Back pain, and what the evidence changed

Low back pain is the leading cause of disability worldwide, and the standard management of it has been substantially wrong for decades in ways that have been well documented.

What the evidence now says

Most low back pain is non-specific, meaning no single structure can be identified as the cause.

The great majority improves substantially within weeks regardless of treatment.

Bed rest delays recovery, and staying active — including continuing work in some form — produces better outcomes.

Imaging in the absence of warning features does not improve outcomes and is associated with worse ones, including more surgery and greater disability, probably because of what the findings do to people's beliefs.

These conclusions appear consistently in guidelines internationally, and practice has been slow to follow them.

The scan problem

Studies imaging people with no back pain at all have found disc degeneration, bulges and other changes in a large proportion, rising steeply with age — such that these findings are better understood as the radiological equivalent of grey hair than as pathology.

A person told they have degeneration, a bulging disc and wear and tear will reasonably conclude their spine is damaged and fragile.

That belief predicts worse outcomes, more avoidance and more disability, independently of the imaging findings themselves.

Which is why guidelines advise against routine imaging and why the language used to report it matters.

When imaging is indicated

The warning features that change the picture.

New bladder or bowel dysfunction, saddle numbness, or progressive weakness — which may indicate cauda equina syndrome and require emergency assessment the same day.

Significant trauma.

A history of cancer.

Fever, night sweats or unexplained weight loss.

Age at onset under twenty or over fifty with new pain.

Immunosuppression or intravenous drug use.

Progressive neurological deficit.

And pain that is unremitting, wakes from sleep and does not vary with position.

What helps

The interventions with evidence, which are less satisfying than people want.

Staying active and returning to normal activity as soon as possible, with gradual increase rather than waiting for pain to resolve.

Exercise of essentially any type, where the evidence favours doing something over the specific choice.

Education about what pain does and does not indicate, which has a measurable effect on outcomes.

Manual therapy as an adjunct in the short term.

Psychological approaches for persistent pain, including cognitive behavioural therapy and acceptance-based approaches, which have evidence for function and distress.

Heat in the acute phase.

And simple analgesia, with the caveat that paracetamol has performed poorly in trials for back pain specifically, that non-steroidal anti-inflammatories have modest effects, and that opioids are not recommended for chronic non-cancer back pain because harms exceed benefits.

Posture and lifting

Where the advice has changed most.

The evidence for a single correct posture preventing back pain is weak, and the association between spinal alignment and pain is much less clear than assumed.

Studies of lifting technique have not consistently shown that the traditional straight-back instruction prevents injury, and interventions based on it have generally failed to reduce back pain in workplaces.

What appears to matter more is variation — not staying in any one position for long — and building capacity to tolerate load rather than avoiding it.

The message that the back is fragile and must be protected has probably done more harm than the postures it warned against.

Persistent pain

Where a different model applies.

Pain persisting beyond the expected healing time involves changes in how the nervous system processes signals, rather than ongoing tissue damage.

This is not the same as pain being imagined; it is a real and measurable phenomenon.

Which is why treatment for persistent pain focuses on function, graded activity, sleep, mood and beliefs, rather than on finding and fixing a structure.

Multidisciplinary pain programmes have the best evidence for this group.

Surgery

Has a defined and narrow role.

It is effective for specific situations: significant nerve compression with corresponding symptoms that has not settled, spinal stenosis with limiting symptoms, and emergencies.

For non-specific low back pain, evidence for fusion surgery compared with structured rehabilitation is weak, and rates vary enormously between regions in a way that reflects practice patterns rather than disease.

Which makes a second opinion reasonable before any elective spinal surgery for pain alone.

General information only, not medical advice. Seek emergency assessment for new bladder or bowel symptoms, saddle numbness or progressive weakness with back pain.