Treatment of insomnia associated with alcohol and opioid use: a narrative review

Abstract

Substance use disorders (SUDs) are associated with profound sleep disturbances, including insomnia, sleep fragmentation, and circadian rhythm dysfunction resulting in serious mental and physical consequences. This minireview presents an overview of the neurocircuitry underlying sleep disturbances in SUDs and elaborates on treatment options with emphasis on alcohol use disorder (AUD) and opioid use disorder (OUD). A PubMed, Embase, CINAHL Plus, Cochrane, and Scopus search were conducted using sleep- and AUD/OUD related keywords from January 1st, 2000, to January 31st, 2023, with preferences for recent publications and randomized-controlled trials. A bidirectional relationship exists between insomnia and addiction with the status of each condition impacting the other in dictating clinical outcome. Existing evidence points to a resurgence of insomnia during detoxification, and unless treated satisfactorily, insomnia may lead to relapse. The discussion summarizes the strengths and limitations of cognitive behavioral therapy and pharmacological treatment for insomnia in SUDs covering evidence from both animal and clinical studies. The assumption of reestablishing normal sleep patterns by attaining and maintaining sobriety is misguided. Comorbid insomnia in patients with SUDs should be approached as an independent condition that requires its own treatment. Future clinical trials are needed with the aim of providing a resource for guiding clinical management of the many patients with insomnia and SUD.

Introduction

Substance use disorders (SUDs) refer to the compulsive intake of mood- or behavior-altering substances such as alcohol or illicit drugs characterized by significant deviation from personal and social norms leading to harmful social, physical, and economic consequences [1]. The condition is considered among the top public health priorities of the United States [2]. In 2018, the National Survey on Drug Use and Health reported that over 20 million over the age of 12 had been diagnosed with substance use disorders with over 74% of those attributed to alcohol abuse [3]. Marijuana and opiate use accounted for 4.4 and 2.0 million individuals, respectively. During that year alone, the mortality rate from substance-related and addictive disorders approached 21 deaths per 100,000 individuals, mostly related to drug overdoses, suicide, and other health complications. The trend has only worsened with the global COVID-19 epidemic. With the unyielding stressful conditions whether from house confinement, limited access to mental health, or loss of loved ones, the death toll of those afflicted with SUDs mounted exponentially, reaching a total of 106,699 deaths in 2021 and resulting in an age-adjusted mortality rate of 32.4 per 100,000 standard population [4]. The socioeconomic implications are of no lesser magnitude. The economic cost of drug abuse in the United States, including healthcare costs, drug-related accidents, and violent crimes, is thought to exceed $442 billion annually with another $90 billion related to loss of productivity and litigation costs [5].

Sleep disturbances are a hallmark of SUDs [6]. Patients with SUDs often report difficulty falling asleep, staying asleep, or advancing through the normal stages of sleep. Both persistent abuse and sustained abstinence disrupt normal sleep patterns which may lead recovering addicts to suffer sleep fragmentation for days or even weeks. This sleep fragmentation, invariably manifested as insomnia, poses not only a risk of perpetuating the habit of substance abuse but also impairs quality of life, increases the risk of relapse, and instigates suicidal behaviors [7, 8]. Insomnia is defined by a spectrum of complaints reflecting dissatisfaction with the continuity, duration, and quality of sleep, which impact not only nighttime, but daytime function in the individual. According to the most recent edition of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition DSM-V, these complaints to become chronic in nature must occur at a frequency of 3 nights per week, for a duration of > 3 months [9]. It is estimated that 20–25% of adults experience symptoms of insomnia on a situational basis and 10–12% on a chronic basis [10]. While insomnia has previously been conceptualized as a symptom of another psychiatric or substance use disorder, recent evidence suggests that chronic insomnia is an independent risk factor for the development of other disorders. When insomnia is comorbid with another psychiatric or substance use disorder, treatment outcomes improve when attending to both insomnia and the comorbid disorder rather than just attending to the comorbid disorder alone [11].

While the diagnostic evaluation of insomnia in SUDs parallels the steps taken in patients without comorbid conditions [12], we provide in this narrative review a brief overview of the underlying pathophysiology of insomnia among patients with substance use with special emphasis on alcohol and opioid use disorders. We also discuss the challenges and existing treatment of insomnia associated with these SUDs.

Methods

Search strategy

An electronic systematic search of PubMed, Embase, CINAHL Plus, Cochrane, and Scopus databases for peer-reviewed studies published between January 1st, 2000 and January 31st, 2023 was performed to identify articles that examined the associations between insomnia and substance use disorders. Preferences were given for recent publications and randomized-controlled trials (RCTs). Medical subject headings and keywords included the terms “insomnia”, “substance abuse”, “illicit substances”, “opioid”, “cocaine”, “heroin”, “alcohol abuse”, and “treatment.” Highly cited papers published before 2000 were also considered. Two authors independently were involved in the literature search, abstract screening, and full-text extraction. References lists of included publications were screened for additional eligible studies not captured through the database searches. Discrepancies were discussed and solved in consensus with the co-authors.

Criteria for study inclusion/exclusion

Eligible studies and reviews met the following criteria: (1) studies that examined the association between insomnia and alcohol or opioid use disorders in a randomized trial, interventional, longitudinal, prospective, cross-sectional, or retrospective design; (2) studies that included adult population ≥ 18 years old with a sample size of ≥ 5; (3) studies that assessed insomnia symptoms through self-reports, clinical interviews, or through standard nosologies based on the World Health Organization Composite International Diagnostic Interview, Version 3 [13], the Diagnostic and Statistical Manual of Mental Disorders, fourth or fifth editions (DSM-IV, DSM-V) [1, 14], the International Statistical Classification of Diseases and Related Health Problems, Ninth or Tenth Revisions (ICD-9, ICD-10) [15, 16], or the International Classification of Sleep Disorders definition [17]. Accepted diagnostic criteria included difficulty in initiating sleep, difficulty in maintaining sleep, early morning awakening with inability to return to sleep, and/or non-restorative sleep, as well as subjective complaints of daytime sleepiness or impaired functioning due to sleep disturbances. Studies that included assessment of insomnia severity, sleep medication use, and the Pittsburg Sleep Quality Index global score [18] as part of insomnia definition were also entertained; and (4) studies in which exposure to any of the following substances: alcohol, illicit substances, heroin, cocaine, or medications used to treat alcohol use disorder (AUD) or opioid use disorder (OUD). Exclusion criteria included studies that were focused on a restrictive population, commentaries, dissertations, editorials, conference presentations, or were not in the English language. Studies that defined insomnia in combination or as a part of another sleep disorder were also not considered. Figure 1 depicts the flow diagram of the selected studies. In total, 126 studies were deemed suitable to be referenced in this review.

Fig. 1.

A flow diagram depicting the selection of studies examining the association between insomnia and substance abuse disorders

Pathophysiology of insomnia-SUD

The relationship between insomnia and substance abuse has been described as reciprocal (or bidirectional) in nature. Existing theoretical models on the pathogenesis of insomnia–SUD complex emphasize self-reinforcing positive feedback loop between addiction and insomnia driven by a hypermotivated state (Table 1). Although a triggering event for the chronic use of substances of abuse may not always be traceable, common themes include a desire to seek pleasure, obtain relief from a depressed mood, overcome a state of anxiety, or alleviate a painful condition. A preexisting biological predisposition in the form of genetic, medical, or psychiatric vulnerability acts as a catalyst in establishing dependence on these substances. Drug consumption evolves into a conditioned behavior via activation of the core reward circuitry mediated by the dopaminergic and the orexigenic systems [19, 20]. As drug consumption turns into addiction, sleep disturbances ensue in the form of insomnia, interrupted sleep, or sleep deprivation. Among the neurochemical pathways identified in promoting wakefulness are inhibition of GABAergic transmission in the dorsal raphe nucleus, and a decrease in acetylcholine concentration in the pontine reticular formation [21]. The state of sleep deprivation coincides with increased activity in the amygdala promoting emotional volatility, erratic behaviors, and suicidal thoughts [22]. Experimentally, sleep-deprived animals exhibit an augmented response to dopaminergic agents, shifting to more aggressive, hyperactive, and hypersexual behavior, compared with control animals [23]. By the same token, substance abuse stimulates the release of orexin, a neuropeptide known also as hypocretin, from glutamatergic neurons in the lateral hypothalamic area (LHA) [24]. Given the ubiquitous projection of orexin-secreting neurons throughout the central nervous system, activation of orexin receptors amplifies reward-seeking behavior [25, 26]. Interestingly, dual orexin receptor antagonists (DORAs) attenuate impulsive behavior and cocaine-mediated dopamine release in rodent models [27]. During oxycodone abstinence, DORA treatment decreases the conditioned reinstatement of oxycodone-seeking behavior and reduces craving primarily via orexin-1 receptors [28]. Further, recent work has uncovered a positive correlation between motivation intensity and the number of orexin cells [29]. In both murine models and human heroin users, a dramatic multiplication in the number of hypocretin-producing neurons was observed in LHA [29]. The density of orexin neurons was preserved 150 days after abstinence and was associated with a greater risk of relapse [30]. Examination of postmortem human brain victims of opioid disorders revealed a 54% increase in orexinergic neurons in the lateral hypothalamus compared with age- and sex-matched controls [29]. There is also evidence of increased cortisol output secondary to dysregulation of the HPA axis during the withdrawal periods which relates to the reduction in total sleep time during recovery [31]. Targeting these pathways offers an opportunity to mitigate adverse treatment outcomes. [32] Additionally, while dorsomedial hypothalamic and perifornical orexin neurons are activated preferentially by putative reward, the orexin neurons in the LHA are activated by the stress of drug withdrawal [33]. As a result of this amplified orexin signaling, patients with SUD undergoing withdrawal often experience increased sleep latency, total sleep time, and poor sleep quality. It is thought that this orexin signaling modulates wakefulness by exciting noradrenergic and histaminergic neurons in the locus coeruleus and the tuberomammillary nucleus [34]. Blocking orexin-2 receptors have been shown to promote sleep and normalize sleep architecture [35].

Table 1.

Pathophysiology of insomnia in substance abuse disorders

Biologic predisposition

Central nervous system hyperarousal

Neurotransmitter imbalance (excitation of dopamine, orexin neurons)

Circadian rhythm disruption (irregular sleep patterns and disruption of sleep schedules)

Sleep architecture alterations (reduction of REM sleep and deep sleep stages)

Psychiatric comorbidities (depression, anxiety)

Behavioral factors (emotional volatility, poor impulse control, risk-taking behavior)

Treatment of insomnia in alcohol abuse disorder

Alcohol abuse disorder (AUD) has been associated with subjective insomnia as well. Cross-sectional studies have estimated the prevalence of insomnia to range from 36 to 74% among this group as compared to 10% prevalence in the general population [36]. Insomnia is usually reported in the acute withdrawal phase (1–2 weeks) and early recovery phase (2–8 weeks after detoxification). During the acute withdrawal phase, insomnia symptoms vary in intensity albeit they gradually improve over the detoxification period [37]. It may continue into the sustained recovery phase (3 months after the detoxification phase) for up to 3 years [38]. Patients with AUD in the early recovery phase often experience increased sleep fragmentation, frequent sleep-stage transitions, and increased awakenings, especially in the second half of the night. With alcohol abstinence, sleep homeostasis is disrupted, including increased sleep-onset latency, shortened NREM to REM cycles, and reduced sleep efficiency. The ensuing insomnia portends a higher risk for suicide and depression [39, 40], and is considered a risk factor for relapse [38, 41, 42].

Guidelines for the management of insomnia in patients with AUD are limited by a scarcity of RCTs, a heterogeneous population with extensive comorbidities, and the absence of extended follow-up. Non-pharmacologic interventions (i.e., behavioral therapy) are considered the preferred approach, but lack of access, cost, high dropout, and relative shortage of skilled providers constitute serious impediments to widespread implementation. Nonetheless, cognitive behavioral therapy for insomnia (CBT-I) was evaluated in five randomized trials involving patients with AUD and comorbid insomnia (Table 2). In four RCTs, CBT-I was delivered individually or in group sessions. [43–47] The results showed greater improvement in sleep latency, sleep efficiency, and number of awakenings, and daytime symptoms, such as fatigue, tiredness, depressive symptoms, and anxiety.[43–46]. Improvements in insomnia symptoms were maintained at 3- and 6-month follow-ups [44, 45]. However, none of the three trials demonstrated an impact on drinking outcomes compared to the control group. The fifth RCT involved a self-paced digital CBT-I referred to as the Sleep Healthy Using the Internet (SHUTi) program [46]. The program entailed an automated and interactive program designed to provide individualized feedback based on patient’s sleep data. SHUTi significantly alleviated symptoms of insomnia compared to patient education, and unlike the first three RCTs, the intervention reduced alcohol consumption. All these trials are limited by small sample sizes, heterogeneity in alcohol consumption, and different treatment durations. Yet, it is reassuring that outcomes converged in favor of CBT-I for insomnia indices.

Table 2.

Randomized-controlled trials of cognitive behavioral therapy for insomnia in individuals with alcohol use disorder

Author, year	N	Intervention	Delivery	Assessment	Outcomes
Currie, 2004	60	CBT-I vs self-help with telephone support vs waiting list	5 sessions delivered individually	Wrist actigraphy, PSQI post treatment	Improved SOL (− 17.9 vs 20.9), sleep efficiency (10.4% vs 0.5%), and sleep quality of life (− 6.6 vs − 2.3) with CBT-I relative to individuals on the waiting list. No impact in preventing relapses
Arnedt, 2011[44]	17	CBT-I vs BPT	8 sessions delivered individually	ISI, sleep diary post treatment	Improved daytime ratings of insomnia severity (− 13.6 vs − 7.6), sleep efficiency (18.8% vs 5.7%), WASO (− 79 vs 7 min), and fatigue with CBT-I compared to BPT. No difference in relapse rate
Chakravorty, 2019	22	CBT-I vs MO	8 sessions delivered individually	ISI, DBAS, SHI post treatment	Greater improvement in ISI (− 12.7 vs − 8.4), sleep latency (− 45 vs 20.6 min), DBAS mean score (− 1.8 vs 0.2), and SHI (− 4.6 vs − 0.8) with CBT-I compared with MO. No difference in percent day abstinent
Verlinden, 2023	70	SHUTi vs PE	6 cores delivered online	ISI, PSQI, PSS post intervention	Reduction in insomnia symptoms (ISI − 10.5 vs − 6.8), improved sleep quality of life (PSQI − 7.4 vs − 4.5) and decreased alcohol consumption
Miller, 2023	67	CBT-I vs sleep hygiene	5 sessions delivered individually	ISI post treatment	Greater decrease in insomnia severity with CBT-I compared to sleep hygiene (ISI − 12.28 vs − 8.3). No difference in abstinence or heavy-drinking frequency

CBT-I cognitive behavioral therapy for insomnia, BPT behavioral placebo treatment, MO monitor-only, WASO wake after sleep onset, DBAS dysfunctional beliefs and attitudes about sleep, SHI sleep hygiene index, ISI insomnia severity index, SHUTi sleep healthy using the internet, PE patient education

While CBT-I is effective in relieving symptoms of insomnia [44, 45], pharmacotherapy is advocated for patients who continue to show symptoms of daytime dysfunction and/or psychosocial distress 4 weeks after alcohol cessation [48]. A list of pharmaceutical agents tested in RCTs for the treatment of AUD is displayed in Table 3. These agents are aimed at either increasing GABAergic transmission or influencing glutamatergic transmission. Benzodiazepine receptor agonists have been traditionally prescribed as the first-line treatment for insomnia; however, the benefits of these agents should be balanced against the risks of abuse, overdose, and alcohol relapse. Trazodone is often prescribed in tandem with AUD, because it is non-addictive, and is not associated with abuse liability, or life-threatening withdrawal syndromes. Two placebo-controlled studies of trazodone initially showed promise in terms of sleep outcomes [49, 50]. These trials enrolled sleep-disturbed alcohol-dependent patients following detoxification. In a small study involving 16 alcohol-dependent patients after detoxification, trazodone reduced awakenings and enhanced sleep maintenance [50]. A larger randomized, double-blind, controlled trial of low-dose trazodone (50–150 mg) administered to alcohol-dependent patients for 12 weeks confirmed that trazodone was associated with reduced awakenings and improved subjective sleep quality as evidenced by the mean change between baseline and 3 months on the Pittsburgh Sleep Quality Index [49]. Nonetheless, the results identified a concerning trend in terms of alcohol abstinence. There was a decrement in abstinence days in the trazodone group associated with an increase in the number of drinks on cessation of the study medication. Although the reasons for these observations are unclear, it was speculated that m-chloro-phenylpiperazine, a trazodone’s metabolite, mimics ethanol and elicits alcohol craving in those undergoing detoxification [51]. For clarification, neither of these two studies used a validated insomnia scale and the observed benefits dissipated the 3- to 6-month interval. For these reasons, trazodone cannot be endorsed as a viable therapy for insomnia during the period after detoxification from alcoholism.

Table 3.

Randomized-controlled trials of pharmacotherapy for insomnia in individuals with alcohol use disorder

Author, year	N	Medication and dosage	Assessment	Outcomes
Anticonvulsants
Malcom, 2007 [55]	75	Gabapentin 1200 for 3 days or 900 mg for 3 days or 200 mg for 3 days vs lorazepam 6 for 3 days	Item 11 of CIWA	No significant difference in insomnia between the gabapentin and lorazepam in insomnia except for rebound increase in those who received lorazepam after discontinuation
Johnson, 2008 [76]		Topiramate (up to 300 mg/day) vs placebo for 14 weeks	Medical Outcomes Study Sleep scale	Reduction of sleep disturbance with topiramate compared to placebo
Brower, 2008 [57]	21	Gabapentin 1500 mg vs placebo for 6 weeks	SPQ, PSG	No improvement in Sleep Problem Questionnaire scores between gabapentin and placebo. PSG-measured sleep parameters, TST, WASO, and SOL did not show improvement from baseline to 3 weeks in either treatment group
Anton, 2009	60	Flumazenil (2 mg for 2 consecutive days) and gabapentin (up to 1200 mg nightly for 39 days) or their inactive placebos	ISI, PSQI, ESS, CIWA	Improved insomnia in patients with low withdrawal symptoms on gabapentin compared versus placebo. Lower propensity for sleepiness in gabapentin group versus placebo
Anton, 2011	150	Naltrexone alone (50 mg/day), naltrexone with gabapentin (up to 1200 mg/day for 6 weeks, or double placebo	ISI, BDS, CIWA	No significant difference between the groups in ISI or BDS although the naltrexone/gabapentin group reported significantly better sleep than either the placebo alone group or the naltrexone-alone group
Mason, 2014	150	gabapentin 900 mg, 1800 mg or placebo for 12 weeks	PSQI, BDS, SAFTEE-GI	Gabapentin showed significant linear dose effects on mood, sleep quality, and relapse-related symptoms of insomnia
Antidepressants
Le Bon, 2003	16	Trazodone 50–200 mg versus placebo for 4 weeks	PSG	Improved sleep efficiency, number of awakenings, intermittent wake sleep time, and non—rapid eye movement sleep in the trazodone group compared to placebo
Friedmann, 2008	173	Trazodone (50–150 mg at bedtime) versus placebo for 12 weeks	PSQI	Improved sleep quality more for the trazodone group compared to the placebo but decreased abstinence days
Antipsychotics
Litten, 2012 [77]	224	Quetiapine XR 400 mg or placebo for 12 weeks	PSQI	Improved sleep quality with quetiapine compared to placebo
Chakravorty, 2014 [75]	20	Quetiapine XR 400 mg or placebo for 8 weeks	ISI, PSQI	Improvement in ISI but not sleep quality
Melatonin receptor agonists
Gendy, 2020 [66]	60	Melatonin 5 mg or placebo for 4 weeks	PSQI	Reduction in the global PSQI score in both groups with no significant drug effect between groups
Other agents
Staner, 2006 [70]	24	Acamprosate 666 mg three times a day versus placebo for 15 days	PSG	No difference in SOL but improved total sleep time and sleep efficiency with acamprosate
Perney, 2012 [71]	592	Acamprosate 2–3 g/day versus placebo for 24 weeks	SSI	Improved sleep disturbances compared to placebo
Petrakis, 2016	96	Prazosin 16 mg or placebo for 13 weeks	PSQI, CPAS	No advantage of prazosin over placebo in treating sleep disturbance, including difficulty falling or staying asleep

CAPS clinician administered PTSD scale, CIWA clinical institute withdrawal assessment-alcohol, PSQI Pittsburg sleep quality index, ISI the insomnia severity index, BDS beck depression scale, SAFTEE-GI systematic assessment for treatment emergent events-general inquiry, SSI short sleep index, SOL sleep onset latency, WASO wake after sleep onset, TST total sleep time

Gabapentin is another agent that has been promoted off-label for the treatment of insomnia in AUD. By blocking voltage-dependent calcium channels via binding to its alpha-2-delta type 1 subunit, it increases brain concentrations of gamma-amino-butyric acid (GABA) leading to a reduction in postsynaptic excitability and a decrease in the release of excitatory neurotransmitters [52]. These effects are thought to ameliorate sleep disturbances among patients with AUD [53]. Initial report derived from an open pilot study established the superiority of gabapentin in treating insomnia of alcoholic patients compared to trazodone [54]. Similarly, gabapentin was found to be comparable to lorazepam for controlling insomnia symptoms in patients seeking treatment for alcohol withdrawal [55]. Subsequent randomized trials cast doubt on the efficacy of gabapentin for the treatment of insomnia in alcoholic patients [56, 57]. The first RCT recruited 21 subjects who met study criteria for alcohol dependence and insomnia [57]. Patients were randomized to either placebo or gabapentin for 6 weeks and titrated over 10 days to 1500 mg or 5 pills at bedtime. Although gabapentin significantly delayed the onset of heavy drinking, it had no effect on effects on sleep as measured by either subjective report or polysomnography. The second RCT found that neither gabapentin nor valproic acid ameliorated the symptoms of depression or sleep disturbances in the post-withdrawal period [56]. In view of these conflicting results, a consensus on gabapentin therapy for AUD could not be established given the small sample size, methodological, and dosing concerns of these preliminary studies. Three subsequent randomized trials established the beneficial effects of gabapentin in alcoholic outpatients with insomnia, with no evidence of tolerance to gabapentin [58–60]. Two of these trials used a combination therapy with gabapentin consisting of intravenous flumazenil or naltrexone [58, 59]. The third trial was a 12-week, double-blind, placebo-controlled trial examining two dosages of gabapentin (900 or 1800 mg) versus placebo in 150 participants with AUD [60]. Gabapentin was effective in a dose-dependent fashion in treating alcohol dependence and reducing relapse-related symptoms of insomnia. Of note, an extended-release formulation of gabapentin (GE-XR) was recently approved by the FDA for the treatment of restless leg syndrome and post-herpetic neuralgia. In an RCT designed to test the efficacy and safety of GE-XR as a treatment for AUD, GE-XR at 600 mg twice a day did not reduce alcohol consumption or elicit significant improvement in sleep symptoms [61]. One study has suggested that gabapentin-treated alcohol-dependent patients with insomnia were less likely to experience residual sleepiness upon awakening than those who received trazodone [54]. However, there is no head-to-head comparison between these two agents to validate this assertion in patients with AUD.

While disruption of circadian rhythms is not routinely considered in the pathogenesis of chronic alcoholism, it is becoming more apparent that a bidirectional association between alcohol abuse and disruption of the circadian clock exists that feeds on the derangement of both entities leading to a self-perpetuating cycle of sleep disruption and alcohol addiction [62]. Interestingly, melatonin secretion is suppressed by alcohol ingestion [63]. With actual evidence suggesting that restoring circadian rhythm is associated with decreased addiction vulnerability, [64] administering melatonin or melatonin receptor agonist could be an appealing choice for the treatment of sleep disturbances of AUD given the safety profile of these agents compared to benzodiazepines or Z-drugs. A preliminary open-label study recruited 5 AUD participants who had initiated abstinence to evaluate the efficacy of ramelteon, a melatonin receptor agonist, on SOL and TST [65]. Subjects received 8 mg nightly for 4 weeks. In comparison to baseline, insomnia was markedly reduced by both subjective and objective measures. However, in a pilot RCT of 60 subjects seeking AUD treatment, administration of melatonin 5 mg 1 h prior to bedtime did not differ from placebo in altering sleep quality as assessed by PSQI, including the insomnia domain [66]. Future RCTs are needed before a firm recommendation on ramelteon is made. Alternatively, the application of light therapy has been suggested as a potential treatment to realign the circadian disruption in patients with AUD. It is speculated that bright light therapy may stabilize sleep architecture and lessen sleep fragmentation during alcohol withdrawal. A pilot study using bright light therapy (BLT) (3000 lx) was administered to 10 alcohol-dependent patients on day 3 of abstinence from 7:00 AM to 9:00 AM and from 5:00 PM to 9:00 PM. [67] Sleep onset latency and time to slow wave sleep onset were significantly reduced, while rapid eye movement (REM) sleep percentage improved on the nights following BLT. However, it should be stated that the results of BLT are considered modest in magnitude. Its indication may be of complementary value with other therapeutic modalities due to its rapid onset of action and safety profile.

Acamprosate, an FDA-approved medication for the treatment of alcoholism, is a glutamatergic inhibitor that has been shown to reduce insomnia. Its mechanism of action involves dampening neuronal hyperexcitability by decreasing the level of glutamate activity in NMDA receptors via modulation of calcium channels [68]. These actions are hypothesized to lead to the reduced positive reinforcement of alcohol consumption and attenuated cravings during withdrawal [69]. A parallel, double-blind placebo-controlled study involving 24 male alcohol-dependent patients who received acamprosate 666 mg three times a day versus placebo showed that acamprosate decreased wake time after sleep onset (WASO) and REM sleep latency but had no effect on sleep-onset latency [70]. In a subsequent RCT of 592 patients, acamprosate (2–3 g per day) resulted in reduced sleep-disturbing symptoms, including insomnia, after alcohol withdrawal compared with placebo [71]. A meta-analysis of 12 studies involving 3508 patients with AUD of whom 59.8% had insomnia at baseline documented approximately twofold reduction in insomnia severity after 6 months over alcohol abstinence alone (change in mean Short Sleep Index by − 45% for acamprosate versus − 26% for placebo) [72]. However, there is no robust evidence to suggest that acamprosate sustains abstinence [73]. Acamprosate is considered a first-line treatment for reducing relapse from alcohol abuse and should be considered the primary agent when comorbid insomnia is present. The drug is devoid of potential abuse or dependence. Reported adverse events include renal impairment, allergic reactions, and gastrointestinal manifestations. Several other studies have tested off-label pharmacologic agents, such as quetiapine, prazosin, or topiramate in treating sleep disturbances of AUD [74–78]. The results of these studies should be interpreted with caution given that the impact of these agents on insomnia was inconclusive, or inferred. Besides, the risks of adverse events (weight gain, diabetes, and orthostatic hypotension) from these drugs may outweigh any potential sleep benefits.

Given the role of orexin-1 receptors in mediating the motivation and craving for alcohol and orexin-2 receptors in modulating wakefulness, dual orexin receptor antagonists have gained significant interest as a potential treatment for insomnia in AUD. A phase II, placebo-controlled, double-blind randomized trial of suvorexant (20 mg once daily) is currently recruiting adult patients between 18 and 75 years of age with a diagnosis of insomnia after being admitted for alcohol withdrawal. Treatment is initiated once the breath alcohol level is undetectable. Benzodiazepines are only administered if the Clinical Institute Withdrawal Assessment for Alcohol scale is greater than 10. Outcome measurements are performed weekly for the first 4 weeks post-inpatient treatment, and then every 4 weeks until week 25. The primary objective of the study is to evaluate the efficacy of suvorexant on sleep parameters during acute inpatient alcohol withdrawal including those derived from objective (actigraphy, portable polysomnography) and subjective (Insomnia Severity Index and the Pittsburgh sleep quality index) instruments. Secondary objectives comprise the efficacy of suvorexant during 6 months after alcohol withdrawal on sleep indices and drug use parameters including craving and abstinence [79].

On a positive note, a recent phase 2 study of an investigational treatment (Sunobinop) for patients with insomnia while recovering from AUD appeared online. Sunobinop is a first-in-class small molecule whose mechanism of action involves binding to the nociceptin/orphanin-FQ peptide receptor (NOP), a protein most recently discovered member of the opioid receptor superfamily that also includes μ, δ, and κ opioid receptor subtypes [80]. The NOP receptor has a wide array of biological functions including modulation of anxiety, response to stress and substance abuse. The randomized, double-blind, multicenter, placebo-controlled, parallel-group clinical study enrolled 114 people assigned at a 1:1:1 ratio to receive sunobinop in either 1 mg or 2 mg doses or placebo every night at bedtime for 21 days [81]. At the end of the study, both doses of sunobinop achieved a significant reduction in wakefulness after sleep onset compared with placebo. Sunobinop was well tolerated across both doses. Adverse events were considered mild to moderate with somnolence being reported in 5% and 26% with 1 mg and 2 mg doses, respectively. No increase in craving was observed in any treatment group.

Treatment of insomnia in opioid use disorder

The term “opiates” is used to refer to compounds derived from the poppy plant (Papaver somniferum) (morphine, codeine). Over the last few decades, the terminology expanded to semisynthetic (dihydromorphine, heroin, and methyldihydromorphine) and synthetic opiates, which are chemically synthesized analgesics whose effects are similar to morphine (propoxyphene, methadone, meperidine, fentanyl analogs, and pentazocine) in interacting with brain receptors in the brain. About 296 million people worldwide (or 5.8% of the global population) reported using drugs at least once in 2021 [82]. Of those, 60 million people experimented with opioids. In the United States, the number of individuals lost to drug overdose surpassed 70,500 in 2019 [83]. Half of these deaths were attributed to synthetic opioids. From 2013 to 2019, the age-adjusted synthetic opioid death rates in the United States increased by 10–40%. The COVID-19 pandemic amplified these statistics due to the widespread mixture of potent synthetic compounds and reduced drug enforcement.

Opioid intake alters sleep architecture even with a single oral dose. Although a short course of opiates for pain control is associated with improved sleep quality and total sleep time, increased sleep latency and time awake after sleep onset are frequent complaints [6]. In chronic opioid users, actigraphy data demonstrate reduced total sleep time, slow wave, and REM sleep when compared with healthy controls [84]. Similarly, acute opioid withdrawal engenders significant changes in sleep patterns. Five-to-seven days after cessation of heroin use, there is an increase in sleep onset, WASO, and a decrease in total sleep time, slow wave, and REM sleep. Sleep disturbances are also commonly observed in patients undergoing opioid detoxification with estimates ranging from 50 to 75% in various samples of patients undergoing treatment for opioid addiction [85]. Sleep complaints persist for months and even years after initiation of medications for opioid use disorder (OUD) [86]. Approximately three-quarters of patients receiving buprenorphine, extended-release naltrexone, or methadone maintenance treatment for OUD (MOUD) experience insomnia [87]. Methadone-maintained patients exhibit reduced sleep efficiency, increased sleep latency, and poor sleep quality [85]. Factors associated with persistent insomnia in OUD include depression, cigarette smoking, and benzodiazepine use [88]. Evaluation of these underlying conditions should be thoroughly sought as more than half of those prescribed methadone maintenance therapy report using illicit drugs to help with sleep despite greater functional impairment [89].

A public forum organized by the FDA identified insomnia as a primary contributor to relapse, persistent opioid cravings, emotional dysregulation, and heightened pain sensitivity [32, 90]. Although CBT-I efficacy in improving sleep-related outcomes has been extensively documented in the general population with insomnia [91], there is only one RCT to our knowledge testing the utility of CBT-I in individuals with OUD [92]. The double-blind randomized trial recruited 22 patients undergoing methadone maintenance therapy. Participants were randomized to either 8 weeks of individual CBT-I sessions or behavioral placebo therapy (BPT). The results showed improvement in sleep duration, sleep efficiency, and daytime dysfunction favoring CBT-I over BPT. However, no significant difference in subjective sleep quality or sleep latency was observed between both groups. The use of an active placebo (BPT) instead of a waiting list and the small sample size could have contributed to the study's shortcomings. It also remains unclear whether behavioral therapy is adequate or efficacious in altering drug use outcomes. Nonetheless, the amelioration in sleep quality and daytime function after CBT-I can be instrumental in strengthening social construct and reducing relapse.

Commonly, sleep medications are prescribed to patients with insomnia receiving treatment for opioid abuse (Table 4). Analysis of combined administrative claims for MOUD, including commercial and multi-State Medicaid databases, showed that prescriptions for hypnotics often consisted of three or more different agents [93]. Benzodiazepines were prescribed for 69.1% of all patients during the 60 days following buprenorphine initiation while 40.9% received Z-drugs. Although benzodiazepines have historically been used for the treatment of insomnia, chronic use of these agents can initiate a vicious cycle whereby tolerance to the hypnotic effect triggers a relapse of opioid dependency which in turn exacerbates insomnia symptoms. While OUD retention in care may initially improve and insomnia symptoms abate in the short term, overall mortality is also increased with concurrent benzodiazepines [94]. The Z-drugs (i.e., zolpidem, eszopiclone) exhibit comparable side effects with BZDs including risk of falls, cognitive impairment, and somnambulism. A recent randomized, double-blind pilot, crossover trial examined 1 week each of zolpidem sustained release (12.5 mg), mirtazapine (30 mg), zolpidem plus mirtazapine (12.5/30 mg), or placebo in 10 methadone-maintained subjects with insomnia [95]. Mirtazapine was superior to the other arms of the study in sleep efficiency, sleep latency, and wake after sleep onset. Zolpidem was consistently reported to have the poorest sleep outcomes of any of the other regimens, including placebo. Given the respiratory depression when co-administered with opiates, and the limited efficacy of benzodiazepines/Z-drugs, GABA receptor agonists are not recommended as a routine modality to improve sleep in this population. Further, the relatively small sample size and the self-reported adherence data preclude a firm recommendation on the use of mirtazapine in this population until a more definitive trial with a larger population is conducted.

Table 4.

Randomized-controlled trials of pharmacotherapy for insomnia in individuals with opioid use disorder

Author, year	N	Medication and dosage	Assessment	Outcomes
Antidepressants
Stein, 2012 [96]	137	Trazodone 50 mg or placebo for 6 months	PSQI, PSG	Trazodone did not improve subjective or objective sleep or sleep quality of life in methadone-maintained persons with sleep disturbance compared to placebo
Stein, 2020 [95]	10	One week each of zolpidem CR 12.5 mg, mirtazapine 30 mg, and zolpidem 10 mg plus mirtazapine 30 mg	Wrist actigraphy	Mirtazapine superior in improving SOL, wake minutes, sleep efficiency, and WASO compared to zolpidem or combination mirtazapine/zolpidem
Melatonin receptor agonists
Ghaderi, 2019 [114]	54	Melatonin 10 mg or placebo for 12 weeks	PSQI	PSQI scores decreased significantly with melatonin supplementation
DORA
Huhn, 2022 [98]	38	Suvorexant 20 mg, 40 mg, or placebo for 10 nights	Wrist actigraphy and three-lead electroencephalography	Increased objective TST and total REM time with no difference in SOL, WASO, or sleep efficiency between suvorexant and placebo during buprenorphine/naloxone taper

SOL sleep onset latency, WASO wake after sleep onset, TST total sleep time, DORA dual-orexin receptor antagonist, PSQI Pittsburg sleep quality index, PSG polysomnography

Trazodone, a triazolopyridine derivative, is among the most commonly prescribed off-label medications for the treatment of insomnia in the United States. It is widely used as a hypnotic drug in sub-therapeutic antidepressant doses of 100 mg or less. Consideration for its use in patients on methadone maintenance therapy has been evaluated in a double-blind, placebo-controlled trial [96]. Subjects were randomized to self-titrating dose of trazodone (50–150 mg) or placebo. At the 6-month follow-up, trazodone had no significant impact on objective or subjective sleep quantity or sleep quality. It neither significantly increased nor decreased illicit drug use relative to placebo [96]. In view of prior investigations suggesting a role of melatonin in suppressing the rewarding behavior associated with cocaine [95] and alcohol [97], 54 patients under methadone maintenance treatment were recruited for a randomized, double-blind, placebo-controlled trial to receive either 10 mg of melatonin or a placebo for 12 weeks. Compared to the control group, melatonin decreased anxiety, depression, and sleep quality of life as evidenced by the difference in PSQI, Beck Depression Inventory (BDI), and Beck Anxiety Inventory scores. The study did not include a specific instrument for insomnia; however, given the favorable safety profile of melatonin, it may represent a safe alternative to hypnotics or antidepressants.

With preclinical evidence suggesting that orexin receptor antagonists reduce opioid withdrawal symptoms [24], targeting the orexin signaling system appears to be the preferred approach to address the symptoms of insomnia which can translate into collateral benefits on OUD treatment outcomes [32]. Suvorexant was recently shown in an escalating dose to alleviate craving, and opioid withdrawal during buprenorphine withdrawal [98]. Patients with OUD stabilized on buprenorphine/naloxone therapy were randomized to either suvorexant 20 mg, 40 mg, or placebo. A 4-day stepwise taper of buprenorphine/naloxone starting on day 1 was initiated followed by a 4-day post-taper observation period. Although insomnia was not included in the inclusion criteria, objective total sleep time was significantly prolonged during both the taper and the post-taper period. In addition, both objective sleep-onset latency and wake after sleep onset did not differ from those receiving a placebo. There were no meaningful differences between the 20 and 40 mg of suvorexant. Interestingly, patients reported on their sleep diaries longer TST and reduced SOL with suvorexant in relation to placebo. There were no serious adverse events and no participant expressed suicidal ideations. Suvorexant exhibited a low risk of abuse and diminished opioid craving. Two other DORAs, lemborexant and daridorexant, have been lately approved by the FDA for the treatment of insomnia [99, 100]. Although their efficacy is yet to be examined in patients with OUD, both are considered to show dose-dependent sleep-promoting effects with low abuse potential [101, 102].

Future directions and opportunities for research

Considering the unabated substance abuse crisis post-COVID-19, there is a pressing need to address sleep disturbances in tandem with other emotional and physical domains if treatment outcomes related to substance abuse disorders are to improve. In clinical practice parameters, a stepped-care approach has been advocated in managing insomnia [103] (Table 5). The success of such an approach in achieving remission after SUD has not been properly quantified, although it is clear from sleep outcome studies tackling insomnia that a one-size-fits-all approach may not be appropriate. Insomnia has long been treated as a homogenous disorder, and traditionally, management has focused on addressing insomnia only after the precipitating factors are removed or resolved. Contemporary evidence suggests otherwise, as insomnia may emerge long after the underlying comorbid condition has been adequately treated. Substance abuse disorders add a third dimension to the complexity of management. (Fig. 2) Not only do drug-seeking behavior and circadian rhythm disruption promote difficulty falling asleep or staying asleep, but abstinence, whether self-imposed or chemically induced, can also lead to flare-ups of insomnia. The worse the sleep disturbance is, the higher the likelihood of relapse. Yet, this complex interaction is still being approached as a binary construct. Current studies invoking insomnia in patients with SUD have focused on a single outcome (either recovery or relapse) by conflating readiness for treatment of SUD with adoption of adequate sleep hygiene often dismissing potent involuntary responses to substance-related cues [104]. Strategies targeting sleep-restoration outcomes that facilitate opioid abstinence, diminish cravings, and prevent relapse may yield more clinically significant outcomes than relying on the premise of complete resolution of addiction. Another notable challenge in achieving meaningful advances in managing comorbid insomnia with SUD is the heterogeneity of the study population. First, the comorbidity of coexisting sleep disorders (i.e., obstructive or central sleep apnea, restless leg syndrome) in patients with comorbid insomnia and SUD is often not considered in the initial evaluation. As an example, a recent systematic review and meta-analysis of 21 studies assessing the relationship between alcohol intake and sleep disorders demonstrated that alcohol consumption increased the risk of sleep apnea by 25% [105]. Failure to account for these disorders introduces inherent variability that leads to misestimation of the tested therapy. Second, previous studies have encompassed insomnia patients at different stages of their addiction. The neurobiological circuitry involved during chronic substance abuse may not necessarily reflect the pathways that are engaged during acute withdrawal or extended abstinence. Third, sleep efficacy trials among patients with SUD have relied on a wide range of behavioral instruments or quality-of-life measurements that made comparison of sleep outcomes between these studies inconsistent and non-generalizable.

Table 5.

Clinical recommendations for managing insomnia in patients with substance abuse disorders

Screen for insomnia among patients diagnosed with SUDs

Conduct a comprehensive assessment of patient’s sleep patterns, substance abuse history, and other comorbid psychiatric conditions

Exclude other causes of sleep dysfunction (e.g., sleep apnea, restless leg syndrome)

Treat insomnia and substance abuse disorders concurrently to improve outcomes of both conditions

Administer CBT-I as the first-line treatment for chronic insomnia

Incorporate mindfulness meditation and guided imagery to reduce arousal and promote sleep quality

Avoid benzodiazepine receptor agonists due to their high potential for abuse and dependence

Selecting a pharmacological agent should take into consideration coexisting psychiatric illnesses, safety of the agent prescribed, and potential for preventing relapse

Schedule regular follow-up appointments to monitor treatment progress, adherence, and side effects

Fig. 2.

Conceptual model linking insomnia to substance abuse disorders during addiction and post-detoxification reflecting the bidirectional associations between sleep disturbances and addiction. SUD substance use disorder

Given these limitations, general guidance on the management of insomnia in SUD is lacking. CBT-I is considered the first line of therapy due to its superior safety profile despite the scarcity of clinical trials in this domain. It is recognized at least in patients with AUD that CBT-I is more effective in ameliorating sleep quality than pharmacotherapy [106]. Although CBT-I may confer self-control skills to assist patients overcome derangement in sleep homeostasis, CBT-I does not appear to influence drug-seeking behavior, prevent relapse, or facilitate sobriety [106]. Based on the existing literature, combination therapy of behavioral and pharmacological interventions has long been recognized as a standard approach in addiction disorders [107, 108]. The efficacy of such a regimen in SUD with comorbid insomnia has not been firmly established. A key element in the approach to combination therapy is the sequence of treatment initiation. Optimal treatment may ultimately be dictated by whether insomnia is present at baseline (acute withdrawal) or whether it emerges during abstinence suggesting that symptoms of insomnia should be discerned during the initial evaluation and monitored over time. Initiating CBT-I and pharmacotherapy treatment concurrently seems more plausible as most patients with SUD who do not exhibit overt insomnia have subclinical symptoms. [109] Theoretically, a short course of a sedative agent could alleviate insomnia while patients are learning to integrate newly developed management skills in addressing residual symptoms. Further, a higher adherence to behavioral treatment sessions may be realized as patients perceive ample opportunities to interact with clinical providers. This increased attention may in turn allow the early evaluation of treatment response.

While CBT-I remains the “gold standard” treatment for insomnia, other forms of psychotherapy have started to garner interest given that adherence to the treatment recommendations in CBT-I can be a challenge and some patients drop out of treatment early [110]. Treatments such as Acceptance and Commitment Therapy (ACT) have recently been applied to insomnia, because they do not include the typical components of treatment that initially reduce total sleep time or quality of sleep (e.g., sleep restriction and stimulus control), and instead focus on mindfulness, acceptance, and cognitive defusion to change the relationship that individuals have with their thoughts and emotions. While ACT has demonstrated efficacy for many psychiatric disorders, limited research exists on its application to insomnia. Research is already underway, however, which will yield important information for the future treatment of insomnia in SUDs.

In terms of insomnia pharmacotherapy for patients with SUD, there is a consensus that selecting the appropriate agent should be influenced by the effects of the drug on sleep parameters and the potential for relapse. An ideal compound would pose no detrimental effect, while it readjusts circadian rhythmicity and improves sleep architecture during abstinence. Benzodiazepine receptor agonists have been the most widely prescribed hypnotic agents for insomnia and are highly effective in patients experiencing acute alcohol withdrawal [111]. However, the use of these agents in SUD is fraught with significant risks given the abuse potential, tolerance to hypnotic effect, and risk of overdose. Similarly, Z-drugs should be avoided in these groups of patients given the nocturnal complex behaviors and potentially fatal overdoses [112]. Although gabapentin has been shown to ameliorate acute alcohol withdrawal by normalizing GABA and glutamate balance, its role in treating insomnia may be limited to those who have low pretreatment withdrawal symptoms [59] or who develop restless leg syndrome. Conversely, the use of gabapentin for the treatment of insomnia in OUD should be avoided due to the risk of opioid-aggravated respiratory depression. At present, the most promising compounds for the management of SUD comprise the DORAs for countering the symptoms of insomnia while alleviating drug cravings and mitigating relapse vulnerability. However, additional investigations are required to explore the optimal timing of administration during addiction cycles and whether the response to these agents may vary with sex given that the orexin system exhibits sexual dimorphism [113].

The strength of this narrative review is the broad and up-to-date review of existing literature, integrating diverse studies and perspective, to assist clinical practitioners in designing a therapeutic plan for insomnia that can be mutually agreed upon with their patients. The majority of selected studies also used well-validated measures of sleep. That said, we acknowledge some limitations. First, non-English publications were excluded due to language constraints. Second, the operational definitions and measurements of insomnia varied across studies, thus limiting the comparability of findings. Third, little is known about the similarities and differences in insomnia phenotypes between AUD and OUD. Most of the published studies exploring the association between sleep disturbances and SUDs have collected sleep data from patients with a single type of SUD. Fourth, synthesizing evidence from a variety of study designs always carries the risk of overgeneralization.

Conclusion

Insomnia is a common symptom of sleep disturbances related to substance abuse disorders and a strong indicator of relapses. Hitherto, there have been relatively few high-quality studies in SUD populations that target insomnia and drug-related outcomes. Substances of abuse induce considerable impairments of sleep homeostatic sleep drive including perturbations of sleep/wake patterns and dysregulation of neurotransmitter systems. In both AUD and OUD, disruption of circadian rhythms and shorter sleep time are associated with impulsive behaviors and social maladaptation. Effective management of insomnia in these cases requires a comprehensive, integrated approach that combines non-pharmacological and pharmacological treatments. Future studies should also seek to recruit SUD patients with insomnia free of other sleep disorders to assess both sleep disturbances and drinking outcomes and investigate which treatments work best for which patients.

Funding

Dr. El-Solh is supported by a grant from the Clinical Science Research and Development (CX001656) of the U.S. Department of Veterans Affairs. The funders had no role in the study design; the collection, analysis, and interpretation of data; the writing of the report; the decision to submit the article for publication.

Declarations

Conflict of interest

None.

Ethical committee permission

Not applicable.