What Research Does and Does Not Show for Manual Therapies

Hands-on therapies have been used for a very long time, and plenty of people say they help. Whether that is the same as showing that they work is a separate question, and it is the one researchers try to answer. This page explains how to read a study, summarises what public health bodies and published reviews say about massage, spinal manipulation and related methods for back and neck pain, and describes what has been published on the Bowen Technique specifically. It also lists what the evidence cannot tell us. Individual situations differ, and your own doctor or qualified health professional has the final word on anything that affects you.

How to read the research

A randomised controlled trial sorts volunteers by chance into groups, so that one group receives the treatment being tested and another receives something else, such as a different treatment, usual care or an inactive comparison. The US National Center for Complementary and Integrative Health (NCCIH) calls a randomised, placebo-controlled trial the "gold standard" for testing interventions in people. A systematic review goes one step further. It searches for every trial that meets set criteria, judges their quality, and sometimes pools the results. When many well-run studies agree, that agreement carries more weight than any one of them. One trial on its own can mislead.

Then there is the placebo effect, which NCCIH defines as a beneficial health outcome resulting from a person's anticipation that an intervention will help. It also notes that the way a provider interacts with a patient can produce a positive response that has nothing to do with the specific treatment. This is not trickery. Expectation and attention are real parts of care. For research, though, they create a problem: a trial has to separate what the technique does from what the experience of being cared for does.

That separation is hard with hands-on methods. A drug trial can give one group a look-alike pill, and neither the volunteers nor the researchers need know who got what. NCCIH's guide advises checking whether a trial was "blinded" or "masked", because knowing who received which treatment can shape how people judge improvement. A person being massaged or manipulated knows it. A sham version of a hands-on treatment is difficult to make convincing, and, as we will see, the reviewers who looked at sham manipulation trials rated them as low quality. That is our reading of why the field struggles, not a claim made by any single source.

Two other limits recur. Small studies are more likely to produce results by chance, and a trial that stops following people after a few weeks cannot say anything about what happens after six months. Both points come up again below. For a practical guide to weighing claims you meet outside journals, see how to read health claims.

Low back pain: what the large reviews found

Low back pain is the condition most studied for hands-on care, so it is the best place to start. A 2019 systematic review in The BMJ by Rubinstein and colleagues pooled 47 randomised trials including a total of 9,211 participants who had chronic (long-lasting) low back pain. It looked at spinal manipulative therapy, a hands-on technique applied to the spine. The authors concluded that it produced effects similar to other recommended therapies, such as exercise, for short-term pain relief, and that it seemed better than non-recommended therapies for improving function in the short term. They also noted that most of the adverse events seen were transient and of mild to moderate severity, and that about half the studies examined them at all.

NCCIH's own fact sheet on spinal manipulation, last updated in September 2022, reaches a similar place in more cautious words. It says spinal manipulation may lead to small improvements in both pain and function. For acute low back pain (no more than 6 weeks), it cites a 2017 analysis of 15 studies with 1,699 participants that found a modest improvement in pain, similar to the benefit from non-steroidal anti-inflammatory drugs. Notice the adjectives: small, modest, similar. Nobody here is describing a cure.

Massage fares less well. A Cochrane review of massage for low back pain found massage improved pain and function in the short term when compared with inactive controls, but not in the long term. Even so, its authors wrote: "We have very little confidence that massage is an effective treatment for LBP". They rated every comparison as low or very low certainty, mainly because the studies were small and had methodological flaws. NCCIH's massage page, last updated in May 2019, makes the same point, describing the evidence for low back pain as weak.

The reader's real worry is usually practical: does any of this mean I should or should not try it? The reviews do not answer that. They describe average results across groups of volunteers, and a person is not an average. Your own doctor can say what is sensible given your history, which is the purpose of the questions in questions to ask your doctor first.

Neck pain and other conditions

For neck pain, NCCIH's page on chronic pain and complementary approaches (last updated January 2023) says that massage may be helpful for neck pain, but the benefits may only last for a short time. It also cites low-to-moderate quality evidence that spinal manipulation can reduce pain and improve function in chronic nonspecific neck pain. Nonspecific means no single identifiable cause was found. The same page notes a 2017 review in which osteopathic or spinal manipulation may have some benefit for chronic pain, and that a guideline on knee and hip osteoarthritis weakly recommends against massage.

The massage fact sheet covers further ground. It says that for knee osteoarthritis the research is limited but suggests short-term benefits, that headache studies are few and their results not consistent, and that massage may help some fibromyalgia symptoms if it is continued for long enough. For people with cancer, it says the evidence for pain and anxiety relief is not strong. Across all of these, the pattern is the same: possible short-term relief, thin or low-quality evidence, little known about the long term.

Soft-tissue and myofascial methods

Massage is one kind of soft-tissue work, meaning methods that act on muscles and the connective tissue around them. Myofascial approaches focus on fascia, the sheet-like connective tissue that wraps muscles and other structures, as explained in fascia and connective tissue, explained. The sources we opened for this page do not review myofascial release as a separate method, so we do not make claims about how well it works. The same is true of many named techniques: they may be taught and used widely without a body of controlled trials behind them. How these methods differ in practice, and who offers them, is covered in how massage, physiotherapy, osteopathy and chiropractic differ.

What has been published on the Bowen Technique

The Bowen Technique is a gentle hands-on method. Its background is set out in what the Bowen Technique is, and where it came from. The research is much thinner than for massage or spinal manipulation.

The most widely cited overview we could open is a systematic review by Hansen and Taylor-Piliae in the Journal of Alternative and Complementary Medicine, published in 2011. According to the record of the paper at the University of Arizona, the authors searched the literature up to September 2009 and found 309 citations, of which 15 met their criteria: one randomised clinical trial, two quasi-experimental studies, three mixed-methods studies, two cross-sectional studies and seven case studies. Of those 15, 53% reported pain reduction and 33% reported improved mobility. The authors wrote that Bowenwork may provide a noninvasive and affordable complementary approach, and also that its scientific evidence is poorly documented and that more systematic testing is needed before widespread recommendations can be given.

That last sentence matters. A case study describes one person or a few. Counting how many such reports mention pain relief says little about whether the method caused it, because pain often changes over time for reasons unrelated to any treatment, which is one reason controlled trials are needed. The review also stops at 2009, so it cannot reflect later trials.

One randomised trial worth knowing about is Marr and colleagues' 2011 study in the Journal of Bodywork and Movement Therapies. 120 people without symptoms were randomised to a control group or to one Bowen treatment, and an assessor who did not know the group measured hamstring flexibility (how far the knee could be straightened) at three points over a week. The Bowen group's flexibility continued to increase over the week, while the control group showed no significant change. This is a real finding, with real limits. The volunteers had no pain, the outcome was a flexibility measure, and the follow-up was one week. It tells us little about people living with pain.

We did not find a source we could open that states a firm conclusion across the full body of Bowen research, so we do not offer one. What the two papers above support is narrow: early signals in small and varied studies, and the authors' own call for more testing.

What the evidence says about safety

Safety is part of the evidence too. NCCIH's massage page says the risk of harm appears to be low, but rare serious events have been reported, including a blood clot, nerve injury or bone fracture, more often with vigorous massage or in people who are at higher risk. For spinal manipulation, NCCIH says mild side effects such as increased pain, stiffness or headache often occur and mostly go away within 24 hours, while serious problems are very rare and there are no accurate estimates of how often they happen. The Cochrane massage review reported that increased pain was the most common adverse event, affecting between 1.5% and 25% of participants across studies.

Anything new, severe or unexplained is a different matter from soreness after a session. The list of symptoms that need a doctor first is in red-flag symptoms that need a doctor.

What we cannot say

We cannot say that any manual therapy treats, cures or prevents a condition. The sources describe small average effects for some methods in some conditions, usually in the short term, and often from low-quality evidence. We cannot say how much of any improvement comes from the technique itself and how much from expectation, attention or the natural course of an ache, because most trials cannot fully separate them. We cannot tell you what is likely to happen in your own body.

We also cannot treat a lack of evidence as proof that something does not work. For many hands-on methods, the trials have simply not been done at scale. It means the question is open, and that claims should be held loosely. For the nervous system's role in pain, which sits behind many of these uncertainties, see pain and manual therapy.

The short version

For low back and neck pain, the best-quality reviews describe small, mostly short-term benefits from spinal manipulation and massage, with low to moderate certainty. For the Bowen Technique, published research is limited, and its reviewers say more testing is needed. Treat any claim stronger than the sources themselves with caution, and take your questions to your own doctor.