RightRep EN · PT Open the app
BLOG

Can You Train With Degenerative Disc Disease? The Worn-Disc Door Audit: The MRI-Report Door the Spine Family Never Opened, Written for the Reader Who Holds a Radiology Sentence and Two Opposite Internet Scripts — Where the Asymptomatic Ledger Prices Disc Degeneration at 37% of Pain-Free 20-Year-Olds Rising to 96% of Pain-Free 80-Year-Olds Across 33 Studies and 3,110 Silent Spines, Where the Twin Spine Study's Loading-Discordant Twins Found Surprisingly Little Occupational Effect on Degeneration While Routine Loading Showed Indications of Benefit, Where the 2026 Meta-Analysis (39 Studies, 4,152 Participants) Found Running the Only Physical Loading Associated With Better Disc Health (g 0.31) With Every Other Loading Subgroup Null, Where the First Randomized Running Trial in Chronic Low Back Pain Cut Pain 15.3 Points With Zero Attrition and Its Pre-Planned Disc Secondary Showing the Most Degenerated Spines Gaining the Most Hydration — and Where the Surgery Column Is Priced on Both Sides: Fritzell's 294-Patient Fusion File at 33%-vs-7% Pain Against Fairbank's BMJ Verdict of No Clear Evidence Over Intensive Rehab

By RightRep AI Assistant · · 23 min read

Degenerative disc disease is the file where the gym's two default scripts — protect the worn disc, stop loading it and it's just aging, push through everything — both fail the reader at the exact door they walked in through, because this reader usually arrives holding a radiology sentence, not a diagnosis: an MRI report that says "degenerative disc disease," "multilevel disc desiccation," or "Modic changes," and a body that may or may not agree with the paper. The protect-it script dies on the asymptomatic ledger — across 33 studies and 3,110 pain-free spines, disc degeneration shows up on imaging in 37% of asymptomatic 20-year-olds and 96% of asymptomatic 80-year-olds [1], on twins whose lifetime loading histories barely moved their discs at all [2], and on a 2026 meta-analysis where running — actual repetitive loading — was the only physical exposure associated with better disc health [8]. The push-through script dies on the same file's fences: Modic changes that do track pain severity and duration [5], red-flag columns every spine door in this library carries, and a surgery literature this reader may already be consulting [14][15]. So the fences come first. First: the red-flag column — loss of bladder or bowel control, saddle numbness, progressive leg weakness, fever with new back pain, or unrelenting night pain with weight loss — is the emergency department today, not a programming question (the cauda equina fence and the cord door own the details). Second: new leg-dominant pain, numbness or true weakness with the back pain routes through the sibling doors — the herniated disc file if a fresh protrusion fits, the sciatica file if the nerve is talking — because the acute disc and the worn disc are cousins, not twins. Third: the surgical consult, if one is on the calendar, is a decision to walk into with the tables from both randomized files in hand — this door prices that column honestly rather than pretending it doesn't exist [14][15]. And fourth: this file assumes the workup has been done — degenerative disc disease is a description of imaging, not an explanation of pain, and the doctor's job before the coach's job is to make sure nothing else is wearing the costume [1]. With the fences standing, the evidence underneath is unusually calm for how loudly the internet argues about it: the disc is a living, mechanosensitive tissue; most loading does nothing dramatic to it either way; one specific kind of loading is associated with better discs; and the programs with the strongest pain evidence look exactly like ordinary training.

The short answer: yes — the disc wants load, and it specifically likes the running kind

Strip the file to its spine. Degenerative findings on imaging are close to a constant of aging, not a pain verdict: disc degeneration in 37% of pain-free 20-year-olds rising to 96% of pain-free 80-year-olds, disc bulge 30%→84%, protrusion 29%→43% [1]. You almost certainly did not train your way into it: the Twin Spine Study compared twins with wildly discordant lifetime loading — different jobs, different sports, different loads — and found surprisingly little disc-degeneration difference, with heredity dominant, smoking a small effect, whole-body vibration null, and routine loading showing indications of benefit [2]; community-level lifestyle and occupational exposures don't survive multivariate analysis for Modic changes either [3], and the lifetime bone-building kind of loading doesn't buy average disc health [4]. The disc is trainable in exactly one direction the meta-analysis can find: across 39 studies and 4,152 participants, upright bipedal loading — mostly running — carried better disc health (g 0.31, 95% CI 0.12–0.50), with every other loading subgroup null and the combined-loading signal honestly mixed [8]. The first randomized running trial in chronic low back pain worked: twelve weeks of progressive run-walk, three 30-minute sessions a week, cut pain 15.3 points and disability 5.2 points with zero attrition [9] — and the pre-planned disc MRI showed the most degenerated spines gaining the most hydration [10]. The floor beneath all of it is old and wide: exercise is first-line for chronic low back pain in the ACP guideline [11], worth a clinically important pain MD of −15.2 across 249 Cochrane trials [12], and a walking-and-education program just prevented recurrence at HR 0.72 in a 701-person Lancet trial [13]. And the surgery column is real but selective: in the one randomized cohort built from exactly this reader — chronic pain plus radiologic disc degeneration — fusion beat nonsurgical care 33%-vs-7% on pain [14], while the UK's 349-patient BMJ trial found no clear evidence fusion beats intensive rehabilitation [15]. That is the audit in one paragraph: this reader needs the report repriced, the red flags cleared, a barbell, and — the part nobody told them — probably a pair of running shoes.

What the condition actually is: a description of imaging, priced against silence

The mechanics in a paragraph: "degenerative disc disease" describes what an MRI sees — discs losing water content and height, annular fissures, bulges, and in about 6% of adults marrow changes in the adjacent vertebrae (Modic changes) — and the radiology sentence became a "disease" mostly because the pictures are so easy to take [1][5]. The reprice this file exists to run: Brinjikji's systematic review pooled every study that imaged pain-free people and found degeneration is near-universal with age — 37% of asymptomatic 20-year-olds, 52% of 30-year-olds, 68% of 40-year-olds, 80% of 50-year-olds, 88% of 60-year-olds, 96% of 80-year-olds carrying disc degeneration, with bulges, protrusions and fissures riding the same curve [1]. The field's own conclusion, written in the review itself: many imaging-based degenerative features "are likely part of normal aging and unassociated with pain," and the findings must be interpreted in clinical context — the sentence on the report is a photograph of your age as much as a photograph of your pain [1]. Which raises the question every reader asks next: if it's not wear from loading, what is it? The twins answer that too.

The guilt column: the twins who broke the wear-and-tear story

For most of the last century, disc degeneration was taught as the output of an injury model: loads in, wear out. The Twin Spine Study — a research program running since 1991 across Canada, Finland and the United States — tested that model on the one population where it can't hide: identical and fraternal twins whose adult lives diverged into genuinely different loading histories. The findings inverted the story: a substantial influence of heredity on disc degeneration; despite "extraordinary discordance between twin siblings in occupational and leisure-time physical loading conditions throughout adulthood, surprisingly little effect on disc degeneration was observed"; a small degeneration effect from heavy smoking discordance; no evidence that whole-body vibration from motorized vehicles accelerates degeneration in these controlled comparisons; and — the sentence the barbell reader should underline — "some indications were found that routine loading may actually have some benefits to the disc" [2]. The community files agree at population scale: in 644 randomly selected Hangzhou adults, 44.7% carried Modic changes, and the lifestyle and lifetime occupational exposures the wear-and-tear story blames did not hold up as pathogenic drivers in multivariate analysis [3]; and in adults with chronic low back pain, the lifetime bone-building kind of heavy loading was not associated with average disc health [4]. Read those together and the guilt question this file's reader arrives with — did squats and deadlifts do this? — has a data-backed answer: the deadlift file prices the movement itself against the same ledger, and the twins price the lifetime loading question at approximately null [2][3][4]. Genetics, age and smoking carry the degeneration file; what the reader does after the report is where all the trainable variables live [7][8][9].

The Modic paragraph: the one finding that talks to pain — quietly

One MRI finding deserves its own paragraph because the report may name it and the internet definitely will: Modic changes — marrow signal changes in the vertebral bodies beside a degenerating disc, Type 1 (inflammatory-looking, edema) through Type 3 (sclerotic). The population census: in 2,449 volunteers, 5.8% carried Modic changes, 83% of them at the lowest two lumbar levels, with smokers who were also overweight running OR 2.18 (1.10–4.30), and lower-level Modic changes associated with a history of low back pain at OR 1.93 (1.05–3.54) — and with the severity and duration of symptoms [5]. So unlike gray-level "degeneration," Modic changes carry a real, modest pain signal — the closest thing on the report to a finding that behaves like tissue talking. Now the question that actually matters for this file: does the Modic finding change the training program? The Danish group ran the subgroup analysis on a rest-versus-exercise randomized trial in chronic low back pain patients with Modic changes, testing five MRI findings as effect modifiers: no interaction reached statistical significance, and the Modic Type 1 interactions — while wide and hypothesis-setting only — trended in the direction opposite the rest prescription (the Modic-1 patients did not benefit more from rest) [6]. The gym translation: the marrow finding upgrades nothing about the fences and downgrades nothing about the program — the reader with Modic changes trains by the same rules as the reader without them, with one honest asterisk: flare pacing and the warm-up earn their keep when the pain pattern is more inflammatory (morning stiffness, pain that eases as the day warms) [5][6].

The loading column: running is the one language the disc answers in

Here is the section this door exists for, because it is the single most misrepresented question in back training. The disc is avascular in adulthood — it feeds by diffusion and osmosis, cyclically loaded and unloaded, which is why "the disc has no blood supply, so protect it" is half a fact wearing a whole conclusion. The human evidence, in order of altitude: Belavý's group showed in 2017 that chronic runners carry better disc composition — hydration and proteoglycan content — and actual disc hypertrophy, with the signal arriving at fast-walking and slow-running accelerations (~2 m/s): not high-impact tasks, not low-intensity strolling, not static positions [7]. Then the 2026 meta-analysis pooled the whole field — 45 reports, 39 studies, 4,152 participants — and gave the loading question its current shape: upright bipedal loading, mostly running, was the only loading subgroup associated with better intervertebral disc health (Hedges' g 0.31, 0.12–0.50, P=0.002, GRADE very low); no other subgroup of physical loading was associated with better or worse disc health; combined physical loading carried one honest mixed signal (greater odds of reduced signal intensity, OR 2.80, 1.53–5.11); and the review's own conclusion: "running was the only physical loading exposure associated with better IVD health" [8]. Certainty tags worn openly: GRADE very low, mostly observational — which is exactly why the randomized column below matters as much as it does. The gym translation is not "everyone must run": it is that the cyclic, moderate, upright loading family — brisk walking, zone-2 run-walk intervals, the cardio-before-weights question's gentle end — is the disc's feeding rhythm, while the heavy-static and high-impact extremes are where the evidence goes quiet rather than negative [7][8]. The knees file prices what running doesn't do to the joint below; this file prices what it plausibly does for the tissue between the vertebrae.

The randomized column: ASTEROID put runners with bad backs on treadmills

The observational file above had one giant hole — no randomized trial had ever prescribed running to back-pain patients — and the Melbourne group closed it. ASTEROID randomized 40 adults aged 18–45 with chronic non-specific low back pain to twelve weeks of a digitally delivered, progressive run-walk program (three 30-minute sessions a week, remotely supported) or waitlist: at twelve weeks, average pain fell 15.30 points more than controls (−25.33 to −5.27, P=0.003), current pain 19.35 points (P=0.003), and disability 5.20 Oswestry points (−10.12 to −0.24, P=0.038) — with zero attrition, 70% adherence (2.1 of 3 weekly sessions), and nine non-serious adverse events, seven of them ordinary lower-limb soreness [9]. Then the part that makes this trial the anchor of the whole loading column: a pre-planned secondary put every participant in the scanner at baseline, six and twelve weeks and quantified whole-disc T2 (hydration) from T11/T12 to L5/S1. The moderating findings: the spines with the highest aggregated Pfirrmann scores — the most degenerated — gained the most (+3.42 ms, 0.22–6.62, P=0.036); cumulative running volumes of 28.6–46.1 km over the trial gained +4.80 ms (2.51–7.10, P<0.001); moderate speeds (10.5–11.7 km/h) and grass surfaces moderated positively [10]. The authors price it themselves: hypothesis-generating, small, twelve weeks — do not read "regenerated discs" into T2 movement [10]. But read what is there: in the population this whole door exists for — adults with chronically painful, degenerating lumbar spines — twelve weeks of graduated running made the pain better, the disability better, and the discs wetter, with the wettest gains in the most degenerated spines. No protective-rest arm of the internet's script produces a column like that [9][10].

The floor: guidelines, 249 trials, and a 701-person walking trial

Beneath the disc-specific columns sits a floor so wide it's easy to forget it's load-bearing. The American College of Physicians' 2017 guideline puts non-drug care first for chronic low back pain, with exercise among the first-line options [11]. The Cochrane exercise review — 249 randomized trials — prices exercise against no treatment, usual care and placebo at a moderate-certainty pain MD of −15.2 points (−18.3 to −12.2) at earliest follow-up, clinically important, with the honest caveat printed beside it — 79% of trials at risk of performance bias from the difficulty of blinding exercise [12]. And the newest floor stone is the WalkBack trial from the Lancet: 701 adults recently recovered from a low back pain episode randomized to an individualised, progressive walking-and-education program or no treatment — activity-limiting recurrence cut at HR 0.72 (0.60–0.85, P=0.0002), median days to recurrence 208 versus 112, at AU$7,802 per QALY with a 94% probability of cost-effectiveness — and the honest leg every program should copy: more lower-extremity adverse events in the walking arm (100 vs 54), because dose still needs ramping even when the medicine is walking [13]. Notice what all three floor stones share with ASTEROID and the running meta-analysis: the effective dose is ordinary, progressive, rhythmic loading — the exact opposite of the protect-the-disc script [9][11][12][13].

The surgery column, priced on both sides

This is the one chronic-condition door where the reader may already hold a fusion consultation date, so this file prices the surgical literature honestly rather than skipping it. The case for: the Swedish Lumbar Spine Study Group randomized 294 patients with at least two years of chronic low back pain and radiologic evidence of disc degeneration at L4-L5, L5-S1 or both — this file's exact cohort — to fusion (three techniques, 222 patients) or nonsurgical physical therapy (72): at two years, back pain fell 33% in the surgical group versus 7% in the nonsurgical (P=0.0002), Oswestry disability fell 25% versus 6% (P=0.015), with 98% follow-up [14]. The asterisks on that case: the gains were greatest in the first six months and then gradually deteriorated; the nonsurgical arm of 2001 was not modern intensive rehabilitation; and the trial's own design predates the psychological-screening era [14]. The case against rushing: the UK MRC trial randomized 349 fusion candidates to surgery or an intensive cognitive-behavioral rehabilitation program — the ODI difference at two years was −4.1 points (−8.1 to −0.1, P=0.045), marginal, just reaching the predefined minimal clinical difference, with no difference in shuttle-walking or any secondary outcome, and the authors' verdict became one of the most quoted sentences in the field: "no clear evidence emerged that primary spinal fusion surgery was any more beneficial than intensive rehabilitation" [15]. The gym translation is not "never" and not "obviously" — it is that the intensive-rehab arm of these trials is a serious comparator: structured, progressive, psychologically informed exercise programs of the exact shape this library programs. A reader walking into a fusion consult with chronic DDD-pattern pain should ask two questions of every table shown to them: what did the nonsurgical arm actually receive, and how recent is it? — and should already be training like the comparator arm regardless [14][15]. The sibling doors price the adjacent surgeries: the back-surgery door for the post-operative reader, the stenosis door and the spondylolisthesis door for the specific neighbors.

The program in gym terms: what trains, what ramps, what stops

Trains always: the whole library, dosed ordinarily. Whole-body progressive strength two-to-three times a week — the Cochrane floor's 249 trials and the fusion comparators were built on this shape [12][14][15] — plus the rhythmic aerobic column the disc column wants: brisk walking daily-ish, progressive run-walk toward zone-2 running three times a week, built exactly like ASTEROID's ramp (intervals, moderate pace, grass where available) [9][10]. The deadlift and the barbell row families are trainable with a degenerating lumbar spine — the twins already priced lifetime loading at approximately null [2][3][4]; what earns respect is the ramp, not the avoidance.

Ramps the run, never launches it: the disc-feeding signal lives at fast-walk/slow-run accelerations [7], the trial's winning volumes were cumulative and modest (28.6–46.1 km over twelve weeks) [10], and WalkBack's honest adverse-event column says even walking needs its progression respected [13]. Week one is walk-run intervals at conversational pace, not a 5k time trial.

Re-doses during flares, never cancels: pain flare = volume and intensity down, pain-free arcs only, keep moving — the exact architecture the herniated disc and sciatica doors run, and the direction the Modic subgroup analysis trended anyway [6]. Morning-stiffness-heavy patterns earn longer warm-ups, not cancelled sessions [5].

Stops and refers: the red-flag column from the intro — bladder/bowel change, saddle numbness, progressive weakness, fever, unexplained weight loss with night pain — is today-level care [1]; new leg-dominant neurological patterns route to the sibling doors and a clinician's exam; and the fusion decision is made at the consult with both randomized tables in hand [14][15].

Keeps one dated ledger: sessions, pain (0–10 at rest and after sessions), and the run-walk progression on one timeline — the instrument that makes every rule above adjustable, and the artifact that makes the next consult concrete [9][13].

The Brazil anchors: the São Paulo census and the honest national gap

The pt-BR twin of this file exists for a Brazilian audience whose scale is finally being measured properly: the São Paulo community census recruited 3,000 adults across randomly selected census sectors — point prevalence of low back pain 9.8% (95% CI 8.8–11.0), one-year prevalence 48.1% (46.3–49.9), lifetime prevalence 62.6% (60.8–64.3), with the prevalence climbing in women, in people with obesity, in the insufficiently active and sedentary, in smokers, and in the stressed and dissatisfied [16]. That is half the city's adults reporting back pain in a given year — and a meaningful share of them carrying, or about to carry, a radiology sentence like the one this file reprices. The honesty file beside it: the systematic review of Brazilian low-back-pain prevalence studies found 18 studies, high risk of bias throughout sampling and case definition, and no representative study providing a generalizable national prevalence [17] — the pt-BR twin says that out loud rather than rounding a prettier national number, because a file whose whole thesis is "read the sentence in context" doesn't get to invent its own context.

So what do you actually do? The practical file

Six rules, every one sourced from the columns above:

  1. Reprice the report before reprogramming your life. Degeneration at 37%→96% of pain-free spines by age means the photograph is not the verdict [1]; the twins price your lifetime loading at approximately null [2][3]. The question the report cannot answer — is this finding the thing that hurts? — belongs to a clinician's exam, and the answer changes nothing about rule 2 anyway.
  2. Clear the red flags and route the cousins. Bladder/bowel, saddle numbness, progressive weakness, fever, night pain with weight loss: today, not Monday [1]. Leg-dominant patterns: the herniated and sciatica doors own them.
  3. Load rhythmically, run progressively. The disc's only positive loading signal is the upright bipedal kind [7][8]; the RCT that tested it in bad backs won at a 3×30-minute progressive run-walk [9]. Start with walk-run intervals; the hydration signal was strongest at modest cumulative volumes and moderate speeds [10].
  4. Lift like the comparator arm of the fusion trials. Twice-weekly whole-body strength is both the Cochrane floor [12] and the intensive-rehab shape that made surgery's edge marginal [15]. The lifetime-loading nulls retire the fear column [2][4]; the ramp retires the injury column [13].
  5. If surgery is on the table, bring both tables. The Swedish file is real (33%-vs-7% pain in the DDD cohort [14]); the BMJ verdict is real (no clear evidence over intensive rehab [15]). Ask what the nonsurgical arm received and how recent it was — and train like that arm while deciding.
  6. Keep the dated ledger. Sessions, pain, progression, one timeline [9][13] — it is the flare instrument, the ramp governor, and the consult artifact at once. The tracking habit is the medical file here.

How RightRep fits into this

The DDD file is, at its core, a dose-finding question — how much rhythmic loading, how much load, at what ramp — and dose-finding is unanswerable from memory:

  • The ramp made measurable. ASTEROID's winning recipe was a progressive program with remote support [9] — logged run-walk sessions with pace and volume make "progressive" a number instead of a vibe, and keep the cumulative volume inside the band where the hydration signal lived [10].
  • The flare governor. Pain logged beside sessions turns "re-dose, don't cancel" into data — which session size your flares actually tolerate is a measurable set point, not a guess [6][12].
  • Form on the loaded spine. Camera-scored reps keep the strength column honest without a coach in the room — calibration, not grinding, is what the comparator-arm programs actually prescribed [14][15].
  • The dated ledger, built in. One timeline of sessions, pain and progression is the artifact that makes the next consult — physio or surgeon — concrete [13][15].

The bottom line

  • The asymptomatic ledger: disc degeneration in 37% of pain-free 20-year-olds rising to 96% of pain-free 80-year-olds; bulge 30%→84%; protrusion 29%→43%; fissure 19%→29% — 33 studies, 3,110 silent spines [1].
  • The guilt column: the Twin Spine Study's loading-discordant twins found surprisingly little degeneration effect from lifetime loading; heredity dominant, smoking small, vibration null, routine loading possibly beneficial [2]; Modic lifestyle/occupational exposures null in multivariate [3]; lifetime bone-loading null for average disc health [4].
  • The Modic signal: 5.8% prevalence, 83% at the lowest two levels, pain association OR 1.93 — real but modest, and it did not steer rest-vs-exercise in the trial that tested it [5][6].
  • The loading verdict: running the only loading associated with better disc health (g 0.31, 0.12–0.50), every other subgroup null, combined signal honestly mixed — 39 studies, 4,152 participants, GRADE very low [8], on Belavý's human first evidence of runner disc composition and hypertrophy [7].
  • The randomized column: ASTEROID — pain −15.3, disability −5.2, zero attrition [9]; disc secondary — the most degenerated spines gained the most T2 (+3.42 ms), volumes of 28.6–46.1 km gaining +4.80 ms [10].
  • The floor: ACP first-line exercise [11]; Cochrane 249 trials, pain MD −15.2 [12]; WalkBack 701 randomized, recurrence HR 0.72, 208-vs-112 median days, 94% cost-effectiveness probability, honest 100-vs-54 lower-limb events [13].
  • The surgery column: Fritzell's DDD cohort — fusion pain −33%-vs-7%, Oswestry −25%-vs-6%, best at six months then fading [14]; Fairbank's BMJ verdict — ODI −4.1 marginal, "no clear evidence" fusion beats intensive rehab [15].
  • The Brazil anchors: São Paulo's 3,000-adult census — point 9.8%, one-year 48.1%, lifetime 62.6% [16]; and the national honesty file — 18 studies, high bias risk, no representative number [17].

Reprice the report, clear the fences, load the spine rhythmically, ramp the running like the trial did — and let the camera count the calm reps at rightrep.app.

This post was written by the RightRep AI Assistant — the same AI that counts and scores your reps at rightrep.app.

Sources

  1. Brinjikji W, Luetmer PH, Comstock B, et al. Systematic literature review of imaging features of spinal degeneration in asymptomatic populations. AJNR American Journal of Neuroradiology, 2015;36(4):811–816. pubmed.ncbi.nlm.nih.gov
  2. Battié MC, Videman T, Kaprio J, et al. The Twin Spine Study: contributions to a changing view of disc degeneration. The Spine Journal, 2009;9(1):47–59. pubmed.ncbi.nlm.nih.gov
  3. Lifestyle and lifetime occupational exposures may not play a role in the pathogenesis of Modic changes on the lumbar spine MR images. The Spine Journal, 2020;20(1):130–138. pubmed.ncbi.nlm.nih.gov
  4. Lifetime physical loading and MRI-derived intervertebral disc health in adults with chronic low back pain: a cross-sectional study. JOR Spine, 2026. pubmed.ncbi.nlm.nih.gov
  5. Mok FPS, Samartzis D, Karppinen J, et al. Modic changes of the lumbar spine: prevalence, risk factors, and association with disc degeneration and low back pain in a large-scale population-based study. The Spine Journal, 2016;16(1):56–64. pubmed.ncbi.nlm.nih.gov
  6. Jensen RK, Kent P, Hancock MJ. Do MRI findings identify patients with chronic low back pain and Modic changes who respond best to rest or exercise: a subgroup analysis of a randomised controlled trial. Chiropractic & Manual Therapies, 2015;23:26. pubmed.ncbi.nlm.nih.gov
  7. Belavý DL, Quittner MJ, Ridgers N, et al. Running exercise strengthens the intervertebral disc. Scientific Reports, 2017;7:45975. pubmed.ncbi.nlm.nih.gov
  8. Samanna CL, Owen PJ, Mitchell UH, et al. The impact of exercise on intervertebral disc health: a systematic review and meta-analysis. Sports Medicine, 2026;56(4):941–965. pubmed.ncbi.nlm.nih.gov
  9. Running is acceptable and efficacious in adults with non-specific chronic low back pain: the ASTEROID randomised controlled trial. British Journal of Sports Medicine, 2025;59. pubmed.ncbi.nlm.nih.gov
  10. Running is associated with intervertebral disc adaptations: a pre-planned secondary analysis of the ASTEROID randomised controlled trial. European Spine Journal, 2026. pubmed.ncbi.nlm.nih.gov
  11. Qaseem A, Wilt TJ, McLean RM, Forciea MA. Noninvasive treatments for acute, subacute, and chronic low back pain: a clinical practice guideline from the American College of Physicians. Annals of Internal Medicine, 2017;166(7):514–530. pubmed.ncbi.nlm.nih.gov
  12. Hayden JA, van Tulder MW, Malmivaara AV, et al. Exercise therapy for chronic low back pain. Cochrane Database of Systematic Reviews, 2021;9:CD012450. pubmed.ncbi.nlm.nih.gov
  13. Hill JC, Ferreira GE, Kamper SJ, et al. Effectiveness and cost-effectiveness of an individualised, progressive walking and education intervention for the prevention of low back pain recurrence (WalkBack): a randomised clinical trial. The Lancet, 2024;404:237–245. pubmed.ncbi.nlm.nih.gov
  14. Fritzell P, Hägg O, Wessberg P, Nordwall A. 2001 Volvo Award Winner in Clinical Studies: Lumbar fusion versus nonsurgical treatment for chronic low back pain: a multicenter randomized controlled trial from the Swedish Lumbar Spine Study Group. Spine, 2001;26(23):2521–2532. pubmed.ncbi.nlm.nih.gov
  15. 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. pubmed.ncbi.nlm.nih.gov
  16. Low back pain prevalence in São Paulo, Brazil: a cross-sectional study. Brazilian Journal of Physical Therapy, 2021;25(6):614–620. pubmed.ncbi.nlm.nih.gov
  17. Nascimento PR, Costa LOP. Low back pain prevalence in Brazil: a systematic review. Cadernos de Saúde Pública, 2015;31(6):1141–1156. pubmed.ncbi.nlm.nih.gov