RightRep EN · PT Open the app
BLOG

Does Exercise Help Sciatica? The Nerve Audit: Why 66% of Herniations Dissolve Themselves, Bed Rest Failed Its Own Trial, Surgery Buys Speed and Not Endings, the Steroid Pill That Fixed the Questionnaire but Not the Pain, and the One Prescription Every Null Keeps Validating

By RightRep AI Assistant · · 14 min read

No diagnosis scares a training person like sciatica. Back pain you can negotiate with; a nerve firing lightning down the back of your leg feels like a structural verdict — stop moving, lie down, protect the disc, start reading about surgery. The folklore around it is the most confident folklore in musculoskeletal medicine, and this library has already cross-examined its neighbours — the back-pain audit with its recurrence file, the arthritis audit with its load-adaptation file, training through soreness — but never the nerve itself. So this is the sciatica audit: what the randomized-trial file actually says about the pinched nerve everyone fears, and where training sits inside it.

One fence before any of it, and it is the most load-bearing paragraph in the post: progressive leg weakness, numbness in the saddle area, or any change in bladder or bowel control is an emergency-room trip, not a training question. That cluster is the cauda equina signature, and speed of evaluation is the outcome that matters. Everything below the fence is about ordinary sciatica — radiating leg pain without those flags — in exactly the populations the trials recruited.

The short answer: the nerve usually un-pinches itself, and moving is what the nulls keep validating

Strip the file to its spine and three findings remain. First, the natural-history engine: pooled across eleven cohorts, 66.66% of lumbar disc herniations managed conservatively spontaneously resorbed — the body dissolved two thirds of the bulges on its own [6]. Second, the treatment nulls: bed rest [2], the steroid pill [3] and the epidural injection [7][8] all failed to beat their comparators on the outcomes that matter at the timepoints that matter, and even early surgery's advantage over conservative care evaporated by the one-year mark, where both arms of the flagship trial reported 95% perceived recovery [1]. Third, the inversion that organizes this whole guide: in a literature where nearly every famous treatment ends in convergence or null, the one instruction that survives every trial — quietly, in both arms — is stay active within tolerance. Not because exercise is a miracle. Because immobility is the one thing the trials kept testing and kept rejecting [2], and because the graded, progressive versions of active care are the arms that actually won when anything did [10][11].

The natural-history engine: the disc that dissolves itself

Start with the finding that reframes everything else. The meta-analysis pooled eleven cohort studies of conservatively managed lumbar disc herniation and found an overall spontaneous-resorption incidence of 66.66% (95% CI 51–69%), with country sub-samples as high as 82.94% in the UK data [6]. The herniation — the thing on the MRI that launched the fear — is, two times out of three, a temporary tenant. The mechanism is mundane immunology: the escaped disc material is avascular tissue where it doesn't belong, and the body's cleanup cells eventually recognize and remove it. This is why the trials below keep converging at one year: the endpoint is being delivered mostly by biology, and every treatment is fighting not each other but the clock [6]. It is also the single most calming number in the file: "does sciatica go away on its own?" is, most of the time, a yes — and the honest treatment question is not what removes the bulge but what makes the waiting livable.

The bed-rest trial: the advice that failed its own test

For most of the twentieth century, sciatica treatment was bed rest. It took a New England Journal trial to check. One hundred eighty-three patients with radicular syndrome severe enough to justify two weeks of bed rest were randomized to either bed rest or watchful waiting — keep living normally, no forced rest [2]. At two weeks, 70% of the bed-rest group reported improvement versus 65% of the people who simply carried on (adjusted odds ratio 1.2, 95% CI 0.6–2.3). At twelve weeks, 87% of both groups had improved — pain intensity, bothersomeness, functional status, absenteeism from work and rates of surgery were all statistically indistinguishable [2]. The trial's verdict, in its own words: bed rest is not a more effective therapy than watchful waiting [2]. Read that as the template for the whole file. The comparison that matters is never "treatment versus nothing" — it is "treatment versus the body's own recovery curve," and against that curve, forced rest adds nothing while subtracting your fitness, your sleep quality and your mood.

The surgery convergence: the knife buys three months, not a different ending

Surgery is the fear at the bottom of every sciatica search, so the file deserves precision. The flagship: 283 patients with severe sciatica of 6–12 weeks' duration, randomized to early microdiskectomy or prolonged conservative treatment with surgery if needed [1]. In the early-surgery arm, 89% were operated at a mean of 2.2 weeks; in the conservative arm, only 39% eventually needed surgery at all, at a mean of 18.7 weeks [1]. Early surgery relieved leg pain faster and reached perceived recovery faster (hazard ratio 1.97, 95% CI 1.72–2.22) — and at one year, perceived recovery sat at 95% in both groups, with no overall difference in disability across the year (P=0.13) [1]. The systematic-review layer confirms the shape: across five randomized trials, early surgery speeds relief, but at one and two years the outcomes are the same [4]. The surgery-versus-physical-activity meta-analysis adds the honest boundary — for disc herniation, surgery's short-term leg-pain edge (−16 points on a 0–100 scale) is real but fades, while for spinal stenosis and spondylolisthesis the surgical advantage persists past two years [5]. So the fair summary is not "surgery doesn't work" — when the fire is unbearable, the knife genuinely buys about three months of faster relief, and post-surgery outcomes across 87 studies of 31,034 patients show large, durable leg-pain improvement [12]. The fair summary is that for the common disc-herniation case, surgery and well-managed conservative care are two staircases to the same room, and 61% of the conservative arm never needed the operation at all [1].

The steroid pill: fixed the questionnaire, not the pain, and doubled the harm

The PRECIS trial is the file's cleanest lesson in reading outcomes carefully. Two hundred sixty-nine adults with MRI-confirmed herniation and acute radiculopathy, randomized 2:1 to a 15-day prednisone taper (total 600 mg) or placebo [3]. The disability score improved more with prednisone: +6.4 ODI points at 3 weeks (95% CI 1.9–10.9, P=.006), a difference that oddly was still there at 52 weeks [3]. But the symptom the drug was prescribed for — leg pain — did not move: 0.3 points on a 0–10 scale at 3 weeks (P=.34), 0.6 at a year (P=.15) [3]. Surgery rates at one year: no difference. And the harm column ran the other way: adverse events in 49.2% of the prednisone group versus 23.9% of placebo (P<.001) [3]. A treatment that improves the paperwork, spares you nothing you actually feel, and doubles the side-effect ledger is the definition of a null wearing a positive abstract. It also explains why "ask your doctor for steroids" is not in the training section of this post.

The injection ceiling: weeks of single-digit relief on a hundred-point scale

Epidural corticosteroid injections are the most-used middle rung of the sciatica ladder, and they have now been through two rounds of placebo-controlled meta-analysis. The Annals review: twenty-three trials; short-term leg pain improved by −6.2 points on a 0–100 scale (95% CI −9.4 to −3.0), disability by −3.1; long-term effects smaller and statistically gone; overall evidence quality rated high — and the authors' own conclusion questions the clinical utility of effects this small [7]. The 2020 Cochrane abridged update, twenty-five trials and 2,470 participants: −4.93 leg pain, −4.18 disability, short-term only [8]. The honest framing for a patient in week six of fire down the leg: an injection may genuinely take the edge off for a few weeks, it will not change where you land at a year [7][8], and those weeks are exactly the window in which staying gently active protects everything the pain is tempting you to surrender [2][11].

The MRI honesty: the scan explains less than the symptom pattern

The most quietly important prognostic paper in the file comes from the surgery trial's own cohort. Three blinded readers parsed baseline MRIs of 379 sciatica patients. Nerve-root compression was visible in 88% of patients whose dominant pain was the leg — but in only 68% of those with disabling back pain alongside (P<0.001) [9]. And prognosis tracked the symptom pattern, not the scan: dominant sciatica with visible compression recovered at 91% at one year, while disabling back pain without compression recovered at 50% (P<0.001), with disabling back pain carrying an odds ratio of 0.32 for recovery [9]. Two lessons, both counterintuitive. The bulge your MRI shows you is not a sentence — two thirds of them resorb [6]. And the pain pattern that predicts a slow year is the one that behaves like back pain, not the one that behaves like a trapped nerve — which is precisely the pattern graded exercise treats best [10][11].

The conservative-care honesty — and the arms that actually won

Now the section this post exists for. When fifty randomized trials (4,920 patients) of chronic sciatica were run through a network meta-analysis, the largest short-term leg-pain reductions belonged to spinal manipulative therapy, exercise plus neural mobilization, and soft-tissue injections — and every one of those estimates sat at very low confidence [10]. The review's summary verdict is the field's own fence, quoted straight: no high-quality evidence confirms the superior effectiveness of any non-surgical intervention for chronic sciatica [10]. That is not a reason to do nothing — it is a reason to read the winners' profiles honestly. The arms that won were active, graded and individualized: the FIFTT trial randomized non-reducible discogenic-pain patients to individualized functional restoration plus guideline advice versus advice alone, and the individualized program took leg pain down another 1.5 points (0–10 scale) and disability another 6.3 ODI points at 10 weeks, with the disability effect still standing at 26 weeks [11]. Graded exposure, progressive loading, nerve-tolerance work — the mechanics of a training program — are the delivery system every surviving conservative effect rides on [10][11]. The same engine already filed under the back-pain audit: exercise plus education cutting recurrence by 45% across 30,850 adults.

So what do you actually do? The graded-active protocol

Assemble the file and the prescription writes itself — and it is the opposite of the couch-and-fear protocol the diagnosis triggers.

1. Walk, early and as tolerated. The bed-rest trial's watchful-waiting arm was the active arm, and it reached 87% at twelve weeks with nothing but normal life [2]. Walking is the minimum viable dose of the one instruction the file never nulls, and it is the entry ramp for everything below.

2. Keep training what doesn't provoke the leg. Pain radiating below the knee is the traffic light to respect; the rest of your body is open for business. Upper-body work, the uninjured side, warm-ups done properly — maintaining fitness, mood and sleep during the waiting period is not filler; it is the treatment envelope [6][11]. Fear-driven total rest is the only intervention in this file with an unambiguous losing record [2].

3. Grade the exposure, don't jump it. The winning arms were individualized and progressive [11] — meaning hip hinges re-enter at a depth and load the nerve tolerates today, and climb from there. That is the ordinary progressive overload machinery this library runs on, applied to nerve tolerance instead of muscle tolerance, with form quality doing the protecting that fear was supposed to.

4. Neural mobilization belongs in the program — with the fence attached. Exercise plus neural mobilization ranked among the top conservative arms, at very low confidence [10]. Translation: a reasonable, low-risk addition guided by a clinician or physiotherapist, not a YouTube-certified cure.

5. Escalate on evidence, not on fear or calendar. The surgery decision in the flagship trial was made at 6–12 weeks of severe, non-improving symptoms — not on day one, and not after two years of suffering [1]. Red flags go to the ER immediately; stubborn cases go to the clinician's timeline [1][3][5].

How RightRep fits into this

The protocol above lives or dies on one skill this app is built around — doing the graded work consistently while the biology runs:

  • The walking dose, tracked. The stay-active prescription [2] is a weekly number before it is anything else; GPS run/walk sessions (with Strava sync) turn "keep moving" into minutes you can see and slowly grow, flare-safe.
  • The graded-loading half, logged honestly. The arms that won were progressive and individualized [11]; camera-counted, rep-scored sets make today's tolerated load explicit on every exercise — the exact dial a nerve-tolerance ladder needs, adjusted by what the log says, not what fear says.
  • The continuity that outlasts the episode. Sixty-six percent of herniations resorb [6] and recurrence is where the exercise file is strongest (the 45% recurrence cut); a session record with dates is the difference between "I recovered" and "I recovered and rebuilt."

The bottom line

  • The engine is natural history: 66.66% of conservatively managed herniations spontaneously resorbed across eleven cohorts — the bulge is usually a temporary tenant [6].
  • Bed rest failed its own trial: 70% vs 65% improved at two weeks, 87% both arms at twelve — rest added nothing [2].
  • Surgery buys speed, not endings: faster relief (HR 1.97) but identical 95% one-year recovery in both arms of the 283-patient trial; stenosis is the honest exception where the surgical edge persists [1][4][5].
  • The pill fixed the questionnaire, not the pain: prednisone moved disability scores (+6.4 ODI) but not leg pain (0.3/10, P=.34), while adverse events doubled to 49.2% [3].
  • The injection ceiling: −6.2 and −4.9 leg-pain points of 100, short-term only, across two placebo-controlled metas [7][8].
  • The scan is not the sentence: recovery 91% with compression + leg-dominant pain vs 50% with back-pain-dominant, no compression — the symptom pattern outranks the MRI [9].
  • Exercise is the delivery system, not a rival pill: no non-surgical treatment wins at high confidence [10], and the arms that did win were individualized, graded, progressive active care [11].

Walk first, load what tolerates, grade the rest — and let the camera keep the comeback honest 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. Peul WC, van Houwelingen HC, van den Hout WB, et al. Surgery versus prolonged conservative treatment for sciatica. New England Journal of Medicine, 2007;356(22):2245–2256. pubmed.ncbi.nlm.nih.gov
  2. Vroomen PC, de Krom MC, Wilmink JT, et al. Lack of effectiveness of bed rest for sciatica. New England Journal of Medicine, 1999;340(6):418–423. pubmed.ncbi.nlm.nih.gov
  3. Goldberg H, Firtch W, Tyburski M, et al. Oral steroids for acute radiculopathy due to a herniated lumbar disk: a randomized clinical trial. JAMA, 2015;313(19):1915–1923. pubmed.ncbi.nlm.nih.gov
  4. Jacobs WC, van Tulder M, Arts M, et al. Surgery versus conservative management of sciatica due to a lumbar herniated disc: a systematic review. European Spine Journal, 2011;20(4):513–522. pubmed.ncbi.nlm.nih.gov
  5. Fernandez M, Ferreira ML, Refshauge KM, et al. Surgery or physical activity in the management of sciatica: a systematic review and meta-analysis. European Spine Journal, 2016;25(11):3495–3512. pubmed.ncbi.nlm.nih.gov
  6. Zhong M, Liu JT, Jiang H, et al. Incidence of spontaneous resorption of lumbar disc herniation: a meta-analysis. Pain Physician, 2017;20(1):E45–E52. pubmed.ncbi.nlm.nih.gov
  7. Pinto RZ, Maher CG, Ferreira ML, et al. Epidural corticosteroid injections in the management of sciatica: a systematic review and meta-analysis. Annals of Internal Medicine, 2012;157(12):865–877. pubmed.ncbi.nlm.nih.gov
  8. Oliveira CB, Maher CG, Ferreira ML, et al. Epidural corticosteroid injections for sciatica: an abridged Cochrane systematic review and meta-analysis. Spine, 2020;45(21):E1405–E1415. pubmed.ncbi.nlm.nih.gov
  9. el Barzouhi A, Vleggeert-Lankamp CL, Lycklama à Nijeholt GJ, et al. Influence of low back pain and prognostic value of MRI in sciatica patients in relation to back pain. PLoS One, 2014;9(3):e90800. pubmed.ncbi.nlm.nih.gov
  10. Zhu Z, Schouten T, Strijkers R, et al. Effectiveness of non-surgical interventions for patients with chronic sciatica: a systematic review with network meta-analysis. The Journal of Pain, 2025;33:105431. pubmed.ncbi.nlm.nih.gov
  11. Chan AYP, Ford JJ, Surkitt LD, et al. Individualised functional restoration plus guideline-based advice vs advice alone for non-reducible discogenic low back pain: a randomised controlled trial. Physiotherapy, 2017;103(2):121–130. pubmed.ncbi.nlm.nih.gov
  12. Rushton A, Heneghan NR, Heymans MW, et al. Clinical course of pain and disability following primary lumbar discectomy: systematic review and meta-analysis. European Spine Journal, 2020;29(7):1660–1670. pubmed.ncbi.nlm.nih.gov