In 2018, The Lancet published a series on low back pain described as one of the most disabling conditions in the world. The series concluded that the management of back pain globally was characterised by widespread use of ineffective and potentially harmful treatments, and that this represented a major public health crisis.
The specific treatments identified as ineffective or harmful included: bed rest, opioid pain medications for non-specific back pain, unnecessary imaging (MRI and CT scans), spinal fusion surgery for non-specific pain, and epidural steroid injections as a first-line treatment. These are, coincidentally, among the most commonly provided treatments for back pain in the UK and globally.
This is not a problem of insufficient medical knowledge. The evidence is clear. The problem is a persistent gap between what evidence shows works and what patients receive - driven by patient expectations, time-pressured consultations, and a healthcare system that finds it easier to investigate and prescribe than to implement the evidence-based approaches that actually help.
What actually causes most back pain
Approximately 85-90% of back pain presenting to primary care is classified as non-specific low back pain - meaning that no specific structural cause can be identified that fully explains the symptoms. This is often profoundly unsatisfying for patients who want a diagnosis, a cause, and a solution. It is, however, the medical reality.
The reason this matters is that common structural findings on MRI - disc bulges, disc degeneration, facet joint arthritis, mild scoliosis, Schmorl's nodes - are also extremely common in completely asymptomatic people. A landmark systematic review published in the American Journal of Neuroradiology (Brinjikji et al., 2015) imaged pain-free volunteers across age groups and found:
- Disc bulge in 30% of 20-year-olds, 60% of 50-year-olds, 84% of 80-year-olds
- Disc degeneration in 37% of 20-year-olds, 80% of 50-year-olds, 96% of 80-year-olds
- Facet joint degeneration in 4% of 20-year-olds, 57% of 50-year-olds
These findings are normal ageing changes. When someone with back pain has an MRI and these findings are reported, they are almost certainly incidental - present before the pain started, present in millions of pain-free people, not the cause of the pain. But the radiology report reads as abnormal. The patient feels validated (there is something wrong) but becomes frightened. They protect the spine, avoid movement, become deconditioned. The pain worsens.
The biopsychosocial model - why back pain is not just about the spine
Modern pain science understands that pain is not simply a signal from damaged tissue. It is a complex output of the central nervous system that is influenced by multiple factors simultaneously: the state of the peripheral tissues, psychological factors (anxiety, depression, catastrophising, fear of movement), social factors (work dissatisfaction, litigation, social support), and the individual's prior experience of pain and beliefs about it.
This does not mean back pain is imagined or exaggerated. It means that treating the spine in isolation - without addressing the psychological and social factors that shape the pain experience - consistently produces worse outcomes than addressing all dimensions.
Case study: Tendai's MRI that made things worse
Tendai, 42, a civil engineer from Zimbabwe based in Bristol, developed sudden, severe lower back pain after lifting a box. He saw his GP, who arranged an MRI.
The MRI report: L4/5 broad-based disc bulge with mild left foraminal narrowing and facet joint hypertrophy; L5/S1 disc degeneration with mild annular fissure and moderate disc space narrowing.
Tendai went home and, despite being told it was probably not serious, could not help Googling. What he found terrified him. He became convinced he had serious, progressive spinal damage. He stopped exercising - something he had done regularly for years. He began moving very carefully, convinced that wrong movement would worsen his spine. He took four weeks off work. He started sleeping on the floor because he read that firm surfaces were better.
His pain did not improve. It got worse.
When he came to see me six weeks after the MRI, I spent 20 minutes explaining what the findings actually meant. I showed him the Brinjikji data. I explained that his MRI findings were essentially normal ageing changes almost certainly present before his pain episode. I drew the pain cycle: pain causes fear, fear causes avoidance, avoidance causes deconditioning and reduced tissue loading, deconditioning worsens pain and disability.
I referred him to physiotherapy. His physiotherapist progressively loaded his back - helping him rebuild confidence in his spine's ability to move and bear load. He returned to his gym programme within 6 weeks. He was running again within 10.
"The scan made things worse, not better," he told me. "It gave me a story about my spine that was not true."
What the evidence actually supports
Stay active: The single most evidence-supported recommendation for acute non-specific back pain. Remain as active as possible. Modify activities that significantly worsen pain, but avoid bed rest. Movement is medicine for back pain. Rest is not.
Exercise-based physiotherapy: For back pain persisting beyond 4-6 weeks, structured exercise-based physiotherapy has strong evidence. The exercises should be active (you do them) rather than passive (things done to you). Manual therapy (manipulation and mobilisation) provides modest short-term benefit but should not be the primary treatment.
Psychological approaches (CBT for pain): For chronic back pain (more than 12 weeks), CBT addressing catastrophising, fear-avoidance behaviour, and activity engagement has robust evidence. Pain management programmes combining physical and psychological approaches are the most effective intervention for chronic, disabling back pain. Ask your GP for referral to a pain management programme, not just physiotherapy.
Simple analgesia: Paracetamol and NSAIDs (ibuprofen if tolerated; check kidney function) help manage acute pain enough to maintain activity. They are tools to enable movement, not treatments in their own right.
What does not help: Bed rest (worsens outcomes), opioids for non-specific chronic back pain (limited evidence, significant harm potential), imaging in the first 6 weeks without red flags (increases fear and does not change management), passive treatments that don't involve active patient participation.
Sources: NICE Clinical Guideline NG59 - Low Back Pain and Sciatica (2016, updated 2023); Brinjikji W et al, American Journal of Neuroradiology 2015 (MRI findings in asymptomatic adults); Hartvigsen J et al, The Lancet 2018 (Low back pain series); Main CJ et al, The Lancet 2008 (biopsychosocial model); Foster NE et al, The Lancet 2018 (prevention and treatment of low back pain); Deyo RA et al, NEJM 2001.
