The Back Pain Nobody Can Name

Why nobody can name it

The Lancet's 2018 low back pain series put the position about as plainly as a medical journal ever does. For nearly all people with low back pain, it is not possible to identify a specific nociceptive cause. Only a small proportion have a well understood pathological cause. The World Health Organization gives the same picture from a different angle: non-specific low back pain is the most common presentation, about 90 percent of cases.

This is not a gap waiting to be closed by a better scanner. The scanner is the problem. A 2015 review pooled imaging from 3,110 people with no back pain at all and found degenerative changes almost everywhere it looked. Disc degeneration appeared in 37 percent of pain-free twenty-year-olds and 96 percent of pain-free eighty-year-olds. Disc bulges ran from 30 percent to 84 percent across the same age range. The authors concluded that many imaging-based degenerative features are likely part of normal ageing and unassociated with pain.

Which is why imaging early, absent specific warning signs, does not help. Six randomised trials covering 1,804 people compared immediate lumbar imaging against usual care and found no meaningful difference in pain or function, in the short term or the long term. Every effect estimate crossed zero. Lumbar imaging for low back pain without indications of serious underlying conditions does not improve clinical outcomes, and that is a direct quote rather than a summary.

The practical consequence is uncomfortable but freeing. If you have been through a scan and been handed a list of degenerative findings, you were probably shown a photograph of an ordinary spine of your age. The list is not a diagnosis and it is usually not a prognosis either.

The tightness is real. The story about it usually is not.

An earlier version of this article told you that the culprit was your hip flexors. That sitting folds the hip, shortens the psoas, and that the shortened psoas then pulls on the lumbar spine where it originates. The back feels tight, we wrote, but the cause is upstream. It is a satisfying story. It is also not supported, and correcting it is worth more to you than anything else on this page.

The chain needs three links to hold. Tight hip flexors, which tilt the pelvis forward, which increases lumbar curve and loads the back. The middle link fails first. A 2014 review comparing lumbo-pelvic movement between people with and without back pain found no difference in lordosis angle across eight studies and no difference in standing pelvic tilt across three. A 2025 meta-analysis of postural asymmetry did find pelvic tilt slightly higher in people with back pain, but at a standardised mean difference of 0.23 with 72 percent heterogeneity, which is a small effect measured inconsistently, and it found no significant difference in lumbar lordosis at all.

The first link fails on the one study built to test it. A 2026 paper measured hip flexor range with the Modified Thomas Test in 118 adults. People with back pain did have less range, 4 degrees against 8, and the difference reached significance. Then the authors adjusted for the obvious confounders, and the association disappeared. Only body mass index remained a significant predictor of hip flexor range. Back pain status was not independently associated with it.

The last link fails on treatment. If tight hips drove the pain, treating the hips would relieve it. Six randomised trials of hip-targeted physiotherapy, pooled in 2019, produced a trivial effect size for pain and no statistically significant reduction in either pain or disability. Meanwhile a 2019 review of hip range of motion concluded the evidence was very low quality throughout, and the only movement significantly associated with back pain was internal rotation, not flexion and not extension.

The counter-evidence deserves a fair hearing, because it exists. A 2024 review of hip biomechanics found people with back pain do show reduced hip range, particularly internal rotation, and weakness in the hip abductors and extensors, and its authors argue clinicians should pay more attention to the hip. A 2023 trial review found hip strengthening added to a general programme helped, though significantly in only three of six pain comparisons. Note the direction, though. The muscles that show up weak are the abductors and extensors. That is close to the opposite of a tight, dominant hip flexor.

So the defensible claim is that the hip is worth assessing and worth training. That is a long way from the hip being the cause. And the tightness you feel at four in the afternoon is real either way. Sensation is not in question. What is in question is the confident anatomical story attached to it.

One more casualty of the same reasoning is the weak core. A 2016 Cochrane review of motor control exercise, the formal version of core training, found it works, and found it is not superior to other forms of exercise. A prospective review looking at what actually predicts developing back pain found abdominal strength, back muscle strength and back muscle endurance all failed to predict it. Core work is a reasonable thing to do. It is not a repair for a defect you have been told you have.

What actually moves the number

The largest synthesis available is a 2021 Cochrane review of 249 trials and 24,486 people. Exercise beat no treatment, usual care or placebo for pain by a mean difference of 15.2 points on a 100-point scale, which the authors call a clinically important difference. It is worth being precise about the second half of that result, because it is usually left out: the effect on function was 6.8 points, which did not meet their own threshold for clinical importance.

And compared against other active treatments rather than against nothing, exercise came out ahead by 9.1 points for pain and 4.1 for function, which the review describes as small and not clinically important when all comparisons are considered together. Exercise is the best-supported thing on the list. It is not a cure, and anyone selling it as one is overselling.

The obvious follow-up is which exercise. The honest answer is that the difference between types is smaller than the difference between doing something and doing nothing. A network meta-analysis of 217 trials found Pilates, McKenzie therapy and functional restoration among the more effective approaches, with effects in the same range as each other. The same authors, in the Cochrane review, concluded that differences between types were small. Their own recommendation is the useful one: people with chronic low back pain should be encouraged to perform the exercise that they enjoy, to promote adherence.

Two specific trials are worth knowing because they are unusually clean. In 2023, the RESTORE trial randomised 492 people across twenty Australian physiotherapy clinics to cognitive functional therapy or usual care, and found an improvement of 4.6 points on the Roland-Morris disability scale at thirteen weeks, sustained at a year, at a lower cost to society. Adding movement biofeedback to it changed nothing.

In 2024, the WalkBack trial took 701 adults who had just recovered from an episode and gave half of them a progressive walking programme with education. The walking group went a median of 208 days before an activity-limiting recurrence, against 112 days for the group given nothing, a hazard ratio of 0.72. Walking is not a fashionable intervention. It performed.

Prevention shows the same shape. Pooling 21 trials and 30,850 people, exercise combined with education cut the risk of an episode by 45 percent. Exercise alone cut it by 35 percent. Education alone did nothing at all, with a risk ratio of 1.03. Back belts did nothing. Shoe insoles did nothing. Being told about your back does not protect your back; moving does.

How long this actually takes

You will read that back pain resolves on its own. For a first acute episode that is broadly true. For pain that has already persisted, it is substantially overstated, and the distinction matters if you are trying to decide whether to be patient or to change something.

1. Baseline

   51 out of 100

   Average pain score in cohorts of people whose pain had already persisted, on a 0 to 100 scale.

2. Six weeks

   33 out of 100

   The steepest improvement happens early, and then the curve flattens sharply.

3. Six months

   26 out of 100

   Progress continues, but slowly. This is the stretch where people conclude nothing is working.

4. One year

   23 out of 100

   Low to moderate pain is still present at twelve months in the persistent-pain cohorts.

A separate review of the clinical course found 33 percent of people recovered within three months, and that a year after onset 65 percent still reported pain. Its authors concluded that the assumption of spontaneous recovery in a large majority of patients is not justified. Hold that alongside the acute-pain data, where scores fall from 52 to 6 within a year, and you have the real picture: the reassuring story is true of new pain and unreliable for old pain.

Recurrence is where published numbers diverge most, and any figure quoted without its definition should be treated with suspicion. One inception cohort followed people for twelve months after recovery and found 69 percent had some recurrence, 40 percent had a recurrence that limited activity, and 41 percent sought care, with a median of 139 days before it happened. A systematic review by the same lead author declined to pool the literature at all, judging it too heterogeneous, and noted the single true inception cohort available at the time reported 33 percent. Both numbers are real. They are counting different things.

When it is not this

Everything above concerns non-specific pain, which is most of it. The small remainder matters enormously, and the point of knowing the warning signs is not to worry but to act quickly on the few that count.

The NHS lists six reasons to call 999 or go to A&E with back pain. All six, not the shorter version this article used to print:

• Pain, tingling, weakness or numbness in both legs.
• A loss of feeling around your genitals or anus.
• Changes in your bladder or bowels, such as difficulty peeing, or peeing or pooing yourself.
• Changes in how your penis or vagina feels during sex, not being able to get or keep an erection, or not being able to orgasm.
• Chest pain.
• Back pain that started after a serious accident, such as a car accident.

The sexual-sensation item is the one that gets dropped, and it is a cauda equina sign in its own right: the nerve roots that serve the bladder and bowel serve sexual sensation too, and a change there can arrive before anything else on the list does. Chest pain belongs on it because back pain is a presentation of problems that are not in the back at all. Neither is a detail.

Urgent same-day advice, from NHS 111 or an urgent GP appointment, is warranted if you feel hot, cold, shivery or generally unwell with the pain, or if severe pain starts suddenly or is getting worse quickly. A routine GP appointment is warranted for pain that does not improve after a few weeks of treating it at home, pain that stops your day-to-day activities, weight you have lost without trying to, a lump or swelling in your back or a change in its shape, pain that does not improve with rest or is worse at night, pain that is worse when sneezing, coughing or pooing, or pain coming from the top of your back between the shoulders rather than the lower back.

Now the part that is usually omitted, and that changes how much comfort you should take from a clear screen. A 2016 review found 46 distinct red flags across sixteen national guidelines, covering cancer, fracture, cauda equina syndrome and infection, and reported that evidence for the accuracy of the recommended red flags was lacking. Most are endorsed by consensus, not by diagnostic accuracy data.

Where accuracy has been measured, the pattern is consistent. Pooling seven studies against MRI, the red flags used to identify cauda equina syndrome had a sensitivity between 0.19 and 0.43 and a specificity between 0.62 and 0.88. In plain terms: if a red flag is present it is worth taking seriously and worth investigating promptly, but the absence of red flags does not rule serious pathology out. If something feels wrong in a way this article does not describe, that is a reason to be seen, not a reason to read more.

None of this is medical advice, and none of it is a substitute for an assessment by a clinician who can examine you. It is a summary of what the published literature currently supports, with its uncertainty left in rather than smoothed out.