- Beta Blockers for POTS: What the Evidence Actually Says - 10 August 2026
- Duloxetine (Cymbalta) for Fibromyalgia: What the Evidence Actually Says - 10 August 2026
- Gabapentin for Fibromyalgia: What the Evidence Actually Says - 7 August 2026
Duloxetine, better known by the brand name Cymbalta, is one of the handful of drugs that actually has a licence for fibromyalgia behind it, and it gets handed out a lot. If you’ve got fibromyalgia you’ve almost certainly been offered it, been put on it, or read a forum thread where half the people call it a lifesaver and the other half describe coming off it as one of the worst experiences of their lives. Both of those can be true at once. The reason they can is buried in the trial data, which almost nobody quotes properly.
Most of what’s written about it sits at one of two extremes. The drug company version and the tidy medical pages tell you it reduces pain, improves function and is well tolerated, which is technically true and quietly misleading. The other extreme, usually on social media, tells you it’s a chemical straitjacket that does nothing but wreck your libido and give you brain zaps. The honest read sits in the middle, and it’s a lot more useful than either.
So this article is about what duloxetine actually does for fibromyalgia, weighed against the best evidence we’ve got rather than the marketing or the horror stories. We’ll go through how it’s meant to work, what the trials really found once you read past the abstract, how it stacks up against the other options, the side effects (including the one people are least warned about), and where a drug like this sensibly fits alongside everything else. What it isn’t is a recommendation to start it, stop it, or change your dose. That’s a conversation for you and your prescriber, and nothing here replaces it. If you’re ready to see what the data honestly says, let’s get into it.
What duloxetine is, and how it’s meant to work
Duloxetine is a serotonin and noradrenaline reuptake inhibitor, an SNRI. In plain terms, it slows down how fast the brain clears away two of its own chemical messengers, so more of them hang around in the gaps between nerve cells. Where it’s licensed depends on where you live. In the US it’s approved specifically for fibromyalgia. In the UK it’s licensed for diabetic nerve pain, depression, generalised anxiety and stress urinary incontinence, and it’s used off label for fibromyalgia. That surprises a lot of people, who assume a drug this common must be formally approved for the thing they were handed it for.
To see why raising serotonin and noradrenaline might help pain, you have to understand what fibromyalgia is actually doing. It’s not a disease of damaged tissue, and while the reasons some people develop it are still being worked out, the mechanism driving the pain is reasonably well described. The current model, laid out clearly in Daniel Clauw’s 2014 review in JAMA, is that fibromyalgia is a state of amplified pain processing in the central nervous system [1]. The volume knob on pain is turned up, and the brain’s own dampening systems are turned down. It’s the same central sensitisation picture that runs through so much of chronic pain, and it’s why we spend so much time explaining what’s actually generating the pain rather than hunting for a structural culprit that often isn’t there.
Serotonin and noradrenaline are the two main chemicals your spinal cord runs on in its descending inhibitory pathways, the top down system that’s meant to quieten pain signals before they ever reach conscious awareness. In fibromyalgia that braking system seems to work less well. The logic of an SNRI is dead simple: give the brakes more of the chemical they run on, and they might grip a bit harder. Clauw’s review names this exact family of drugs, alongside the tricyclics and the gabapentinoids, as the options with the most evidence behind them [1]. That’s the theory, and it’s a reasonable one. The interesting part is what happens when you actually test it.
What the evidence actually shows
The good news first, and it’s genuine: duloxetine is one of the better evidenced drugs in fibromyalgia. Being one of the better evidenced options in this field is a low bar, mind, because the competition isn’t exactly stiff. Most of what gets prescribed for fibromyalgia rests on thin or messy data, so against that backdrop duloxetine stands out.
Held to an absolute standard, though, the effect is real but modest, and it helps a minority of people rather than most.
It starts with the pivotal trials. The first big multicentre trial comparing duloxetine with placebo in fibromyalgia was published by Lesley Arnold and colleagues in 2004 [2]. A follow up in 2005 tightened the design and studied 354 women, split between 60mg once a day, 60mg twice a day, and placebo, over 12 weeks [3]. Both duloxetine groups improved their pain scores significantly more than placebo, with a bit over half of the people on the drug counting as responders against a third on placebo [3]. That’s a real difference. And it’s worth sitting with one detail in particular: the pain benefit held up even in the women who didn’t have major depression, which tells you the drug is doing something to the pain directly rather than just lifting mood and dragging the pain along behind it [3].
Those trials got duloxetine its US fibromyalgia licence in 2008. Since then the evidence has been pooled and pooled again, and this is where you have to read carefully, because the way a benefit gets described can make it sound far bigger than it really is.
The responder numbers, told honestly
The most useful way to talk about a pain drug isn’t the average change on a scale, because averages hide the people it does nothing for. The better question is how many people get a response worth having. The 2014 Cochrane review by Lunn and colleagues pulled the fibromyalgia trials together, 6 studies and 2,249 people, and looked at meaningful thresholds [4]. At the standard 60mg dose, for a substantial result of at least 50% pain relief, the number needed to treat was 8, with a wide confidence interval running from 4 to 21 [4]. For a more modest 30% or more of pain relief, it was 6 [4].
A number needed to treat of 8 means that for every 8 people who take duloxetine, roughly 1 more gets that big 50% improvement than would have on a dummy pill. The other 7 get a smaller benefit, no benefit, or side effects that make them stop. That’s not really a criticism of the drug, it’s just how pain drugs tend to work. It’s a lever that nudges the odds for some people, not a switch that fixes most.
The broader Cochrane review of all the SNRIs in fibromyalgia, from Welsch and colleagues, makes the same point on a bigger canvas: 18 studies and 7,903 people across duloxetine and its cousin milnacipran [5]. There, 31% of people on the drug reported at least 50% pain relief, against 21% on placebo [5]. The drug moved the needle, then, but a big chunk of the response you see in the clinic is response people would have got on a placebo anyway. The most current summary, an overview of 21 Cochrane reviews covering 87 trials and 17,631 people, published in 2025, puts a plain number on it: across the whole field of fibromyalgia drugs, only about 1 person in 10 with moderate or severe pain gets their pain cut by at least half [6]. Duloxetine, milnacipran and pregabalin are the three named with moderate to good evidence, and even for them the honest figure is that modest [6].
Does it help anything besides pain?
This is where a lot of the marketing quietly overreaches. Fibromyalgia is never just pain. It’s fatigue, poor sleep, brain fog and low mood all tangled together, so the obvious hope is that a drug touching serotonin and noradrenaline would help across the board. The Welsch review looked at exactly that, and the results are sobering. For fatigue, the effect was tiny and not clinically meaningful. For sleep, no significant difference from placebo at all. For quality of life and for mood, the effects were small enough that the reviewers judged them not substantial [5].
That matters if you were sold it as a fix for the whole picture. On the current evidence, duloxetine is best thought of as a pain drug that happens to be an antidepressant, rather than an all rounder that’ll lift your energy and sort your sleep. If fatigue and broken sleep are your worst problems, a drug aimed mainly at pain can leave the parts of the day you dread most largely untouched, which is worth knowing before you pin your hopes on it. Sleep in particular tends to need its own approach, and we’ve written separately about the tangled relationship between rest, napping and fibromyalgia. Same goes for the other threads people hope one tablet will catch, from the headaches to skin that hurts to the lightest touch to the widespread ache that settles in the back, each of which really needs its own plan rather than one drug carrying the lot.
The dose question
One genuinely practical thing the trials tell you is where the useful dose sits. In the Lunn review, the 20mg and 30mg doses were negative on essentially every measure, so a very low dose is unlikely to be doing much beyond acting as a rung to build up from [4]. Sixty milligrams a day is where the benefit actually shows up. Going higher, up to 120mg, didn’t add meaningful benefit over 60mg in the pooled data [4]. The 2023 network analysis of antidepressants in chronic pain landed on the same conclusion from a different angle, finding the standard dose about as effective as the high dose across most outcomes [7].
The takeaway is worth holding onto. If 60mg isn’t helping, pushing to 120mg is more likely to buy you extra side effects than extra relief. That’s a handy thing to have in your back pocket when a prescriber reaches for a dose increase as the automatic next move.
How long does it keep working?
This is the gap in the evidence that gets glossed over most often, and it’s a big one. Almost all the trials run for 12 weeks or so. The Lunn review found some support out to 28 weeks, but the numbers behind that were thin [4]. The 2025 overview is blunt about it: there’s no efficacy evidence beyond six months for any of the three drugs, duloxetine included [6]. The 2023 antidepressant analysis says the same, that there’s no reliable evidence for the long term efficacy of any antidepressant in chronic pain, and none on their long term safety either [7].
Fibromyalgia is, by definition, a long term problem. A drug that’s been properly tested for three months and only patchily beyond that is being asked to do a job it was never really trialled for. That doesn’t mean the benefit vanishes at six months. Plenty of people report steady help for years. It does mean that if someone tells you the long term evidence is solid, they’re telling you something the data doesn’t back up.
How duloxetine compares to the other drugs
If you’ve got fibromyalgia you’ve probably been round the medication carousel, or watched other people ride it, and we’ve covered the wider medication picture in its own guide. Amitriptyline, pregabalin, gabapentin, an SSRI, maybe low dose naltrexone, and more recently a newer bedtime formulation that got a lot of headlines. Some of these matter for the conditions that travel with fibromyalgia too, like restless legs, which is worth knowing if your nights are already broken. So where does duloxetine actually rank? The most rigorous answer comes from the 2023 Cochrane network analysis, which compared the antidepressants against each other for chronic pain rather than each one against placebo on its own [7]. Its conclusion was strikingly clean: duloxetine was the only antidepressant they were confident about, the one with moderate certainty evidence for reducing pain, with an odds ratio of 1.91 for substantial pain relief across 16 studies [7]. For every other antidepressant, the evidence was too weak to draw a firm conclusion [7].
Read that carefully, because it cuts both ways. It’s a real point in duloxetine’s favour against the other antidepressants, including the old standby amitriptyline, whose evidence is genuinely shaky for all that it gets handed out. It’s also a quiet indictment of how little we can say about most of what gets prescribed. Being the best antidepressant for pain is a meaningful title. It’s just a smaller trophy than it sounds once you see how bare the rest of the shelf is.
Against the gabapentinoids, the comparison is closer than the tribal online debates suggest. The 2025 overview lumps duloxetine, milnacipran and pregabalin together as the three drugs with moderate to good evidence, all clustering around that same modest one in ten figure for a big response [6]. Duloxetine’s theoretical edge is that it works on both pain and mood through the same mechanism, which is appealing given how tightly the two run together in chronic pain, and it tends to be weight neutral where a drug like gabapentin can pile weight on. None of that makes it a clear winner. It makes it a reasonable first choice for some people and the wrong choice for others, which is about as decisive as fibromyalgia pharmacology ever gets.
The European guidelines put all of this in its proper place. The 2017 EULAR recommendations reviewed the whole toolkit and handed out exactly one strong recommendation, and it went to exercise, not to any drug [8]. Duloxetine and the rest of the medications sit in the weaker, conditional tier: worth considering for specific problems in specific people, not the foundation the whole plan is built on [8]. That ordering isn’t an anti drug position, it’s just what the evidence supports when you line everything up honestly.
Side effects, and the part people aren’t warned about
Every drug that does something has a cost, and duloxetine’s are reasonably well mapped. The common early side effects are nausea, dry mouth, constipation, drowsiness or trouble sleeping, dizziness and a smaller appetite. Nausea in particular is common in the first couple of weeks and can be genuinely rough, though for most people it settles, and starting low and building up slowly helps. In the SNRI Cochrane review, 19% of people on the drug dropped out because of side effects, against 10% on placebo [5]. In the duloxetine trials across its various uses, around 1 in 8 people, 12.6%, stopped because of adverse effects [4]. Serious harms were rare in the trials [5].
Then there’s the one that ambushes people, because it isn’t really a side effect of taking the drug at all, it’s what happens when you stop. Duloxetine has a short half life, and stopping it, even tapering it, can set off a withdrawal reaction that’s well recognised and frankly nasty for some people: dizziness, nausea, irritability, anxiety, vivid dreams, and the strange electric shock sensations in the head that people call brain zaps. It isn’t addiction in the street sense of the word, and it isn’t the drug doing something sinister, it’s physical dependence, a different and far more mundane thing, but it’s real and it’s badly under discussed.
This matters because it shapes the whole decision. A drug you can walk away from cleanly is a lower stakes experiment than one you might have to taper off slowly and miserably over weeks or months. None of that argues against ever taking it. It argues for going in with your eyes open, knowing that stopping is often the hard part, and never coming off it abruptly on your own. If you decide duloxetine isn’t for you, that exit gets planned with your prescriber, at a pace your body can actually handle.
So who might it actually suit?
Pull the evidence together and a few reasonable patterns fall out. This is my read of the data, not a prescribing instruction, and your prescriber knows your full picture in a way a blog never can.
Duloxetine looks like a more sensible bet when pain and low mood or anxiety are travelling together, because it has a genuine shot at both through the one mechanism, and the pain benefit doesn’t hang on you being depressed in the first place [3]. It’s a weaker bet if your dominant problems are fatigue and unrefreshing sleep, since those are the outcomes it barely shifted in the trials [5]. Anxiety is common alongside fibromyalgia and hypermobility. If that’s a big part of your picture, it’s worth reading how anxiety and a sensitive nervous system feed into pain, and even how hard your nervous system is running can show up in measures like heart rate variability. Work out what you’re actually trying to treat before you start.
The practical test is refreshingly simple. Because the trial benefit shows up within weeks rather than months, you’ll have a fair sense of whether it’s doing anything reasonably quickly. If you’ve been at 60mg for six to eight weeks and honestly can’t point to a difference, it’s probably not your drug, and grinding on at a higher dose in hope is exactly the move the data argues against [4]. Give it a proper, fair go, then judge it on results rather than on the promise it came with.
Where a drug like this actually fits
This is the part the leaflet leaves out. At its best, duloxetine turns the pain volume down a notch. What it can’t do is teach your nervous system to feel safe moving again, rebuild the strength and control you’ve lost, or unpick the fear that quietly grows around movement when everything hurts. Those are the things that actually shift the course of chronic pain, and no tablet delivers them. If you want the bigger picture, our guide to living with fibromyalgia pulls the whole approach together.
Remember the mechanism. Fibromyalgia is a sensitised, over protective nervous system with its pain brakes worn down [1]. A drug that tops up the chemicals those brakes run on can help, but the brakes get genuinely stronger through use, through graded, gradual, well paced movement that teaches the system it’s safe. That’s why exercise is the single strongest recommendation in the guidelines, sitting above every drug on the list [8], not because movement is virtuous but because it’s the one thing that changes the underlying system instead of just muffling its output. None of that means punishing yourself in a gym. It can be as gentle as getting outdoors and moving a little, and it’s worth clearing up the myths around stretching before you start, because a lot of the standard advice there is flat wrong.
The useful way to think about duloxetine, then, is as a lever, not a destination. For the right person it can quieten the pain enough to make the real work possible, and getting moving again in a way that doesn’t flare you is the real work. Used like that, as a bit of breathing room while you rebuild, it earns its place. Used as the whole plan, it’ll almost always disappoint, because it was never built to be the whole plan. Careful pacing and a slow, deliberate return to activity do the heavy lifting a tablet simply can’t.
Frequently asked questions
Does duloxetine actually work for fibromyalgia?
For some people, yes, and the evidence for it is better than for most fibromyalgia drugs. The honest scale is the bit that matters. At 60mg the number needed to treat for at least 50% pain relief is around 8, and across the whole field of fibromyalgia drugs only about 1 person in 10 gets that big a response [4][6]. So it genuinely helps a meaningful minority and does little for the majority. That’s worth a try for a lot of people, as long as you go in expecting a useful nudge rather than a cure.
How long does duloxetine take to work?
The benefit in the trials showed up over weeks rather than months, with most of the pivotal studies running at around 12 weeks [3]. A fair trial is roughly six to eight weeks at 60mg. If you genuinely can’t notice a difference by then, it’s probably not going to be your answer, and the data doesn’t support pushing to a higher dose in the hope that more will help [4].
Can you just stop taking it?
No, and this is the bit people most need to hear. Duloxetine has a short half life, and stopping it suddenly can set off a withdrawal reaction that includes dizziness, nausea, anxiety, vivid dreams and the electric shock sensations known as brain zaps. It’s physical dependence rather than addiction, but it’s real and it can be grim. Any plan to come off it should be a slow taper agreed with your prescriber, never a cold stop on your own.
Duloxetine or pregabalin, which is better?
Neither is clearly better across the board. Both sit in the small group of fibromyalgia drugs with moderate to good evidence, and both land around that same modest response rate [6]. Duloxetine may suit people whose pain travels with low mood, and it tends to be weight neutral. Pregabalin may suit people whose sleep is badly broken. The right one depends on your particular mix of symptoms and what side effects you can live with, which is a conversation for you and your prescriber rather than a league table.
Is duloxetine an antidepressant or a painkiller?
It’s both, and that’s not a contradiction. It’s an antidepressant by class, but its effect on fibromyalgia pain works directly on the nervous system’s pain pathways and doesn’t depend on you being depressed, which the trials showed clearly [3]. Being prescribed an antidepressant for pain isn’t a hint that anyone thinks the pain is in your head, it’s a reflection of the chemistry those drugs happen to act on [1].
The honest bottom line
Duloxetine is a legitimate option for fibromyalgia with better evidence behind it than most of the alternatives, and for a meaningful minority it takes enough edge off the pain to matter. It’s not a cure, it barely touches fatigue and sleep, its long term evidence is thin, and coming off it can be genuinely hard. Weigh all that together and it becomes what it always was: a reasonable tool for some people, in the right situation, used alongside the things that actually change the course of chronic pain rather than instead of them.
The point of all this isn’t to talk you out of medication, or into it. It’s to hand you the real numbers, so that whatever you and your prescriber decide, you’re deciding with the honest picture in front of you instead of the sales version or the scare version. Never start, stop or change a dose on the strength of an article, this one included. Bring the questions, ask for the plan, and put the rest of your effort into the work a tablet can’t do for you.
– Adam –
References
[1] Clauw DJ. Fibromyalgia: a clinical review. JAMA. 2014;311(15):1547-1555. doi: 10.1001/jama.2014.3266
[2] Arnold LM, Lu Y, Crofford LJ, et al. A double-blind, multicenter trial comparing duloxetine with placebo in the treatment of fibromyalgia patients with or without major depressive disorder. Arthritis and Rheumatism. 2004;50(9):2974-2984. doi: 10.1002/art.20485
[3] Arnold LM, Rosen A, Pritchett YL, et al. A randomized, double-blind, placebo-controlled trial of duloxetine in the treatment of women with fibromyalgia with or without major depressive disorder. Pain. 2005;119(1-3):5-15. doi: 10.1016/j.pain.2005.06.031
[4] Lunn MP, Hughes RAC, Wiffen PJ. Duloxetine for treating painful neuropathy, chronic pain or fibromyalgia. Cochrane Database of Systematic Reviews. 2014;(1):CD007115. doi: 10.1002/14651858.CD007115.pub3
[5] Welsch P, Üçeyler N, Klose P, et al. Serotonin and noradrenaline reuptake inhibitors (SNRIs) for fibromyalgia. Cochrane Database of Systematic Reviews. 2018;(2):CD010292. doi: 10.1002/14651858.CD010292.pub2
[6] Moore A, Bidonde J, Fisher E, et al. Effectiveness of pharmacological therapies for fibromyalgia syndrome in adults: an overview of Cochrane Reviews. Rheumatology. 2025;64(5):2385-2394. doi: 10.1093/rheumatology/keae707
[7] Birkinshaw H, Friedrich CM, Cole P, et al. Antidepressants for pain management in adults with chronic pain: a network meta-analysis. Cochrane Database of Systematic Reviews. 2023;(5):CD014682. doi: 10.1002/14651858.CD014682.pub2
[8] Macfarlane GJ, Kronisch C, Dean LE, et al. EULAR revised recommendations for the management of fibromyalgia. Annals of the Rheumatic Diseases. 2017;76(2):318-328. doi: 10.1136/annrheumdis-2016-209724
