Clomipramine (Anafranil): Independent Evidence on OCD & Risks

Table of contents

• Evidence summary
• What clomipramine 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 clomipramine
• 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

Independent evidence and credibility scorecard

Independence tiers: 1 = regulator or government body; 2 = academic or Cochrane team with no declared commercial funding for the work; 3 = independent authors analysing mostly manufacturer-run trials, or minor declared industry ties; 4 = manufacturer-funded or manufacturer-authored. Credibility runs from A (highest) to D.

Most of the independent sources cluster in the UK and US. The core OCD trials are mostly 30 to 35 years old; we did not re-check the dates of the trials in the 2024 review (Cohen 2024).

What clomipramine is

Clomipramine hydrochloride is a tricyclic antidepressant of the dibenzazepine group, ATC code N06AA04 (UK SmPC 5.1). The US label calls it "an antiobsessional drug" from the tricyclic class (FDA label).

• Brand name: Anafranil. In the US, Anafranil capsules (25, 50 and 75 mg) are made in Canada by Patheon for SpecGx LLC; the label states that Anafranil "is a trademark of Mallinckrodt LLC" (FDA label).
• US approval: NDA 019906, approved on 29 December 1989 as a new molecular entity under priority review (Drugs@FDA; openFDA).
• UK products: 10 mg, 25 mg and 50 mg capsules from Generics (UK) Ltd, trading as Viatris.
• Status: Prescription-only in the US and UK.
• Veterinary versions: DailyMed also lists clomipramine tablets for dogs. These are separate animal products and are not covered here.

How it works

Clomipramine blocks the reuptake of both noradrenaline and serotonin, "with inhibition of 5-HT reuptake being the more important" (UK SmPC 5.1). The US label is more cautious: the mechanism in OCD "is unknown", but serotonin reuptake inhibition "is thought to be important" (FDA label).

The serotonin link is supported indirectly in humans. Other tricyclics appear to lack efficacy in OCD, and the selective serotonin reuptake inhibitors (SSRIs) do work (Pigott and Seay 1999). That is consistent with the theory, but it is not direct proof.

Clomipramine also blocks several other receptors: alpha-1 adrenergic, muscarinic (anticholinergic), histamine and some serotonin receptors (UK SmPC 5.1). These actions explain most of its side effects rather than its benefits: dry mouth, constipation and urinary problems (anticholinergic), drowsiness and weight gain (antihistamine), and dizziness on standing (alpha-1).

Pharmacokinetics that matter in practice

• Active metabolite: It is turned into desmethylclomipramine, which is also active. After a 150 mg dose, the half-life averaged 32 hours for clomipramine and 69 hours for the metabolite (FDA label).
• Non-linear build-up: Blood levels can rise disproportionately at higher doses. The US label warns of possibly "dramatically higher" levels above 150 mg a day, which may increase dose-dependent risks, "in particular seizures".
• Wide variation: At 75 mg a day, steady-state levels ranged from about 20 to 175 ng/ml between patients (UK SmPC 5.2).
• Genes and enzymes: Clearance depends heavily on the liver enzyme CYP2D6. About 7% to 10% of white people are "poor metabolisers", and some drugs make normal metabolisers behave like poor ones (FDA label).
• Smoking: In smokers, clomipramine blood levels were about half those in non-smokers (UK SmPC 4.5).

What it is prescribed for

Licensed uses

Where guidelines place it

• OCD (NICE): SSRIs and CBT with exposure and response prevention come first. If there has been no response to an SSRI, a different SSRI or clomipramine should be offered. Clomipramine should be considered after at least one SSRI has been ineffective or poorly tolerated, or if the person prefers it or has responded to it before (NICE CG31, 1.5.4.3–1.5.4.4). Other tricyclics should not normally be used for OCD.
• NICE safety advice for OCD: prescribe only small amounts at a time to people at significant risk of suicide "because of its toxicity in overdose", and do an ECG and blood pressure check first in people at significant cardiovascular risk (NICE CG31, 1.5.4.11–1.5.4.12).
• Panic disorder (NICE): offer an SSRI first. If an SSRI is not suitable or there is no improvement after 12 weeks, imipramine or clomipramine may be considered (off-label in the UK) (NICE CG113, 1.3.26).
• Treatment-resistant OCD: after specialist review, NICE lists adding an antipsychotic to an SSRI or clomipramine, or combining clomipramine with citalopram, as options (NICE CG31, 1.5.4.7). That combination raises interaction concerns, which is why it is reserved for specialist teams.

What works and what does not

Benefits by claim

Obsessive-compulsive disorder

The case for clomipramine in OCD starts with the US registration trials. Their sizes and percentage falls were not confirmed in our check, so the independent meta-analysis below carries more weight (Clomipramine Collaborative Study Group 1991).

The best independent test is the publicly funded network meta-analysis by Skapinakis and colleagues (funder as stated in the article; not re-checked here): 54 trials and 6,652 participants. Clomipramine reduced Y-BOCS scores by 4.72 points more than placebo (95% credible interval −6.85 to −2.60) (Skapinakis 2016). The Y-BOCS runs from 0 to 40, so a 4 to 5 point extra fall is a real but moderate gain for most people.

Is clomipramine stronger than an SSRI?

This question has been argued for 30 years. Evidence on each side:

• For a clomipramine edge: A 1995 meta-analysis (Greist et al.) is cited for the case that clomipramine is more effective; we did not re-check its design or quoted wording (Greist et al. 1995). A 2002 meta-regression by Ackerman and Greenland is cited for this question, but we did not re-check its findings (Ackerman and Greenland 2002). A 2024 bias-adjusted meta-analysis of placebo-controlled OCD trials also found clomipramine more effective than SSRIs. The same review found signs of publication bias across OCD drug trials, and correcting for it shrank the overall drug effect (Cohen et al. 2024).
• Against: In direct comparisons, 5 of 6 head-to-head trials of fluoxetine or fluvoxamine against clomipramine found similar efficacy with fewer side effects on the SSRI (Pigott and Seay 1999). In a 406-person, 12-week trial, paroxetine was of "comparable efficacy" to clomipramine and had better tolerability (Zohar and Judge 1996). The 2016 network meta-analysis found no significant difference, and the gap closed further when old clomipramine trials that analysed only completers were removed (Skapinakis 2016, full text).

Why the comparisons disagree. The placebo-controlled clomipramine trials were mostly run before SSRIs existed, so patients were less likely to have already failed a serotonergic drug. Cohen and colleagues tested study quality, sponsor status and publication year as possible explanations (Cohen 2024). A further possibility, which we could not find tested in any trial, is unblinding: clomipramine's obvious side effects (84% dry mouth, 54% tremor) may have let patients and raters guess who was on the active drug, which can inflate apparent benefit. We flag this as an untested concern, not a finding. Our reading: clomipramine is at least as effective as an SSRI for OCD, possibly slightly more, but the extra benefit is uncertain and comes with much heavier side effects.

Depression

In the largest independent comparison of antidepressants, all 21 drugs beat placebo for response in adults with major depression. It stood out for acceptability: it was the only drug with more dropouts than placebo (OR 1.30, 95% CrI 1.01 to 1.68), and certainty across comparisons ranged from moderate to very low (Cipriani 2018). The UK licence frames it for depression "especially where sedation is required" (UK SmPC). In practice, its overdose toxicity makes it a poor first choice for anyone at risk of self-harm.

Panic disorder and phobias

A 2023 Cochrane network meta-analysis of 70 trials listed clomipramine among the medicines with the strongest effect on treatment response versus placebo, and among the strongest for agoraphobia symptoms. We did not re-check the review's risk-of-bias ratings (Guaiana et al. 2023). The UK SmPC warns that many people with panic disorder feel "intensified anxiety" in the first days of antidepressant treatment, usually settling within two weeks (UK SmPC 4.4).

Cataplexy in narcolepsy

Clomipramine is licensed in the UK as an add-on for cataplexy, at 10 to 75 mg a day (UK SmPC 4.2). The Cochrane review (2008) found no reliable evidence that antidepressants help cataplexy. Its only clomipramine trial is a small crossover against fluvoxamine; the participant numbers reported elsewhere (18 enrolled, 11 analysed) were not confirmed in our check, and the reviewers judged the trial unsuitable for drawing a conclusion (Vignatelli 2008, full text). Stopping suddenly can worsen cataplexy, including status cataplecticus (UK SmPC 4.4).

Risks and all side effects

Common side effects (US OCD trials)

Pooled placebo-controlled OCD trials: 322 adults on Anafranil (FDA label).

About 20% of 3,616 patients in US premarketing trials stopped because of an adverse event, most often drowsiness. The UK SmPC adds restlessness, increased appetite, micturition problems and blood sugar changes as very common (UK SmPC 4.8).

Serious and rare risks

Overdose toxicity compared with other antidepressants

A UK study linked deaths from single-antidepressant poisoning in England and Wales (2000 to 2006) to UK prescribing (Hawton et al. 2010). Its published abstract reports case fatality ratios of 13.8 for tricyclic antidepressants as a class and 0.5 for SSRIs. We could not confirm clomipramine's own toxicity index against other tricyclics in this check, so we do not rank it here. The class-level figures are why tricyclics, clomipramine included, sit in the high-risk group for overdose.

Withdrawal

Stopping suddenly can cause nausea, vomiting, abdominal pain, diarrhoea, insomnia, headache, nervousness and anxiety (UK SmPC 4.8). The US label adds dizziness, malaise, hyperthermia and irritability, and possible worsening of psychiatric symptoms (FDA label). Both labels say to reduce the dose gradually under supervision. Neither gives a fixed taper, and we do not either.

All interactions

Clomipramine has two kinds of interaction: drugs that raise its blood level (mainly through CYP2D6, CYP1A2 and CYP3A4), and drugs that add to its effects on serotonin, the heart, sedation or anticholinergic load.

Who should avoid clomipramine

Contraindications

• US: hypersensitivity to clomipramine or other tricyclics; use with or within 14 days of a psychiatric MAOI; starting during linezolid or intravenous methylene blue; the acute recovery period after a heart attack (FDA label).
• UK: hypersensitivity; recent heart attack, any degree of heart block or other arrhythmias; severe liver disease; MAOIs (including moclobemide); narrow-angle glaucoma; urinary retention; mania (UK SmPC 4.3).

Use with extreme care

Epilepsy or other seizure risks (brain injury, alcohol or benzodiazepine withdrawal); heart disease; prostate enlargement or bladder outflow problems; raised eye pressure; overactive thyroid or thyroid medicine; adrenal tumours such as phaeochromocytoma; chronic constipation; bipolar disorder; schizophrenia; liver or kidney disease; and contact lens wearers, because reduced tears can damage the cornea (UK SmPC 4.4; FDA). Before planned surgery, the anaesthetist must know.

Older adults

The specific Beers Criteria and UK SmPC wording on clomipramine in older adults was not confirmed in our check, so older-adult advice should be checked against the label and with the prescriber (AGS Beers Criteria 2023).

Pregnancy and breastfeeding

• Pregnancy: Swedish registry figures on first-trimester exposure were not re-checked in this review. The UK SmPC says clomipramine is "not recommended during pregnancy" (UK SmPC 4.6). An earlier Swedish case-control study flagged clomipramine among drugs associated with infant heart defects, while noting confounding and multiple testing (Källén and Otterblad Olausson 2003).
• Near delivery: Babies exposed until birth have had jitteriness, tremor and seizures (FDA), and breathing difficulty, lethargy, irritability and blood pressure changes (UK SmPC 4.6).
• Breastfeeding: It passes into breast milk. The UK SmPC advises stopping the medicine or stopping breastfeeding; the US label asks for an individual decision (UK SmPC; FDA).

Pregnancy decisions should weigh the risks of untreated OCD or depression too. This needs a specialist conversation, not a blanket stop.

Children and adolescents

In the US, the label covers children and adolescents with OCD (its paediatric trial enrolled ages 10 to 17), with a lower maximum dose. Safety and effectiveness for other uses in young people have not been established (FDA label). The UK SmPC says it should not be used for depression, phobias or cataplexy under 18 (UK SmPC 4.4).

Dosage and how to take it

Your prescriber sets the dose. The ranges below are official label doses, shown for context only. They are not a guide to self-dosing.

How it is taken

At first it is usually taken in divided doses with meals to reduce stomach upset; later the whole daily dose may be taken at bedtime to reduce daytime drowsiness (MedlinePlus). Because the drug and its metabolite take 2 to 3 weeks to reach steady levels, the US label suggests waiting 2 to 3 weeks between later dose changes (FDA label). Low potassium should be corrected before starting (UK SmPC 4.2).

How long before it works

"It may take several weeks or longer" to feel the full benefit (MedlinePlus). The OCD trials measured results at 10 weeks.

Driving and dangerous activities

Because of drowsiness, blurred vision and seizure risk, the US label advises caution with driving, complex machinery, swimming and climbing (FDA label). In the UK it falls under drug-driving law; the SmPC sets out the statutory defence (UK SmPC 4.7).

Product quality

We found no recalls or contamination findings specific to clomipramine in the sources we reviewed. Some UK capsules contain lactose and the dye sunset yellow (UK SmPC, 10 mg).

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

Who made it, and who sells it now

• Origin: A review of "me-too" medicines lists clomipramine (Anafranil) as a Geigy product dating from 1968 (Aronson and Green 2020); another review dates its European introduction to 1970 (López-Muñoz et al. 2022). Geigy merged with Ciba in 1970, and Ciba-Geigy merged with Sandoz in 1996 to create Novartis, based in Basel, Switzerland (Novartis company history).
• US approval: FDA approval followed on 29 December 1989 (Drugs@FDA).
• US brand today: NDA 019906 is now held by SpecGx LLC, and the label says Anafranil "is a trademark of Mallinckrodt LLC" (FDA label). We did not find a public record of when or how the US rights passed from Novartis to Mallinckrodt.
• US generics: openFDA lists 22 clomipramine applications; 13 were marked as currently prescription products, including makers such as Sandoz, Taro, Lupin, Zydus, Aurobindo, Alembic, Micro Labs, Mankind and Jubilant Cadista (openFDA).
• UK: The SmPCs listed on the electronic medicines compendium are held by Generics (UK) Ltd, trading as Viatris (formerly Mylan) (UK SmPC).

Documented integrity event: the clomipramine price-fixing allegations

In May 2019, Connecticut led a coalition of US states in suing Teva and 19 other generic makers over alleged price-fixing of more than 100 generic drugs (Connecticut Attorney General, 2019). The complaint includes a section on clomipramine. The states allege that Taro, Sandoz and Mylan agreed to raise the price of clomipramine (States' complaint (2019)).

These are allegations. In August 2026, the states announced a $400 million settlement with Sandoz (about $469 million including earlier state deals), backed by a coalition of 43 states and territories, not yet final, because every participating state must sign and covering numerous generic drugs, not only clomipramine. It would also resolve the states' allegations that Sandoz's former parent Novartis AG and other Sandoz affiliates took part in the conduct and moved assets to avoid liability. The settlement resolves allegations rather than making a finding of guilt (Connecticut Attorney General, August 2026). We did not find a clomipramine-specific resolution for Taro or Mylan in the records we checked. This case is about pricing, not about the medicine's safety or its trial results.

Who funded the evidence

• Approval trials: The US registration trials, reported by the Clomipramine Collaborative Study Group, underpin the FDA label's efficacy claim; their size and centre count were not confirmed in our check (CCSG 1991). We could not see the paper's funding statement. As registration trials for the maker's application, we treat them as manufacturer evidence.
• The "clomipramine is stronger" case: The 1995 pooled comparison's design and funding were not confirmed in our check (Greist 1995). The 2024 meta-analysis's grant and competing-interest statements were not verified in our check (Cohen 2024).
• The independent network meta-analyses: Skapinakis 2016 (declarations) is described in the article as NIHR-funded; this review could not re-check its full-text funding statement or individual author declarations. Cipriani 2018 (full text) is described as funded by the NIHR Oxford Health Biomedical Research Centre and the Japan Society for the Promotion of Science, with some authors declaring drug-company fees; not re-checked here.
• Safety data: Hawton 2010 funding was not re-checked in this review (Hawton 2010). The Beers Criteria sponsor statement was not verified (AGS 2023).
• Cochrane's rule: Cochrane reviews "cannot be funded or conducted by commercial sponsors" with a vested interest in the findings (Cochrane policy).

What this means: The proof that clomipramine works for OCD comes mainly from its maker's registration trials, and independent re-analyses broadly confirm it. The claim that it beats SSRIs rests mostly on comparisons between separate company programmes; independent head-to-head and network evidence is less convinced. The safety warnings come from regulators and publicly funded researchers. Because clomipramine is now a cheap generic, no company has a reason to fund new trials, so the evidence base is unlikely to improve.

Related research

For wider, evidence-graded context, see these Pure City Research guides:

• Obsessive-compulsive disorder: assessment, treatment evidence and safety
• Exposure and response prevention (ERP)
• Body dysmorphic disorder
• Trichotillomania (hair-pulling disorder)
• Panic disorder
• Narcolepsy
• Antidepressant withdrawal and dependence
• Torsades de pointes

Sibling medicine reviews:

• Fluvoxamine
• Fluoxetine
• Sertraline
• Paroxetine
• Amitriptyline
• Nortriptyline
• Dosulepin
• St John's wort (interacts with clomipramine)

Frequently asked questions

Sources and funding notes

What we could not verify in this check (PubMed and NICE pages were blocked or rate-limited, and search results gave abstracts only): the funding statements of the 1991 Clomipramine Collaborative Study Group registration trials (treated as manufacturer evidence, sponsor not seen), Greist 1995, Ackerman and Greenland 2002, Pigott and Seay 1999 and Zohar and Judge 1996; the design of Greist 1995 and the findings of Ackerman and Greenland 2002; the registration-trial sample sizes and percentage falls; the placebo column of the label table (only the active-arm figures for dry mouth, drowsiness, tremor, dizziness and ejaculation failure were confirmed); the Beers Criteria and UK SmPC older-adult wording; the Swedish pregnancy registry figures; the 2013 timing and details of the price allegations against Taro, Sandoz and Mylan, and whether those companies have resolved them; the 2019 complaint paragraphs, which were not read directly; and the date and route of Novartis' transfer of US Anafranil rights to Mallinckrodt. Corporate history, approval dates and the remaining label dosing, overdose and interaction details were not re-checked item by item. Cipriani 2018 reported clomipramine's efficacy odds ratio in a figure we could not extract as text, so we report only its acceptability result. We found no trial that tested whether clomipramine's side effects unblinded patients or raters, so we present that only as an untested concern.

1. FDA-approved prescribing information: Anafranil (clomipramine hydrochloride) capsules, SpecGx LLC (DailyMed, Rev 09/2024) — US regulator-approved label; efficacy data from the original sponsor's trials.
2. Drugs@FDA: NDA 019906 (Anafranil) — US FDA database; approved 29 December 1989.
3. openFDA Drugs@FDA query: clomipramine hydrochloride applications — US government data.
4. Clomipramine 25 mg Capsules, UK SmPC (Generics (U.K.) Ltd t/a Viatris; revised July 2026) — MHRA-approved product information.
5. Clomipramine 10 mg Capsules, UK SmPC (Viatris) — MHRA-approved; excipients.
6. MedlinePlus: Clomipramine (last revised 15 March 2026) — US National Library of Medicine; text licensed from ASHP.
7. NICE CG31: Obsessive-compulsive disorder and body dysmorphic disorder — UK government-funded guideline.
8. NICE CG113: Generalised anxiety disorder and panic disorder in adults — UK government-funded guideline.
9. Skapinakis P et al. Interventions for OCD in adults: network meta-analysis. Lancet Psychiatry 2016 (full text) — UK/Greece; funding as stated in the article, full-text declarations not re-checked.
10. Cipriani A et al. Comparative efficacy and acceptability of 21 antidepressants. Lancet 2018 (full text) — UK/Japan/multinational; NIHR Oxford Health BRC and JSPS; some authors declared pharma fees.
11. Cohen SE et al. Study characteristics, rigour and publication bias in OCD pharmacotherapy trials. BMJ Ment Health 2024 — Netherlands; grant and competing interests not verified here.
12. Greist JH et al. Efficacy and tolerability of serotonin transport inhibitors in OCD: a meta-analysis. Arch Gen Psychiatry 1995 — US; citation confirmed; design and funding not verified here.
13. Ackerman DL, Greenland S. Multivariate meta-analysis of controlled drug studies for OCD. J Clin Psychopharmacol 2002 — US; citation confirmed; funding not verified here.
14. Pigott TA, Seay SM. Review of SSRI efficacy in OCD. J Clin Psychiatry 1999 — US; funding not verified.
15. Zohar J, Judge R. Paroxetine versus clomipramine in OCD. Br J Psychiatry 1996 — multinational trial; sponsor not verified.
16. Clomipramine Collaborative Study Group. Clomipramine in OCD. Arch Gen Psychiatry 1991 — US registration trials; funding statement not seen.
17. Guaiana G et al. Pharmacological treatments in panic disorder: network meta-analysis. Cochrane 2023 — Cochrane; author declarations not re-checked here.
18. Vignatelli L et al. Antidepressant drugs for narcolepsy. Cochrane 2008 (full text) — Italy; funding and declarations not verified.
19. Hoffman J et al. Pharmacotherapy for trichotillomania. Cochrane 2021 — Cochrane review; author declarations not verified here.
20. Hawton K et al. Toxicity of antidepressants. Br J Psychiatry 2010 (full text) — UK; funding not re-checked in this review.
21. American Geriatrics Society 2023 Beers Criteria — US professional society; sponsor statement not verified.
22. Källén BA, Otterblad Olausson P. Maternal drug use and infant cardiovascular defect. Reprod Toxicol 2003 — Sweden; registry study; funding not verified.
23. Aronson JK, Green AR. Me-too pharmaceutical products. Br J Clin Pharmacol 2020 — UK; history of tricyclic introductions.
24. López-Muñoz F et al. Serendipity in the discovery of classical antidepressants. World J Psychiatry 2022 — history; dates of introduction.
25. Novartis: company history — company source; Geigy, Ciba-Geigy and the 1996 merger.
26. Connecticut Attorney General press release, May 2019 — US state government.
27. State of Connecticut et al. v. Teva Pharmaceuticals USA et al., complaint (redacted public version, 2019) — court filing; allegations only.
28. Connecticut Attorney General: settlement with Sandoz (August 2026) — US state government; not yet final.
29. Cochrane commercial sponsorship policy — Cochrane.

Educational research reviewed 6 October 2026. This article is not medical advice; talk to a qualified clinician or pharmacist before starting, stopping or combining treatments.