How effective are antidepressants for musculoskeletal pain?

By Nicole MacKee
Antidepressants are probably ineffective for back pain and uncertainties remain about their role in sciatica and osteoarthritis (OA), Australian researchers report in The BMJ.

In a systematic review and meta-analysis, 33 trials with more than 5300 participants were evaluated. For pain and disability to be considered clinically meaningful, the researchers required a 10-point difference on a scale of 0 (no pain or disability) to 100 (worst pain or disability). They found moderate-certainty evidence that serotonin–noradrenaline reuptake inhibitors (SNRIs) reduced back pain by 5.3 points at three to 13 weeks but this was ‘unlikely to be clinically important’.

A stronger effect was found for SNRIs in OA, with an average pain scale difference of 9.7 points at three to 13 weeks, and the researchers reported that a ‘clinically important effect could not be excluded’. All trials evaluated were in knee OA.

The researchers found low to very-low certainty evidence showing that tricyclic antidepressants (TCAs) were not effective in back pain and related disability, but wide confidence intervals meant benefit could not be excluded. Also, they found that SNRIs and TCAs might reduce sciatic pain, but only low and very-low certainty evidence was available on this indication. SNRIs were found to be effective at two weeks, but not at three to 13 weeks, while TCAs were only effective at three to 13 weeks.

Professor Rachelle Buchbinder, Rheumatologist and Director of the Monash Department of Clinical Epidemiology, Melbourne, said the analysis clarified the current status of the evidence for the use of antidepressants in chronic back pain and OA.

She noted that the 2018 Royal Australian College of General Practitioner’s Guideline for the Management of Knee and Hip Osteoarthritis (Second Edition) included a ‘conditional recommendation’ for the use of the SNRI duloxetine for OA.

‘The guideline ... considered only duloxetine and made a conditional recommendation based upon moderate certainty evidence that it could be considered for knee OA but it would need to be considered experimental as it is currently not registered for OA – or back pain – by the TGA,’ she said.

‘There were no trials in hip OA but the guideline made a conditional recommendation as well for hip OA under similar circumstances to knee OA. The guideline did not consider other antidepressants.’

Professor Buchbinder said conditional recommendations were made when the evidence for or against use of a treatment was uncertain.

She said developing ‘living guidelines’ for questions that were uncertain would help to ensure that clinical practice changed in line with the latest evidence as it became available.

In the meantime, Professor Buchbinder recommended a shared decision-making approach – taking into account the evidence and patient circumstances and preferences – to prescribing antidepressants for any of these conditions.

‘All patients should be provided with a clear summary of the evidence for its efficacy and safety, and its certainty,’ she said. ‘Other considerations, such as concomitant depression, comorbidities and severity, may also be factors in any decision. If a decision is made to trial antidepressants, there should be a clear plan for reviewing the outcome and stopping therapy if it is ineffective.’
BMJ 2021;372:m4825 http://dx.doi.org/10.1136/bmj.m4825.