Some antidepressant medications work for some pain conditions, but the majority are either ineffective or the evidence is incomplete, according to a review of the most recent evidence published in The BMJ.

Researchers called for a more nuanced approach when prescribing antidepressants for pain. The use of antidepressants doubled in OECD countries from 2000 to 2015 and their off-label (unapproved) use to treat common pain conditions such as fibromyalgia, persistent headaches, and osteoarthritis is thought to be part of this increase.

To explore this further, a team of researchers led by Giovanni Ferreira at the University of Sydney carried out an overview of the effectiveness, safety, and tolerability of antidepressants for pain according to condition.

They searched databases for systematic reviews comparing any antidepressant with a placebo for any pain condition in adults and found 26 eligible evidence reviews published between 2012 and 2022 involving 156 separate trials and over 25,000 participants.

These reviews reported on the effectiveness of eight classes of antidepressants covering 22 pain conditions (42 distinct antidepressants versus placebo comparisons). Almost half (45%) of the trials in these reviews had ties to the industry.

Using data from each review, the researchers estimated relative risks of pain or average differences in pain between groups on a 0-100 point scale, taking account of dose, treatment duration, and the number of trials and participants.

They also assessed safety and tolerability (withdrawals due to adverse events), the certainty of the evidence, and the risk of bias. Findings were then classified from each comparison as effective, not effective, or inconclusive.

No review provided high certainty evidence on the effectiveness of antidepressants for pain for any condition. Nine reviews provided evidence that some antidepressants were effective compared with placebo for nine conditions in 11 distinct comparisons.

For example, moderate certainty evidence suggested that serotonin-norepinephrine reuptake inhibitors (SNRIs) were effective for back pain (average 5.3 points lower on the pain scale than placebo), postoperative pain, fibromyalgia, and neuropathic pain.

Low certainty evidence suggested that SNRIs were effective for pain linked to breast cancer treatment, depression, knee osteoarthritis, and pain related to other underlying conditions.

Low certainty evidence also suggested that selective serotonin reuptake inhibitors (SSRIs) were effective for people with depression and pain related to other conditions; and those tricyclic antidepressants (TCAs) were effective for irritable bowel syndrome, neuropathic pain, and chronic tension-type headache.

For the other 31 comparisons, antidepressants were either not effective (five comparisons) or the evidence was inconclusive (26 comparisons). Most safety and tolerability data were imprecise, indicating that the safety of antidepressants for several conditions is still uncertain.

This was a well-designed review based on a thorough literature search and the researchers took steps to minimise the impact of issues such as differences in study design and quality, imprecision, and publication bias.

But they acknowledge that most comparisons had a limited number of trials, and results may not apply to antidepressants prescribed for symptoms linked to pain conditions, such as fatigue or sleep disturbance.

Caution is also needed in interpreting these findings because 45% of the trials forming the evidence for this review had ties to industry, they add.

In conclusion, they said: "Some antidepressants were efficacious for some pain conditions; however, efficacy appears to depend on the condition and class of antidepressant. The findings suggest that a more nuanced approach is needed when prescribing antidepressants for pain."

These findings suggest that for most adults living with chronic pain, antidepressant treatment will be disappointing, say researchers in a linked editorial.

They acknowledge that clinicians continue to prescribe medicines for which the evidence is poor because they see that some people respond to them, albeit modestly, but say other less potentially harmful options, such as exercise and support with mobility and social isolation, can help people to live well with pain.

For people with pain, compassionate and consistent relationships with clinicians remain the foundations of successful care, they write.

Studies must also involve people living with pain to ensure, among many other things, that pain research is meaningful to those living with pain and helps them and their clinicians make better-shared decisions about treatments, they conclude.