Your MRI Is Lying to You | The Pained Brain, Chapter 3

Low back pain imaging · MRI

Imaging for Low Back Pain and the Metabolic Workup the MRI Skips

Early imaging does not help uncomplicated low back pain: six randomized trials pooling 1,804 patients found no gain in pain or function, and the ACR reserves imaging for suspected cauda equina syndrome, cancer, fracture or infection, or for failure of six weeks of treatment.

Guideline-concordant imaging for low back pain is well defined and poorly followed. The larger gap is what goes unordered: the metabolic and vascular data that separate a symptomatic finding from an incidental one.

Imaging for low back pain has one of the clearest evidence bases in primary care and one of the weakest adherence records. The less discussed problem is substitution: the MRI answers a structural question the patient rarely needed answered, while the metabolic and vascular data that distinguish a symptomatic finding from an incidental one go unmeasured.

The argument is laid out in Dr. Padda’s video Your MRI Is Lying to You, drawn from Chapter 3 of The Pained Brain (Dr. Gurpreet Singh Padda, MD, MBA, MHP, and Dr. KrisJay Fucanan, MD). For a practice considering Measura [Cardiometabolic and Autonomic Health Analysis], the relevant questions are narrower: which back and joint pain patients warrant a terrain workup, what a positive finding changes, and how to make the order reproducible.

What are the guidelines for imaging for low back pain?

Six randomized trials pooling 1,804 patients without indications of serious disease found that immediate imaging did not improve pain or function; the long-term pain effect was -0.04. The ACR Appropriateness Criteria describe uncomplicated acute low back pain as not warranting imaging, reserving it for suspected cauda equina syndrome, malignancy, fracture or infection, or for failure of up to six weeks of medical management and physical therapy.

Practice diverges in both directions. Across 45 studies, 24.8% of primary care and 35.6% of emergency department back pain visits ended in imaging, and the CT and MRI share rose 53.5% between 1995 and 2015. Roughly a third of lumbar referrals lacked a red flag, while 65.6% of patients who had one were never imaged. The flags themselves perform poorly: among 1,172 consecutive primary care patients, 0.9% had serious pathology and 80.4% carried at least one flag. The Cochrane review of flags for vertebral fracture found most not useful, with age over 70 the exception at a positive likelihood ratio of 11.19.

Does an early MRI for back pain lead to more procedures?

Nonadherent imaging shows up as downstream procedures. In 3,022 workers’ compensation claims, less severely injured workers with an early MRI had a relative risk of 27.40 for injections and 28.35 for surgery against the no-MRI referent, and comparing early with timely MRI inside the same severity band still left a surgical relative risk of 4.38. Among 405,965 matched veteran episodes without red flags, a setting where fee-for-service incentive is largely absent, lumbar surgery within a year ran 1.48% after an inappropriate early MRI against 0.12%, and prescription opioids 35.1% against 28.6%. The last recorded pain score was marginally higher in the scanned group, 3.99 against 3.87.

Report language compounds the effect. A stepped-wedge trial that inserted age-specific prevalence benchmarks into 238,886 spine imaging reports moved twelve-month spine-related utilization by -0.7%, a null result. In a small randomized trial, patients given a descriptive finding-by-finding account of their MRI catastrophized more and functioned worse at six weeks than patients given a clinical explanation. A printed sentence of context does not neutralize the image; the explanation delivered by the clinician is the lever with trial evidence behind it.

Why do some spine MRI findings hurt and others do not?

The most direct case for a terrain workup comes from the Wakayama Spine Study. Among 451 community residents with moderate radiographic canal stenosis on whole-spine MRI, symptomatic and asymptomatic residents did not differ by age, sex, smoking or body mass index (23.7 against 23.1). Diabetes separated them at an odds ratio of 3.92, present in 25.0% of the symptomatic group against 6.4%, and a low ankle-brachial index separated them at 1.36 per standard deviation. In severe stenosis nothing separated the groups. The symptomatic cell is small and the design is cross-sectional, but the selection implication is clear: at intermediate structural severity, the discriminating data sit in the blood and the arteries.

The degenerative findings track the same terrain. Pooled Mendelian randomization estimates put genetically predicted body mass index and waist circumference at 1.26 each for disc degeneration, triglycerides at 1.08 and type 2 diabetes at 1.05. Observationally, diabetes carried 1.68 times the odds across 2,881,170 adults, settling at 1.47 once code-based database studies were removed. In Framingham, aortic calcification anterior to a vertebral level predicted disc deterioration at that level over 25 years at an odds ratio of 1.5, consistent with a disc that depends on diffusion from segmental vessels. Outside the spine, metabolic syndrome carried a relative risk of 1.72 for enlarging medial bone marrow lesions in the knee, and diabetes carried odds of 3.69 for frozen shoulder and 2.24 for rotator cuff tendinopathy.

Which back pain patients need a metabolic workup, and which tests?

A workable selection frame for pain, primary care and geriatrics panels: back or joint pain where imaging severity and symptoms are discordant; moderate stenosis with neurogenic or vascular claudication in the differential; pain patients with central adiposity, known dysglycemia or dyslipidemia; and any patient whose workup has been imaging-heavy and laboratory-light. The criteria are summarized in selection criteria.

Laboratory panels: HbA1c, fasting glucose, triglycerides and HDL cholesterol, the variables with the most direct support above.
Ankle-brachial index: the vascular variable that separated symptomatic from asymptomatic moderate stenosis in Wakayama.
Bioimpedance body composition: in 223 knee replacement candidates, bioimpedance fat mass correlated with lower pressure pain thresholds at every site, including the forehead, while body mass index correlated with none.
Arterial stiffness and endothelial function: a functional read on the vascular bed that avascular tissue depends on.

What a finding changes in management

A metabolic or vascular finding does not replace imaging when a red flag is present, and Measura performs no imaging. It changes three things. Sequencing: history, examination and laboratory data come first, the image is read last and against all three, and a needle placed for a reason the examination supplied is framed as a bridge that buys time for terrain repair rather than as the plan. Explanation: the patient hears that the report describes shape while the measured values describe risk that can be tracked. Follow-up: a repeatable baseline turns a lifestyle recommendation into a monitored intervention.

Sleep belongs in the same conversation, because it is a behavioral input with a measurable effect on the pain threshold. Two nights of fragmented sleep lowered heat pain thresholds in healthy adults, with 34.9% of the effect mediated by monocyte inflammatory output. Work schedules and caregiving load drive that input as much as any clinical variable does, and none of it appears on a spine report.

Documentation and workflow

The workup is reproducible only when it is protocolized rather than remembered. A standing order keyed to a discordant-imaging or symptomatic-stenosis trigger removes the dependence on individual recall. The same data serve the annual wellness visit, where cardiovascular risk factors are already documented, and quality programs such as MIPS and HEDIS where cardiometabolic measures apply. For older patients, in whom age over 70 is the flag that carries weight for vertebral fracture, file results where the fall-prevention and memory workup already lives, as described in cognitive assessment and fall prevention. Reports return to the ordering physician, and interpreting the report covers how to read them. The evidence behind each figure is in the book companion for Chapter 3.

Frequently asked questions

Does a metabolic workup replace imaging for low back pain?

No. Suspected cauda equina syndrome, malignancy, fracture or infection, and failure of about six weeks of conservative care, remain indications for imaging, and Measura performs no imaging. The workup answers a different question: why a finding common in asymptomatic adults is symptomatic in this patient. In moderate stenosis, diabetes and a low ankle-brachial index answered it when canal dimensions did not. Clinical rationale.

Which laboratory values have the most support in degenerative spine and joint pain?

Diabetes status or HbA1c, triglycerides, HDL cholesterol and waist circumference carry the most direct evidence. Diabetes separated symptomatic from asymptomatic moderate stenosis at an odds ratio of 3.92, and genetically predicted triglycerides and waist circumference raised the odds of disc degeneration. Fasting insulin is the upstream variable that glucose-based values miss, which is the subject of the companion piece on hyperinsulinemia. Screening for hyperinsulinemia.

How should imaging findings be explained to the patient?

Clinically rather than descriptively. In a small randomized trial, a factual finding-by-finding explanation produced more catastrophizing and worse function at six weeks than a clinical explanation, and printing prevalence benchmarks into reports did not change utilization. Pairing the report with measured, repeatable values gives the patient something concrete to act on. A patient-facing version of this material is available. What degenerative disc disease means.

Can the results support the annual wellness visit?

Yes, as documentation of cardiovascular risk factors the visit already addresses. Laboratory values, body composition and ankle-brachial index results can be recorded within the same encounter structure, and repeat measurements at later visits show whether the terrain is changing. The integration steps, including how results reach the chart, are described on the practice pages. Annual wellness visit integration.

Where do quality measures fit?

MIPS and HEDIS include cardiometabolic measures that the same laboratory and vascular data can support as documentation. The clinical rationale should lead: test the patients whose imaging and symptoms are discordant or whose pain sits on a metabolic background, and let documentation follow from work that was clinically indicated rather than the reverse. Quality measures that cardiometabolic testing supports.

When should you get an MRI for lower back pain?

When a red flag points to serious disease: suspected cauda equina syndrome, cancer, fracture or infection. Imaging also fits when up to six weeks of medical care and physical therapy have not helped. Outside those cases, six randomized trials pooling 1,804 patients found that scanning early did not improve pain or function, and an early MRI was followed by more injections, surgery and opioids.

What is a red flag in low back pain?

A red flag is a sign that back pain may come from serious disease, such as cauda equina syndrome, cancer, fracture or infection. On their own, the flags sort patients poorly: among 1,172 primary care patients, 80.4% carried at least one flag, yet only 0.9% had serious pathology. For vertebral fracture, age over 70 is the flag that carries real weight.

See how the protocol fits your workflow

Learn how Measura testing is added to a practice through standing orders, report interpretation and chart integration.

Send a message

4477 Woodson Rd, Suite 201, St. Louis, MO 63134. Monday to Friday, 9:00 a.m. to 5:00 p.m. Please do not send symptoms, diagnoses or images through a web form — a website form is not a secure medical channel. Send your name and number and we will call you back.

References

Chou, R., Fu, R., Carrino, J. A., & Deyo, R. A. (2009). Imaging strategies for low-back pain: systematic review and meta-analysis. Lancet, 373(9662), 463–472. https://doi.org/10.1016/S0140-6736(09)60172-0
Downie, A., Hancock, M., Jenkins, H., Buchbinder, R., Harris, I., Underwood, M., Goergen, S., & Maher, C. G. (2020). How common is imaging for low back pain in primary and emergency care? Systematic review and meta-analysis of over 4 million imaging requests across 21 years. British Journal of Sports Medicine, 54(11), 642–651. https://doi.org/10.1136/bjsports-2018-100087
Jenkins, H. J., Downie, A. S., Maher, C. G., Moloney, N. A., Magnussen, J. S., & Hancock, M. J. (2018). Imaging for low back pain: is clinical use consistent with guidelines? A systematic review and meta-analysis. The Spine Journal, 18(12), 2266–2277. https://doi.org/10.1016/j.spinee.2018.05.004
Webster, B. S., Choi, Y., Bauer, A. Z., Cifuentes, M., & Pransky, G. (2014). The cascade of medical services and associated longitudinal costs due to nonadherent magnetic resonance imaging for low back pain. Spine, 39(17), 1433–1440. https://doi.org/10.1097/BRS.0000000000000408
Jacobs, J. C., Jarvik, J. G., Chou, R., Boothroyd, D., Lo, J., Nevedal, A., & Barnett, P. G. (2020). Observational Study of the Downstream Consequences of Inappropriate MRI of the Lumbar Spine. Journal of General Internal Medicine, 35(12), 3605–3612. https://doi.org/10.1007/s11606-020-06181-7
Henschke, N., Maher, C. G., Refshauge, K. M., Herbert, R. D., Cumming, R. G., Bleasel, J., York, J., Das, A., & McAuley, J. H. (2009). Prevalence of and screening for serious spinal pathology in patients presenting to primary care settings with acute low back pain. Arthritis and Rheumatism, 60(10), 3072–3080. https://doi.org/10.1002/art.24853
Maeda, T., Hashizume, H., Yoshimura, N., Oka, H., Ishimoto, Y., Nagata, K., Takami, M., Tsutsui, S., Iwasaki, H., Minamide, A., Nakagawa, Y., Yukawa, Y., Muraki, S., Tanaka, S., Yamada, H., & Yoshida, M. (2018). Factors associated with lumbar spinal stenosis in a large-scale, population-based cohort: The Wakayama Spine Study. PLoS One, 13(7), e0200208. https://doi.org/10.1371/journal.pone.0200208
Wang, X., Gao, Z., Chen, K., Huang, C., & Li, Y. (2024). Diabetes Mellitus and Intervertebral Disc Degeneration: A Meta-Analysis. World Neurosurgery, 188, e81–e92. https://doi.org/10.1016/j.wneu.2024.05.043
Kauppila, L. I., McAlindon, T., Evans, S., Wilson, P. W., Kiel, D., & Felson, D. T. (1997). Disc degeneration/back pain and calcification of the abdominal aorta. A 25-year follow-up study in Framingham. Spine, 22(14), 1642-1647; discussion 1648-1649. https://doi.org/10.1097/00007632-199707150-00023
Meert, L., Vervullens, S., Heusdens, C. H. W., Smeets, R. J. E. M., Meeus, M., & Mertens, M. G. C. A. M. (2024). Unravelling relationships between obesity, diabetes, and factors related to somatosensory functioning in knee osteoarthritis patients. Clinical Rheumatology, 43(8), 2637–2645. https://doi.org/10.1007/s10067-024-07022-2