Does massage help back pain? What the studies show

Start with the bodies that have looked hardest at this. The Agency for Healthcare Research and Quality (AHRQ), in a 2016 evaluation of nondrug therapies for low-back pain, examined 20 studies comparing massage to usual care or other treatments. It found evidence that massage helped chronic low-back pain, but rated the strength of that evidence as low. The same review looked at 6 studies pitting different massage styles against each other and found the evidence too thin to say whether any one type beats another.

Then there is the 2015 Cochrane review already mentioned. Across both acute and chronic low-back pain, patients had short-term improvements in pain after massage. The catch is the recurring one: the studies were low quality, so confidence in the finding is weak. NCCIH does not hedge about the hedging. Its plain-language read is that evaluations of massage for low-back pain have found "only weak evidence that it may be helpful."

One trial deserves a closer look because it was better designed than most. A 2011 randomized controlled trial of 401 adults with nonspecific chronic low-back pain compared two kinds of massage, structural and relaxation, against usual care. At 10 weeks, both massage groups were doing meaningfully better on function and symptoms than the usual-care group. The relaxation-massage group scored 2.9 points lower on the Roland Disability Questionnaire, a standard measure where lower means less disability. Benefits on function still showed up at 52 weeks, though by then they were small.

And structural massage, with its targeted, technical approach, was no better than plain relaxation massage. That undercuts the idea that you need a specialized method to get the effect — a finding that will come up again when we ask what massage is actually doing.

Read enough of these summaries and one phrase repeats: short-term. Massage tends to help in the days or weeks right after a course of treatment, and the relief fades. The 2015 Cochrane review — the one carrying 3,096 patients behind it — said it directly: the long-term effects of massage for low-back pain have not been established.

A trial cited in the American Massage Therapy Association's position statement shows the pattern in a different setting. Researchers randomized 605 veterans recovering from major surgery to one of three tracks — routine care, brief individual attention, or evening back massages for up to 5 days. The massage group had faster short-term drops in pain intensity, pain unpleasantness, and anxiety. What massage did not change: opioid use, length of stay, or complications. The relief was real and it was bounded — exactly the shape the back-pain reviews keep describing.

Short-term relief is still real relief, especially during a flare-up that is keeping you off your feet. But it shapes how you should hear the question. Does massage help back pain? Yes — for a stretch of days or weeks. Frame it as a temporary reset rather than a cure, and the evidence and the experience line up.

The doctors with the most credibility in this area have weighed in, and the split is telling. The American College of Physicians issued clinical practice guidelines in 2017 covering acute, subacute, and chronic low-back pain. Massage made the list, but only for acute and subacute pain, and based on low-quality evidence. For chronic low-back pain, the ACP left massage off its recommended options entirely. The message: massage has a place for a recent, short-lived backache, but the ACP does not endorse it as a tool for the long, grinding kind. That distinction — acute versus chronic — turns out to be the most practical line in the whole literature, and it is worth knowing which side of it your pain sits on before you book anything.

The wider trend still works in massage's favor. The Centers for Disease Control and Prevention's 2022 guideline on prescribing opioids for pain concluded that multiple noninvasive, nondrug treatments improve chronic pain and function, "with small to moderate effects in specific pain conditions," and are "not associated with serious harms." Massage therapy is named in that guideline as appropriate for some types of chronic pain, alongside acupuncture, tai chi, yoga, and mindfulness. Read plainly, that is a modest endorsement: small to moderate effects, low risk. A reason to try massage as part of a plan, not a promise it will solve the problem.

The neck and shoulder evidence looks much like the back evidence. A 2013 review of 12 studies (757 participants) found massage more helpful for neck and shoulder pain than inactive comparisons, but no better than other active treatments. For shoulder pain specifically, the benefit was short-term only. A 2016 review of 4 U.S. trials found massage could give short-term relief of neck pain, with a condition attached: the sessions had to be long enough and frequent enough.

That dose question got a careful test. A 2014 trial randomized 228 people with chronic nonspecific neck pain to different "doses" of massage over four weeks. The result was specific and a little surprising. Sessions of 30 minutes, no matter how often, were no better than sitting on a wait list. What worked was 60-minute massages two or three times a week. Those longer, more frequent sessions raised the odds of a clinically meaningful improvement in neck function by roughly 3 to 5 times, and the odds on pain by about 2 to 3 times.

A 2012 Cochrane review of 15 neck-pain trials is the dissent worth noting. It concluded that no recommendations for practice could be made, because the effectiveness of massage for neck pain remained uncertain. Even inside the neck literature, the experts do not fully agree.

The mechanism is less settled than the marketing implies, and the honest sources admit it. The American Massage Therapy Association, in its position statement, gathers a long list of effects researchers have measured — reduced anxiety and depression, lower blood pressure and heart rate, decreased pain — and points to studies showing physiological changes from a single Swedish massage, including shifts in stress hormones and immune markers. A 2012 study found that massage of muscle damaged by exercise dialed down inflammatory signaling inside the tissue. That is laboratory evidence that touch does something measurable in the body.

Measurable is not the same as explained. NCCIH, describing premature-infant research in which massaged babies gained more weight, states flatly that the mechanism "is not well understood." A clue comes from that better-designed back-pain trial: structural massage did no better than relaxation massage. If a specific manipulation were correcting a specific problem, you would expect the technical version to win. It didn't. That points toward general effects — relaxation, reduced muscle tension, a calmer nervous system — rather than anything mechanically precise.

A meta-analysis cited in AMTA's statement describes massage's largest effects as reductions in trait anxiety and depression, of a magnitude similar to psychotherapy. The most consistent thing massage does may be to calm you down, and calmer people tend to hurt less. None of that makes the relief less real. It just relocates where the benefit comes from.

Here massage has a genuine edge over many pain treatments. NCCIH's read: the risk of harmful effects "appears to be low." For a healthy person getting massage from a trained therapist, that is most of the story. NCCIH's broader chronic-pain guidance adds that people with medical conditions and people who are pregnant may need to modify or skip some approaches. Set against the risks of long-term opioid use, a low-probability bruise or strain is a different category of concern — and that favorable safety profile is part of why massage keeps appearing in guidelines despite effectiveness evidence that sits somewhere between weak and modest.

Notice what massage is almost always measured against in these studies: usual care. It is tested as something you add, not something you swap in. The comparisons that exist are humbling. Across the 12 neck and shoulder studies in the 2013 review, massage beat doing nothing but tied with other active treatments. The ACP's 2017 guideline treats massage as one option among many for acute back pain, not the centerpiece, and leaves it off the chronic list, where the guidance points instead toward approaches like exercise and tai chi. That is not an indictment of massage; it is a job description — the supporting role, played well.

A fair reading of the evidence turns "does massage help back pain" from a yes-or-no question into a plan — and one thing massage will not do is reward you for waiting. NCCIH's advice is blunt: don't use massage to postpone seeing a clinician about a medical problem. With that said, here are the steps the research supports, in order.