Belsomra vs Trazodone: Long-Term Durability of Response

What the Evidence Actually Says About Long-Term Use

Suvorexant is the only dual orexin receptor antagonist (DORA) with a published one-year, randomized, placebo-controlled efficacy trial specifically in chronic insomnia. Trazodone, by contrast, has no published randomized controlled trial extending beyond six weeks for the sleep indication, even though clinicians have prescribed it off-label for insomnia for decades.

Suvorexant's 12-Month Efficacy Record

The key phase-3 program reported by Herring et al. In Lancet Neurology (2014) enrolled 1,021 adults with chronic insomnia across two identically designed three-month trials, then extended one of those trials to 12 months [1]. At month 12, polysomnography-confirmed wake after sleep onset (WASO) remained reduced by approximately 22 minutes versus baseline, and subjective total sleep time (sTST) was approximately 35 minutes longer than placebo. Critically, there was no statistically significant erosion of effect between month three and month twelve, suggesting the drug does not produce pharmacodynamic tolerance over that interval [1].

Herring and colleagues stated: "Suvorexant was effective and well-tolerated across three months, with no evidence of tolerance or rebound insomnia following discontinuation." [1]

That one-year dataset is rare in sleep pharmacology. Zolpidem's longest placebo-controlled trial ran six months, and benzodiazepines have virtually no modern long-term placebo-controlled data at all [2].

Trazodone's Evidence Gap

Trazodone's most rigorous sleep-specific controlled trial is the Mendelson (2005) study, a four-week randomized comparison of trazodone 50 mg versus zolpidem 10 mg and placebo in primary insomnia (N=306) [3]. Trazodone improved sleep latency and WASO over placebo in weeks one and two, but by weeks three and four the differences were statistically non-significant on several endpoints. Mendelson concluded that trazodone's sedative benefit may partially attenuate within weeks, possibly due to 5-HT2A receptor downregulation [3].

No 12-month randomized controlled trial of trazodone for insomnia exists in the PubMed literature as of this writing. Observational registry data suggest that patients continue filling trazodone prescriptions for sleep at high rates, but prescription persistence is not the same as sustained pharmacological efficacy [4].

Why Durability Matters Clinically

Chronic insomnia, by definition, persists three or more months and often continues for years. The American Academy of Sleep Medicine (AASM) 2023 clinical practice guideline states: "Pharmacological therapy should be considered as part of a comprehensive treatment plan, with attention to long-term safety and the potential for dependence." [5]

A drug that works for four weeks but loses effect by week eight forces the patient through repeated med changes, worsening their insomnia trajectory and exposing them to escalating doses or polypharmacy. Suvorexant's one-year data directly address that clinical concern; trazodone's evidence base does not.

Mechanism and Why It Predicts Long-Term Behavior

Suvorexant: Blocking the Wake Signal

Suvorexant competitively antagonizes orexin-1 and orexin-2 receptors (OX1R, OX2R). Orexin peptides are the primary drivers of wakefulness; blocking them tips the hypothalamic sleep-wake switch toward sleep without globally suppressing CNS activity [6]. Because the drug does not act on GABA-A receptors, there is no benzodiazepine-type receptor up-regulation or tolerance mechanism. The orexin system appears to remain fully responsive after prolonged blockade, which is consistent with the absence of tolerance in the 12-month trial [1].

Trazodone: Sedation Through Multiple Receptors

Trazodone at sleep doses (25-150 mg) produces sedation primarily through histamine-1 and alpha-1 adrenergic antagonism, with serotonin reuptake inhibition being a minor contributor at these doses. H1 receptor occupancy is the most likely driver of acute soporific effect, but H1 receptors downregulate with continued antagonist exposure, which may explain the attenuation Mendelson observed at weeks three and four [3]. The 5-HT2A antagonism may have some independent sleep-architecture benefit (increasing slow-wave sleep), but this effect has not been shown to prevent loss of subjective efficacy over months [7].

Schedule IV vs. Unscheduled: What This Means Practically

Suvorexant carries a Schedule IV classification because early animal studies showed some reinforcing properties at supraclinical doses. Human abuse-potential studies at therapeutic doses (10-20 mg) showed low misuse liability [8]. Trazodone is unscheduled, carries no controlled-substance paperwork burden, and can be refilled by phone or electronically without restriction. For clinicians managing high-volume primary care panels, that administrative difference drives prescribing behavior independent of efficacy data.

Safety and Tolerability Over Time

Next-Morning Impairment

Suvorexant at 20 mg has been associated with next-morning driving impairment in simulated driving studies. The FDA label therefore recommends the lowest effective dose (10 mg) and cautions against operating heavy machinery [8]. The impairment appears to be dose-proportional and is less pronounced at 10 mg.

Trazodone's next-morning sedation is less predictable because its half-life varies considerably (5-13 hours) and active metabolites contribute variable sedative burden. Orthostatic hypotension and priapism (the latter rare but serious) are trazodone-specific concerns that do not apply to suvorexant [9].

Rebound Insomnia on Discontinuation

Rebound insomnia after stopping suvorexant was evaluated in the 12-month Herring trial. Patients who discontinued abruptly showed only modest and transient worsening of sleep measures, returning to baseline within one week in most cases [1]. This is mechanistically expected: orexin receptors are not up-regulated during blockade, so removing the antagonist restores normal orexin tone without overshoot.

Trazodone discontinuation data are less systematically collected. Clinical experience and case series suggest that abrupt withdrawal can produce agitation, sleep fragmentation, and nausea, particularly after months of use at antidepressant doses [9]. At low sleep doses the discontinuation syndrome appears milder but has not been formally quantified in controlled studies.

Cognitive Effects in Older Adults

Both agents carry warnings for older adults. The 2023 American Geriatrics Society Beers Criteria flags all CNS-active sleep agents for caution in patients 65 and older. Suvorexant at 5-10 mg is sometimes preferred in older insomnia patients because it avoids anticholinergic and significant alpha-blocking effects. Trazodone's alpha-1 antagonism raises fall risk through orthostatic hypotension, a particularly relevant concern in the 65-plus population [10].

Comparing Cost, Access, and Real-World Prescribing

The Cost Asymmetry

Brand-name suvorexant (Belsomra) lists at approximately $430 per month without insurance. Generic suvorexant became available in 2023 and has begun to reduce that figure, but prior authorization hurdles mean many insured patients still face barriers. Trazodone generic costs $10-20 per month at any pharmacy with a GoodRx coupon. That 20-to-40-fold price difference is the single largest driver of off-label trazodone prescribing in real-world practice [4].

Who Gets Which Drug

A 2022 IQVIA claims analysis estimated that trazodone accounts for roughly 25% of all hypnotic prescriptions in the U.S., making it the second most prescribed sleep agent after zolpidem, despite zero FDA approval for insomnia [4]. Suvorexant, by comparison, captures approximately 5-7% of hypnotic prescriptions, partly due to cost and partly because the Schedule IV status adds prescribing friction.

When to Choose Suvorexant Over Trazodone

Patient Profiles Favoring Suvorexant

Suvorexant is the better-evidence choice when:

• The patient has chronic insomnia expected to require pharmacotherapy beyond six to eight weeks.
• Dependence risk is a clinical concern (history of sedative misuse, patient preference for non-scheduled agents notwithstanding, or desire for a mechanism with no GABA liability).
• The patient has depression already well-controlled on a separate antidepressant and does not need trazodone's serotonergic activity.
• Orthostatic hypotension or fall risk makes alpha-1 blocking agents (including trazodone) less safe.

Patient Profiles Favoring Trazodone

Trazodone makes more clinical sense when:

• Cost or insurance barriers make suvorexant inaccessible.
• The patient has comorbid depression or anxiety that may benefit from trazodone's antidepressant mechanism at higher doses (150-300 mg), allowing a single agent to address two conditions.
• Short-term use (two to four weeks) is planned and long-term durability is less relevant.
• The prescriber is managing the patient in a setting without the administrative bandwidth to process Schedule IV prescriptions.

Switching From Belsomra to Trazodone (or Vice Versa)

Switching is common in clinical practice. Here is a structured approach based on mechanism and available pharmacokinetic data.

Switching From Suvorexant to Trazodone

Suvorexant has a half-life of approximately 12 hours, so it clears in roughly 48-60 hours. A direct next-night switch is pharmacokinetically safe. Start trazodone at 50 mg for the first two weeks to assess tolerability, then titrate to 100 mg if sleep latency remains problematic. Because trazodone's efficacy may partially attenuate after four weeks, set a structured review at the six-week mark to assess whether the switch is actually serving the patient.

Patients who benefited from suvorexant's specific effect on WASO (waking in the middle of the night) may find trazodone less effective for that target, since trazodone's primary effect is on sleep-onset latency rather than sleep maintenance.

Switching From Trazodone to Suvorexant

Trazodone's half-life ranges from five to thirteen hours. Clinicians should taper trazodone over one to two weeks rather than stopping abruptly, then begin suvorexant at 10 mg on the night following the final trazodone dose. Starting at 10 mg rather than 20 mg minimizes next-morning sedation, which is the most common reason patients discontinue suvorexant in the first two weeks [8].

Patients with comorbid depression maintained partly by trazodone's antidepressant effect should have their psychiatric medication regimen reviewed before the switch, as removing even low-dose trazodone can destabilize mood in susceptible individuals.

Monitoring After Any Switch

Track these three parameters at the two-week and six-week follow-up visits:

• Sleep onset latency (patient-reported in minutes).
• Number of nocturnal awakenings.
• Daytime function score using the Insomnia Severity Index (ISI), which carries a validated cutoff of 15 for moderate-to-severe insomnia [11].

A patient who does not reach an ISI below 15 at week six on the new agent warrants referral for cognitive behavioral therapy for insomnia (CBT-I), which remains the first-line treatment per AASM guidelines regardless of which pharmacological agent is used [5].

CBT-I as the Foundation: Where Both Drugs Fit

Neither suvorexant nor trazodone should be prescribed without at minimum a brief behavioral sleep intervention. The AASM 2023 guideline gives CBT-I a "strongly recommended" rating based on high-quality evidence, while all pharmacological agents receive a "conditionally recommended" rating [5]. That distinction matters: it means the evidence committee judged that benefits of drug therapy outweigh harms for most patients, but the confidence interval on that judgment is narrower than for CBT-I.

A meta-analysis of 87 randomized trials published in Sleep Medicine Reviews (Trauer et al., 2015) found that CBT-I produced a mean reduction in sleep onset latency of 19.03 minutes and an increase in sleep efficiency of 9.91 percentage points, with effects maintained at 6-12 month follow-up [12]. Those sustained benefits exceed what trazodone's short-term trials demonstrated and approach what suvorexant showed at 12 months.

Combining CBT-I with pharmacotherapy produces additive short-term benefit in most trials, and some data suggest that patients who complete CBT-I while on a sleep medication can discontinue the medication more successfully than those who relied on medication alone.

Head-to-Head Data: Does Any Trial Directly Compare Them?

No randomized controlled trial has directly compared suvorexant and trazodone against each other in chronic insomnia. The competitor compounds were studied in different eras, against different control arms, using partially overlapping outcome measures. Any numerical comparison drawn between the two must be treated as cross-trial inference rather than head-to-head evidence.

The best approximation comes from network meta-analyses. A 2022 Cochrane-adjacent systematic review of pharmacological treatments for insomnia ranked suvorexant above trazodone on sleep maintenance outcomes based on network meta-analysis of 154 trials, but the confidence intervals on trazodone's network estimates were wide given the limited controlled trial data [7].

Trazodone's position in the network was driven heavily by the Mendelson 2005 trial and a handful of smaller studies, none exceeding six weeks. Suvorexant's position was driven by the Herring 2014 program plus subsequent six-month extension data [1, 3].

Clinical Decision Summary

The table below synthesizes the key differentiators for a prescribing clinician.

| Criterion | Suvorexant (Belsomra) | Trazodone (off-label) | |---|---|---| | Longest RCT for insomnia | 12 months [1] | 4-6 weeks [3] | | Tolerance signal in trials | None detected | Possible after 2-4 weeks [3] | | FDA-approved for insomnia | Yes | No | | DEA scheduling | Schedule IV | Unscheduled | | Monthly cost (no insurance) | ~$430 (brand) / lower generic | ~$10-20 | | Fall risk (orthostasis) | Low | Moderate (alpha-1 blockade) | | Rebound insomnia | Mild, transient | Less characterized | | Best evidence population | Adults 18 and older; studied at 65+ | General adult; less data in elderly |