Spinal manipulation for chronic low back pain: what the evidence shows

A common treatment, and a fair question

If your lower back has hurt for months, someone has probably suggested "getting an adjustment" — a chiropractor, a physiotherapist, or a friend who swears by it. Spinal manipulation is one of the most widely used hands-on treatments for back pain, and it's reasonable to wonder what it actually does. The honest answer, drawn from the better-quality research, is that spinal manipulative therapy tends to produce a small improvement in pain and function — roughly in the same range as other active treatments people are commonly offered, rather than clearly better or worse than them. This piece lays out what the studies found, how certain that evidence is, what's known about safety, and where the major guidelines land, so you can have a more informed conversation with a regulated health professional. We're not here to talk you into it or out of it.

What spinal manipulation is, and who provides it

"Spinal manipulative therapy" (SMT) is an umbrella term for hands-on treatment of the spine. It usually covers two related techniques: mobilisation, which uses slower, gentler movements to ease a joint through its range, and manipulation, a quick, controlled thrust to a joint that sometimes produces an audible "pop" [2]. Practitioners often combine these with advice, reassurance, and exercise rather than using them alone.

Several kinds of regulated professionals provide it. Chiropractors, whose training centres on the spine, are the most closely associated with manipulation; physiotherapists may use manipulation and mobilisation as part of a broader movement-based program; and osteopaths use manual techniques including spinal manipulation [2]. Scope of practice, training, and regulation differ by profession and by province or country, which is one reason a proper local assessment matters more than any general summary.

A note on scope: the evidence below is about chronic nonspecific low back pain — pain lasting roughly three months or longer with no single identifiable cause such as a fracture, infection, or tumour, which describes the large majority of long-standing back pain. It is not about back pain with "red flag" features (for example new bladder or bowel problems, saddle numbness, progressive leg weakness, fever, or a history of cancer), which need prompt medical assessment rather than manual therapy.

What the evidence shows

The most useful way to read this research is to notice that the answer depends heavily on what SMT is being compared to.

The largest and most current synthesis is a 2026 Cochrane systematic review by de Zoete, Rubinstein, and colleagues, which pooled 76 trials involving 11,866 people with chronic low back pain [2]. Cochrane reviews are generally regarded as the most rigorous tier of evidence because of how carefully they assess study quality. This one used a threshold of about 10 points on a 100-point scale as the mark of a difference likely to matter to a patient. Its headline findings:

• Compared with other conservative treatments (the kind of care a person might otherwise receive), SMT produced only a small difference — around 4.7 points for pain and 4.9 for function on the 100-point scales, both below the 10-point threshold. In the review's own words, SMT "may result in little to no difference in pain and a small improvement in functional status" [2].
• Compared with fake ("sham") manipulation, the difference was about 7.0 points for pain and 8.8 for function — again short of the threshold usually considered clinically meaningful [2].
• Compared with no treatment, SMT looked best: roughly 14 points for pain and 12.9 for function [2]. The catch, as with any "no treatment" comparison, is that it can't separate the manipulation itself from simply receiving hands-on care and attention.

Crucially, the review rated all of this as low-to-very-low-certainty evidence, largely because the trials were conducted in different settings and populations, using different techniques, doses, and treatment frequencies, with results that varied widely [2].

A second high-quality synthesis points in the same direction. A 2019 network meta-analysis in The BMJ, led by Rubinstein, compared SMT specifically against treatments that guidelines recommend versus those they don't [1]:

• Versus recommended therapies (such as exercise and standard care), SMT produced a similar effect on pain — a difference of about 3 points on the 100-point scale that was not statistically significant (mean difference −3.17, 95% CI −7.85 to 1.51, moderate-quality evidence) — and a small additional improvement in function (standardised mean difference −0.25, 95% CI −0.41 to −0.09) [1].
• Versus non-recommended therapies, SMT showed a small benefit for pain (mean difference −7.48, 95% CI −11.50 to −3.47, high-quality evidence) and a small-to-moderate one for function (SMD −0.41, 95% CI −0.67 to −0.15) [1]. These patterns held at intermediate and longer-term follow-up.

Read together, the two reviews tell a consistent story: SMT delivers a small, real improvement, and against the active treatments people are actually choosing between, its effect is roughly comparable — not a clear winner, not clearly worse. It is worth being explicit that "comparable to other treatments" is the finding; nothing here supports ranking SMT above exercise, physiotherapy, or the other reasonable options.

Safety and limitations, honestly

On safety, the picture is reassuring but not unqualified. The Cochrane review reported that no serious adverse effects related to SMT were observed; the common downsides were minor and short-lived — muscle soreness or a temporary increase in pain [2]. The BMJ meta-analysis found that most reported adverse events were musculoskeletal, transient, and mild to moderate; one adequately powered trial found no increased risk of adverse events compared with sham manipulation (relative risk 1.24, 95% CI 0.85 to 1.81), and across all the trials a single serious adverse event was judged by a safety board as possibly related to treatment [1].

Two honest caveats sit alongside that. First, both reviews noted that fewer than half of the trials reported on adverse events at all, so the safety data are thinner than the benefit data [1][2]. Second, serious complications from spinal manipulation are considered rare, but rare is not the same as never, which is part of why a qualified assessment beforehand matters.

The larger limitation is certainty. The Cochrane authors rated the evidence low-to-very-low, and the reasons are worth understanding [2]:

• The effects are small. Against other active treatments, most measured differences fell below the threshold usually considered meaningful to patients [1][2]. That is a modest foundation, and a poor one for strong claims in any direction.
• The trials are heterogeneous. "Spinal manipulation" is not one standardised thing — technique, number of sessions, practitioner training, and patient populations all vary from study to study, which makes pooling difficult and downgrades confidence [2].
• Blinding is hard. A patient can sometimes be kept unaware of whether they received real or sham manipulation, but the practitioner always knows, which introduces a recognised risk of bias [1].
• Durability is unclear. Effects were often measured in the short term, and the longer-term comparisons tended to show even smaller differences [1].

What the guidelines say

Clinical guidelines, which weigh this kind of evidence for practising clinicians, generally treat spinal manipulation as a reasonable option within a broader plan rather than a stand-alone answer.

• The UK's National Institute for Health and Care Excellence (NICE) advises clinicians to "consider manual therapy (spinal manipulation, mobilisation or soft tissue techniques such as massage) for managing low back pain … but only as part of a treatment package including exercise, with or without psychological therapy" (recommendation 1.2.7) [3]. Exercise itself is recommended as a first-line option (recommendation 1.2.2) [3]. In other words, NICE supports manipulation, but tied to an active program rather than on its own.
• The American College of Physicians (ACP) includes spinal manipulation among the non-drug options to consider first for chronic low back pain, as part of its broader recommendation to try non-drug approaches before medication — while rating the supporting evidence for manipulation as low-quality [4].

Both bodies, then, place SMT among several acceptable non-drug choices, emphasise exercise and active care, and stop well short of singling manipulation out as superior. That framing matches the effect sizes: several reasonable options, none a clear front-runner.

What this means for you

If you're living with chronic low back pain, here's a fair summary. Spinal manipulation appears to offer a small improvement in pain and function; against the other active treatments you might be weighing, its effect looks broadly comparable rather than better; the certainty of that evidence is low; serious harms appear rare, with minor soreness the usual downside; and guidelines treat it as one reasonable option, best combined with staying active [1][2][3][4]. None of that is a promise, and it isn't a substitute for a proper assessment.

The right next step is a conversation with a licensed, regulated health professional — a family physician, physiotherapist, chiropractor, or osteopath — who can take your history, examine you, rule out the small number of causes that need different care, and help you weigh spinal manipulation alongside options like exercise and other therapies. They can tell you whether it's a sensible fit for you, and how to combine it with the active care that the evidence most consistently supports — questions the research alone can't answer about your particular back.