Your MRI Isn't
Telling You What
You Think It Is

What's Actually Going On

80–97% of Adults Over 40 Would Fail Their Own MRI

High-resolution imaging is extraordinarily sensitive. It detects every structural change your body has made in response to decades of living — and in the absence of actual danger, almost all of those changes are meaningless.

In 1994, Jensen and colleagues published a landmark paper in the New England Journal of Medicine showing 52% of completely asymptomatic people had disc bulges on lumbar MRI. Three decades later, Brinjikji's systematic review of 3,110 pain-free individuals confirmed the pattern holds — and gets more pronounced with age.

This isn't a flaw in imaging technology. It's doing exactly what it's designed to do: detect structural changes. The flaw is in assuming those changes explain the pain.

Asymptomatic MRI Findings by Region

How to Identify It — Imaging Triage

The Clinical Decision Framework

The question isn't "should we scan?" — it's "will this scan result change what we do?" For most MSK presentations, the answer is no. For specific red-flag presentations, imaging is non-negotiable.

ACR Appropriateness Criteria 2021: Advanced imaging medically necessary only after history, physical exam, and failed conservative management — NOT as a first-line diagnostic tool for non-specific MSK pain.

Red Flags

When Imaging Is Urgent or Emergency

This is not an anti-imaging argument. Missing serious pathology is catastrophic. These presentations require immediate imaging — no delay, no conservative trial first.

The Debate

What the Evidence Overturned

Honest Limitations

Why Overuse Persists Despite the Evidence

The data is clear. The behaviour hasn't changed. Here's why.

What Works

Managing the Patient Who Has a Scan Result

Return to Training

When to Continue Training After a Scan Result

A scan result alone is not a reason to stop training. Use these criteria to guide the decision:

• Red flags explicitly ruled out by clinical assessment (not just scan review)
• Patient understands the benign prevalence context of their finding
• Pain during activity ≤3/10 NRS and returns to baseline within 24 hours
• No progressive neurological deficit on assessment
• Loading program commenced — patient tolerating progressive load

Key principle: Load should be guided by symptom response, not radiological appearance. Tissue adapts to load. Avoidance leads to deconditioning → increased pain sensitivity → more pain. The scan did not change the tissue; it just described it.

The Nuance

What the Simple Answer Misses

The argument here is not that imaging is bad. It's that timing and indication determine whether it helps or harms.

There is a specific window where MRI demonstrably changes outcomes: the patient with genuine structural pathology (true disc compression, labral tear in a surgical candidate, stress fracture not visible on X-ray) who has failed a supervised conservative trial. In these cases, imaging does what it's supposed to — it identifies a structural problem that a targeted intervention can fix.

The harm occurs when this tool is used outside its intended window: as a first-line diagnostic response to generic pain, or as a way to validate a patient's suffering rather than guide their treatment.

The other underappreciated nuance is the communication problem. Radiology reports use precise anatomical language that is accurate within its context. "Disc herniation at L4/L5 with thecal sac contact" is a factual description. But to a patient reading it alone on a Tuesday night, it reads as catastrophe. The nocebo harm isn't from the imaging itself — it's from the absence of clinical context around that report.

What would change this protocol: A large multi-center RCT showing immediate MRI for non-specific acute MSK pain (no red flags) produces superior long-term outcomes without inflating surgical rates. Currently no such evidence exists (Chou 2009 systematic review; ACR 2021).

Sources

Key References

• Brinjikji W et al. (2015) AJNR — Systematic review, 3,110 asymptomatic subjects. Degenerative spine findings are ubiquitous: disc degeneration 37–96% across age groups. STRONG
• Jensen MC et al. (1994) NEJM — 52% of asymptomatic subjects had disc bulge; 27% had protrusion on lumbar MRI. Landmark study establishing the pain-pathology disconnect.
• Horga LM et al. (2020) — 97% of asymptomatic knees (median age 44) showed at least one MRI abnormality on 3.0T scan. STRONG
• Culvenor AG et al. (2019) BJSM — Meniscal tears present in 4% under-40s, 19% over-40s in asymptomatic populations. Cartilage defects: 43% over-40.
• Jacobs JC et al. (2020) — N=405,965 VHA Cohort — Early MRI for LBP: surgery 1.48% vs 0.12% (12.7×), opioids 35.1% vs 28.6%, cost $8,082 vs $5,560. No outcome improvement. STRONG
• Sihvonen R et al. (2013) NEJM — FIDELITY trial, N=146 — APM = sham surgery at 12 months for degenerative meniscal tears. No difference in Lysholm score or knee pain. STRONG
• Moseley JB et al. (2002) NEJM — N=180 — Arthroscopic debridement/lavage = sham surgery at 24 months for knee OA. Clinically equivalent. STRONG
• Webster BS et al. (2021) — N=677 — Survey: 59–71% viewed "disc degeneration" and "Modic changes" as serious conditions; 42–57% became fearful of movement from terminology alone. MODERATE
• Darlow B et al. (2013) — Qualitative, N=23 — Pathoanatomical language from clinicians strongly induced fear-avoidance beliefs, worry, and the belief that pain equals real damage.
• ACR Appropriateness Criteria (2021) — Advanced imaging medically necessary only after failed conservative management and clinical evaluation. No routine early imaging for non-specific MSK pain.