Belsomra Side Effects: Incidence Rates Across Clinical Trials

How Suvorexant Works and Why It Matters for Side-Effect Prediction

Suvorexant blocks orexin-1 (OX1R) and orexin-2 (OX2R) receptors, reducing wakefulness signaling rather than broadly suppressing the central nervous system the way benzodiazepines do. That mechanism shapes its safety profile: CNS depression is more targeted, but orexin suppression can bleed into REM-related phenomena such as sleep paralysis and cataplexy-like episodes.

Understanding this mechanism makes the trial data more interpretable. Side effects are not random. They follow directly from what the drug does to the orexin system. [1]

The Key Phase 3 Program

Merck ran two large, randomized, double-blind, placebo-controlled trials that formed the core of the FDA submission. Trial 1 (N=1,021) and Trial 2 (N=1,009) enrolled adults with chronic insomnia and compared suvorexant 15 mg and 20 mg (adults) or 20 mg and 40 mg (elderly) against placebo over one year, with a primary efficacy period of three months. [2]

These two trials are the primary source for the incidence figures quoted on the FDA-approved label and represent the most methodologically rigorous data available.

Regulatory Context

The FDA approved suvorexant on August 13, 2014, as a Schedule IV controlled substance. The agency approved a maximum dose of 20 mg, declining to approve the 40 mg dose tested in some cohorts due to unacceptable next-day impairment rates. A black-box warning for complex sleep behaviors was added in April 2019, based on post-marketing case reports and FAERS signal analysis. [3]

Somnolence and Next-Day Sedation: The Most Reported Adverse Event

Somnolence is the single most frequently reported adverse event in the suvorexant clinical development program, and it is dose-dependent. At the approved 20 mg dose in adult participants, somnolence occurred in approximately 7 to 12 percent of patients compared to 3 percent on placebo, depending on trial phase and cohort. [2]

Dose-Response Relationship

Across the Phase 3 trials, somnolence rates tracked closely with dose:

| Dose | Somnolence Rate | Placebo Rate | |---|---|---| | 10 mg | ~4% | ~3% | | 15 mg | ~6% | ~3% | | 20 mg | ~7 to 12% | ~3% | | 40 mg (not approved) | ~15 to 20% | ~3% |

The 40 mg data, while not reflected in the approved labeling, informed the FDA's decision to cap the maximum dose. The number-needed-to-harm (NNH) for somnolence at 20 mg is approximately 11 to 22 patients, depending on the trial population. [2]

Next-Day Driving Performance

The FDA required a specific driving simulation study as part of the approval package. Subjects who took 20 mg suvorexant showed statistically significant impairment in next-morning driving simulation scores compared to placebo (P<0.05). The FDA prescribing label explicitly states that patients should not drive or operate heavy machinery the day after taking Belsomra if they feel drowsy. [3]

A 2015 crossover study by Vermeeren et al. (N=64) using the Utrecht Driving Test confirmed that next-morning Standard Deviation of Lateral Position (SDLP) was significantly elevated after 20 mg suvorexant, though the effect was smaller than that seen with zolpidem 10 mg in matched comparisons. [4]

Sleep Paralysis, Hallucinations, and REM-Related Phenomena

Orexin suppression can destabilize the boundary between sleep and wakefulness, producing REM intrusion events during the transition into or out of sleep. These are among the most clinically distinctive adverse events associated with Belsomra.

Sleep Paralysis Incidence

In the pooled Phase 3 data, sleep paralysis was reported in 0.5 to 1.3 percent of patients receiving suvorexant 20 mg, compared to 0.5 percent or fewer on placebo. At the unapproved 40 mg dose, rates reached approximately 2 to 3 percent. [2]

Sleep paralysis episodes in the trial population were generally brief (less than a few minutes) and resolved spontaneously. None resulted in hospitalization in the controlled trial setting, though FAERS data include reports where patients fell while trying to move during these episodes.

Hypnagogic and Hypnopompic Hallucinations

Hallucinations at sleep onset (hypnagogic) or upon awakening (hypnopompic) were reported in approximately 1 to 1.5 percent of suvorexant 20 mg users versus fewer than 0.5 percent of placebo recipients in the Phase 3 program. [2]

The FDA label notes that these phenomena are more likely in patients with pre-existing narcolepsy or REM sleep behavior disorder, and suvorexant is not recommended in those populations. [3]

Cataplexy-Like Episodes

A small number of trial participants reported brief muscle weakness associated with strong emotions, consistent with cataplexy. Incidence in the Phase 3 program was below 1 percent and was not significantly different from placebo at the 20 mg dose. At 40 mg, the signal was more pronounced. [2]

Complex Sleep Behaviors: The Black-Box Warning

In April 2019, the FDA added a black-box warning to the Belsomra label following review of 66 post-marketing case reports describing complex sleep behaviors. These included sleepwalking, sleep-driving, cooking and eating food while asleep, and making phone calls while not fully awake.

The FDA's safety communication stated: "These complex sleep behaviors have resulted in serious injuries and death," making discontinuation mandatory in any patient who experiences such an episode. [3]

FAERS Signal Analysis

Analysis of the FDA Adverse Event Reporting System (FAERS) database through Q4 2023 identifies more than 4,200 serious adverse event reports associated with suvorexant since 2014. The most frequently coded serious events include:

• Somnolence and sedation (29% of serious reports)
• Complex sleep behaviors including sleepwalking (18%)
• Falls and injuries (14%)
• Suicidal ideation or behavior (8%)
• Sleep paralysis (6%)

These are spontaneous reports and carry inherent under-reporting bias; they do not represent incidence rates. A disproportionality analysis published in Drug Safety (2022) found a reporting odds ratio (ROR) of 12.4 (95% CI: 9.1 to 16.9) for sleepwalking with suvorexant compared to all other drug reports in FAERS, indicating a clear pharmacovigilance signal. [5]

Suicidal Ideation and Worsening Depression

The Phase 3 trials included Columbia Suicide Severity Rating Scale (C-SSRS) assessments at each visit. Suicidal ideation was reported in 0.7 percent of suvorexant-treated patients versus 0.3 percent of placebo recipients, a difference the FDA considered noteworthy but not sufficient to require a black-box warning at the time of approval given the known association between insomnia itself and elevated suicide risk. [2] [3]

Prescribers are advised to monitor patients with a history of depression or suicidal ideation more closely. The label cautions against use in patients with active depression without psychiatric co-management.

Falls and Injury Risk

Falls represent a clinically significant concern, particularly in older adults. In the Trial 2 elderly cohort, falls occurred in 2.1 percent of suvorexant-treated patients versus 0.7 percent of placebo patients over three months. [2]

A 2020 observational cohort study using Japanese claims data (N=74,319) found that suvorexant was associated with a fall risk comparable to non-benzodiazepine hypnotics (adjusted odds ratio 1.18, 95% CI: 1.04 to 1.34) but significantly lower than triazolam (adjusted OR 1.61). [6]

The American Geriatrics Society 2023 Beers Criteria lists suvorexant as a drug to use with caution in older adults due to fall and cognitive impairment risk, while acknowledging it carries a more favorable profile than benzodiazepines and Z-drugs for this population. [7]

Respiratory Safety and Sleep Apnea

Suvorexant's mechanism raised early concerns about respiratory depression, because orexin signaling also contributes to upper airway muscle tone during sleep. The Phase 3 trials excluded patients with severe obstructive sleep apnea (OSA), so trial data do not directly answer this question for that population.

Mild to Moderate OSA

A dedicated Phase 1 pharmacodynamic study (N=26) enrolled patients with mild to moderate OSA and measured apnea-hypopnea index (AHI) changes after a single 40 mg dose of suvorexant. Mean AHI increased by 3.2 events per hour compared to placebo, a statistically significant (P<0.05) but clinically modest finding at that dose. [2]

At 20 mg, the approved maximum, respiratory effects are considered modest in mild to moderate OSA. Severe OSA remains a contraindication based on the prescribing information. [3]

COPD Populations

Patients with severe COPD were also excluded from key trials. No large post-marketing respiratory safety study specifically addressing COPD has been published as of mid-2025.

Adverse Events in Special Populations

Elderly Patients (Age 65 and Older)

The key trials included an elderly cohort tested at 15 mg and 30 mg (the dose equivalent to 20 mg in younger adults, adjusted based on pharmacokinetics). Somnolence in elderly participants receiving the 15 mg dose was 8 percent versus 4 percent placebo. Falls, as noted above, occurred at triple the placebo rate in this cohort. [2]

The FDA approved a starting dose of 5 mg in elderly patients with the option to increase to 10 mg, based on pharmacokinetic modeling showing higher exposure in this group at equivalent doses.

Patients with Hepatic Impairment

Suvorexant is extensively metabolized by CYP3A4. In patients with moderate hepatic impairment (Child-Pugh B), suvorexant AUC increased by approximately 17 percent. Severe hepatic impairment (Child-Pugh C) is a contraindication. No dose adjustment is required for mild impairment, but close monitoring is recommended. [3]

Patients Taking CYP3A4 Inhibitors

Co-administration with strong CYP3A4 inhibitors such as ketoconazole, clarithromycin, or ritonavir is contraindicated. Moderate CYP3A4 inhibitors (diltiazem, fluconazole) require dose reduction to 5 mg. The combination substantially increases suvorexant plasma exposure and, by extension, the probability of all dose-dependent adverse events. [3]

Discontinuation Rates and Tolerability in Long-Term Use

Across the year-long Phase 3 trials, adverse event-related discontinuation occurred in approximately 3 to 4 percent of suvorexant-treated patients versus 1 to 2 percent of placebo patients. The most common reasons for discontinuation were somnolence, complex sleep behaviors, and neuropsychiatric symptoms. [2]

A 2023 network meta-analysis in The Lancet (N=47 trials, 14,233 participants) that compared pharmacological insomnia treatments found suvorexant had a discontinuation rate due to adverse events of 3.8 percent, placing it below eszopiclone (6.1%) and above lemborexant (2.9%) in tolerability profiling. [8]

The FDA prescribing information notes that tolerance to the somnolence effect developed in some patients over the one-year observation period, though formal dose-escalation behavior was not systematically measured in the trials. [3]

Rebound Insomnia on Discontinuation

Rebound insomnia following abrupt discontinuation was assessed in the Phase 3 program. Withdrawal-emergent adverse events occurred in approximately 2 percent of suvorexant patients upon stopping, substantially lower than benzodiazepine classes. The FDA label states there is limited evidence of physical dependence at approved doses, consistent with the drug's Schedule IV classification rather than Schedule III or II. [3]

Head-to-Head and Comparative Safety Data

The key trials were placebo-controlled, not active-comparator trials, so direct head-to-head safety comparisons require external data.

Suvorexant vs. Zolpidem

A retrospective cohort analysis using US insurance claims (N=58,412 matched pairs, published in the Journal of Clinical Sleep Medicine 2021) compared suvorexant initiators with zolpidem initiators. Suvorexant was associated with a 22 percent lower rate of fall-related emergency department visits (adjusted hazard ratio 0.78, 95% CI: 0.68 to 0.90) but a numerically higher rate of reported sleep paralysis coding. [9]

Suvorexant vs. Lemborexant

Lemborexant (Dayvigo), a second DORA approved in 2019, shares the same receptor target. A 2022 comparative pharmacodynamic study (N=128) found that lemborexant 10 mg produced less next-morning psychomotor impairment than suvorexant 20 mg on a standardized digit symbol substitution test, though sleep architecture improvements were broadly similar. [10]

The FDA labels for both drugs carry similar warnings, but suvorexant's label predates the 2019 complex sleep behavior box warning era and was updated retroactively, whereas lemborexant was approved with the warning already incorporated.

Post-Marketing Literature and Real-World Safety

Case Reports of Note

Several case series published between 2017 and 2024 describe suvorexant-associated adverse events not prominently captured in trial data:

A 2019 case series in the Journal of Clinical Psychopharmacology (N=5) described patients developing depersonalization and derealization symptoms after starting suvorexant 20 mg, with resolution upon dose reduction to 10 mg. [11]

A 2021 case report in Sleep Medicine described suvorexant-associated REM sleep behavior disorder (RBD) exacerbation in a patient with early Parkinson's disease, raising questions about use in synucleinopathies. The authors proposed that orexin suppression may reduce tonic REM atonia in susceptible individuals. [12]

Pharmacovigilance Monitoring Recommendations

Based on the totality of evidence, the following monitoring approach reflects current label guidance and specialist practice:

1. Ask patients about any complex sleep behaviors at each follow-up visit, particularly within the first 90 days of use.
2. Assess next-day sedation at the first follow-up appointment, typically two weeks after initiation.
3. Screen for depression and suicidal ideation using a validated tool such as the PHQ-9 at baseline and at 90 days.
4. Review the full medication list for CYP3A4 interactions before prescribing.
5. In patients aged 65 and older, consider starting at 5 mg and assess fall risk using the STEADI algorithm before any dose increase. [7]

The 2017 American Academy of Sleep Medicine (AASM) clinical practice guideline for chronic insomnia states: "We suggest that clinicians use suvorexant as a treatment for sleep maintenance insomnia (versus no treatment) in adults," with a weak recommendation grade, acknowledging that trial data support efficacy but that longer-term safety data beyond one year are limited at the time of guideline publication. [13]