In two Norwegian randomized trials, 124 people with chronic low back pain and disc degeneration were assigned to one of two paths: instrumented spinal fusion, or a structured program of cognitive intervention plus exercise. Then researchers followed them for four years. And then for nine. What they found is worth understanding carefully — including what it does not apply to, which matters just as much as what it does.
This is a topic where it’s easy to be irresponsible in either direction. Overstate it and you talk someone out of surgery they genuinely need. Ignore it and you leave people uninformed about a decision that permanently changes their spine. So let’s be precise about scope first, then look at what the data actually shows.
This research is about chronic non-specific low back pain with disc degeneration — pain lasting a year or more without a clear structural emergency. It says nothing about the situations where spine surgery is clearly indicated and sometimes urgent, including:
- Cauda equina syndrome (loss of bowel or bladder control, saddle numbness) — a surgical emergency
- Progressive or significant neurological deficit — worsening weakness, foot drop
- Spinal fracture, tumor, or infection
- Significant instability or spondylolisthesis
- Severe, progressive spinal stenosis unresponsive to care
If any of those apply to you, this article is not about your situation — see a physician promptly. Nothing here should delay appropriate surgical evaluation.
This isn’t our editorial caution — it’s how the study itself was built. The researchers screened 234 patients and excluded 84 of them, specifically ruling out people with radicular pain, spinal stenosis with neurological signs, disc herniation with radiculopathy, previous spinal fracture, previous fusion, inflammatory disease, and generalized disc degeneration. The people who would most clearly benefit from surgery were deliberately kept out of the trial.[1]
The Study
The paper is Brox and colleagues, published in Annals of the Rheumatic Diseases in 2010 — the four-year follow-up of two merged randomized controlled trials conducted in Norway.[1]
Brox et al. (2010) — Study Design
What They Found at Four Years
The mean treatment effect for the primary outcome was 1.1 (95% CI −5.9 to 8.2) by intention-to-treat analysis, and −1.6 (95% CI −8.9 to 5.6) as-treated. Both confidence intervals comfortably span zero.
The authors’ conclusion: in patients with chronic low back pain with and without previous surgery for disc herniation, lumbar fusion was not superior to cognitive intervention and exercises at relieving symptoms, improving function, or returning to work at four years. The results were consistent across both analyses.[1]
Source: Brox JI, Nygaard ØP, Holm I, Keller A, Ingebrigtsen T, Reikerås O. Ann Rheum Dis. 2010;69(9):1643-1648. PubMedAn ODI difference of around 1 point is not a small win for surgery — it’s statistically and clinically indistinguishable from no difference at all. For context, a change of roughly 10–15 points on the ODI is generally considered the threshold for a meaningful improvement. The trial was measuring a gap that, in practical terms, wasn’t there.
Both groups did improve substantially — ODI dropped from roughly 44 to about 29 in the fusion group and 43 to 27 in the conservative group. Back pain fell in both. The question was never whether these patients got better. It was whether surgery got them better more.
It’s the obvious objection, and the authors addressed it directly. The confidence intervals for the treatment effects fell within the 10-point ODI difference the trial was designed to detect — meaning the study was precise enough to rule out the size of benefit it set out to find. Lack of statistical power is unlikely to explain the result.[1]
Worth noting too: patients in both arms believed more strongly in surgery than in non-surgical treatment at baseline (belief scores around 70 for surgery versus roughly 40–45 for conservative care). Expectation was tilted toward the operation. The outcomes still landed in the same place.
Then They Kept Following Them
Four years is a reasonable follow-up. Nine is unusual — and the extended follow-up published in European Spine Journal is where this gets more interesting.[2]
At nine years, the intention-to-treat analysis again found no difference between groups. But one finding in the as-treated analysis deserves attention: more operated patients used pain medication and were out of work.[2]
Read that carefully, because it’s easy to over-interpret. This is an as-treated comparison, not a randomized one — the people who ended up operated may have been sicker or more disabled to begin with, which is exactly why they were operated. It isn’t proof that surgery caused worse outcomes. But it’s also not the pattern you’d expect if fusion were quietly delivering long-term benefits the ODI wasn’t capturing.
At four years: no difference.
At nine years: still no difference.
The Crossover Data Cuts Both Ways
Here’s where honest reading matters, because this study doesn’t say “never have surgery” — and pretending it does would be misrepresenting it.
At four years, 14 patients (24%) assigned to cognitive intervention and exercises had also undergone surgery. Meanwhile, 15 patients (23%) assigned to fusion had undergone re-surgery.[1] By nine years, roughly a third of each group had crossed that line.
Two honest conclusions follow:
- Conservative care isn’t a guaranteed solution. Roughly a quarter of the non-surgical group eventually chose surgery. Some people don’t get where they need to go without it, and that’s a real finding, not a footnote.
- Fusion isn’t a one-and-done fix. The re-operation rate in the surgical group was essentially the same as the crossover rate in the conservative group. The authors noted this directly: the number of re-operations in the surgery group was similar to the number operated in the non-surgical group.
The re-operation figure is worth stating precisely: 25% of those operated needed a further operation — and that rate was identical whether they’d been randomized to surgery or crossed over from the conservative group. The reason given was persistent complaints or deterioration. The authors called the high re-operation rate an argument against surgery, and noted that outcomes after surgery for back pain may be less predictable than after surgery for leg pain.[1]
So the practical framing isn’t “surgery bad, exercise good.” It’s: for this specific population, surgery didn’t produce better outcomes — while carrying surgical risk, recovery time, permanent structural change, and a one-in-four chance of needing another operation.
In the as-treated analysis, patients who had surgery were more likely to be taking pain medication daily or weekly (58% vs 35%; adjusted OR 2.3) and more likely to be on a disability pension (53% vs 32%; adjusted OR 2.5). Those look damning — but in the intention-to-treat analysis, neither difference remained statistically significant (p=0.14 and p=0.21).[1]
The authors offered their own alternative explanations rather than claiming surgery caused the harm: patients on more medication may genuinely hurt more, or may have become habituated to it. And on disability pensions, they suggested a claims adjuster may treat fusion as the end stage of treatment, making a disability claim easier to approve. That’s a systems explanation, not a surgical one — and the fact that the authors raised it themselves is a mark of an honest paper.
One Thing Did Differ — And It’s the Most Interesting Finding in the Paper
The trial measured a long list of secondary outcomes: general function, back pain, leg pain, emotional distress, life satisfaction, work status, patients’ own rating of success. Across essentially all of them, the two groups landed in the same place.
Except one.
The single treatment effect the researchers found in the secondary outcomes was a reduction in fear-avoidance beliefs — favoring cognitive intervention and exercises.
Fear-avoidance beliefs about physical activity: mean treatment effect −3.5 (95% CI −5.8 to −1.1) by intention-to-treat.
Fear-avoidance beliefs about work: −4.3 (95% CI −8.3 to −0.2).
Both favored the non-surgical group, and both held up in the as-treated analysis. The authors noted that the reduction in avoidant behavior seen at one year was still maintained at four.[1]
Fear-avoidance beliefs measure how strongly someone believes that movement and activity will damage their back. High scores predict disability, guarding, and deconditioning — people stop moving because they’re convinced moving is dangerous.
Surgery matched education and exercise on almost everything.
It couldn’t match it on teaching people their back wasn’t fragile.
And that wasn’t incidental — it was the explicit target. The researchers described the aim of the non-surgical intervention as giving patients the understanding that ordinary daily activity would not harm their back, and to build that confidence, patients were deliberately encouraged into physical activities that had previously been discouraged.[1]
That’s the intervention that went toe-to-toe with spinal fusion. Not a machine. Not a passive modality. Progressive, deliberate exposure to movement, paired with a genuine change in what patients believed about their own spine.
The Detail Most Summaries Miss
If you only take one thing from this article, make it this. The comparison arm in these trials was not “wait and see,” and it was not just a stack of exercises. It was a structured program of cognitive intervention plus exercise.
That distinction turns out to be decisive. Compare two trials:
Fritzell et al. (2001) — the Swedish Lumbar Spine Study, 294 patients — found that lumbar fusion did reduce pain and disability more effectively than the control condition at two years. But the control was described as unstructured community care within the primary health system.[3]
Brox (2003, 2006, 2010) and Fairbank (2005, the MRC Spine Stabilisation Trial, 349 patients) compared fusion to structured, intensive rehabilitation programs — and found no clear advantage for surgery.[1],[4]
The pattern: fusion beats doing very little. It does not beat doing something good.
And “something good” here had a specific composition. The cognitive component addressed beliefs, fear, and understanding about the spine — reassurance that the back is not fragile, that movement is safe, that hurt does not equal harm. The exercise component was supervised and progressive. Together. Not education alone, not exercise alone.
What the Authors Themselves Said About Selection
Good researchers state the limits of their own findings, and this discussion section does something worth repeating: the authors explicitly acknowledged that their results do not rule out fusion being appropriate for carefully selected patients. But they paired that with a pointed observation — that widening indications have driven the rise in fusion rates.[1]
Then they raised the problem that makes “careful selection” so difficult in practice:
Despite substantial effort to improve selection criteria, the authors note there is no agreed-upon valid tool to diagnose discogenic pain — and that even discography and MRI are not reliable for selecting patients for this surgery.[1]
That’s a significant statement. It means an MRI showing disc degeneration doesn’t establish that the disc is your pain generator, or that fusing it will help. Degenerative changes are extremely common on imaging in people with no back pain at all.
They did cite work suggesting some predictors of a better surgical result — lower psychological distress and lower disc height predicted functional improvement, while younger age and shorter sick leave predicted return to work.[1] Useful signals, but a long way from a reliable selection tool.
The Wider Evidence Agrees
This isn’t a single surprising trial standing against a consensus. It sits inside one.
A meta-analysis of four RCTs covering 666 patients found the mean ODI reduction favoring fusion was −2.91 (95% CI −6.66 to 0.84) — statistically and clinically insignificant. The authors’ conclusion was unusually direct: there is strong evidence that lumbar fusion is not more effective than conservative treatment in reducing disability from chronic low back pain in patients with degenerative spinal disease, adding that further research would be unlikely to change that conclusion.[5]
An earlier meta-analysis reached the same place, noting that fusion favored only a marginal ODI improvement that was neither statistically significant nor clinically important, that surgery carried a significant risk of complications, and that the cumulative evidence “does not support routine surgical fusion for the treatment of chronic low back pain.”[6]
This has moved into policy. The UK’s National Institute for Health and Care Excellence (NICE) recommends not offering spinal fusion for adults with low back pain except as part of a randomized clinical trial.[7]
Where the Evidence Actually Stands
What This Means If You’re Facing the Decision
-
Ask What Category You’re Actually In
Fusion for instability, spondylolisthesis, fracture, or nerve compromise is a genuinely different question than fusion for chronic non-specific pain with degenerative changes. Ask your surgeon directly which one describes you — it’s the single most important question in the conversation.
-
Ask What “Conservative Care Failed” Actually Meant
The trials suggest the quality of the non-surgical arm determines the outcome. If your “failed conservative care” was a handful of passive treatments and a home exercise sheet, that’s not the intervention that matched fusion in these studies.
-
Understand That the Cognitive Piece Isn’t Optional
The winning comparator wasn’t exercise alone. It paired progressive loading with education that reframed the spine as robust rather than fragile. Programs that skip that half aren’t replicating what the research tested.
-
Weigh the Asymmetry
Structured rehab that doesn’t work leaves your options open. Fusion that doesn’t work leaves you with a permanently altered spine and a roughly one-in-four chance of another operation. That asymmetry is a legitimate part of the decision.
We’re not anti-surgery, and we’d be doing you a disservice to pretend spine surgery is never the right answer — sometimes it clearly is, and part of our job is recognizing when to refer. What this research shapes is the order of operations for chronic non-specific back pain: a genuinely good conservative program deserves a real attempt before an irreversible one, because the evidence says it performs comparably without the surgical risk.
It also tells us what that program has to include. Assessment to find the actual driver. Progressive loading, not passive treatment. And honest education that stops treating your spine as fragile. That’s not a consolation prize while you wait for surgery — in these trials, it was the treatment that matched surgery.
The Bottom Line
Across two randomized trials and 124 patients followed for four years and then nine, instrumented lumbar fusion was not superior to cognitive intervention plus exercise for chronic low back pain with disc degeneration. Multiple meta-analyses agree, and NICE now advises against fusion for low back pain outside a trial. Roughly a quarter of patients in each arm ended up having surgery or re-surgery, so neither path is a guarantee.
And the one measure where the groups genuinely separated is the one worth remembering: the conservative group ended up less afraid to move. Surgery could match education and exercise on pain and disability. It couldn’t match it on changing what people believed their spine could tolerate — and four years on, that change was still holding.
None of that means surgery is never right. It means that for this particular problem, a well-built conservative program is not the lesser option — it’s the comparable one, and it doesn’t cost you a permanently fused segment to find out.
Considering Surgery for Chronic Back Pain?
Before an irreversible decision, it’s worth knowing whether you’ve had a genuinely thorough conservative attempt — assessment, progressive loading, and honest education. Let’s find out what’s actually driving your pain.
Book an Assessment ?References
- Brox JI, Nygaard ØP, Holm I, Keller A, Ingebrigtsen T, Reikerås O. Four-year follow-up of surgical versus non-surgical therapy for chronic low back pain. Annals of the Rheumatic Diseases. 2010;69(9):1643-1648. doi:10.1136/ard.2009.108902. (124 patients; mean treatment effect 1.1, 95% CI −5.9 to 8.2 ITT; 24% of the conservative group underwent surgery, 23% of the fusion group underwent re-surgery.) PubMed · PMC
- Froholdt A, Holm I, Keller A, Gunderson RB, Reikerås O, Brox JI. No difference in 9-year outcome in CLBP patients randomized to lumbar fusion versus cognitive intervention and exercises. European Spine Journal. 2012. (Mean adjusted ODI treatment effect 1.9, 95% CI −7.8 to 11.6; one-third crossover and one-third re-operation; as-treated, more operated patients used pain medication and were out of work.) European Spine Journal
- Fritzell P, Hägg O, Wessberg P, Nordwall A. 2001 Volvo Award Winner: Lumbar fusion versus nonsurgical treatment for chronic low back pain — a multicenter randomized controlled trial from the Swedish Lumbar Spine Study Group. Spine. 2001. (294 patients; fusion superior to unstructured community care at 2 years.)
- Fairbank J, Frost H, Wilson-MacDonald J, et al. Randomised controlled trial to compare surgical stabilisation of the lumbar spine with an intensive rehabilitation programme for patients with chronic low back pain: the MRC Spine Stabilisation Trial. BMJ. 2005;330:1233. (349 patients; no clear evidence fusion was more beneficial than intensive rehabilitation.)
- Lumbar fusion compared with conservative treatment in patients with chronic low back pain: a meta-analysis. International Journal of Rehabilitation Research. 2014. (666 patients across four RCTs; mean ODI reduction −2.91, 95% CI −6.66 to 0.84; concluded strong evidence that fusion is not more effective than conservative treatment.) NCBI
- Ibrahim T, Tleyjeh IM, Gabbar O. Surgical versus non-surgical treatment of chronic low back pain: a meta-analysis of randomised trials. International Orthopaedics. 2008. (Marginal, non-significant ODI difference of minimal clinical importance; significant complication risk; cumulative evidence does not support routine surgical fusion for chronic low back pain.) PubMed
- National Institute for Health and Care Excellence (NICE). Low back pain and sciatica in over 16s: assessment and management (NG59). 2016. (Recommends not offering spinal fusion for adults with low back pain except as part of a randomized controlled trial.) NCBI Bookshelf
- Mannion AF, Brox JI, Fairbank JC. Comparison of spinal fusion and nonoperative treatment in patients with chronic low back pain: long-term follow-up of three randomized controlled trials. The Spine Journal. 2013;13(11):1438-1448.
This article is for general educational purposes and does not constitute medical advice, and it is not a recommendation for or against surgery in any individual case. Surgical decisions should be made with a qualified spine specialist who knows your imaging, examination, and history. Seek prompt medical care for any red-flag symptoms described above.
