Can the Orexin Antagonist Suvorexant Preserve the Ability to Awaken to Auditory Stimuli While Improving Sleep?

Abstract

Study Objectives:

The safety profile of the dual orexin receptor antagonists (DORAs) are currently unknown with regard to nocturnal responsivity among people with insomnia. We compared the auditory awakening thresholds (AATs) of the DORA suvorexant (10 and 20 mg) versus placebo in 12 individuals with DSM-5 insomnia.

Methods:

The study used a double-blind, placebo-controlled, three-way crossover design. Participants were randomly assigned to a treatment sequence that included placebo, suvorexant 10 mg, and suvorexant 20 mg. At the time of maximum drug concentration, auditory tones were played during stable stage N2 sleep. Tones increased by 5-decibel (db) increments until the participant awakened. The db at awakening was recorded as the AAT and compared between conditions. The proportion of awakenings higher than 85 db was also compared between conditions. Finally, sensitivity analyses were also conducted using surrounding thresholds (80 db and 90 db).

Results:

The mean AAT did not differ significantly between either dose of suvorexant compared to placebo. Moreover, the proportions of individuals who remained asleep at the AAT 85 db cutoff did not differ across conditions. In addition, wake after sleep onset decreased and total sleep time increased in the suvorexant 20 mg condition compared to placebo.

Conclusions:

Suvorexant (10 and 20 mg) preserved the ability to respond to nocturnal stimuli, whereas the 20-mg dose improved the sleep of people with insomnia. This suggests that DORAs such as suvorexant can effectively treat insomnia while allowing patients to awaken to nocturnal stimuli in the environment.

Clinical Trial Registration:

Registry: ClinicalTrials.gov; Title: A Phase IV 3-Way Double-blind, Randomized, Crossover Study to Compare the Awakening Threshold Effects (Responsivity) of Belsomra 10 mg and 20 mg to Placebo in Non-elderly Insomniacs; Identifier NCT03312517; URL: https://clinicaltrials.gov/ct2/show/NCT03312517.

Citation:

Drake CL, Kalmbach DA, Cheng P, Roth T, Tran KM, Cuamatzi-Castelan A, Atkinson R, SinghM, Tonnu CV, Fellman-Couture C. Can the orexin antagonist suvorexant preserve the ability to awaken to auditory stimuli while improving sleep? J Clin Sleep Med. 2019;15(9):1285–1291.

BRIEF SUMMARY

Current Knowledge/Study Rationale: This study was performed to investigate the safety of the dual orexin receptor antagonist, suvorexant, in terms of nocturnal responsivity to auditory stimuli.

Study Impact: By showing that suvorexant improves sleep while maintaining the ability to awaken to auditory stimuli, this study indicates that it has a relatively safe nocturnal responsivity profile.

INTRODUCTION

Safe and effective pharmacological treatment of insomnia requires the use of therapeutic approaches that improve the ability to initiate and/or maintain sleep while minimizing alterations in natural sleep including responsivity. The most commonly prescribed hypnotic medications are benzodiazepine receptor agonists (BzRAs).¹ Despite the efficacy of this class of hypnotic drugs, BzRAs suppress the ability to awaken during the night in response to stimuli.^2,3 Indeed, an early study showed that 50% of individuals taking the hypnotic drug triazolam were unresponsive to multiple 60-second 78-decibel (db) alarms, whereas all individuals in the placebo group routinely awoke to the auditory stimuli.⁴ In a subsequent study, 64% of young, healthy, good-sleeping men who were administered 10 mg of zolpidem, the most commonly used hypnotic drug, were unresponsive to brief auditory stimuli of 105 db during sleep.⁵ Because many individuals with insomnia taking BzRAs do not wake up readily to nocturnal auditory stimuli, commonly prescribed hypnotic medications carry significant safety risk for the potential impaired ability to awaken to external stimuli such as smoke alarms, as well as other stimuli (eg, caregiver needs, intruders, etc.) and even everyday stimuli (eg, alarm clocks).

The newest class of Food and Drug Administration (FDA)-approved hypnotic medications are the dual orexin receptor antagonists (DORAs).^6–8 Although BzRAs target gamma-aminobutyric acid-ergic receptors throughout the central nervous system (CNS), acting as CNS depressants on a variety of functions, DORAs activate a more focused set of lateral hypothalamic neurons, suggesting the potential for less widespread CNS suppression.⁹ Although numerous studies demonstrate the impairing effects of BzRAs on nocturnal responsivity,^2–4 the effects of DORAs on nocturnal responsivity of humans have not been determined. Therefore, this potentially important aspect of their nocturnal risk-profile remains unknown.

Despite the absence of human studies, the nocturnal responsivity profile of DORAs to auditory stimuli in nonhuman mammals suggests that the awakening threshold may be less affected relative to other hypnotic drugs.^10–12 For example, male rhesus monkeys administered an orexin antagonist (DORA-22: 3 mg, 10 mg, 30 mg) awoke more easily to emotionally salient auditory stimuli (tones that were previously paired with food) than to neutral auditory stimuli (unpaired tones). In contrast, all tested doses of a BzRA hypnotic drug and of a benzodiazepine suppressed the ability to awaken to salient stimuli.¹² Importantly, salient stimuli woke a similar percentage of monkeys following treatment with DORA-22 compared to placebo (∼75% for both). Other results showing similar nocturnal responsivity to emotionally salient stimuli between DORA-22 and placebo have been reported in dogs.¹¹ In the most recent study investigating the effects of DORAs on nocturnal responsivity in animals, investigators found that the latency to awakening in response to an auditory stimulus was similar between DORA-22 and placebo in mice.¹⁰ Taken together, these animal models provide consistent evidence that DORA-22 does not impair nocturnal responsivity in nonhuman mammals during the sleep period.

The current study sought to determine potential differences in nocturnal responsivity to auditory stimuli between an FDA-approved DORA (ie, suvorexant) and placebo in adult patients with insomnia. Specifically, we compared auditory awakening thresholds (AATs) in 12 adult patients with insomnia for placebo, suvorexant 10 mg, and suvorexant 20 mg. Given that animal models show similar nocturnal responsivity profiles between orexin antagonists and placebo, we hypothesized that the AAT would not differ among placebo and both 10-mg and 20-mg doses of suvorexant. In addition, as cognitive arousal (eg, worry, rumination) is associated with greater nocturnal wakefulness in insomnia,¹³ we conducted exploratory analyses to test whether levels of nocturnal worry and rumination were associated with AAT levels on each of the treatment nights.

METHODS

Participants

Patients were 5 males and 7 females, ages 24 to 60 years old, recruited from local advertisements, previous clinical trials, and the sleep medicine clinic of the Henry Ford Health System. The presence of insomnia was determined by clinical interview performed by a physician board certified in sleep medicine who also performed physical examinations to ensure individuals were in good health prior to and after participation. Participants also completed sleep disorder symptoms questionnaires^14,15 to rule out other potential sleep disorders. Study inclusion criteria were a Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) diagnosis of insomnia¹⁶ and an Insomnia Severity Index score > 10.¹⁷ Exclusion criteria included a body mass index > 35 kg/m², current untreated or unstable medical conditions, any clinically significant abnormal finding on physical examination or clinical laboratory tests, apnea-hypopnea index or periodic limb movement arousal index > 10 on polysomnography (PSG), known pharmacological sensitivity/intolerance to suvorexant, history of alcohol or drug abuse, a positive Breathalyzer test for alcohol, a positive urine drug screen (amphetamines, barbiturates, benzodiazepines, cocaine, opiates, or cannabinoids), presence of any hearing abnormalities, smoking more than 15 cigarettes per day (to avoid extended smoke breaks), current use of any CYP3A inhibitor, a history of epilepsy or serious head injury, use of prescribed or over-the-counter CNS-acting medications that may interfere with study drug within 30 days of screening, and a positive pregnancy test at screening.

During the initial screening, participants underwent a hearing test where three 1-second, 1000-Hz, 25-db tones were presented in succession with random interstimulus intervals. Participants were required to make a keyboard response each time they heard a tone. If the participant failed to respond to any of the three tones, the procedure was repeated. Participants were excluded from the study if they failed the hearing test a second time. One participant was excluded from study participation due to failing the hearing test. Another participant discontinued the study prior to completion of the protocol due to adverse events that were experienced during the second treatment night (20 mg suvorexant) and was excluded from analyses. Participants who met all study entry criteria were eligible to enter the PSG screening period. All study procedures were reviewed and approved by the Institutional Review Board at the Henry Ford Health System. All individuals completed informed consent before enrolling in the study.

Procedures

The PSG screening period consisted of 1 night of an 8-hour recording period to rule out sleep apnea and periodic limb movement disorder. Bedtime was scheduled according to the participant’s self-reported average habitual bedtime over the past 2 weeks. Participants were instructed to maintain their habitual bedtime and wake time throughout the study. For all study nights, participants completed brief surveys of worry and rumination before bedtime, and the total PSG recording duration was 8 hours to ensure participants had adequate sleep opportunity. All PSG tests were performed using Grass Technologies, Comet Plus PSG recording equipment (Natus Medical Inc., Pleasanton, California, United States) and were scored by a registered sleep technologist using established standardized procedures.¹⁸ Sleep measures included percentage of N1, N2, N3, non-rapid eye movement, and rapid eye movement (REM) sleep; wake after sleep onset (WASO); latency to sleep, and latency to REM sleep, as determined by standard PSG assessment. PSG sleep was recorded similarly for baseline and treatment nights with the exception that recording of respiration and leg movements did not occur during the treatment nights.

The study used a double-blind, placebo-controlled, three-way crossover design. Participants were randomly assigned to a treatment sequence that included placebo, suvorexant 10 mg, and suvorexant 20 mg with order of treatment determined according to a Latin Square design with each treatment separated by 6 to 12 nights. Auditory awakening procedures were scheduled to occur at the approximate time of maximum drug concentration (∼T-max of suvorexant), which is 2.5 hours following drug administration. Thus, study drug was administered in a single dose 30 minutes prior to bedtime, and awakening procedures commenced 2 hours after bedtime, immediately following the first period of 5 minutes of stable (ie, consecutive) N2 sleep. If participants did not reach stable N2 sleep within a 20-minute window, the AAT procedures were administered from the stage of sleep present at that time.

Auditory awakening procedures followed protocol identical to previous studies in our laboratory.⁵ Arousability was assessed by determining the AAT during each of the three treatment nights. Headphones were calibrated at the beginning of each night to ensure precision. At the appropriate time (following 5 minutes of N2 sleep at the approximate T-max of suvorexant), a 1000-Hz tone with a starting db of 30 was played for 3 seconds through standard earpieces. Time of awakening was determined by behavioral response. Participants were required to verbally respond to the tone presented (“I am awake” or similar response) for an awakening to be recorded. If participants did not respond following the first tone, a second tone 5 db greater in volume was played following a 15 second inter-stimulus interval. This process proceeded in 5-db increments until either the participant responded or until the maximum db level (110 db) was reached. Any participant who did not awaken after receiving the initial 110-db tone received additional 110-db stimuli in 15-second increments for up to 2 minutes; if the participant remained asleep, they were physically awakened by a laboratory technician at that time. The decibel level required to awaken an individual was taken as the AAT for that condition. If the participant did not awaken to any of the tones presented, the maximum 110 db tone was used as the AAT. Once the AAT was determined, participants were instructed to fall back asleep. Adverse events were assessed following each treatment night and were monitored throughout the study. A final study visit was performed, including a second physical evaluation, after each participant had completed all three treatment periods.

Outcome Measures and Statistical Analyses

The primary study endpoint was the AAT in db for each treatment period in a completers dataset. Secondary outcomes tested for significance included the standard PSG-derived sleep parameters of WASO, latency to persistent sleep, and total sleep time (TST). Planned comparisons using pairwise t tests were performed to determine statistical significance between each of the two suvorexant doses and placebo for each of the study outcome measures. Given that standards exist for sound levels of environmental danger warnings (ie, fire alarms), and to test for the robustness of the study results, sensitivity analyses were undertaken using the generalized linear mixed model approach with a logit link function to assess for differences in odds of sleeping through an 85-db stimulus (the equivalent of a standard fire alarm)¹⁹ by condition.

Last, we conducted zero-order correlations between measures of nocturnal worry and rumination and AATs; analyses were run separately for each treatment night. Nocturnal worry and rumination were measured using a brief 5-item version of the Penn State Worry Questionnaire²⁰ and the Ruminative Response Scale’s²¹ brooding scale. Instructions for both measures were modified to specifically assess nocturnal cognitive arousal: “Please read each of the items below and indicate to what degree you’ve been thinking or feeling these things TONIGHT.” Statistical power was computed a priori based on a previous study using zolpidem 10 mg in younger men where a difference of 20 db in AAT was observed. We estimated the power of the current study to be higher than 80% for detecting a mean difference of 10 db between suvorexant and placebo given a sample of 12 individuals.

RESULTS

Only one of the participants was taking a prescription or over-the-counter sleep aid (melatonin) prior to study entry (melatonin was discontinued 3 weeks prior to the screening PSG). Two participants were excluded based on PSG criteria. After a PSG screening to rule out exclusion criteria, a total of 12 participants completed the protocol (age = 36.50 ± 10.96 years; 41.7% male, 58.3% female). Most of the participants identified as white (n = 7, 58.3%), whereas 4 participants identified as black (33.3%), and 1 participant identified as other (8.3%). Mean TST reported by the participants at baseline was 5.74 ± 1.01 hours with an average time in bed of 7.85± .88 hours and a mean sleep efficiency of 73.53% ± .14. The mean Insomnia Severity Index for the sample was 18.25 ± 5.33.

Auditory Awakening Threshold

Means and standard deviations for the auditory awakening threshold for each condition are shown in Table 1. A repeated-measures analysis of variance comparing within-subject differences in AAT across conditions was not statistically significant (F = 1.13; P =.34; Figure 1). Each of the AAT trials was intended to be performed during stable N2 sleep. However, because of the limited time window to capture assessments during probable peak plasma concentrations of study drug, a total of 4/36 AAT trials occurred outside of N2 sleep in 4 separate participants (placebo: REM and N3; suvorexant 10 mg: REM; suvorexant 20 mg: N1). Excluding these observations from AAT analyses did not alter the results. Given the small number of AAT trials performed outside of N2 sleep, no trends in AAT threshold between sleep stages could be observed.

Table 1.

AAT by condition.

Figure 1. Mean decibel level for AAT task in each condition.

AAT = auditory awakening threshold, dbA = decibel, PBO = placebo, SD = standard deviation, SUV-10 = suvorexant 10 mg, SUV-20 = suvorexant 20 mg.

Threshold Differences

In order to explore potential effects of suvorexant on responsivity thresholds, we compared the odds of individuals sleeping through an 85-db stimulus in each condition (Figure 2). A repeated- measures logistic regression was conducted using a generalized linear mixed model with a logit link function to estimate binary outcomes (1 = ATT ≥ 85 db, 0 = ATT < 85 db). Results indicated that neither the 10 mg (odds ratio = 0.42, 95% confidence interval [0.50–2.61]) nor the 20 mg (odds ratio = 1.00, 95% confidence interval [0.15–6.58]) suvorexant conditions were significantly different from placebo. The percentage of participants who awakened to 85 db or less was 50% for both placebo and suvorexant 20-mg conditions, and 33% for the suvorexant 10-mg condition. Although 85 db was chosen as a threshold as it is the standard db level of a home fire alarm, sensitivity analyses were also conducted using 80-db and 90-db thresholds to determine whether arousability differed across conditions at different thresholds. No differences were found.

Figure 2. Percentage of individuals who did not wake up to the 85-db tone.

dbA = decibel, PBO = placebo, SUV-10 = suvorexant 10 mg, SUV-20 = suvorexant 20 mg.

Treatment Effects on PSG Parameters

Descriptive statistics for all PSG-defined sleep variables by condition appear in Table 2. WASO was significantly lower for suvorexant 20 mg (22.85 ± 18.50) compared to placebo (55.69 ± 40.00), t = 2.609, P < .024. TST was higher for suvorexant 20 mg (446.35 ± 23.94) compared to placebo (401.09 ± 51.29), t = 2.606, P = .024. The remaining pairwise comparisons were not statistically different across study conditions.

Table 2.

Polysomnography sleep data.

Effects of Nocturnal Worry and Rumination on AAT

On the placebo night, higher levels of nocturnal worry were significantly correlated with lower AATs (r = −.632, P = .037), whereas an association between nocturnal rumination and AATs in the same direction approached significance (r = −.568, P = .068). On the suvorexant 10 mg night, higher levels of both nocturnal worry (r = −.61, P = .036) and rumination (r = −.62, P = .032) were again significantly linked to lower AAT levels. In contrast, neither nocturnal worry (r = −.395, P = .203) nor rumination (r = −.509, P = .091) was significantly related to AATs on the suvorexant 20 mg treatment night.

Safety

No clinically significant findings in laboratory assessments, vital signs, or physical examinations were found. One participant reported adverse events (total of five: dry mouth, feeling upset, anxious, dizziness, and tingling in arm) during the study, and all adverse events for this participant occurred during the suvorexant 20 mg condition. The first four of the reported five adverse events were considered to be drug-related and all were considered mild and resolved within 24 hours. No other participants reported any adverse events.

DISCUSSION

The primary aim of this randomized, double-blind, 3-way crossover study was to compare the auditory awakening threshold effects of FDA approved doses of suvorexant (10 and 20 mg) to placebo in non-elderly individuals with DSM-5 insomnia at approximate maximum drug concentrations (∼2.5 hours postdosing). There were no statistically or clinically significant differences in the level of sound required to awaken participants following either 10 mg or 20 mg of suvorexant compared to placebo. In addition, the proportions of participants who remained asleep at the AAT 85 db cutoff (approximate sound level of a residential smoke alarm) did not differ across conditions. The absence of differences in AAT between drug and placebo conditions was observed in conjunction with significant improvement in WASO and TST with the 20-mg dose of suvorexant. This is the first study to investigate the effects of a DORA on AAT in humans and suggests that suvorexant improves sleep in those with insomnia without impairing the ability to respond to external stimuli. In addition, we found that higher levels of nocturnal worry and rumination were associated with lower arousal thresholds during sleep for the placebo night and suvorexant 10 mg night. However, neither worry nor rumination was linked to nocturnal responsivity on the suvorexant 20 mg night, offering preliminary evidence that higher doses of suvorexant may potentially dampen cognitive arousal-related sensitivity to external stimuli during sleep.

We employed AAT procedures consistent with those used by Keefe and colleagues to facilitate comparison with previous studies. Notably, our placebo results are consistent with a mean AAT of 76 db found by these investigators during N2 sleep²² and a previous study using an identical protocol in our laboratory where the mean N2 placebo AAT ranged from 77.9 to 84.7.⁵ In the current study, the mean AAT for both suvorexant doses combined was 78.96 db, suggesting a remarkably similar awakening threshold both in terms of comparison to other previous studies during N2 sleep and to the placebo of the present study (ie, 79.17 db). Importantly, mean AAT for each of the suvorexant doses tested was within 5 db of the placebo AAT. This indicates that if statistical differences could be achieved across conditions using a much larger sample, the differences in AAT between placebo and suvorexant doses are likely to be limited in terms of clinically meaningful responsivity (ie, < 5 db).

The current results are consistent with several previous animal studies demonstrating a lack of impairment in awakening response latency¹⁰ and ability to awaken to salient stimuli during sleep^11,12 following administration of the dual orexin receptor antagonist DORA-22. Although the methodology in those protocols differed from the current study—including drug doses, auditory awakening procedures, and species—the consistency of results across studies with respect to a lack of impairment in nocturnal auditory responsivity during sleep with DORAs support our interpretation that suvorexant does not significantly impair auditory awakening threshold. Future replications using an active control are needed to establish this conclusively. In contrast, previous AAT studies have shown that response thresholds are significantly impaired following administration of most currently available hypnotic drugs that have been studied to date including triazolam,²³ flurazepam,⁴ and zolpidem.⁵

Using a nearly identical laboratory protocol, we previously compared AATs across zolpidem 10 mg, doxepin 6 mg, and placebo at drug T-max during N2 sleep in healthy sleepers.⁵ In that study, mean AAT for zolpidem was 103.2 db, which was higher than both 85.2 db for doxepin and 77.9-85.2 db for placebo. By extension, the AAT profile for suvorexant at 10 mg and 20 mg appears more similar to doxepin 6 mg and placebo, but approximately 20 db lower than the AAT for zolpidem at T-max. This is consistent with reports that BzRA hypnotic drugs can increase AAT by 20 db or more.^2,3,23 These effects are also consistent with the hypothesis that CNS- depressing medications suppress arousal threshold whereas wake inhibiting medication (eg, histamine and orexin antagonists) do not.

Along these lines, an early study on the effects of the benzodiazepine hypnotic drug triazolam on nocturnal arousal to a 78-db (at pillow level) smoke detector showed that all participants on placebo awoke to smoke detector alarms during N2 and slow wave sleep, whereas 17% of participants on triazolam did not wake up to the alarm during N2 and 50% of those on triazolam did not awaken during slow wave sleep.⁴ Although we cannot speak to any effects of suvorexant on arousability to a smoke alarm (the tone of which will differ in Hz and duration of exposure than tones used in our study), it is worth noting that we observed no threshold differences between placebo and suvorexant at 80, 85, or 90 db during N2. These data suggest that preserved nocturnal responsivity may potentially be a safety benefit of DORAs over other hypnotic drugs known to increase AAT levels. However, future research is needed to investigate whether DORAs affect AAT in other stages of sleep and whether they affect arousability to a smoke alarm relative to placebo and other sleep aids that act by inhibiting wake.

Last, we ran exploratory analyses examining the associations of presleep worry and rumination with nocturnal responsivity. Overall, when people with insomnia took placebo or suvorexant 10 mg, those who reported higher levels of worry and rumination at night exhibited lower AAT levels. These findings indicate that high presleep cognitive activity may increase one’s vulnerability to waking to external auditory stimuli at lower sound levels. It is possible that people with insomnia with high levels of worry and rumination may therefore be at higher risk for nocturnal awakenings due to greater sensitivity to stimuli. However, findings also showed that worry and rumination were no longer associated with nocturnal responsivity when people with insomnia took 20 mg of suvorexant. Although preliminary, one interpretation is that higher therapeutic doses of suvorexant may protect ruminators/worriers from overresponsivity to neutral nocturnal stimuli.

Despite these promising results, it remains unclear whether suvorexant also preserves nocturnal discrimination to salient stimuli, such as awakening to one’s own name or the cries of an infant versus neutral stimuli. Animal models show that DORAs, unlike BzRAs, preserve the ability to preferentially awaken to salient stimuli while remaining asleep during neutral stimuli.^10–12 Human research is needed to determine whether these discriminatory processes are similarly preserved in people with insomnia receiving DORAs such as suvorexant. In addition, although people with insomnia with high cognitive arousal are vulnerable to waking to neutral stimuli at lower sound levels, the effects of rumination and worry on nocturnal responsivity to salient stimuli remain unknown. Therefore, it is important for future research to determine whether suvorexant attenuates overresponsivity to neutral nocturnal stimuli while preserving appropriate levels of responsivity to salient stimuli in cognitively aroused people with insomnia.

Several study limitations should be addressed. First, the primary limitation was the small sample size, which was partially offset by the crossover design allowing for intraindividual comparisons. Although the current study did not employ an active comparator for comparison to suvorexant, our primary interest was the level of AAT relative to no medication as that is a potentially important comparison in terms of safety for a hypnotic drug. That is, ideally in terms of this safety metric, hypnotic drugs should not produce blunting of the arousal response. Further, interindividual statistical tests were underpowered—for instance, near-significant trends for associations between rumination and AAT may have been type II errors. Second, although efficacy on some sleep measures was demonstrated despite the small number of individuals tested in this study, the lower 10-mg dose of suvorexant did not produce statistically significant effects on sleep. As previous studies show mixed results on sleep with 10 mg doses of suvorexant,⁷ this may be explained either by the small number of participants or a lack of reliable effect at this dosage. Another possibility is that study participants were not required to meet objective sleep disturbance criteria for study entry typical of hypnotic drug clinical trials. Therefore, potential drug effects on PSG parameters may have been somewhat attenuated in this sample relative to previous studies with PSG inclusion criteria. Even so, the absence of PSG study entry criteria enhances the generalizability of the current results with respect to clinical insomnia populations compared to previous trials. Future studies may benefit from increasing sample size to better detect effects on sleep onset latency using low doses of suvorexant. Third, given night-to-night variability in sleep, respiration, and leg movements, we cannot rule out the possibility that full PSG assessment for multiple nights may have identified sleep disorders not picked up on a single night of PSG in our sample. Finally, we focused on suvorexant’s effects on AAT levels during N2 at estimated drug T-max. Although BzRA hypnotic effects on AAT tend to be strongest early in the sleep period when blood concentration is highest, AAT effects change across the night and have also been investigated during slow wave sleep.^2,3,23 Research is needed to determine whether DORAs affect AAT in other sleep stages across the night.

CONCLUSIONS

This study has safety implications in hypnotic medication use, suggesting that suvorexant in the treatment of insomnia does not pose a risk in terms of awakening threshold compared to placebo. Our findings indicate that a 20-mg dose of suvorexant improves sleep in people with insomnia without impairing the ability to awaken to an auditory stimulus up to and above the decibel level of a typical household smoke alarm. Next steps should be aimed at determining whether this is true in other insomnia populations where responses to auditory stimuli may already be affected by population-specific factors (eg, elderly, insomnia patients with mild cognitive impairment). Future studies should also determine if the lack of effect on AATs generalize to other DORAs and hypnotic drugs that act to dampen wake-promoting systems.

DISCLOSURE STATEMENT

This study was funded by an Investigator Initiated Grant (PI: Drake) from Merck Sharp and Dohme Corp (MISP 57338). Dr. Drake has received research funding from Merck, Jazz, Eisai, and Harmony Biosciences; Dr. Roth has received research funding from Merck and is a consultant for Merck, Eisai, Purdue, Jazz, Teva, Neurim, Avadel and SEQ; Dr. Kalmbach has received research support from Merck; Dr. Cheng has received research funding from Harmony Biosciences and is a consultant for Pfizer. The other authors report no conflicts of interest.

ABBREVIATIONS

AAT

auditory awakening threshold

BzRAs

benzodiazepine receptor agonists

db

decibels

DORA

dual orexin receptor antagonist

PSG

polysomnography

TST

total sleep time

WASO

wake after sleep onset