Low-Dose Doxepin (Silenor): Independent Evidence on Insomnia & Side Effects

Table of contents

• Evidence summary
• What low-dose doxepin is
• How it works
• What it is prescribed for
• What works and what does not
• Benefits by claim
• Risks and all side effects
• All interactions
• Who should avoid low-dose doxepin
• Dosage and how to take it
• Follow the money: who makes it and who funded the evidence
• Related research
• Frequently asked questions
• Sources and funding notes

Evidence summary

The table ranks the main claims about low-dose doxepin for sleep by the strength of the human evidence. Independent reviews are listed before company-run trials.

Independent evidence and credibility scorecard

Independence tier: 1 means no money from the product's sellers; 4 means paid for by the seller. Credibility runs from A (highest) to D.

What low-dose doxepin is

Doxepin is a tricyclic antidepressant (TCA). Somaxon's filings note it "has been marketed and used for over 35 years at dosages from 75 mg to 300 mg per day" for depression and anxiety (Somaxon 10-K for 2011). Silenor repackages the same molecule as 3 mg and 6 mg tablets for insomnia. That is between 10 and 100 times less than antidepressant doses (Silenor label).

• Brand name (sleep): Silenor 3 mg and 6 mg tablets (US). openFDA lists eight approved generic (ANDA) applications for 3 mg/6 mg doxepin tablets, from Zydus, Strides, Taro, Ajanta, Aurobindo, MSN, Actavis and RK Pharma (openFDA).
• US approval for insomnia: March 2010 (NDA 022036). It went on sale in September 2010 (openFDA Drugs@FDA, Somaxon 10-K).
• Developer of the low-dose product: Somaxon Pharmaceuticals, San Diego, California, USA. Somaxon in-licensed the low-dose patents from ProCom One, Inc. in 2003. The original developer of doxepin itself was not verified in the sources reviewed.
• Current US holder: Currax Pharmaceuticals LLC, Brentwood, Tennessee (Silenor label).
• Controlled status: "Doxepin is not a controlled substance" (Silenor label). It is prescription-only.
• UK status: we found no licensed 3 mg or 6 mg doxepin product on the UK electronic medicines compendium. Oral doxepin in the UK comes as 10 mg, 25 mg and 50 mg capsules (e.g. Sinepin, Marlborough Pharmaceuticals). These are licensed for "symptoms of depressive illness in adults, especially where sedation is required", at 25–300 mg daily (Sinepin SmPC, Doxepin 10 mg SmPC). Any UK use of doxepin for insomnia is therefore off-label.

How it works

At antidepressant doses, doxepin blocks the reuptake of serotonin and noradrenaline and also acts on several other receptors. That is why high doses cause dry mouth, constipation, urinary retention, dizziness on standing and weight gain.

At 3–6 mg the picture is narrower. The Silenor label says the sleep mechanism "is unclear; however, doxepin's effect could be mediated through antagonism of the H₁ receptor". It notes that doxepin "has high binding affinity to the H₁ receptor (Ki < 1 nM)" (Silenor label). Histamine is a wakefulness signal in the brain. A very low dose can block it while hardly touching the other receptors responsible for tricyclic side effects. This is the theory behind the "ultra-low-dose" approach (Rojas-Fernandez & Chen, 2014).

How long it stays in the body explains its profile (all figures from the Silenor label):

• Slow to peak. Peak levels arrive about 3.5 hours after a 6 mg dose on an empty stomach. This is one reason it helps the second half of the night more than falling asleep.
• Food. Taking it with a high-fat meal raises total exposure by 41% and delays the peak by about 3 hours. Hence the rule not to take it within 3 hours of a meal.
• Half-life. The half-life (time for blood levels to halve) is 15.3 hours for doxepin and 31 hours for its active breakdown product, nordoxepin.
• Metabolism. The liver breaks it down mainly through CYP2C19 and CYP2D6, and to a lesser extent CYP1A2 and CYP2C9. People who are "poor metabolisers" through these enzymes may have higher levels.

What it is prescribed for

Licensed uses.

• US (Silenor): "treatment of insomnia characterized by difficulty with sleep maintenance". The supporting trials lasted up to 3 months (Silenor label).
• US (capsules, 10 mg and above): depression and/or anxiety (capsules label).
• UK: licensed only for depressive illness (Sinepin SmPC).

Where guidelines place it.

What works and what does not

Benefits by claim

Staying asleep: the guideline meta-analysis

The AASM 2017 guideline pooled four or five Somaxon-era trials for each outcome. Compared with placebo:

• Total sleep time (sleep lab): +26.14 minutes at 3 mg (95% CI +18.49 to +33.79) and +32.27 minutes at 6 mg (95% CI +24.24 to +40.30). The task force judged both increases clinically significant.
• WASO (sleep lab): −22.17 minutes at 3 mg and −23.14 minutes at 6 mg. Both were clinically significant.
• Sleep efficiency (share of time in bed spent asleep): +6.78% at 3 mg and +7.06% at 6 mg.
• Sleep latency (time to fall asleep): −2.30 minutes at 3 mg and −5.29 minutes at 6 mg. Both were below clinical significance.
• Number of awakenings: not reduced (+0.53 and +0.44).

The task force judged that "the quality of evidence in the meta-analytic data for the majority of variables was moderate to low due to industry sponsorship and, in some cases, imprecision", and that "the overall quality of evidence for the doxepin data is considered very low". Even so, it concluded that benefits outweighed harms.

Independent reviews

• Cochrane. The Cochrane review of antidepressants for insomnia analysed six tricyclic trials (812 participants), five of them doxepin. Results compared with placebo:
  • subjective sleep quality SMD −0.39 (moderate quality);
  • sleep efficiency +6.29 percentage points;
  • total sleep time +22.88 minutes (95% CI 13.17 to 32.59);
  • sleep latency −4.27 minutes, not significant.
  Its conclusion was cautious: "There may be a small improvement in sleep quality with short-term use of low-dose doxepin and trazodone compared with placebo." It also found no evidence at all "for long-term antidepressant use for insomnia".
• Yeung et al. (Hong Kong). This team asked companies and the FDA for unpublished data. They found "low-dose doxepin had a small to medium effect size against placebo for sleep maintenance and sleep duration but not for sleep initiation". They warned that firm conclusions on short-term benefits, risks and withdrawal were not possible "due to the small number of studies" (Yeung et al., 2015).
• Rojas-Fernandez & Chen (Canada). This review, which declared no conflicts, reached the same pattern: better sleep maintenance and duration, no significant effect on falling asleep (Rojas-Fernandez & Chen, 2014).
• Lancet NMA. The publicly funded Lancet network meta-analysis did not list doxepin among drugs with clear acute efficacy on its main sleep-quality outcome. It concluded: "Doxepin, seltorexant, and zaleplon were well tolerated, but data on efficacy and other important outcomes were scarce and do not allow firm conclusions."

The company trials

Silenor's approval rested on six randomised, double-blind studies with 1,423 participants aged 18 to 93. The main ones, all co-authored by Somaxon Pharmaceuticals staff, were:

• Older adults, 12 weeks. 240 people aged over 65. The 3 mg dose improved WASO, total sleep time and sleep efficiency on night 1, and these gains were maintained at night 85 (Krystal et al., Sleep 2010).
• Adults, 35 days. 221 adults. Both doses improved WASO, and there was "no evidence of rebound insomnia" after stopping (Krystal et al., Sleep 2011).
• Older adults, 4 weeks, outpatient. 254 people. The 6 mg dose increased self-reported total sleep time. There was no difference in time to fall asleep (Lankford et al., Sleep Med 2012).
• Adults, crossover. 67 adults each took 1, 3 and 6 mg and placebo in turn (Roth et al., Sleep 2007).

Affiliations list Somaxon Pharmaceuticals, San Diego, for several authors of each trial (Krystal 2010, Krystal 2011, Roth 2007).

Is the benefit clinically meaningful?

Roughly 20–30 extra minutes of sleep and 20 fewer minutes awake in the night is a modest, noticeable change for someone who wakes repeatedly. It is not a cure.

• Placebo responses in insomnia trials are large. One meta-analysis estimated "63.56% of the drug responses are achieved even in the placebo groups" across sleep-drug trials in general (Winkler & Rief, Sleep 2015).
• The honest summary is a small benefit, mainly in the second half of the night, with a favourable short-term side-effect record at 3–6 mg.
• That record comes from a thin evidence base, and all the trials that make it up were paid for by the company.

Risks and all side effects

No boxed warning at the sleep dose. Silenor does not carry a boxed warning. Its label does carry warnings about complex behaviours, suicide risk, CNS depression and severe sleep apnoea (Silenor label). By contrast, doxepin capsules used for depression carry the class boxed warning that antidepressants "increased the risk compared to placebo of suicidal thinking and behavior (suicidality) in children, adolescents, and young adults" (capsules label).

Common side effects at 3–6 mg

Dropouts. In the development programme, 0.6% on placebo stopped because of an adverse reaction, compared with 0.4%, 1.0% and 0.7% on 1, 3 and 6 mg (Silenor label).

Serious and important risks

Dependence, withdrawal and rebound

Unlike Z-drugs and benzodiazepines, doxepin "is not a controlled substance" and "is not associated with abuse potential in animals or in humans". In a discontinuation assessment "no symptoms indicative of a withdrawal syndrome were observed", and the 35-day study found no rebound insomnia at 3 or 6 mg. The one signal was nausea and vomiting in 5% of people stopping 6 mg (Silenor label). These findings come from company studies lasting up to 3 months. Yeung and colleagues said withdrawal effects could not be firmly judged from the small evidence base (Yeung 2015).

All interactions

Always give your prescriber and pharmacist a full list of medicines, including over-the-counter sleep aids and allergy tablets. Many contain sedating antihistamines.

Who should avoid low-dose doxepin

• People taking an MAOI, or who stopped one within the past two weeks (Silenor label).
• Untreated narrow-angle glaucoma or severe urinary retention, both contraindications. People prone to urinary retention should start at 3 mg.
• Hypersensitivity to doxepin or other dibenzoxepines.
• Severe sleep apnoea: "ordinarily not recommended". It has not been studied in obstructive sleep apnoea.
• Pregnancy: studies have not shown a raised risk of major birth defects. Newborns exposed to tricyclics late in pregnancy have needed "prolonged hospitalization, respiratory support, and tube feeding" because of poor neonatal adaptation.
• Breastfeeding: "breastfeeding is not recommended". Excess sedation and breathing depression have been reported in breastfed infants.
• Children and teenagers: safety and effectiveness have not been evaluated.
• Older adults: start at 3 mg. Beers supports doses of 6 mg or less as comparable to placebo for safety, but says avoid anything higher. It also lists tertiary tricyclics among drugs to avoid in people with a history of syncope (Beers 2023).
• Liver impairment: start at 3 mg and monitor for daytime effects. Kidney impairment is not expected to matter much.
• People with depression, bipolar disorder or suicidal thoughts: needs specialist oversight. The low dose is not an antidepressant treatment.

Dosage and how to take it

Your prescriber sets the dose. The ranges below are the official US licensed ranges for Silenor, given for information only. They are not personal dosing advice.

Source: Silenor prescribing information. For comparison, antidepressant doses are 75–150 mg a day and up to 300 mg (US capsules label), or 25–300 mg a day in the UK (Sinepin SmPC).

Antidepressant-strength capsules are not a sleep dose. The licensed sleep product comes as 3 mg and 6 mg tablets. UK doxepin capsules start at 10 mg, which is already above the 6 mg Beers threshold. Any other dose or formulation is a decision for the prescriber.

How quickly it works. In the trials, benefits on night-time waking appeared from the first night (Krystal 2010). Because peak levels arrive about 3.5 hours after the dose, it acts mostly on the middle and later part of the night.

How to stop. Company studies found no withdrawal syndrome or rebound insomnia at 3–6 mg, though 5% had nausea or vomiting after stopping 6 mg (Silenor label). Plan any change with your prescriber. People on higher antidepressant doses need a supervised taper.

Follow the money: who makes it and who funded the evidence

From an old antidepressant to a new sleep brand

• The licence. Doxepin was a long-established generic antidepressant. In August 2003, the newly formed Somaxon Pharmaceuticals (San Diego, California) took an exclusive worldwide licence from ProCom One, Inc. to patents "to develop and commercialize low dosages of doxepin for the treatment of insomnia". Somaxon acknowledged that it did "not hold composition of matter patents" on doxepin, only use and formulation patents. Those included a patent covering 0.5–20 mg doses for transient insomnia, due to expire in February 2020 (Somaxon 10-K for 2011).
• Launch and partners. The FDA approved Silenor in March 2010, and it launched in September 2010. Somaxon then signed:
  • a co-promotion agreement with Procter & Gamble (August 2010), and a possible deal on over-the-counter rights;
  • a licence with Paladin for Canada, South America, the Caribbean and Africa (June 2011).
• Sales against spending. Silenor net product sales were $1.38 million in 2010 and $16.16 million in 2011. Selling, general and administrative expenses were $36.6 million and $69.8 million in those years (Somaxon 10-K). In other words, the company spent several times more on selling and running the business than the product earned.
• Patent challenges. Actavis, Mylan, Par and Zydus each filed generic applications with paragraph IV patent challenges, and Somaxon and ProCom sued them (Somaxon 10-K).
• Pernix. On 6 March 2013, Somaxon merged into Pernix Therapeutics Holdings and was renamed Pernix Sleep, Inc. (SEC Form 8-K).
• Bankruptcy and Currax. Pernix filed for Chapter 11 bankruptcy in Delaware on 18 February 2019. Under an April 2019 asset purchase agreement, its assets, including "the Silenor Assets", were sold to Currax Holdings LLC (formerly Phoenix Top Holdings LLC). The agreement's notice address for the buyer was "c/o Highbridge Capital Management, LLC", an investment firm (Pernix–Currax asset purchase agreement).
• Today. The US label is held by Currax Pharmaceuticals LLC, Brentwood, Tennessee (Silenor label). Eight generic 3 mg/6 mg tablet applications are now approved (openFDA).
• Not verified. We did not verify current ownership of Currax beyond the 2019 agreement, or any recent Silenor revenue figures.

Who funded the evidence

• Every low-dose trial was company-funded. An independent systematic review states it directly: "All low-dose studies were industry-sponsored" (Yeung et al., 2015). Author affiliations in the main trials include Somaxon Pharmaceuticals staff (Krystal 2010, Krystal 2011, Roth 2007).
• The US guideline and its author ties.
  • The AASM guideline downgraded the doxepin evidence for "industry sponsorship" and still issued a weak "suggest".
  • One of its four authors, Andrew Krystal, was first author of two Somaxon doxepin trials. He declared consultancy for Pernix, Silenor's owner from 2013, and for many other companies (AASM 2017 disclosures).
  • This was disclosed. It does not show the recommendation was wrong: independent reviews reached similar effect estimates.
• Independent checks:
  • the harm-focused Beers panel (no sponsor), which exempts doses of 6 mg or less;
  • the Hong Kong and Canadian systematic reviews;
  • the publicly funded Lancet NMA, which found doxepin well tolerated but its efficacy data "scarce";
  • the Cochrane review, which found only "a small improvement".

Our reading. The independent sources converge on the same conclusion: a small, real benefit for staying asleep, with a good short-term tolerability record at 3–6 mg. There is no independent long-term trial.

Related research

• Sleep: prevention and management guide: CBT-I, sleep scheduling and non-drug first steps
• Sleep supplements: independent evidence
• Melatonin · Magnesium · L-theanine
• Stress, anxiety and depression prevention guide · Stress, anxiety and depression supplements
• Sibling medicine reviews: Eszopiclone · Trazodone · Mirtazapine · Zolpidem · Amitriptyline

Frequently asked questions

Sources and funding notes

• Silenor (doxepin) tablets, US prescribing information, DailyMed / openFDA. Manufacturer-written label, approved and regulated by the FDA. Holder: Currax Pharmaceuticals LLC (USA).
• Doxepin hydrochloride capsules, US prescribing information (DailyMed). Antidepressant-dose label with a boxed suicidality warning. Repackager label, FDA-regulated.
• openFDA Drugs@FDA, NDA 022036 and doxepin tablet ANDA listing. US government data.
• Sinepin 50 mg capsules SmPC and Doxepin 10 mg capsules SmPC, electronic medicines compendium. MHRA-approved texts (UK). Holders: Marlborough Pharmaceuticals; Morningside Healthcare.
• Sateia MJ et al. AASM clinical practice guideline for the pharmacologic treatment of chronic insomnia. J Clin Sleep Med 2017 (PubMed). Funded by the AASM (USA). Krystal declared consultancy for Pernix, Merck, Jazz, Janssen and others, and research support from Sunovion; Neubauer consulted for Purdue.
• Everitt H et al. Antidepressants for insomnia in adults. Cochrane Database Syst Rev 2018. Academic (University of Southampton, UK). Baldwin declared grants and advisory boards with several antidepressant makers; Malizia is a former SmithKline Beecham director; Wilson had pharmaceutical lecture honoraria (none since 2009); others none. We could not access the review's sources of support.
• Yeung WF et al. Doxepin for insomnia: a systematic review of randomized placebo-controlled trials. Sleep Med Rev 2015. Academic (University of Hong Kong). Funding and conflicts not stated in the abstract.
• Rojas-Fernandez CH, Chen Y. Ultra-low-dose (≤6 mg) doxepin for insomnia in older people. Can Pharm J 2014. Academic (University of Waterloo, Canada). No conflicts declared.
• De Crescenzo F et al. Pharmacological interventions for insomnia: network meta-analysis. Lancet 2022. Funded by the NIHR Oxford Health BRC (UK public). The first author was employed by Boehringer Ingelheim; the senior author leads Janssen-sponsored seltorexant trials; some authors had Angelini fees.
• Riemann D et al. The European Insomnia Guideline 2023. J Sleep Res. European Sleep Research Society (pan-European). We could not verify author conflicts.
• American Geriatrics Society 2023 Beers Criteria. "There was no sponsor for this paper" (USA). Panel disclosures are listed in the paper.
• Winkler A, Rief W. Placebo response in primary insomnia. Sleep 2015. Academic (University of Marburg, Germany).
• Krystal AD et al. Doxepin 1 mg and 3 mg in elderly chronic insomnia, 12 weeks. Sleep 2010 (PMC). Manufacturer-linked: co-authors affiliated with Somaxon Pharmaceuticals.
• Krystal AD et al. Doxepin 3 and 6 mg in adults, 35 days. Sleep 2011 (PMC). Manufacturer-linked: Somaxon co-authors.
• Lankford A et al. Doxepin 6 mg in elderly adults, 4 weeks. Sleep Med 2012. Pivotal registration trial; the author list overlaps with the Somaxon trials. Funding statement not retrieved.
• Roth T et al. Doxepin 1, 3 and 6 mg in adults. Sleep 2007 (PMC). Manufacturer-linked: Somaxon co-author.
• NHS, Insomnia (reviewed 19 March 2024). UK public health service.
• Somaxon Pharmaceuticals, Annual report (Form 10-K) for 2011. Company SEC filing (USA). Regulated financial disclosures.
• Pernix Sleep (formerly Somaxon), Form 8-K, merger completion (March 2013). SEC filing.
• Amended and restated asset purchase agreement, Pernix Therapeutics and Currax Holdings LLC (April 2019). SEC filing exhibit.

Last reviewed: 26 September 2026. Educational information from Pure City Research, not medical advice. Medicine decisions belong with a qualified prescriber.