duloxetine (Cymbalta)

the boxed warning: suicidal thoughts and behaviors, by age

the boxed warning says antidepressants increased the risk of suicidal thoughts and behavior in children, adolescents and young adults in short-term studies, that the studies did not show an increase above age 24, that there was a reduction in risk at 65 and older, and that everyone started on one should be watched closely for worsening [label, boxed warning and 5.1]. the number behind it is pooled placebo-controlled trials of antidepressants generally: 24 pediatric trials of 9 drugs in more than 4,400 patients and 295 adult trials of 11 drugs in more than 77,000. the label states the difference against placebo per 1,000 people treated: under 18, 14 additional cases; 18 to 24, 5 additional; 25 to 64, 1 fewer; 65 and older, 6 fewer [label 5.1, table 1]. no suicides occurred in any of the pediatric duloxetine trials; there were suicides in the adult trials, too few to conclude anything about the drug’s effect on suicide itself [label 5.1].

so the warning is specific: thoughts and behaviors, not deaths; the first months, not years (the label says the risk beyond several months is unknown); and an age boundary printed in the label itself, above which the pooled trials showed no increase [label 5.1; age-ceiling]. the label adds that anyone with depressive symptoms should be screened for bipolar disorder before starting, that duloxetine is not approved for bipolar depression, and that prescriptions should be for the smallest quantity of capsules consistent with good management, to reduce the risk of overdose [label 5.1]. one more line matters for duloxetine in particular: it is approved for anxiety in children from age 7 and fibromyalgia from 13, so the boxed warning applies to children taking it for those conditions too, and the label says pediatric patients on it for any reason should be monitored [label 8.4].

the liver warning, with the label’s numbers

warning 5.2 is the one that separates duloxetine from the ssris. there have been reports of hepatic failure, sometimes fatal, in people taking it, presenting as hepatitis with abdominal pain, an enlarged liver and liver enzymes more than twenty times the upper limit of normal, with or without jaundice; cholestatic jaundice with only small enzyme rises has also been reported [label 5.2]. the label says duloxetine should be stopped in anyone who develops jaundice or other evidence of significant liver dysfunction and not restarted unless another cause is found.

the numbers: in the placebo-controlled adult trials, the liver enzyme alt rose to more than three times the upper limit of normal in 1.25% of people on duloxetine (144 of 11,496) against 0.45% on placebo (39 of 8,716), with a dose-response relationship in the fixed-dose trials; across the whole development program, 0.3% (92 of 34,756) stopped because of a transaminase rise, and the median time to detecting it was about two months [label 5.2]. because duloxetine and alcohol may interact to cause liver injury, the label says it "should not be prescribed to patients with substantial alcohol use or evidence of chronic liver disease", and separately that use should be avoided in chronic liver disease or cirrhosis, where exposure was about fivefold higher, and in severe kidney impairment [label 5.2, 2.7, 8.9, 8.10]. three people in the trials database had liver injury with obstruction, each with substantial alcohol use at the time [label 7.15]. the patient counselling section names the signs to report: itching, right upper belly pain, dark urine, yellow skin or eyes [label 17].

the discontinuation syndrome, in the label’s words

warning 5.7 says discontinuation symptoms "have been systematically evaluated" for duloxetine, and lists the ones that occurred at 1% or more and significantly more often than after stopping placebo, after abrupt or tapered stopping: dizziness, headache, nausea, diarrhea, pins and needles, irritability, vomiting, insomnia, anxiety, sweating and fatigue [label 5.7, 2.8]. it adds the class list reported with other ssris and snris, particularly after abrupt stopping: low mood, agitation, electric-shock sensations, confusion, lethargy, emotional lability, hypomania, tinnitus and seizures, "generally self-limiting" but with some reported as severe [label 5.7]. seizures upon discontinuation, and tinnitus upon discontinuation, are among the post-marketing reports [label 6.2].

the instruction is the part worth quoting. "a gradual reduction in the dose rather than abrupt cessation is recommended whenever possible. if intolerable symptoms occur following a decrease in the dose or upon discontinuation of treatment, then resuming the previously prescribed dose may be considered. subsequently, the healthcare provider may continue decreasing the dose but at a more gradual rate" [label 5.7]. in children, the most common symptoms after stopping were headache, dizziness, insomnia and abdominal pain [label 6.1]. what the label does not give is a schedule or a rate; two regulators and a systematic review that do address those are in the research section below [nice-ng215; rcpsych-2019; fava-2018].

the other warnings, in plain words

the warnings section runs from 5.1 to 5.16. beyond the boxed warning, the liver and the discontinuation syndrome:

• low blood pressure, fainting and falls [5.3]: orthostatic hypotension and fainting tend to occur in the first week and after dose increases; across all placebo-controlled trials people on duloxetine fell more often than people on placebo, the risk rose with the size of the blood-pressure drop, with age, and at doses above 60 mg, and falls with fractures and hospitalisation have been reported. the label asks prescribers to consider lowering the dose or stopping if these occur.
• serotonin syndrome [5.4]: a potentially life-threatening excess of serotonin signalling, with agitation or confusion, racing heart, sweating, fever, tremor, muscle jerking and overactive reflexes. the risk rises with other serotonergic drugs (the label names triptans, tricyclics, fentanyl, lithium, tramadol, meperidine, methadone, tryptophan, buspirone, amphetamines and st john’s wort) and the combination with an maoi, including the antibiotic linezolid and intravenous methylene blue, is contraindicated [4]. it can also happen on duloxetine alone. a 2003 study of 2,222 overdoses defined the seven signs clinicians use [dunkley-2003]; a 2018 records study found the triptan combination the label names produced two definite cases in 30,928 person-years [orlova-2018].
• bleeding [5.5]: drugs that block serotonin reuptake impair platelets. the label reports gastrointestinal bleeding in case-control and cohort studies, bruising, nosebleeds and life-threatening haemorrhage across the class, and a post-marketing study showing more postpartum haemorrhage in mothers on duloxetine: an adjusted relative risk of 1.53 among 955 women exposed in the last month of pregnancy against more than 4 million unexposed [5.5; 8.1]. aspirin, nsaids and anticoagulants add to the risk; warfarin levels were not changed in a study, but the label asks for careful monitoring when duloxetine starts or stops [7.4].
• severe skin reactions [5.6]: erythema multiforme and stevens-johnson syndrome, at a reporting rate the label says exceeds the background of 1 to 2 cases per million person-years. stop at the first blister, peeling rash or mouth sore if no other cause is found.
• mania [5.8]: 0.1% of people on duloxetine in the depression trials (4 of 3,779) against 0.04% on placebo (1 of 2,536); none in the pain or anxiety trials.
• glaucoma [5.9], seizures [5.10]: pupil dilation can trigger an angle-closure attack in a susceptible eye; seizures occurred in 0.02% (3 of 12,722) against 0.01% (1 of 9,513), with people with seizure disorders excluded from the trials.
• blood pressure [5.11]: mean rises of 0.5 mmhg systolic and 0.8 diastolic against small falls on placebo, no difference in sustained high readings, but at about 3.3 times the maximum dose in a pharmacology study blood pressure rose 4.7 to 7 mmhg. the label asks for blood pressure to be measured before starting and periodically during treatment. hypertensive crisis is among the post-marketing reports [6.2].
• interactions [5.12, 7.1, 7.2, 7.9]: duloxetine is cleared by two liver enzymes, cyp1a2 and cyp2d6. potent cyp1a2 inhibitors should be avoided: with fluvoxamine, exposure rose about sixfold; ciprofloxacin, enoxacin and cimetidine are named too. potent cyp2d6 inhibitors such as paroxetine and fluoxetine raise duloxetine levels by about 60%. duloxetine itself is a moderate cyp2d6 inhibitor and tripled the exposure of the tricyclic desipramine, so tricyclics, phenothiazines and certain heart-rhythm drugs need caution, and thioridazine "should not be co-administered". smokers have about a third less duloxetine in the blood, with no dose change recommended [8.7].
• low sodium [5.13]: hyponatremia, with cases below 110 mmol/l, usually from inappropriate antidiuretic hormone secretion, reversible on stopping; older people and people on diuretics are at greater risk. signs are headache, poor concentration, confusion, weakness and unsteadiness; severe cases have included seizures, coma and death.
• diabetes [5.14]: in the three diabetic nerve-pain trials, where people had had diabetes for about 12 years, fasting glucose rose a small amount over 12 weeks and, in the extensions of up to 52 weeks, rose 12 mg/dl on duloxetine while falling 11.5 mg/dl in the routine-care group; hba1c rose 0.5% against 0.2%. the medication guide puts it plainly: duloxetine "makes it harder for some people with diabetes to control their blood sugar".
• urinary retention [5.15], sexual dysfunction [5.16]: hesitancy and retention, some needing catheterisation, have been reported. sexual dysfunction has its own warning: delayed or failed ejaculation, low desire and erectile dysfunction in men, low desire and delayed or absent orgasm in women, with the instruction to ask before starting and during treatment because people do not volunteer it. the numbers are in the side-effects section below.
• pregnancy and breastfeeding [8.1, 8.2]: besides the postpartum haemorrhage finding, a cohort of 2,532 women exposed in the first trimester showed no clinically meaningful increase in major birth defects. babies exposed late in pregnancy have needed prolonged hospital care, breathing support and tube feeding, with symptoms the label says are consistent with either a toxic effect or a withdrawal syndrome. women who stop antidepressants in pregnancy relapsed more often in a study of 201 women the label cites. in breast milk the infant dose was under 1% of the mother’s, with reports of sedation, poor feeding and poor weight gain.

the five approved uses, their trials, and the trials that failed

depression rests on four fixed-dose trials of 8 or 9 weeks in adult outpatients, with 86 to 139 people per arm. duloxetine beat placebo on the 17-item hamilton scale by 4.9 points in one trial and by 2.2 to 3.6 in the other three [label 14.2, table 8]. the label states twice that "there is no evidence that doses greater than 60 mg/day confer additional benefits" [label 2.2, 14.2]. a maintenance trial took 533 people through 12 weeks of open-label duloxetine, randomised the 278 who had responded to keep taking it (136) or switch to placebo (142) for 6 months, and found a longer time to relapse on the drug; as on every such page, that design cannot separate relapse from withdrawal [label 14.2]. in children and adolescents, two 10-week trials in 800 patients with depression failed: neither duloxetine nor an approved comparator beat placebo, so there is no pediatric depression indication [label 8.4].

generalized anxiety disorder rests on three adult trials of 9 to 10 weeks (about 160 to 170 per arm; 2.2 to 4.4 points on the hamilton anxiety scale), a 26-week maintenance trial in 429 responders, one 10-week trial in 291 people aged 65 and over (4.2 points), and one 10-week trial in 272 children aged 7 to 17 (2.7 points on the pediatric anxiety rating scale) [label 14.3, table 9]. again, 120 mg was no better than 60.

the pain uses. diabetic peripheral neuropathic pain: two 12-week trials of 457 and 334 people (791 in all, 75% completing), 60 mg once or twice daily beating placebo on a 0-to-10 pain scale and on the share with at least 50% pain relief; 20 mg was also tested [label 14.4]. fibromyalgia: one 3-month trial of 354 women and one 6-month trial of 520 people (95% women), 874 in all with 62% completing, about a quarter with depression as well; 60 and 120 mg beat placebo, "neither trial demonstrated a benefit of 120 mg compared to 60 mg", and in a separate 60-week trial people who had not responded to 60 mg by 8 weeks were no more likely to respond if blindly moved to 120 [label 14.5]. the adolescent fibromyalgia trial of 184 reached a p-value of 0.052 on its pre-specified primary analysis and 0.011 to 0.020 on sensitivity analyses, which the label reports as "improvement over placebo" [label 14.5]. chronic musculoskeletal pain: in low back pain, two 12- to 13-week trials of 236 and 401 people were positive and one of 404 people, testing 20, 60 and 120 mg, was negative on every dose; in knee osteoarthritis, one 13-week trial of 256 was positive and one of 231 was negative [label 14.6]. the label presents the failed trials by name, which is to its credit; the indication rests on three positive trials out of five.

the arithmetic that matters: the efficacy trials ran 8 to 13 weeks for every indication except the two fibromyalgia and two maintenance trials, which reached 6 months; about 4.1 million americans a year take it; and the label itself says 60 mg is the ceiling of demonstrated benefit for every use.

side effects, with the label’s numbers

the label pools 8,100 adults on duloxetine against 5,655 on placebo across all five approved conditions, and lists the reactions at 5% or more and more common than placebo: nausea 23% against 8%, headache 14% against 12%, dry mouth 13% against 5%, somnolence 10% against 3%, fatigue 9% against 5%, insomnia 9% against 5%, constipation 9% against 4%, dizziness 9% against 5%, diarrhea 9% against 6%, decreased appetite 7% against 2%, sweating 6% against 1%, abdominal pain 5% against 4% [label 6.1, table 2]. nausea, constipation, dizziness, decreased appetite and sweating were dose-related in the fixed-dose trials, and dry mouth was significantly more common at 120 mg than at 60 [label 6.1].

dropouts for side effects, by condition, are the label’s own measure of tolerability: 8.4% against 4.6% in depression, 13.7% against 5% in generalized anxiety, 12.9% against 5.1% in diabetic nerve pain, 17.5% against 10.1% in fibromyalgia, 15.7% against 7.3% in osteoarthritis pain and 16.5% against 6.3% in low back pain, with nausea the leading drug-related reason in every one [label 6.1].

sexual side effects, in the depression and anxiety trials (4,797 against 3,303): erectile dysfunction 4% against 1%, decreased libido 3% against 1%, delayed ejaculation 2% against 1%, abnormal orgasm 2% against under 1% [label 6.1, table 3]. when the label’s four depression trials asked with a questionnaire instead of waiting for complaints, men on duloxetine had significantly worse total scores and significantly more difficulty reaching orgasm than men on placebo, and women did not differ from placebo [label 6.1, table 5]. the research section has what the questionnaire studies across antidepressants found.

in children (476 against 362 in the anxiety and depression trials): nausea 18% against 8%, headache 18% against 13%, abdominal pain 13% against 10%, clinically significant weight loss 14% against 6%, somnolence 11% against 6%, decreased appetite 10% against 5%, vomiting 9% against 4%, dizziness 8% against 4% [label 6.1, table 6]. children lost a mean 0.1 kg over 10 weeks while children on placebo gained 0.9 kg, 16% against 6% lost 3.5% or more of their weight, and over studies up to 9 months their height percentile fell a mean 1%; the label asks for regular weight and height monitoring [label 6.1, 8.4]. in the adolescent fibromyalgia trial: nausea 25% against 15%, decreased appetite 15% against 3%, vomiting 15% against 5%, weight loss 15% against 5% [label 6.1, table 7].

post-marketing reports include aggression and anger "particularly early in treatment or after treatment discontinuation", hallucinations, hypertensive crisis, galactorrhea, extrapyramidal disorder, restless legs, and seizures and tinnitus on stopping [label 6.2]. in overdose, deaths have been reported with duloxetine alone, including one at 1,000 mg [label 10.1]. the label says duloxetine has not been systematically studied in humans for abuse potential and showed no dependence-producing potential in rats [label 9.2, 9.3].

what the research says, in both directions

in favour. in the network meta-analysis of 522 trials covering 116,477 adults with depression, duloxetine beat placebo for response like all 21 antidepressants studied [cipriani-2018]. for pain, the evidence is stronger than for most antidepressants: a 2023 cochrane network meta-analysis of 176 trials in 28,664 adults with chronic pain ranked 60 mg duloxetine highest, roughly doubling the odds of substantial pain relief against placebo (odds ratio 1.91) with moderate-certainty evidence, and found higher doses no better for most outcomes [birkinshaw-2023]. an earlier cochrane review of 18 trials in 6,407 people found 60 mg achieved 50% or greater pain reduction more often than placebo in diabetic neuropathy (risk ratio 1.73, one extra responder per 5 treated) and fibromyalgia (1.57, one per 8), and that 20 mg did not work [lunn-2014]. for generalized anxiety, a network meta-analysis of 89 trials noted that antidepressants such as duloxetine, venlafaxine and escitalopram rest on a larger evidence base than buspirone [slee-2019].

against, or smaller than it sounds. in the same depression network meta-analysis, duloxetine was among the seven drugs with the highest dropout rates in head-to-head comparisons [cipriani-2018]. a cochrane review of 16 trials in 5,735 people found it no more effective for depression than other antidepressants and harder to stay on than escitalopram (odds ratio 1.62 for stopping) or venlafaxine (1.56) [cipriani-2012]. in children and adolescents, the network meta-analysis of 34 trials found duloxetine, with venlafaxine and imipramine, caused more dropouts for side effects than placebo, with very low evidence quality throughout [cipriani-2016]. the pain reviews carry the same tolerability cost: about one in six stopped because of side effects [lunn-2014]. and in a european indication the fda never granted, stress urinary incontinence, a meta-analysis of the manufacturer’s own clinical study reports in 1,913 women found one in eight much improved, one in seven stopping for adverse events, and one in seven with activation-type events the authors linked to potential suicidality risk [maund-2017].

the warnings, measured. sexual dysfunction: when it is asked about with a questionnaire, a meta-analysis found treatment-emergent sexual dysfunction in 25.8% to 80.3% of people across antidepressants, with duloxetine eighth of ten drugs that were significantly worse than placebo, below sertraline, venlafaxine, citalopram, paroxetine and fluoxetine and above escitalopram [serretti-2009]; in 2019 the european regulator ordered every maker of duloxetine and nine other ssris and snris to warn that sexual dysfunction has in some cases continued after stopping, a warning the us label does not carry [prac-2019]. discontinuation: a systematic review of 61 reports found withdrawal symptoms after stopping any snri, most often venlafaxine, typically starting within days and lasting weeks, and appearing even with gradual tapering [fava-2018]; how often withdrawal happens across antidepressants is disputed between about 15% and 56% depending on method, and that it happens is not [davies-read-2019]. the uk’s nice guideline says withdrawal can take months, that a medicine should not be stopped abruptly except in exceptional circumstances, and that the taper should be a slow, stepwise reduction proportionate to the current dose [nice-ng215]; the royal college of psychiatrists in 2019 asked for tapering over potentially several months at a rate the patient can tolerate, while stating that antidepressant use is not dependence in the addiction sense [rcpsych-2019].

questions people ask

what does the cymbalta (duloxetine) black box warning mean?

it is the antidepressant class warning. in pooled placebo-controlled trials of antidepressants, suicidal thoughts and behaviors increased in people under 25: 14 extra cases per 1,000 treated under 18 and 5 extra at 18 to 24, against 1 fewer at 25 to 64 and 6 fewer at 65 and over. no suicides occurred in the pediatric duloxetine trials. the label says to watch everyone starting the drug for worsening, especially in the first months and after dose changes, and to screen for bipolar disorder first. because duloxetine is approved for anxiety from age 7 and fibromyalgia from 13, the warning covers children taking it for pain as well as for mood. it is not an instruction to stop; it describes who should be watched, and for what.

does duloxetine damage the liver?

the label carries a liver warning most antidepressants do not: hepatic failure, sometimes fatal, has been reported, usually as hepatitis with abdominal pain and liver enzymes more than twenty times normal. the numbers are small but not zero: alt rose to more than three times normal in 1.25% of people on duloxetine against 0.45% on placebo, rising with dose, and 0.3% of the 34,756 people in the development program stopped for a liver enzyme rise, typically detected around two months in. the label says duloxetine should not be prescribed to people with substantial alcohol use or chronic liver disease, and should be stopped in anyone who develops jaundice. signs to report: itching, right upper belly pain, dark urine, yellow skin or eyes.

what happens if you stop duloxetine suddenly?

the label lists what occurred significantly more often than after stopping placebo: dizziness, headache, nausea, diarrhea, pins and needles, irritability, vomiting, insomnia, anxiety, sweating and fatigue; across the class it adds electric-shock sensations, confusion, low mood and, rarely, seizures. it recommends gradual reduction rather than abrupt cessation whenever possible, and says that if intolerable symptoms follow a dose decrease, resuming the previous dose and then tapering more gradually may be considered. a systematic review found snri withdrawal symptoms typically start within days, last weeks, and can appear even with tapering. the uk’s nice guideline says to taper in steps proportionate to the current dose, so the cuts get smaller near the end. the schedule is a prescriber conversation.

is duloxetine approved for pain, and does it work?

yes, for three pain conditions: diabetic peripheral neuropathic pain, fibromyalgia (adults and adolescents from 13) and chronic musculoskeletal pain, all at 60 mg a day. the label’s trials: two 12-week trials in 791 people with diabetic nerve pain, two trials of 3 and 6 months in 874 people with fibromyalgia, and five trials in back and knee pain of which three were positive and two were negative, which the label reports by name. a 2023 cochrane network meta-analysis of 176 trials in 28,664 adults with chronic pain ranked 60 mg duloxetine highest of the antidepressants, roughly doubling the odds of substantial pain relief, with moderate-certainty evidence; an earlier cochrane review put it at one extra responder per 5 people treated in diabetic neuropathy and per 8 in fibromyalgia, with about one in six stopping for side effects. in every trial, 120 mg was no better than 60.

what are the most common duloxetine side effects?

across 8,100 adults in the label’s trials: nausea 23% vs 8% on placebo, dry mouth 13% vs 5%, somnolence 10% vs 3%, constipation 9% vs 4%, dizziness 9% vs 5%, fatigue 9% vs 5%, insomnia 9% vs 5%, decreased appetite 7% vs 2%, sweating 6% vs 1%. dropouts for side effects ran from 8.4% in depression to 17.5% in fibromyalgia against 4.6% to 10.1% on placebo, with nausea the leading reason. sexual side effects: erectile dysfunction 4% vs 1%, decreased libido 3% vs 1%, and on a questionnaire men, though not women, scored significantly worse than placebo. in children: nausea 18% vs 8% and clinically significant weight loss in 14% vs 6%.

does duloxetine raise blood pressure?

a little, on average. the label reports mean rises of 0.5 mmhg systolic and 0.8 diastolic against small falls on placebo, with no difference in the share with sustained high readings, and larger rises at about three times the maximum dose in a pharmacology study. it asks for blood pressure to be measured before starting and periodically during treatment, and hypertensive crisis is among the post-marketing reports. the same drug can also drop blood pressure on standing, especially in the first week and after dose increases, and the label links that to falls and fainting, rising with age and above 60 mg.

is duloxetine addictive?

the label says duloxetine has not been systematically studied in humans for abuse potential, showed no drug-seeking behaviour in the trials, and did not produce dependence in rat studies; it is not a controlled substance. the discontinuation syndrome above is a separate matter: the royal college of psychiatrists states that antidepressant withdrawal is real and can be prolonged, and that antidepressant use is nonetheless not dependence in the addiction sense. a plan for reducing or stopping belongs with a prescriber.

sources for this section

the library rows link to our one-page summary of each study, with the original paper, its funding and our evidence grade on that page.

1. duloxetine delayed-release capsules prescribing information, fda label on dailymed (set id 00628e5e-4c5b-2573-e063-6294a90a0e3b, version 4, effective august 20, 2026; breckenridge, repackaged by golden state medical supply): boxed warning; sections 1, 2.2 to 2.8, 4, 5.1 to 5.16 with table 1, 6.1 tables 2 to 7, 6.2, 7.1 to 7.15, 8.1, 8.2, 8.4, 8.5, 8.7, 8.9, 8.10, 9.2, 9.3, 10.1, 14.2 to 14.6 with tables 8 and 9, 17, medication guide, recent major changes
2. drugs@fda, nda 021427 (cymbalta), approved august 3, 2004
3. ahrq medical expenditure panel survey, prescribed medicines public-use files 2014 (hc-168a) and 2024 (hc-254a): people with at least one duloxetine purchase, weighted; computed by veisund, method on the usage section of this page
4. the antidepressant suicidality warning’s age ceiling, from the labels themselves, as a library row (library row)
5. cipriani a, et al. comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis. lancet 2018 (library row)
6. cipriani a, et al. duloxetine versus other anti-depressive agents for depression. cochrane database of systematic reviews 2012 (library row)
7. cipriani a, et al. comparative efficacy and tolerability of antidepressants for major depressive disorder in children and adolescents: a network meta-analysis. lancet 2016 (library row)
8. birkinshaw h, et al. antidepressants for pain management in adults with chronic pain: a network meta-analysis. cochrane database of systematic reviews 2023 (library row)
9. lunn mp, hughes ra, wiffen pj. duloxetine for treating painful neuropathy, chronic pain or fibromyalgia. cochrane database of systematic reviews 2014 (library row)
10. maund e, guski ls, gøtzsche pc. considering benefits and harms of duloxetine for treatment of stress urinary incontinence: a meta-analysis of clinical study reports. cmaj 2017 (a european indication, not a us one) (library row)
11. slee a, et al. pharmacological treatments for generalised anxiety disorder: a systematic review and network meta-analysis. lancet 2019 (library row written up for buspirone) (library row)
12. serretti a, chiesa a. treatment-emergent sexual dysfunction related to antidepressants: a meta-analysis. journal of clinical psychopharmacology 2009 (library row)
13. pharmacovigilance risk assessment committee, european medicines agency. prac recommendations on signals adopted at the 13-16 may 2019 meeting, section 1.3: snris and ssris, persistent sexual dysfunction after drug withdrawal (ema/prac/265212/2019) (library row)
14. fava ga, et al. withdrawal symptoms after serotonin-noradrenaline reuptake inhibitor discontinuation: systematic review. psychotherapy and psychosomatics 2018 (library row written up for venlafaxine; covers every snri) (library row)
15. davies j, read j. a systematic review into the incidence, severity and duration of antidepressant withdrawal effects. addictive behaviors 2019 (with the henssler estimate the row sets beside it) (library row)
16. national institute for health and care excellence. medicines associated with dependence or withdrawal symptoms: safe prescribing and withdrawal management for adults (ng215), 2022 (library row)
17. royal college of psychiatrists. position statement on antidepressants and depression (ps04/19), 2019 (library row)
18. dunkley ej, et al. the hunter serotonin toxicity criteria: simple and accurate diagnostic decision rules for serotonin toxicity. qjm 2003 (library row)
19. orlova y, rizzoli p, loder e. association of coprescription of triptan antimigraine drugs and ssri or snri antidepressants with serotonin syndrome. jama neurology 2018 (library row)
20. the duloxetine legal and safety record, a separate page on this site: court and regulator records, kept apart from the label and trial facts here

Most-reported reactions

• Nausea17,925
• Fatigue15,979
• Drug ineffective14,131
• Headache13,688
• Dizziness13,379
• Pain12,377
• Diarrhoea10,017
• Off label use9,952
• Insomnia9,797
• Anxiety9,678