Reviews/Chronic Pain Pharmacotherapy/SNRIs reduce pain scores in fibromyalgia
Ant · CPP-ANT-03 · v2
Several large, carefully blinded trials tested two of these drugs, duloxetine and milnacipran, against a dummy pill in people with fibromyalgia, and most found that the drug lowered pain more than placebo [1][7][10][17][22]. Two combined re-analyses of the trial data pointed the same way [9][12]. The effect, though, was fairly small, and it was not seen everywhere: three well-run trials did not beat placebo on their main pain measure [6][14][18], and a lower dose failed where higher doses worked [14]. Almost all of the usable evidence comes from just those two drugs, so extending the finding to the whole class is a stretch. The panel treated the direction of the finding as well supported while holding back from calling it settled, because the effect is modest, uneven, and narrow in scope.
Multiple double-blind placebo-controlled trials show a drug-versus-placebo pain reduction: duloxetine 120 mg cut Brief Pain Inventory average pain -2.31 vs -1.39 (P<0.001) [1], duloxetine reduced BPI average pain severity (P=0.008) [22], milnacipran beat placebo on composite pain responder criteria (P<0.001) [7] and (P=0.0003) [10], and milnacipran 200 mg gave an adjusted pain difference of -18.4 mm (P=0.02) [17]. Two pooled analyses of primary trial data reinforced the signal [9][12]. Pain is the directly measured outcome, so there is no surrogate step. Against this, three adequately sized trials missed their primary pain endpoint [6][14][18], and duloxetine monotherapy did not beat placebo on daily pain in a crossover trial [23]. The contested point: The division is whether a modest, uneven, two-drug signal supports a class-level claim. The same trials that are significant [1][7][10][17][22] sit alongside three that failed their primary pain endpoint [6][14][18] and a dose that did not work [14], and the only venlafaxine evidence is a single uncontrolled study of 15 patients [32].
All 39 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Chronic Pain Pharmacotherapy, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
Three adequately powered FM RCTs failed their primary pain endpoint — Chappell [6] (BPI P=0.053), Arnold 30 mg [14] (-2.04 vs -1.70, P=0.202), and Pickering [18] ('Milnacipran did not display a significant analgesic effect after 1-month treatment') — so the effect is not uniform across the included primary trials.
The effect is dose-dependent and small: duloxetine 30 mg fails [14] while 60/120 mg succeed, and in the crossover trial duloxetine monotherapy did not significantly beat placebo on daily pain [23] (only the combination did). A class claim that 'SNRIs reduce pain scores' obscures that only certain drugs at certain doses do so.
Where it remains uncertain. Three adequately powered trials missed their primary pain endpoint [6][14][18], the effect is dose-dependent and small [14], and class evidence rests almost entirely on two drugs with venlafaxine resting on one uncontrolled N=15 study [32].
Aleatory
Even in positive trials only a minority reach meaningful pain relief and three trials missed the primary endpoint [6][14][18], so response varies across patients.
Epistemic
Venlafaxine rests on one uncontrolled N=15 study [32] and no eligible primary desvenlafaxine trial is in the corpus, so class-wide behaviour is unmeasured.
The load-bearing papers from a corpus of 39, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Large FM RCT showing duloxetine 120 mg reduced BPI pain significantly more than placebo.
“those patients treated with duloxetine 120 mg/day improved significantly more on the co-primary outcome measures at 3 months (change in BPI score [-2.31 vs -1.39, P<0.001] and PGI-I [2.89 vs 3.39, P=0.004])”
Panel auditInternal validity: high · scope of inference: ACR fibromyalgia patients with or without MDD, duloxetine 60-120 mg/day over 6 months.Key limitations: 60 mg benefit less consistent at 6 months · industry-relevant outcome instruments
Large milnacipran FM RCT with significant pain-score improvements vs placebo.
“Milnacipran-treated patients also demonstrated significantly greater improvements from baseline on multiple secondary outcomes, including 24-hour and weekly recall pain score, PGIC score, SF-36 PCS and mental component summary scores, average pain severity score on the Brief Pain Inventory, Fibromyalgia Impact Questionnaire total score (all P < 0.001 versus placebo)”
Panel auditInternal validity: high · scope of inference: Fibromyalgia patients on milnacipran 100 mg/day.Key limitations: primary endpoint was composite responder, pain a secondary · high nausea rate
Pooled analysis of 4 duloxetine FM trials confirming significant pain reduction.
“After 12 weeks of treatment, pain was significantly reduced in patients treated with duloxetine (P < .001) compared with placebo.”
Panel auditInternal validity: medium · scope of inference: ACR fibromyalgia patients on duloxetine 60-120 mg/day over 12 weeks.Key limitations: LOCF imputation · pooled, single sponsor's trials
The abstract reports MCID thresholds anchored to PGI-I, not a drug-versus-placebo pain comparison.
Panel auditInternal validity: medium · scope of inference: Measurement properties of BPI pain in ACR fibromyalgia patients.Key limitations: no efficacy contrast reported · methodological, not comparative
4 of 39 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
Converging significant primary RCTs establish a real but small and heterogeneous pain-reduction effect, so the claim is well supported directionally but not Established.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
No controlled evidence for SNRIs beyond duloxetine and milnacipran. The class claim cannot be tested for venlafaxine or desvenlafaxine from this corpus; venlafaxine rests on one uncontrolled N=15 study [32]. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: Adequately powered double-blind placebo-controlled RCTs of venlafaxine and desvenlafaxine in fibromyalgia with a directly measured pain primary endpoint.
Open question: Do SNRIs other than duloxetine and milnacipran reduce pain scores in fibromyalgia? Why it is still open: The only non-duloxetine, non-milnacipran evidence is an uncontrolled N=15 venlafaxine study [32] and no eligible desvenlafaxine RCT is present.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19