Axial Spondyloarthritis

Your back pain might be inflammatory, not mechanical, and there's a screen that takes two minutes.

If you have chronic back pain that started before age 45 and your morning stiffness lasts over 30 minutes, ask your physio or GP to screen you for inflammatory back pain — not just mechanical back pain. That two-minute conversation is the single biggest move against the 6.7-year diagnostic delay.

Here's what's really happening: this is an immune-mediated disease that targets the joints of the spine and pelvis, not a mechanical problem.

What most people get wrong: a normal HLA-B27 or a normal CRP does not rule it out, especially in the non-radiographic form.

Start here: ask for inflammatory back pain screening and a rheumatology referral if you have three or more of the pattern features.

Mechanical back pain is like a strained hinge that creaks when you load it and rests quietly. Inflammatory back pain is the opposite: the hinge swells when it sits still and loosens up when you move it. If your back hurts more from sitting and gets better from walking, your body is telling you it's a different mechanism — and a different treatment pathway.

Red Flags — When to Refer

Safety first. If any of these are present, the referral is more important than any exercise prescription.

Three or more inflammatory back pain features in a patient under 45 with chronic back pain

Morning stiffness over 30 minutes, improvement with activity not rest, alternating buttock pain, second-half-of-night pain, NSAID response, prior anterior uveitis, psoriasis, IBD, dactylitis, enthesitis, family history of any SpA-spectrum condition. → Refer to GP or rheumatology with the feature list documented.

Anterior uveitis history, psoriasis, or inflammatory bowel disease + chronic back pain

Refer regardless of how many other features are present. → GP or rheumatology.

Flat response to a competent 6-week mechanical-style PT trial in a back-pain client under 45 with any IBP feature

Re-screen for inflammatory back pain. → Refer with documented feature list.

New eye pain, redness, photophobia, or vision changes in a known axSpA patient

Anterior uveitis is sight-threatening if untreated. → Same-day ophthalmology referral.

Severe new back pain after low-energy trauma in established ankylosing spondylitis

Ankylosed spine is brittle. Transdiscal fractures can be missed on plain film. → Emergency department for CT imaging.

New persistent fever, productive cough, or atypical infection symptoms on a biologic

Real-world serious infection rate is 1.68 per 100 patient-years on these drugs. → Same-day GP or rheumatology contact; pause physiotherapy until cleared.

New neurological deficit (cauda equina pattern, progressive weakness, sensory loss)

→ Emergency referral.

The Verdict

Your back pain might be inflammatory, not mechanical, and the screen for it takes two minutes.

Mechanical back pain is like a strained hinge that creaks when you load it and rests quietly. Inflammatory back pain is the opposite: the hinge swells when it sits still and loosens up when you move it. If your back hurts more from sitting still and gets better when you walk around, your body is telling you it's a different mechanism — and a different treatment pathway.

Here's what's really happening:This is an immune-mediated disease that targets the joints of the spine and pelvis. It is not a mechanical problem and it does not respond to mechanical fixes.
What most people get wrong:A normal HLA-B27 blood test or a normal CRP does not rule it out, especially in the non-radiographic form — sacroiliac inflammation can be present with clean blood markers.
Start here:Ask for inflammatory back pain screening and a rheumatology referral if you have three or more of the pattern features above.

Best for

Anyone under 45 with chronic back pain whose stiffness lives in the morning rather than the evening, and improves with movement rather than rest.

Skip if

Your back pain has a clear single-event onset (you lifted something and felt it pop) and is improving predictably with mechanical PT.

Want the full evidence? Keep scrolling.⌄

What's Actually Going On

Cinematic anatomy of the sacroiliac joint and lumbar spine

Axial spondyloarthritis is an HLA-B27-associated immune-mediated inflammatory disease centred on the sacroiliac joints and the spine. The dominant tissue lesion is enthesitis — inflammation where tendons and ligaments attach to bone — rather than the joint-lining synovitis you see in rheumatoid arthritis. The IL-23/IL-17 axis and TNF-α are the validated drug targets, which is why the biologic drugs in this disease are TNF inhibitors, IL-17 inhibitors, and JAK inhibitors.

Two forms sit under the umbrella. Non-radiographic axSpA shows active sacroiliac inflammation on MRI but the plain X-rays are still clean. Ankylosing spondylitis is the radiographic form — plain-film sacroiliitis is established and over years can progress to spinal syndesmophyte formation and ankylosis (fusion). The ASAS classification framework introduced in 2009 is what made early non-radiographic diagnosis possible — before that, patients waited for plain-film changes that take years.

Extra-articular manifestations share the immune biology. Anterior uveitis, psoriasis, and inflammatory bowel disease cluster in the same patients and the same families. That clustering is part of the recognition pattern, not a coincidence.

How to Identify It

Cinematic anatomy of the spinal assessment focus

Recognition is done from the history, not from provocative tests. No bedside MSK examination cluster has published sensitivity and specificity for distinguishing inflammatory back pain from mechanical chronic back pain. BASMI components measure spinal mobility for outcome tracking — they do not diagnose.

The pattern that matters is in the patient's own words.

Insidious onset of back pain before age 45 (no single inciting event).
Symptom duration over 3 months.
Morning stiffness lasting over 30 minutes.
Improvement with activity, not rest.
Alternating buttock pain (switches sides over weeks).
Second-half-of-night pain that wakes them and is relieved by getting up.
Clinically meaningful response to NSAIDs.
Personal or first-degree family history of uveitis, psoriasis, IBD, dactylitis, enthesitis, or any SpA-spectrum condition.

NICE NG65 codifies the referral rule. Chronic back pain over 3 months with onset before age 45 plus three or more SpA features earns HLA-B27 testing through the GP and rheumatology referral. One feature plus uveitis or IBD history also earns referral. The MSK clinician's job is to populate the feature count, not to diagnose.

Note on the tests that exist:

Modified Schober (lumbar flexion ROM): used for outcome tracking. Sn/Sp: data unavailable for axSpA diagnosis

Sacroiliac provocation cluster (FABER, compression, distraction, thigh thrust): designed for mechanical SIJ pain, not inflammatory disease. Sn/Sp: data unavailable for axSpA diagnosis

The Debate

Where the field is moving — older guidance versus what recent RCTs have shown.

Older view

Diagnosis required radiographic sacroiliitis (modified New York criteria, 1984). Patients waited years for plain-film changes.

Current view

ASAS criteria (2009) allow non-radiographic diagnosis from the first MRI-positive presentation. Non-radiographic axSpA is now a treatable diagnosis from day one.

Use the ASAS framework. Early diagnosis is the field's direction.

Older framing

TNF inhibitors framed as a uniform class — drug choice was largely interchangeable for spinal disease.

Recent NMA

Bechman 2024 (44 RCTs): anti-TNF monoclonal antibodies (adalimumab, infliximab, golimumab, certolizumab) outperform etanercept for anterior uveitis prevention specifically.

Class effects exist for spinal disease activity; for uveitis prevention, prefer anti-TNF mAbs over etanercept.

Older framing

Stretching and passive modalities as core therapy — heat, ultrasound, massage as standard treatment.

APTA CPG

Supervised exercise conditionally preferred over passive modalities. Mind-body exercise (Tai Chi, Pilates, Yoga, Qigong) produces comparable function, pain, and disease-activity changes (Wang J 2024, GRADE moderate-to-low).

Supervised exercise is the Tier 1 conservative pillar. Mind-body is a real alternative, not a consolation prize.

Honest Limitations

What the research can and can't tell us about real-world axSpA care.

Diagnostic delay is the field's failure to translate

Research: Zhao 2021 meta-analysis of 64 studies puts mean diagnostic delay at 6.7 years (95% CI 6.2–7.2) with no improvement when stratified by publication year.

Real-world: Despite a codified NICE referral rule, screening for inflammatory back pain features is not routinely performed in primary care or MSK physiotherapy.

Adjustment: Every chronic back pain client under 45 gets two minutes on the IBP feature screen. The single highest-leverage MSK-clinician action against the delay.

Exercise meta-analyses pool heterogeneous interventions

Research: Zhang 2025 reports moderate improvements but with high heterogeneity for BASMI (I²=72%) and pain (I²=66%), pooling supervised + home, land + aquatic, and mixed-modality designs.

Real-world: Pooled magnitudes are direction-correct but not precise effect sizes for any specific protocol. Adherence to home programs collapses over 6–12 months without a maintenance structure.

Adjustment: Use the meta-analytic numbers as a floor expectation, not a guarantee. Build a maintenance structure into the initial prescription, not as an afterthought.

BASDAI / BASFI are noisy in patients with comorbid pain conditions

Research: Son 2022 (7 comparative studies) shows axSpA + fibromyalgia patients have significantly higher BAS scores than axSpA without fibromyalgia, with no difference in inflammatory markers.

Real-world: Rising BASDAI in a patient with comorbid fibromyalgia does not necessarily mean uncontrolled axSpA — it may mean fibromyalgia is amplifying patient-reported scores.

Adjustment: When patient-reported scores climb but objective markers (CRP, ASDAS-CRP, SIJ MRI) are stable, flag the fibromyalgia overlap to rheumatology rather than escalating physical therapy or biologic re-evaluation by default.

The Nuance

Cinematic anatomy of the spine in long-term axial spondyloarthritis

AxSpA is a 30–50-year disease. Three nuance points matter most.

The biologic-versus-exercise question is a false binary. Modern axSpA care is biologic or NSAID therapy PLUS supervised exercise PLUS lifestyle. Neither replaces the other. Exercise reduces disease-activity scores and improves function in patients who are already managed pharmacologically. It does not substitute for biologic therapy in active disease, and it does not slow radiographic progression in any clinically meaningful way that has been demonstrated. The pharmacological pipeline transformed long-term outcomes; the exercise programme compounds them.

Surgery is rare and specialist. Surgery enters the conversation for late structural complications: severe spinal kyphotic deformity correction (three-column osteotomy in low-volume centres), hip arthroplasty in coxitis, and urgent spinal stabilization after fracture in an ankylosed spine. None of these are part of MSK physical therapy's decision tree.

De-escalation is a rheumatology decision, not a physio decision. TNFi tapering preserves response better than withdrawal (taper RR 1.45 vs withdrawal RR 2.28, Lawson 2021 + Uhrenholt 2022). If a known axSpA client's BASDAI and BASFI start climbing post-taper, that is a rheumatology message — not "do more exercise."

Cardiovascular comorbidity is part of long-term care. LV diastolic dysfunction OR is 3.43 in axSpA (Romand 2022, 28-study meta-analysis). Higher BMI tracks higher disease activity. Cardiovascular risk modification — BP, lipids, glucose, smoking, BMI — is a standard part of the long-term plan and the supervised aerobic exercise component is itself CV-protective.

Differentials at a Glance

Cinematic anatomy of the spinal differential focus

The most common differentials that share back-pain presentation with axSpA but route to different care:

Mechanical chronic low back pain — pain worse with activity, better with rest; no morning stiffness over 30 min; no EAM cluster.

Fibromyalgia — widespread pain, sleep disturbance, fatigue, cognitive symptoms; female predominance; non-inflammatory pattern. Note: can co-exist with axSpA in about 10–15% of patients.

Rheumatoid arthritis with cervical involvement — symmetrical peripheral synovitis dominant; positive RF or anti-CCP.

Psoriatic arthritis (axial form) — same SpA spectrum; refer rheumatology regardless.

Diffuse idiopathic skeletal hyperostosis (DISH) — older age, flowing anterior ossifications without sacroiliitis on imaging.

Spinal malignancy / metastatic disease — night pain not relieved by movement, weight loss, prior cancer history, age over 50 with new back pain.

Sources

Zhao SS et al. 2021 — Diagnostic delay in axSpA: meta-analysis of 64 studies; pooled mean delay 6.7 years (95% CI 6.2–7.2). Rheumatology (Oxford). PMID 33428758.

de Winter JJ et al. 2016 — EAM prevalence AS vs nr-axSpA meta-analysis (N=2,236 AS, N=1,242 nr-axSpA): uveitis 23% AS vs 16% nr-axSpA; HLA-B27 ~78% both. Arthritis Res Ther. PMID 27586785.

Zhang M et al. 2025 — Exercise therapy in axSpA: meta-analysis of 20 RCTs N=1,670; BASDAI WMD −0.78, BASFI WMD −0.49, ASDAS WMD −0.44, VO2peak +3.2. Arch Phys Med Rehabil. PMID 38942347.

Wang J et al. 2024 — Mind-body exercise for axSpA: 15 studies, GRADE-assessed moderate-to-low; BASFI MD −0.76, BASDAI MD −0.76; no serious AEs. J Orthop Surg Res. PMID 39342350.

Bechman K et al. 2024 — Uveitis incidence with biologics in axSpA: network meta-analysis of 44 RCTs; anti-TNF mAb best, etanercept worst. Arthritis Rheumatol. PMID 38116697.

Lawson DO et al. 2021 — TNFi dose reduction: 6 RCTs N=747; reduced dose loses response (RR 0.62 ASAS40); more flares (RR 1.73). Arthritis Care Res. PMID 32166872.

Uhrenholt L et al. 2022 — b/tsDMARD tapering vs withdrawal flare risk: 22 RCTs; taper RR 1.45 vs withdrawal RR 2.28. Rheumatology (Oxford). PMID 34864896.

Tian H et al. 2023 — CRP and MRI-detected inflammation: 18 studies N=1,392; CRP correlates with spinal MRI but NOT SIJ MRI. Clin Rheumatol. PMID 37336841.

Son SM et al. 2022 — Fibromyalgia in axSpA: 7 comparative studies; concomitant fibromyalgia inflates BAS scores. J Clin Rheumatol. PMID 33298817.

Romand X et al. 2022 — Echocardiography in axSpA: 28 studies; LV diastolic dysfunction OR 3.43. Joint Bone Spine. PMID 35259478.

Delcourt C et al. 2025 — Treatments and fatigue in axSpA: 34 studies; physical activity + anti-TNF reduces fatigue more than anti-TNF alone. Rheumatology (Oxford). PMID 39388256.

Verhoeven F et al. 2019 — Aerobic exercise vs standard PT: 6 RCTs; aerobic improves within-group but not vs supervised PT between-group. Int J Rheum Dis. PMID 30187695.

NICE 2017 (reviewed 4 March 2025) — Spondyloarthritis in over 16s: diagnosis and management (NG65). [cite-unverified — preflight-sourced NICE guidance, not PubMed-indexed]

APTA / ACR / SAA / SPARTAN — Recommendations for the Treatment of AS and Non-Radiographic axSpA. [cite-unverified — preflight-sourced CPG summary]

Sieper J et al. — Referral strategy for early recognition of axial spondyloarthritis. [cite-unverified — preflight-sourced landmark referral framework]