Prescribing Suvorexant (Belsomra)

Why Suvorexant Matters

Suvorexant was the first drug of a genuinely new class of hypnotics, a dual orexin receptor antagonist (DORA), when the FDA approved it in August 2014. Every hypnotic before it worked by pushing the brain down: benzodiazepines and z-drugs flood GABA receptors, antihistamines block histamine, ramelteon nudges melatonin receptors. Suvorexant does something different. It doesn't sedate the brain broadly; it removes one specific wake signal by blocking orexin, the neuropeptide that keeps you awake. The result is a hypnotic that lets sleep happen rather than forcing it, and, importantly, one that comes without the two liabilities that make the older sleeping pills so troublesome: tolerance/dependence and complex sleep behaviors at anything like the rate seen with z-drugs.

That mechanism is the whole reason to know this drug. Suvorexant is not more effective than zolpidem: head-to-head, all hypnotics land within a narrow band, shaving 12 to 22 minutes off sleep latency and adding roughly 10 to 20 minutes of total sleep versus placebo. Nobody prescribes suvorexant because it works better. You prescribe it because it works more safely in exactly the patients where the older agents are most dangerous: the elderly, and people with substance use histories. It is Beers-list-acceptable in older adults, it showed no increased fall risk in trials enrolling patients up to age 93, and it carries far lower abuse liability than the scheduled GABAergic hypnotics.

The catch is cost. Suvorexant runs several hundred dollars a month, it's frequently branded-only or step-therapy-gated, and most insurers demand documented failure of cheaper agents first. That single fact, not any clinical shortcoming, is why it remains a second- or third-line choice in most practices rather than a go-to.

This guide is deliberately tighter than a guide to a complex, high-monitoring drug would be. Suvorexant is, by design, a low-maintenance medication: no titration to a level, no lab monitoring, no taper, no boxed warning. The clinical skill is mostly in patient selection, counseling on next-day impairment, and managing the handful of odd REM-intrusion side effects that come with turning off orexin.

Part 1: Indications, Who Is Suvorexant For?

FDA-Approved Use

• Insomnia, characterized by difficulties with sleep onset and/or sleep maintenance, in adults.

That's the entire label. Note that it covers both falling asleep and staying asleep, unlike, say, zaleplon (onset only) or low-dose doxepin (maintenance only). The ~12-hour half-life is what buys the maintenance benefit; it also, as we'll see, is what drives the next-day-impairment caution.

Where It Actually Earns Its Place

Suvorexant is not a first-line hypnotic in any guideline, and it shouldn't be: CBT-I is first-line for chronic insomnia, full stop, and cheaper drugs are tried before it. Suvorexant's real niche is defined by the patients in whom the standard agents are contraindicated or risky:

The elderly. This is the strongest case for suvorexant. Older adults are the population most harmed by z-drugs and benzodiazepines: falls, fractures, next-day confusion, delirium, complex sleep behaviors. Suvorexant belongs to the short list of hypnotics considered acceptable in this group (alongside low-dose doxepin, ramelteon, melatonin, and lemborexant). It is not flagged as potentially inappropriate the way most hypnotics are on the Beers Criteria, and trials showed no increased fall risk even in very old patients.

Patients with a substance use history, or at risk for one. Benzodiazepines and z-drugs are off the table (or should be) in anyone with addiction risk. Suvorexant is Schedule IV, so it is not risk-free (high-dose studies in people with sedative-abuse histories did show some misuse potential), but its abuse liability is meaningfully lower than the GABAergic hypnotics, and animal data show low reinforcing effects. For a patient you don't want to give a controlled GABA agonist, suvorexant is a reasonable prescription hypnotic.

Patients who need a hypnotic long-term. Because suvorexant shows no tolerance and no rebound insomnia on discontinuation, it's a more honest long-term choice than a z-drug, whose effect fades and whose withdrawal rekindles the very insomnia you were treating.

Patients with a history of complex sleep behaviors on z-drugs. If a patient has sleep-walked, sleep-driven, or sleep-eaten on zolpidem, you cannot re-prescribe that class. Suvorexant's complex-sleep-behavior rate is far lower (on the order of 0.6% versus 3 to 15% for z-drugs), making it a rational switch, though not a guarantee, since the behavior is a class warning for suvorexant too.

Off-label / investigational. There is interest in orexin antagonism for insomnia comorbid with PTSD and other psychiatric conditions, and some use in insomnia that hasn't responded to long-term benzodiazepine therapy. These uses are not well-established; the evidence base outside primary insomnia is thin, and that thinness is a legitimate barrier to broader psychiatric use.

Part 2: Before You Start, Candidacy and Workup

Suvorexant requires no baseline labs: no metabolic panel, no CBC, no ECG. What it requires instead is a short screening conversation.

Screen for these before prescribing

Poor candidates

• Narcolepsy (absolute)
• Patients who cannot reliably allow ≥7 hours in bed before needing to be alert
• Significant untreated sleep apnea or respiratory failure
• Patients for whom cost/prior-authorization will simply prevent adherence, a real-world disqualifier

Part 3: How to Start and Dose

This is refreshingly simple. There is no titration to a target and no monitoring level: you pick a dose, take it at bedtime, and adjust once based on response and tolerability.

Formulations

Tablets: 5 mg, 10 mg, 15 mg, 20 mg.

Starting dose and range

• Start: 10 mg once nightly. This is the recommended starting dose for most adults.
• Range: 10–20 mg nightly.
• Maximum: 20 mg/night. The FDA capped approval at 20 mg specifically because higher doses produced unacceptable next-day driving impairment. (Merck originally proposed starting non-elderly adults at 20 mg and going up to 40 mg; the FDA rejected that as too impairing, a useful reminder that the ceiling here is a safety ceiling, not an efficacy one.)
• If 10 mg is tolerated but insufficient, you may increase toward 20 mg, but counsel harder about morning grogginess and driving as you go up, since daytime somnolence and fatigue are dose-dependent.

How to take it, the counseling that makes or breaks it

• Take within 30 minutes of getting into bed.
• Take only when you can devote ≥7 hours to sleep before you need to be up and functional. This is the core instruction: with a ~12-hour half-life, cutting sleep short is what produces the dangerous morning.
• Take on a relatively empty stomach. A large or high-fat meal delays the onset of effect (the drug is absorbed more slowly), so a bedtime dose taken right after a heavy dinner may not work when the patient wants it to, and a delayed onset can mean residual drug at the wrong hour. Don't take it with or immediately after a big meal.
• One dose per night. No re-dosing if the patient wakes at 3 a.m.: the half-life means there's plenty of drug still on board, and a second dose guarantees a wrecked morning.

Part 4: Monitoring

There is no laboratory monitoring for suvorexant: no levels, no renal/hepatic/thyroid panels, no ECG, no CBC. What you monitor is clinical, and it's short:

• Next-day impairment: Ask at follow-up about morning grogginess, driving, and cognitive fog, especially in patients on 20 mg, the elderly, and anyone on a CYP3A4 inhibitor. Reassess at ~1 month and periodically thereafter.
• Complex sleep behaviors and amnesia: Ask explicitly at each visit about any sleep-walking, sleep-driving, sleep-eating, phone/email use, or sexual activity with no memory of it. If any occurs, however mild, stop the drug.
• REM-intrusion phenomena: Ask about sleep paralysis, vivid/disturbing dreams, and hypnagogic (falling-asleep) or hypnopompic (waking) hallucinations, and any daytime leg weakness suggestive of cataplexy-like events.
• Depression/suicidality: Continue routine screening, as with any hypnotic.
• Ongoing need: Because insomnia is often comorbid and self-limited, periodically ask whether the drug is still needed at all.

Part 5: Side Effects and How to Manage Them

Suvorexant is generally well tolerated: in the pivotal trials, fewer than 5% discontinued for adverse events. The side-effect profile splits cleanly into common-and-mild and rare-and-strange, the latter being the REM-boundary phenomena that fall out of blocking orexin.

Common (roughly 2–8%)

• Somnolence / next-day sedation, the most important, and dose-dependent. This is the one that matters clinically, because it translates into driving and functional impairment.
• Headache
• Abnormal / vivid dreams
• Dry mouth

Management: For next-day sedation, lower the dose (20 → 10 mg), reinforce the ≥7-hours-in-bed rule, confirm the patient isn't re-dosing mid-night, and check for an added CYP3A4 inhibitor that's raising exposure. Most of the mild effects are manageable with reassurance and dose adjustment.

Next-day driving impairment, the headline safety issue

This is the effect that shaped the drug's entire regulatory story. Impairment of next-morning driving and alertness is real, is dose-related, and is present even at the approved 20 mg dose. It's why the FDA capped the dose at 20 mg and required Merck to keep studying driving performance.

Management:

• Counsel every patient that they may be impaired the next morning and should not drive or do anything requiring full alertness if they feel at all groggy: this warning is not optional.
• Prefer the 10 mg dose where it controls symptoms.
• Avoid alcohol and other CNS depressants, which stack with it.
• Be especially cautious in the elderly and in anyone on a CYP3A4 inhibitor (higher blood levels lead to more impairment).

Rare but characteristic, the REM-intrusion phenomena (<1%)

Blocking orexin loosens the boundary between wake and REM sleep, so a small minority of patients get REM phenomena bleeding into wakefulness:

• Sleep paralysis, a transient inability to move or speak for seconds to minutes at sleep-wake transitions.
• Cataplexy-like leg weakness, brief weakness that can occur day or night.
• Hypnagogic / hypnopompic hallucinations, vivid, sometimes frightening perceptions while falling asleep or waking.

Management: These are usually benign but alarming. Reassure, and lower the dose or discontinue if they persist or distress the patient. Their existence is exactly why narcolepsy is an absolute contraindication: a patient already prone to REM intrusion should not have orexin blocked on top of it.

Complex sleep behaviors

Sleep-walking, sleep-driving, sleep-eating, and other complex behaviors performed while not fully awake and with no memory afterward are a class warning for all hypnotics, suvorexant included. The reassuring news is that the rate is much lower than with z-drugs: on the order of 0.6% for suvorexant versus 3 to 15% for zolpidem and the other z-drugs.

Management:

• Discontinue permanently if any complex sleep behavior occurs, however mild. This is the same rule that governs the z-drugs.
• Reduce risk by having the patient get into bed immediately after dosing, avoid alcohol and other GABAergic/CNS depressants, and avoid taking it on a full stomach (delayed onset raises the risk).

What you don't see

• No tolerance: the hypnotic effect does not wear off over months, unlike z-drugs and benzodiazepines.
• No rebound insomnia or withdrawal on stopping (see Discontinuation).
• No clear next-day anterograde amnesia of the type z-drugs are notorious for.
• No routine metabolic, hematologic, or cardiac toxicity requiring monitoring.

Part 6: Overdose, Toxicity, and Warnings

Boxed warning

Suvorexant does not carry a boxed (black-box) warning. Its labeling follows the standard hypnotic-class warnings: complex sleep behaviors, next-day impairment, worsening of depression/suicidal ideation, and the possibility of sleep paralysis and related REM phenomena. (Some sources loosely describe the DORA-class depression/suicidality caution as a "black-box"-style warning; treat the practical message, screen and monitor, as what matters, rather than the label formatting.)

Overdose

There is no specific antidote; management is supportive. Relative to benzodiazepines, orexin antagonists carry less respiratory-depression risk in overdose, which is part of their safety appeal, but the danger rises sharply when combined with alcohol, opioids, or other CNS depressants, where additive sedation and respiratory compromise become the real hazard. Expect prolonged sedation given the ~12-hour half-life.

Abuse and scheduling

Suvorexant is a Schedule IV controlled substance, the same schedule as zolpidem and temazepam. The scheduling reflects some abuse potential, not a high one: animal studies found low reinforcing effects and no withdrawal, but high-dose human studies in people with sedative-abuse histories did detect misuse potential. In practice its abuse liability is low relative to the GABAergic hypnotics, which is why it's a preferred prescription hypnotic in addiction-prone patients, but it still warrants the ordinary controlled-substance judgment about quantities and refills.

Part 7: Drug Interactions

The interaction profile is driven by two things: CYP3A4 metabolism and additive CNS depression.

CYP3A4, the pharmacokinetic axis

Suvorexant is metabolized primarily by CYP3A4. That makes its blood level sensitive to CYP3A4 inhibitors and inducers.

Pharmacodynamic, additive CNS depression

• Alcohol: avoid. Additive sedation and impairment; raises complex-sleep-behavior risk.
• Opioids and other CNS depressants: additive respiratory depression and sedation; use caution, and combine only with clear justification.
• Other GABAergic sedatives (benzodiazepines, barbiturates, and the herbal sedatives valerian, kava, skullcap): additive sedation and increased complex-sleep-behavior risk. Notably, adding suvorexant to an existing benzodiazepine has produced excess sedation, one reason cross-tapering, rather than piling on, is preferred when switching.

What's clean

• Most antidepressants have no major pharmacokinetic interaction (watch nefazodone, a 3A4 inhibitor).
• No meaningful renal-clearance interactions of the kind that plague renally handled drugs; suvorexant is hepatically metabolized.

Part 8: Special Populations

Elderly

This is suvorexant's best population, not its worst. It is Beers-acceptable, showed no increased fall risk in trials up to age 93, and has a low complex-sleep-behavior rate. That said:

• Older patients clear the drug more slowly and are more sensitive to next-day impairment; favor the 10 mg dose, and be alert to added CYP3A4 inhibitors in a polypharmacy setting.
• It sits on the short list of genuinely reasonable hypnotics for older adults, alongside low-dose doxepin, ramelteon, and melatonin.

Pregnancy

Human data are limited; there is no established safety in pregnancy, and it is not a preferred agent. Non-pharmacologic measures (CBT-I, sleep hygiene) are first-line, and the decision to use any hypnotic in pregnancy is an individualized OB-coordinated trade-off between the risks of the drug and the risks of untreated insomnia. If a hypnotic is judged necessary, agents with more reproductive data are generally chosen preferentially; use suvorexant only after that weighing. The source notes do not provide reliable pregnancy-specific data for suvorexant; do not assume safety.

Lactation

Lactation data are limited. As a lipophilic, hepatically metabolized drug with a ~12-hour half-life, some excretion into breast milk is plausible. If used, monitor the infant for sedation and poor feeding, and consider timing/feeding logistics. Prefer better-characterized options where possible.

Hepatic impairment

Because clearance depends on hepatic CYP3A4, hepatic impairment raises exposure.

• Mild–moderate hepatic impairment: use cautiously; a lower dose and closer attention to next-day sedation are prudent.
• Severe hepatic impairment: not recommended; clearance is too compromised and levels too unpredictable.

Renal impairment

Suvorexant is hepatically metabolized, so renal impairment is far less of an issue than for renally cleared drugs. No major dose adjustment is generally required for renal impairment, though, as with any sedative in a medically compromised patient, start low and watch for increased CNS sensitivity.

Children and adolescents

Not established. There are no FDA-approved pediatric indications. Insomnia in children is managed first with behavioral measures and sleep hygiene; melatonin is the commonly used off-label agent. Suvorexant is not a pediatric drug.

Part 9: Discontinuation

This is one of suvorexant's genuine advantages, and it's short: you can just stop it.

• No tolerance, no dependence, no rebound insomnia, no withdrawal has been demonstrated in the trials, including long-term (1-year) data.
• No taper is required. A patient can discontinue abruptly without a withdrawal syndrome or a rebound flare of insomnia, a sharp contrast with benzodiazepines (which need a slow taper) and z-drugs (which can rebound).

The main caveat is switching to suvorexant from a GABAergic hypnotic: the underlying benzodiazepine or z-drug still needs its own appropriate taper, and because adding suvorexant to an existing benzo can cause excess sedation, a cross-taper with monitoring is usually cleaner than an abrupt swap. Also set expectations: patients used to the GABA "knockout punch" sometimes find the DORA subtler and may need coaching not to abandon it prematurely.

Part 10: Suvorexant vs. the Alternatives

Efficacy across hypnotics is broadly similar; the choice is almost entirely about safety profile, cost, and mechanism.

vs. Z-drugs (zolpidem, zaleplon, eszopiclone). Similar efficacy. Suvorexant wins decisively on safety: far fewer complex sleep behaviors (~0.6% vs. 3 to 15%), no tolerance/withdrawal, no notable next-day amnesia, and lower abuse potential. Z-drugs win only on cost (cheap generics). If money and formulary weren't factors, suvorexant would displace z-drugs in most patients; in the real world, cost keeps the z-drugs first.

vs. Benzodiazepines. Similar efficacy, and suvorexant is much safer: no meaningful tolerance/dependence/withdrawal, fewer falls, less cognitive impairment, lower abuse liability, and safer in the elderly. Benzos retain a role when insomnia is entangled with anxiety, but for insomnia as such, suvorexant is the safer tool.

vs. Lemborexant (Dayvigo). The other well-established DORA. Similar half-life (~17 to 19 hours, with an active metabolite) and similar efficacy; network meta-analyses have suggested lemborexant 10 mg carries the largest effect size among orexin antagonists but with a somewhat higher daytime-somnolence signal. Suvorexant has the longer real-world track record (since 2014). Largely interchangeable; choose on formulary coverage and tolerability.

vs. Daridorexant (Quviviq). The newest DORA, with a shorter half-life (~8 hours), theoretically the least next-day carryover of the three, which is attractive when morning impairment is the dominant concern. Dose-dependent efficacy; the 50 mg dose also improved daytime functioning in trials.

vs. Ramelteon (Rozerem). Ramelteon is even safer on abuse (non-scheduled, no abuse potential at all) and has no next-day impairment to speak of, but it is generally less potent and works best for sleep-onset/circadian problems. Suvorexant is the stronger hypnotic and better for sleep maintenance; ramelteon is the choice when abuse liability must be zero or the problem is circadian.

vs. Low-dose doxepin (Silenor). Doxepin (3 to 6 mg) is cheap, Beers-acceptable, and good for sleep maintenance, but it doesn't help sleep onset and carries mild anticholinergic caveats. Suvorexant covers both onset and maintenance.

vs. Trazodone. Trazodone (50 to 100 mg) is the cheap workhorse, especially in depressed patients, but brings orthostatic hypotension, next-day sedation, and (rarely) priapism. Suvorexant is more specific, with fewer of those liabilities, at far greater cost.

Mechanism: The Short Version

Orexins (also called hypocretins) are hypothalamic neuropeptides that act as a master "stay awake" switch, driving arousal by stimulating the release of the classic wake-promoting transmitters: histamine, dopamine, norepinephrine, serotonin, acetylcholine. Suvorexant is a dual orexin receptor antagonist, blocking both OX1 and OX2 receptors. By silencing the orexin wake signal, it lets the brain transition into sleep without broadly depressing CNS function the way GABAergic drugs do, which is why it lacks the amnesia, the marked complex-sleep-behavior risk, and the tolerance/dependence of the older agents.

The flip side of the same mechanism explains its signature side effects: orexin also stabilizes the boundary between wakefulness and REM sleep, so blocking it can let REM phenomena (sleep paralysis, hypnagogic/hypnopompic hallucinations, cataplexy-like weakness) intrude into the waking state. And because narcolepsy is itself a disease of orexin deficiency, pharmacologically removing orexin in a narcoleptic patient is contraindicated.

Pharmacokinetics: rapid onset (~30 minutes, delayed by a fatty meal), half-life ~12 hours (the source of both its sleep-maintenance benefit and its next-day-impairment risk), and metabolism primarily by CYP3A4.

The Bedside Cheat Sheet