The APA pain guideline backs exercise, then declines to say which kind is better

The American Psychological Association's clinical practice guideline for chronic musculoskeletal pain has been in circulation since its council approved it in August 2024. On 20 August a summary of it appeared in American Psychologist, which is the version most clinicians will actually read (American Psychologist, 2026). The recommendations are worth knowing. The tier underneath them is worth more.

The panel worked from three systematic reviews and meta analyses, and it was not a psychology monoculture. It included professionals from psychology, occupational medicine, nursing, social work and physical therapy, alongside community members who identified as living with chronic musculoskeletal pain. Its conclusions are sorted into first line recommendations, worded as recommend, second line recommendations, worded as suggests, and a third tier for interventions where the evidence was not sufficient to recommend for or against.

The first line list is short. For chronic musculoskeletal pain generally, the panel recommends offering multicomponent self management interventions over no treatment or usual care, and Cognitive Behavioral Therapy over treatment as usual or another active intervention. For chronic low back pain it recommends exercise for short term management, and psychological therapy across short, intermediate and long term. For osteoarthritis knee pain it recommends exercise over usual care (APA, 2024). Second line, worded more tentatively, sit spinal manipulation, mindfulness based stress reduction, acupuncture for short term relief, and multidisciplinary rehabilitation.

One second line entry deserves quoting closely. For chronic low back pain the panel suggests exercise over yoga, citing a slight risk of harms in yoga, and then immediately adds that if the patient prefers yoga, yoga should be offered, because there is "essentially no difference in outcomes and only quite low risk associated with yoga." The stated reason for the ordering is not efficacy at all. It is that exercise is usually supervised by a physiotherapist or exercise physiologist with training in that domain.

What the panel could not resolve

Then comes the third tier, and this is where a practitioner should slow down. Under other treatments reviewed, the guideline lists the comparisons where evidence was insufficient to recommend one intervention over the other. For chronic low back pain: psychological therapy versus exercise, Qi Gong versus exercise therapy, massage versus exercise, spinal manipulation versus exercise. For chronic neck pain: relaxation training versus exercise, massage versus exercise, Alexander Technique plus usual care versus usual care alone, and basic body awareness therapy versus exercise. For osteoarthritis knee pain: Tai Chi versus attention control, and pain coping skills training versus exercise.

Read that list as a block rather than line by line. Nearly every head to head question between one movement approach and another came back the same way, and the panel's instruction in each case is identical: decisions should be based on shared decision making with the patient and consideration of available resources.

That is a more interesting result than it first looks, and it is easy to misreport. Insufficient evidence does not mean these approaches were shown to be equivalent. It means the trials needed to separate them were not there, or were too small, or too heterogeneous to pool. But it does mean that as of this guideline, no one gets to claim a hierarchy among them and cite the evidence for it. Yoga, Pilates, Tai Chi, qigong, Alexander Technique and the Feldenkrais Method® reach the body through different mechanisms, and each has its own strengths. What the guideline records is that the question of which mechanism does more has not yet been asked well enough to answer.

If efficacy cannot decide it, something else has to

The next variable is the one that quietly mattered all along, which is what the person keeps doing between appointments. The guideline's own background section is blunt about the size of that problem. Analysis of the National Ambulatory Medical Care Survey found that 21.5 percent of new visits for chronic musculoskeletal pain included an opioid prescription, while only 10.0 percent included a prescription for a guideline recommended nonpharmacological option such as physical therapy. In a separate survey of 31,916 adults living with chronic pain, 3.8 percent reported using psychological therapies at all (APA, 2024).

So the constraint is rarely that the right modality was unavailable. It is that whatever was prescribed did not survive the week. For the between session slot specifically, the property that matters is not which method would score highest in a trial that has not been run. It is which format someone will still open on a Tuesday evening when they are sore and tired. Awareness Through Movement® is delivered by voice, usually lying down, at low effort, and it asks for attention rather than exertion. That is a claim about format and adherence, not about superiority, and this guideline is a useful reminder that superiority is not currently anyone's to claim.

The practical version for a caseload is a licence to split the decision. The evidence supports being directive about the tier: get them moving, give them a framework, treat self management as a real intervention rather than a discharge instruction. It does not support being directive about the flavour. If someone has been referred to supervised exercise three times and stopped three times, nothing in this document obliges a fourth referral. It rather explicitly permits asking what they would repeat. Our Feldypedia entries on chronic lower back pain and neck and shoulder tension go into what that conversation tends to surface.

None of this displaces clinical care, and it is not meant to. Feldenkrais-based work sits alongside what a physiotherapist has already given someone, and the guideline's first line recommendations stay exactly where they are. The shift is smaller than a new treatment and more useful than most. For years the honest answer to which of these should I do was a shrug dressed up as a preference. A panel with three systematic reviews behind it has now written the shrug down and made it the recommended position. That makes it a much better question to ask out loud.