Massage for chronic pain: where it actually fits

Chronic pain does not spread evenly. NCCIH notes it grows more common with age, partly because pain-causing conditions like osteoarthritis pile up over the years. Military veterans are at higher risk: national survey data show both general pain and severe pain are more common among veterans than nonveterans. People in rural areas report chronic pain more often than people in cities.

Prevalence also varies by race and Hispanic origin. An analysis of 2010–2017 survey data found the highest rate among non-Hispanic American Indian and Alaska Native respondents and the lowest among non-Hispanic Asian respondents. Among people who do have a diagnosis, low-back pain and osteoarthritis lead, followed by rheumatoid arthritis, migraine, carpal tunnel syndrome, and fibromyalgia. But plenty of people have chronic pain with no clear diagnosis at all. Sometimes the problem sits in the nervous system itself, which clinicians call neuropathic pain. Sometimes no one can name the cause.

The pain rarely shows up alone, either. It tends to drag disturbed sleep, restricted movement, anxiety, and depression along with it. Any honest accounting of what helps has to reckon with that whole cluster, not just the ache.

The distinction does more work than it looks like. A complementary approach is something you use alongside conventional medical care. An alternative approach is something used instead of it. NCCIH is blunt about the rule that runs underneath all of its guidance: do not use an unproven product or practice to put off seeing a clinician about your pain. That rule is not anti-massage; it is anti-delay. The worry is not the massage table — it is the diagnosis that never happens while you are on it.

NCCIH has also changed how it labels these methods. What it used to group as "mind and body practices" is now sorted by how the therapy is delivered: psychological approaches such as meditation and music therapy, physical approaches such as acupuncture, massage, and spinal manipulation, and combined approaches such as yoga and tai chi. The relabeling is a useful corrective. A "complementary" approach can be as physical and specific as a needle or a spinal manipulation. It is not a vague cloud of good vibes.

Massage for chronic pain is the approach most people assume is best proven. The evidence is thinner than that assumption, and it helps to be precise about where it holds and where it doesn't. For low-back pain, several reviews found only weak evidence of benefit. The Agency for Healthcare Research and Quality reviewed 20 studies in 2016 and concluded massage helped with chronic low-back pain, but rated the strength of that evidence as low. A 2015 Cochrane review found massage may give short-term relief, again on evidence that wasn't high quality, with long-term effects unestablished. The guidelines reflect that wobble: the American College of Physicians, in its 2017 guideline, listed massage as an option for acute and subacute low-back pain but left it off the list for chronic low-back pain.

The clearest signal comes from neck and shoulder pain, and even there the benefit fades. A 2013 review of 12 studies (757 participants) found massage outperformed inactive therapies for neck and shoulder pain, though it was no better than other active treatments, and the shoulder benefit was short-term only. Dose seems to matter: a 2014 trial of 228 people with chronic neck pain found that 60-minute sessions given several times a week worked better than shorter or less frequent ones. For osteoarthritis, a 2017 review of 7 trials (352 participants) found low-to-moderate quality evidence that massage reduced pain and improved function, with most of the research done on knee osteoarthritis.

Fibromyalgia produced one of the more specific findings. A 2014 review of 9 studies (404 participants) concluded that massage improved pain, anxiety, and depression — but only if it continued for at least five weeks, and it did nothing for sleep disturbance. The duration is part of the prescription, not a footnote. A single session is unlikely to move the needle. That specificity is useful, though: it turns "try massage" into an actual plan with a duration attached.

Here the sources openly disagree, and that disagreement is worth keeping rather than smoothing over. NCCIH's chronic pain review notes that low-to-moderate quality evidence suggests massage may reduce arthritis pain, yet the guideline from the American College of Rheumatology and the Arthritis Foundation weakly recommends against massage for hip or knee osteoarthritis, citing weaknesses in the evidence. Both conclusions draw from the same shallow pool — the 2017 review's 7 trials and 352 participants, most of them on knees. When careful reviewers split like that, the effect, if it exists, is probably small. On safety, the risk of harm from massage looks low; NCCIH points to rare reports of serious problems — a blood clot, a nerve injury, a bone fracture — often tied to vigorous techniques like deep tissue work or to people already at higher risk of injury.

If you are spending limited time and money, a few approaches carry more weight than massage for chronic pain. Acupuncture has the strongest case. A large meta-analysis of individual patient data found acupuncture more effective than no treatment and more effective than sham acupuncture for back and neck pain. The gap over no treatment was wider than the gap over sham, which tells you a real share of the effect comes from everything around the needle, not just the needle. There is moderate-quality evidence it reduces migraine frequency, and acupuncture is the one approach NCCIH flags as having evidence it can reduce a patient's need for opioids. In studies comparing acupuncture against migraine drugs, the people getting acupuncture were far less likely to drop out over side effects.

Tai chi earns the strongest guideline language of anything here. For knee or hip osteoarthritis, the American College of Rheumatology and the Arthritis Foundation strongly recommend it; people who practiced tai chi saw improvements in pain, stiffness, balance, and physical function. The ACP's 2017 guideline also lists it as an option for chronic low-back pain. Yoga holds up for low-back pain in both the short and intermediate term, with effects similar to other forms of exercise, and that same 2017 guideline includes it as an option for chronic — not acute — low-back pain.

For chronic low-back pain specifically, the 2017 ACP guideline also points to mindfulness-based stress reduction, progressive muscle relaxation, and biofeedback as reasonable first-step, nondrug options. On the psychological side, studies found mindfulness-based interventions and cognitive behavioral therapy both lowered pain intensity and improved function, with no important difference between them, and CBT remains the prevailing psychological treatment for chronic pain.

The frame that ties this together comes from the CDC. Its 2022 opioid prescribing guideline concluded that multiple noninvasive, nondrug treatments improve chronic pain and function, with small to moderate effects in specific conditions, and are not associated with serious harms. That one sentence is the frame the rest of this guide hangs on: modest effects, low risk, best used early and in combination rather than late and alone.

Honesty cuts both ways: several popular approaches do not hold up. For osteoarthritis, the rheumatology guideline strongly recommends against glucosamine at the hand, hip, or knee, because the best-designed studies show it works no better than placebo. The same guideline advises against chondroitin and combination glucosamine-chondroitin products for hip or knee, and against fish oil and vitamin D for osteoarthritis generally. Reiki has no high-quality research showing it relieves pain. Static magnets, the kind sold in bracelets and shoe insoles, have no conclusive evidence behind them for any type of pain.

Cannabis products sit in a murkier zone. Short-term studies suggest some formulations with THC and CBD may reduce chronic pain, though with side effects like dizziness and drowsiness, and little is known about long-term use or other formulations. Hypnosis may help, but the studies that found a benefit involved at least 8 sessions. That is a real commitment, not a single appointment. Neither finding is a reason to write these options off entirely — but both are reasons to price in the commitment and the unknowns before you start.

The realistic expectation, pulled straight from the evidence, is modest and worth holding onto. Massage for chronic pain — like the other better-supported complementary approaches — tends to deliver the small to moderate benefit the CDC described in 2022, often short-term, usually as one piece of a larger plan that still includes conventional care. That is not a reason to wave them off. A small, low-risk benefit you can stack on top of other treatments is a genuinely good deal, as long as you walk in knowing that is what you are getting.

The most useful instruction in all of NCCIH's guidance is also the least dramatic: make these decisions with the people treating you. In practice, that comes down to a few concrete steps.