerability (10, 66). The term “social jet lag” has been coined from the circadian desynchrony resulting from social, academic, and work schedules, and this form of desynchrony is associated with increased addiction in adolescents (67).

Taking opioids over a long period can lead to tolerance and dependence. A person who is dependent on opioids will experience symptoms of withdrawal should they reduce or suddenly stop taking opioids. Signs of withdrawal are similar for all opioids and can include nausea, vomiting, diarrhea, insomnia, anxiety, tachycardia, hypertension, muscle and bone pain, hyperthermia, sweating, and chills (68). Sleep disturbance is common and often severe in this setting during opioid withdrawal. For example, persons with OUD undergoing supervised withdrawal report significant sleep disturbances. including increased sleep onset latency, reduced total sleep time, and poor sleep quality (69). This reduced sleep quality often persists into the postwithdrawal period and has been linked to increased drug craving (70).

Many patients report significant sleep disturbance upon entering recovery programs for MOUD (71) and continue to report poor sleep quality during treatment (15). In one study of patients undergoing OUD, 90% of patients experienced poor sleep quality defined as a Pittsburgh Sleep Quality Index of more than 5, and nearly half had excessive daytime sleepiness defined as and Epworth Sleepiness Score of more than 10 (15). Whether sleep deficiency contributes to OUD relapse is the focus of several ongoing mechanistic studies, observational cohorts, and mechanistic clinical trials studies funded by the NIH Helping to End Addiction Long-term initiative. Importantly, there are several plausible biologic mechanisms whereby sleep deficiency may lead to OUD relapse, which are described in detail in the following section. Finally, most opioid-related deaths occur during sleep owing respiratory failure, where opioid-induced respiratory depression, sleep-related loss of respiratory drive, and loss of behavioral protective mechanisms that control breathing cooccur.

Mechanisms of Sleep Deficiency Leading to OUD

Sleep deficiency may impact OUD treatment outcomes through its influence on a range of neurocognitive mechanisms linked to addiction, such as brain circuits (connectomes) involved in executive function, incentive salience, and reward processing (Figures 1 and 5) (9). On the other hand, sleep deficiency among patients with OUD may lead to neuropsychiatric mechanisms such as stress, pain, negative affect, or other substance use (e.g., nicotine, stimulants, and alcohol) that may increase the likelihood of illicit opioid use (outlined below).

Figure 5.

Schematic overview of the multiple brain regions implicated in opioid use disorder, many of which overlap with those involved with sleep deficiency. Multiple brain regions interact dynamically to influence a range of complex cognitive processes relevant to OUD that can be influenced by sleep deficiency. OUD = opioid use disorder. Adapted by permission from Reference 72.

Neurocognitive Mechanisms: Executive Function, Incentive Salience, Negative Emotionality, Disrupted Circadian Rhythms, and the Orexin System

Multiple brain regions interact dynamically to influence a range of complex cognitive processes relevant to OUD that can be influenced by sleep deficiency (Figure 5). One accepted neuroscience-based framework for addictive disorders focuses on three key cognitive domains (executive function, incentive salience, and negative emotionality) that are tied to different phases in the cycle of addiction and form the core functional elements of addictive disorders (9). The executive function domain broadly includes processes related to organizing behavior toward future goals. Certain subdomains of executive function are particularly relevant to addiction, including attention, response inhibition, working memory, behavioral flexibility, and valuation of future events. These deficits in executive function cause loss of top-down control and result in preoccupation, anticipation (craving), and map to the prefrontal cortex (9, 72). The construct of incentive salience can be defined as psychological processes that transform the perception of stimuli and make them attractive (e.g., drug cue reactivity). This results in binging and intoxication, which is associated with activity in the basal ganglia (9, 72). Finally, increases in negative emotional responses to various stimuli and overall self-reported dysphoria are common in individuals with addictive disorders, and the reduction in negative affect (e.g., self-medication) has long been held up as a primary driver for the consumption of addictive substances mapping to the extended amygdala (9, 72). Importantly, all three of these cognitive domains impaired in addictive disorders are also impacted by sleep deficiency, forming a feed-forward allostatic framework (73).

Neuropsychiatric Mechanisms: Stress, Pain, Mood Disorders, and Other Substances

Periods of sleep deficiency lead to heightened blood pressure (74) and elevated cortisol (75) and promote sympathetic activation (76). Both acute and chronic stress are associated with the onset and progression (77) of OUD, and tonic stress and stress reactivity contributes to relapse in patients with addiction (78). Thus, evidence points toward a bidirectional (and mutually enforcing) relationship between sleep deficiency and chronic stress among patients with OUD.

Sleep deficiency produces “hyperalgesia” (increased pain sensitivity to noxious stimuli) in healthy subjects and clinical samples (26). Slow-wave sleep deprivation appears to exert this effect in humans, and recovery of slow-wave sleep increases pain tolerance (26). This may lead to a vicious cycle of opioid dose escalation, whereby opioids impair sleep, which lowers pain threshold and leads to greater opioid use and worsened pain (Figure 2) (26).

Among adults with mood disorders, sleep deficiency is critical for cognitive functioning, compromises health, and may contribute to substance use comorbidity and suicidality (79). This evidence has triggered a shift away from viewing sleep deficiency as an epiphenomenon to an important but underrecognized mechanism in the multifactorial cause and maintenance of the various mood disorders. In contrast to the current rates of depressive and anxiety disorders in the general population (2–5% and 6–10%) (80), between 4% and 24% of treatment-seeking individuals with OUD meet current criteria for a depressive disorder and between 5% and 17% meet criteria for an anxiety disorder (81–84). Ongoing mood disorders are important to monitor as they are associated with continued substance use, poorer retention, and lower quality of life (85–89).

Other substance use is common in OUD and may also contribute to sleep deficiency (90–92). For example, the rate of cigarette smoking among individuals with OUD far exceeds that of the general population (93). Nicotine lengthens sleep onset latency and decreases total sleep duration, particularly during deeper sleep stages (92). Likewise, alcohol and cocaine use disorders are highly comorbid in OUD (94) and are associated with poorer OUD treatment outcomes (95). Alcohol, a depressant, may promote the initiation of sleep and maintenance of sleep during the first half of sleep (e.g., decreased sleep onset latency), but it can be disruptive during the second half of sleep (e.g., increased wake after sleep onset, decreased slow wave sleep) (96, 97) and contribute to SDB (98).

Treatments targeting SDB

Positive Airway Pressure Therapy

Few studies have systematically investigated positive airway pressure (PAP) therapy for people with opioid-induced SDB. The existing data indicate variable success with PAP and adaptive serv-ventilation approaches (24, 99–105). Variable responses likely reflect, at least in part, the different manifestations of opioid-induced SDB (i.e., central vs. obstructive) and the extent to which PAP elicits lung reflexes which may further inhibit breathing (106). Nonetheless, effective treatment of SDB with PAP may improve health and well-being in this population (107).

In a recent trial in which over 200 people on opioids for chronic pain were systematically evaluated for sleep apnea, 120 (∼60%) had an apnea–hypopnea index (AHI) of >5 events/h (99). More than 50% declined attendance for sleep clinic review or treatment. Of the 20 (∼17%) people with varying degrees of obstructive and central apnea (average AHI > 40 events/h) who went onto PAP therapy, 17 were prescribed continuous positive airway pressure and/or automatic positive airway pressure, two adaptive servoventilation, and one BiPAP. At 1 year after therapy, similar to nonopioid SDB populations (108), 55% were adherent with PAP, which remained efficacious at reducing the AHI. Interestingly, as is the case in many opioid populations (63, 101), these patients were typically not obese. As most nonobese people with SDB have a low respiratory arousal threshold (64), this may be a physiological contributor that makes PAP therapy challenging (109) in these individuals.

Dose Reduction and Other Approaches

Given the dose–response relationship between opioids and SDB (34, 35), although dose reduction can be challenging to achieve, it is highly effective in reducing SDB severity and, thus, should be prioritized (32, 37, 99, 110). Other non-continuous positive airway pressure treatment strategies, such as positional therapy, can also be beneficial in this population (99). Although we are not aware of any studies that have investigated mandibular advancement splint therapy in people with opioid-related SDB, conceptually, this treatment approach should be equally effective for obstructive predominant SDB as for nonopioid SDB populations.

Emerging Pharmacotherapies

New emerging pharmacotherapies to treat OSA also have respiratory stimulatory effects (111–113). As such, these approaches may be beneficial for opioid-related SDB. This requires investigation. The use of ampakines, modulators of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) glutamatergic receptors, with opioids, offer promise that it may be possible to maintain analgesia while preventing the unwanted respiratory depression that accompanies opioids (114–118). Intranasal leptin can augment hypercapnic and hypoxic sensitivity, which may be beneficial for certain people with opioid-induced SDB (i.e., those with major hypoventilation) (119, 120). However, this work has not yet been translated to humans. To date, the effects of the opioid receptor antagonist naloxone on sleep apnea have been mixed (121, 122). Although caution is warranted and this approach is not recommended clinically without further research, certain hypnotics increase the arousal threshold and may reduce SDB risk in opioid users with a low arousal threshold endotype (63, 123).

Treatments targeting other sleep deficiency domains

Previous Trials

Given concerns for abuse with benzodiazepines, particularly among patients with OUD, previous trials of non-PAP treatments to improve sleep among patients with OUD have largely focused on nonbenzodiazepine medications. To date, we are aware of only three published trials. The first was a randomized, double-blind, placebo-controlled trial that compared trazadone to placebo with 6 months of follow-up in 137 patients recruited from methadone maintenance programs. Trazadone did not improve subjective or objective sleep, nor did it significantly increase or decrease illicit drug use relative to placebo (124). A more recent pilot trial of n = 10 methadone maintenance patients compared mirtazapine (30 mg), zolpidem sustained release (12.5 mg), mirtazapine (30 mg) + zolpidem (10 mg), and placebo using a within-subject, crossover design with a 1-week washout between drugs. The mirtazapine arm alone improved total sleep time (23 min), sleep latency (23 min), and sleep efficiency (3%), surpassing all the other regimens (125). Finally, a randomized, double-blind, placebo-controlled trial of 54 patients receiving methadone maintenance comparing melatonin 10 mg with placebo for 3 months resulted in significant improvement in subjective sleep quality, depression symptoms, and anxiety symptoms versus placebo (126).

Targeting the Orexin Neurotransmitter System

Orexin (a.k.a., hypocretin)-producing neurons, which are primarily located in the lateral hypothalamus, project to several subcortical and brainstem regions. These neurons are responsible for regulating wakefulness and arousal, diurnal neuroendocrine stress signaling and food, drink, and even drug consumption (127–130). Evidence from preclinical models of OUD indicates that increased orexin signaling contributes to arousal and stress reactivity (one of the neurobiological hallmarks of OUD) and orexin receptor antagonists (commonly prescribed for insomnia) attenuate opioid withdrawal symptoms (129–131). Thus, this system may be an important intermediate target in OUD. In support of this concept is the fact that hyperarousal is a hallmark feature of insomnia (132), a sleep disorder that is common in OUD characterized by difficulty initiating and maintaining sleep and other daytime impairments.

In 2018, the National Institute of Drug Abuse Division of Therapeutic and Medical Consequences listed its “10 most wanted” medication development priorities in response to the opioid crises (133). At the top of this list were the antagonists or negative modulators of the hypothalamic orexin (hypocretin) neuropeptides. This includes the already U.S. Food and Drug Administration–approved dual orexin receptor antagonist (DORA), suvorexant (Belsomra), and another recently U.S. Food and Drug Administration–approved DORA, lemborexant, both for the treatment of insomnia. In addition to insomnia therapeutic benefits, there is preclinical evidence that DORAs reduce opioid withdrawal and drug seeking (129, 130). This may, in part, be mediated through normalizing sleep disturbances present in 75% of patients with OUD, as sleep disturbances can worsen OUD outcomes (71). Alternatively, DORAs might improve OUD through other mechanisms and comorbidities such as stress and anxiety (134). Given that these are schedule IV controlled substances, this drug needs to be rigorously tested in clinical trials, several of which are currently underway (NCT03412591, NCT03897062, NCT03789214, NCT03937986, NCT03789214, and NCT03657355).

Future Research

Further work is needed to investigate the mechanistic knowledge gaps by which sleep deficiency may contribute to OUD and how treatments that improve sleep deficiency may improve OUD outcomes. Although by no means an exhaustive list, key knowledge gaps and future priority areas are outlined in Table 2.