RightRep EN · PT Open the app
BLOG

Can You Train With Spinal Stenosis? The Narrow-Canal Audit: The Flexion Door the Back Library Skipped — Where SPORT's Randomized Cohort Prices Decompression at 7.8 Bodily-Pain Points ITT With Zero Physical-Function or ODI Edge and 43% of the Nonsurgical Arm Crossing to the Knife, Where Delitto's Pittsburgh RCT Ties Surgery to a Six-Week PT Regimen at a 0.9-Point Function Difference (−7.9 to 9.6) With the 57% Crossover Surviving Causal-Effects Analysis, Where the Cochrane Files the Whole Surgical Door at No Clear Benefit With Side Effects 10–24% and the Fusion Upsell Dies Twice (ODI 27-vs-24 in the Swedish RCT, 63%-vs-63% at Five Years in Nordsten-DS) While Complex Fusion Grew 15-Fold at Complication OR 2.95 — and the Real Training File Is the Canal's Own Geometry: the Dural Sac Measuring 96-vs-74 mm² Flexed-Versus-Extended on Dynamic CT, Which Is Why Cycling Beats the Treadmill, Why the Shopping Cart Works, and Why the Component Analysis of 13 Trials/1,440 Participants Puts Flexion, Strength, Supervised Fitness Work and Psychologically Informed Loading in the Winning Programs

By RightRep AI Assistant · · 21 min read

The back library has audited the doors around this one — the herniated disc, the sciatica pair, back pain itself, the deadlift — and skipped the door the reader over 60 is most likely to be standing in: the bony canal that has narrowed around the nerves, so that standing upright and walking downhill fire legs that go quiet the moment you sit, lean, or push a shopping cart. Three fences outrank everything below. First: cauda equina is an emergency room today, not a training question — saddle numbness, loss of bladder or bowel control, or rapidly progressive weakness in both legs means you stop reading and go. Second: the claudication file has a vascular twin — artery disease claudicates under its own load rules and belongs to a physician workup, not to a posture program; the stenosis reader who is undiagnosed gets the diagnosis before the program, always. Third: a fresh decision about decompression or fusion is a shared-decision-making conversation with a surgeon, and this file's job is to put the honest numbers on that table — not to make it for you [2][3]. With the fences respected, the internet's two scripts both fail: the "rest until the operation, then rest some more" script dies on what conservative care actually delivers [2][7], and the "no pain no gain, walk it off standing tall" script dies on the canal's own geometry — the one condition in this family where extension is the enemy and flexion is the door [11].

The short answer: train — flexion opens the door the canal closed

Strip the file to its spine. Surgery works, modestly, and the honest price tag is narrower than the waiting-room lore says: SPORT's randomized cohort — 289 surgical candidates — showed decompression beating usual nonsurgical care by 7.8 points of bodily pain at two years on intention-to-treat (1.5–14.1), with no significant physical-function or disability difference, while 43% of the nonsurgical arm crossed to the knife anyway [1]; only the pooled as-treated analysis — everyone who actually operated, both arms — favored surgery across every primary outcome [1]. And a structured exercise program ties the operation when the comparison is run head-to-head: Delitto's Pittsburgh randomized trial gave 169 surgical candidates either decompression or a physical-therapy regimen and measured physical function at two years: +22.4 points surgical, +19.2 points PT, difference 0.9 (−7.9 to 9.6) — and the 57% of the PT arm that crossed to surgery still didn't separate the groups under causal-effects sensitivity analysis [2]. The Cochrane's verdict on the entire surgical-versus-conservative file: five trials, 643 patients, "no clear benefits were observed with surgery versus non-surgical treatment" — against surgical side-effect rates of 10–24% and not one reported side effect from conservative care [3]. Meanwhile the modality that carries the training door is mechanical, measurable, and free: on dynamic CT the stenotic dural sac averages 96.3 mm² flexed against 73.6 mm² extended at L3-4 (72.3 against 61.0 at L4-5) [11] — your spine is measurably roomier in the posture the symptoms have been pushing you toward all along. The verdict writes itself: the canal is narrowed, not closed; the operation is one door among several, priced honestly; and the program that fits the geometry — flexion-biased, strength-first, cycling-conditioned, supervised — is the same program the component analysis of 1,440 trial participants keeps finding in the winning arms [5].

The verdict column: what decompression actually buys

Start with the file that prices the operation your surgeon is offering, because it is fairer to surgery than the internet is. SPORT — the Spine Patient Outcomes Research Trial, stenosis without spondylolisthesis — randomized 289 patients to decompression or usual nonsurgical care and watched a further 365 in a parallel observational cohort: at two years, 67% of the randomized surgical arm had operated and 43% of the nonsurgical arm had too, the intention-to-treat analysis showed a bodily-pain edge of 7.8 points (1.5–14.1) for surgery with no significant physical-function or Oswestry difference, and the confounder-adjusted as-treated analysis — pooled across both cohorts — favored surgery on all primary outcomes from three months through two years [1]. That is a real benefit; it is also a benefit measured against usual nonsurgical care, not against a structured program. Put the structured program in the other arm and the edge vanishes: Delitto's trial — 169 patients aged 50+, all surgical candidates, all consenting to surgery — randomized to decompression or to twice-a-week physical therapy, and at two years the masked testers measured SF-36 physical function improving 22.4 points (16.9–27.9) with surgery against 19.2 (13.6–24.8) with PT, an intention-to-treat difference of 0.9 (−7.9 to 9.6) — no difference — and the crossover problem (57% of the PT arm choosing surgery) was handled with causal-effects methods that still found no separation [2]. The Cochrane pooled the whole randomized file — decompression with or without fusion versus multimodal conservative care — and priced it at ODI −3.66 (−10.12 to 2.80) at six months, −6.18 (−15.03 to 2.66) at one year, −4.43 (−7.91 to −0.96) at two years: a between-trials mosaic of nulls and small late effects summarized as no clear benefit, low confidence throughout, and the asymmetry that should anchor every shared-decision conversation — complications in 10–24% of surgical patients, none reported in any conservative arm [3]. The sciatica file runs the same spine one door down the nerve; this door's verdict is not "surgery never" — it is "surgery is priced, and so is the program."

The fusion column: the upsell door, priced twice

The most expensive question in the stenosis consult is not whether to decompress but whether to fuse, and the randomized file answers it twice, in two countries, in the same decade. The Swedish trial: 247 patients aged 50–80 with one- or two-level stenosis, randomized to decompression-plus-fusion or decompression alone — Oswestry at two years 27 versus 24 (P=0.24), six-minute walk 397 versus 405 meters (P=0.72), no clinical difference at five years, hospitalization 7.4 versus 4.1 days, and additional lumbar surgery over 6.5 years in 22% versus 21% [9]. The Norwegian five-year file (Nordsten-DS): 267 stenosis patients with degenerative spondylolisthesis, decompression alone versus decompression with instrumented fusion — 63% versus 63% achieving the primary outcome of ≥30% Oswestry reduction (difference 0.4 points, CI −11.2 to 11.9), identical −17.8-point ODI improvement in both arms, reoperations 16% versus 18% [10]. Fusion bought nothing the decompression didn't, twice, at the price of longer operations, longer stays, more blood and more money [9][10]. What makes this column urgent is the market file: in Medicare between 2002 and 2007, complex fusion for stenosis grew fifteen-fold — 1.3 to 19.9 per 100,000 beneficiaries — while life-threatening complications ran 2.3% after simple decompression and 5.6% after complex fusion, an adjusted odds ratio of 2.95 (2.50–3.49) [15]. The surgery rate itself barely moved; the mix migrated up the invasiveness ladder, and the randomized trials say the upper rungs of that ladder buy nothing [9][10][15]. For the trainee, the fusion column is the freedom clause: a patient who kept the strength and the walking while deciding is a better candidate for the smaller operation — or for no operation — and a spine that stayed trained recovers from whichever door is chosen faster.

The decade column: what ten years actually say

The short-term columns understate the training door, because stenosis is a decade-scale condition and the decade files are where conservative care earns its tie. Oslo, ten years, one hundred patients: Amundsen's prospective cohort treated 19 severe patients surgically and 50 moderate patients conservatively, then followed everyone for a decade — at four years, excellent-or-fair results in half the conservative patients against four-fifths of the surgical ones; the conservative patients who failed and crossed to delayed surgery got essentially the initial-surgery result; and the final six years showed no significant deterioration in either arm — the verdict, written in 2000 and never overturned: "an initial conservative approach seems advisable for many patients because those with an unsatisfactory result can be treated surgically later with a good outcome" [7]. Maine, eight to ten years, 148 patients: back-pain relief, predominant-symptom improvement and satisfaction converged between the surgical and nonsurgical arms, leg-pain relief and back-specific function still favored surgery — and the reoperation ledger read 23% of the surgical arm reoperated by year ten, while 39% of the nonsurgical arm eventually operated [8]. Stack the decade files under the two-year files and the shape emerges: the operation buys speed and leg-pain relief; the program buys a floor; neither column decays into the other's regret [1][7][8] — and the reader who trains through the decision keeps access to both doors [7].

The geometry column: the canal's own rule

This is the section no other file in the back family can carry, because stenosis is the one condition whose mechanics invert the standard advice. The disc files warn against flexion; the stenotic canal is roomier in flexion. Kanbara's dynamic study put 100 stenosis patients in a CT scanner after myelography, in flexion and in extension, and measured the dural sac — the bag the nerves travel in — at every disc level: average cross-sectional area 96.3 mm² flexed against 73.6 mm² extended at L3-4, and 72.3 against 61.0 at L4-5, with anteroposterior diameter running 9.2 against 7.4 mm at L3-4 [11]. Roughly a fifth to a third more room, level after level, in the bent-forward posture. That single measurement explains the entire symptom pattern the reader already knows from the supermarket aisle: sitting relieves (the spine is flexed), leaning on the cart relieves (the spine is flexed), walking upright on cobblestones is the worst block of the week (the spine is extended and loaded) — and it writes the program's geometry rules directly: prefer the positions and exercises that hold the lumbar spine in neutral-to-flexion, treat loaded end-range extension as the one position to program around (the deep arch, the standing back-bend, the heavy overhead press braced into extension), and pick cardio whose posture is flexed. Hence the cycling column below [5][11]. The São Paulo gait lab adds the cautionary twin to the geometry file: in 14 patients walking to symptom onset on a treadmill, overall lower-limb gait quality fell as pain rose and the walkers adopted protective cadence and speed strategies — while the canal's cross-sectional area on MRI correlated with nothing [13]. The image prices the canal; the walk prices the patient; the program trains the second and respects the first [13][14].

The program column: what goes in the block

The exercise file for stenosis is bigger than most clinicians assume, and its newest synthesis is the most useful. The Leeds/Oxford component analysis — 13 randomized trials, 23 exercise interventions, 1,440 participants — dissected the winning programs component by component: most featured supervision and flexion-based exercise, and the components overrepresented in successful interventions were stretches, strength and trunk-muscle work, fitness exercise — especially cycling — and psychologically informed approaches [5]. That last clause is not decoration: the fear-avoidance spiral of a patient whose legs quit two blocks from home is itself a treatment target, and the programs that address it win more often [5]. The dose file: a Japanese meta-analysis of three trials, 244 participants found that adding supervised exercise to nonsurgical care improved pain immediately (SMD −0.94, −1.60 to −0.29), added 298–534 meters of walking distance, lifted quality of life — and left walking distance and quality of life still higher at long-term follow-up — with the authors' own caveat that the trial count is small [6]. And the head-to-head that prices program content: Whitman randomized 58 patients to six weeks of manual therapy plus body-weight-supported treadmill walking plus exercise against flexion exercise plus walking plus subtherapeutic ultrasound — perceived recovery favored the richer program at six weeks (P=0.0015, NNT 2.6) and held at one year, 62% recovered versus 41%, with disability, satisfaction and treadmill-walking measures following [4]. Read the three files together: the block is supervised (worth real effect sizes), flexion-biased (worth canal geometry), strength-and-trunk-loaded (the only column that compounds), cycling-conditioned (the aerobic door that doesn't require extension), and honest about fear [4][5][6].

The counter-file: where caution wins, honestly

Four honesties keep this door from becoming a pep talk. The red-flag fence is load-bearing, not legal boilerplate: progressive leg weakness, saddle numbness, bladder or bowel change, or gait collapse is a same-day physician door, and no program in this file outranks it. The claudication differential is a real trap: arterial claudication coexists with canal stenosis at an age where both are common, worsens under different rules, and needs a vascular answer before a posture answer — the undiagnosed reader gets the workup first, this file second. The imaging file is not the patient file: radiological stenosis appears in 11% (5–18%) of asymptomatic spines in pooled prevalence data, clinical stenosis in ~11% of the general population — the scan alone neither explains the legs nor writes the program [14][13]. And the operation's relief is real where it's real: leg-pain relief and back-specific function genuinely favored the surgical arm at Maine's decade mark [8], SPORT's as-treated column held every primary outcome for two years [1], and the Brazilian satisfaction ledger below reads the same direction [12] — the honest file prices both doors and lets the reader walk through either with strength in the legs.

The program: blocks that open the canal

The shape the trials and the geometry actually support:

  1. Get the diagnosis priced before the program. Symptoms that fire on extension and quiet in flexion, in a reader over 50, earn a physician workup — the vascular differential and the red flags are ruled out before anything is trained [3][14].
  2. Train flexion-biased by default. Neutral-to-flexed lumbar posture under load: hip-hinge patterns, seated and supported variations, curls over the pelvis — the canal's 96-vs-74 mm² is the operating manual [11].
  3. Program around loaded end-range extension — don't worship it. The standing back-bend, the over-arched overhead press, the cobblestone-walk posture: reduce, modify, or swap; extension stretching as gentle symptom-testing is a different animal from extension under load [11][4].
  4. Cycle for the engine. The component analysis's own flag — fitness exercise, especially cycling, overrepresented in winning programs [5] — and the geometry column's explanation: the flexed riding posture opens the canal the treadmill posture closes [11].
  5. Walk with a strategy, not a grimace. Body-weight-supported treadmill walking carried Whitman's winning arm [4]; flat-to-gentle routes, a forward lean on the poles or the cart when symptoms rise, intervals that respect the symptom clock rather than fighting it [4][6].
  6. Load the trunk and legs progressively. Strength and trunk-muscle work is a winning component [5] — the deadlift-geometry file owns the hinge pattern, scaled to the reader's spine, in the flexion-biased groove this canal wants.
  7. Buy supervision for the first block. The supervised column carries the meta-analytic edge (pain SMD −0.94, walking +298–534 m) [6] and Whitman's richer program beat the pamphlet-style arm [4] — supervision is the highest-yield line item in the budget.
  8. Log walking tolerance as the program's KPI. Distance-to-onset and distance-to-stop, dated — the column every trial in this file measured and the surgeon will ask for [1][2][6].

The fences, in order

  1. The cauda-equina fence. Saddle numbness, bladder/bowel change, progressive bilateral weakness — emergency room today; everything else in this file waits.
  2. The differential fence. Undiagnosed claudication earns the vascular workup before the posture program; the two conditions coexist at this age [14].
  3. The shared-decision fence. Surgery is priced, not banned — 7.8 bodily-pain points ITT, leg-pain relief at a decade [1][8] — and the decision belongs in a conversation carrying the honest two-column table [2][3].
  4. The fusion-question fence. If the consult jumps from decompression to fusion, the two randomized nulls [9][10] and the complication ladder [15] belong in that conversation, explicitly.
  5. The extension fence. Symptom-flaring under load means the dose or the geometry is wrong — modify the position, not the commitment [11].
  6. The osteoporosis-and-age fence. The stenosis decade is also the bone-density decade; the loaded-flexion rules here assume a bone-aware progression, and the reader with both conditions programs both files.

The Brazilian anchors

The pt-BR reader owns one of the cleanest surgical-versus-training files in the entire door. The UNIFESP trial (Rodrigues & Natour, Escola Paulista de Medicina, São Paulo): 63 patients randomized to surgery — or to no surgery — with both arms receiving the identical twelve weeks of twice-weekly physiotherapy: pain did not differ between groups over time (P=0.145), disability favored surgery at the medium term (P=0.006), the short-term walking-function differences were driven by the surgical arm worsening before improving, and satisfaction still favored the operated group [12] — the anywhere-adds-training design most stenosis trials never run. The USP gait lab (Garbelotti et al., USP/UNINOVE/Einstein): 14 patients to treadmill symptom-onset with 3D capture before and after — gait quality falling as pain rose, protective strategies emerging, and the MRI canal area predicting none of it [13]. Between them they carry the door's two quiet verdicts for the Brazilian reader: a fisioterapia é o chão comparativo que a cirurgia precisa vencer and o exame não anda — o paciente anda — the program above is those two sentences with a bar and a bike attached [12][13].

So what do you do in practice? The practical file

Seven rules, all sourced from the files above:

  1. Run the fences first — red flags and the vascular differential are physician doors before any set is loaded [3][14].
  2. Price your own doors honestly — surgery: 7.8 bodily-pain points ITT, no function edge versus a real PT program, leg-pain relief that holds a decade [1][2][8]; training: 62%-vs-41% one-year recovery rates under the right program, no side effects priced anywhere in the file [3][4].
  3. Hold the lumbar spine neutral-to-flexed under load — the canal is a fifth to a third roomier there [11].
  4. Ride the bike for the engine, walk the flats with a lean for the tolerance — the fitness modality the winning programs share [5][11].
  5. Load trunk and legs progressively, supervised at first — the strength component plus the supervision edge [4][5][6].
  6. Treat the fusion question as its own question — two randomized nulls and a 15-fold complication ladder say so [9][10][15].
  7. Log distance-to-onset, dated — the one number every door in this file reads [1][2][6].

How RightRep fits into this

The narrow-canal door's currency is the dated tolerance record — the artifact every trial in this file measured and no stenosis patient is ever handed:

  • The walking-tolerance ledger. Distance-to-onset and distance-to-stop on a single timeline — the column SPORT, Delitto and the supervised-exercise meta all treated as the outcome that matters [1][2][6], made visible session by session.
  • The form that holds the geometry. The same AI that counts and scores every repetition is the one that flags the drift into the over-extended lumbar posture this canal can't afford [11] — the coaching cue that matters most, on every rep.
  • The cycling and strength columns together. Both modalities the component analysis prices as winning components [5], logged as one program instead of two habits.
  • The symptom-versus-load record. Which session, which dose, which symptom response — the shared-decision table for the surgeon consult, assembled as a side effect of training [2][3].
  • The decade artifact. Stenosis is a ten-year file [7][8] — the streak, the gaps and the trajectory are how "managed intelligently" stops being a promise and becomes a log.

The final tally

  • The verdict column: SPORT — 289 randomized, ITT bodily pain 7.8 (1.5–14.1), no PF/ODI ITT edge, as-treated favored all primaries, 67%/43% operated by 2 years [1]; Delitto — 169 patients, SF-36 PF +22.4 vs +19.2, ITT difference 0.9 (−7.9 to 9.6), 57% crossover surviving causal-effects analysis [2]; Cochrane — 5 RCTs/643, ODI nulls at 6 and 12 months, −4.43 at 24, no clear benefit, side effects 10–24% vs none [3].
  • The fusion column: Swedish RCT — 247 patients, ODI 27 vs 24 (P=0.24), 6MWT 397 vs 405 m, LOS 7.4 vs 4.1 days, reoperation 22% vs 21% [9]; Nordsten-DS — 267 patients, 63% vs 63% at 5 years (0.4 pp, CI −11.2 to 11.9), ODI −17.8 both, reoperations 16% vs 18% [10]; Medicare — complex fusion ×15, life-threatening complications 2.3%→5.6%, adjusted OR 2.95 (2.50–3.49) [15].
  • The decade column: Oslo — 100 patients, conservative half vs surgical four-fifths good at 4 years, delayed surgery ≈ initial surgery, stable final 6 years, conservative-first verdict [7]; Maine — 148 patients, 8–10-year convergence on pain relief and satisfaction, leg pain/function favoring surgery, reoperation 23% surgical arm / 39% nonsurgical crossover to surgery [8].
  • The geometry column: dynamic CT — dural sac 96.3/73.6 mm² (L3-4) and 72.3/61.0 mm² (L4-5) flexed/extended, AP 9.2/7.4 mm [11]; São Paulo gait — quality falling with pain, canal area predicting nothing [13].
  • The program column: component analysis — 13 RCTs/23 interventions/1,440 participants, winners: supervision, flexion, stretches, strength/trunk, fitness esp. cycling, psychologically informed [5]; supervised-exercise meta — 3 trials/244, pain −0.94, walking +298–534 m, long-term walking/QoL up [6]; Whitman — 58 patients, NNT 2.6 at 6 weeks, 62% vs 41% at 1 year [4].
  • The Brazil file: UNIFESP — 63 patients, identical-PT-both-arms design, pain P=0.145, ODI P=0.006, satisfaction favoring surgery [12]; USP gait lab [13].
  • The prevalence file: clinical LSS 11% (4–18%) general population, radiological 11% (5–18%) of asymptomatic spines [14].
  • The fences: cauda equina, vascular differential, shared decision, the fusion question, extension under load, bone-aware progression — in that order.

Train in the posture the canal prefers, log the tolerance the surgeons ask for, and let the record — not the fear — decide when the next door opens, at rightrep.app.

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

Sources

  1. Weinstein JN, Tosteson TD, Lurie JD, et al. Surgical versus nonsurgical therapy for lumbar spinal stenosis. N Engl J Med, 2008;358(8):794–810. pubmed.ncbi.nlm.nih.gov
  2. Delitto A, Piva SR, Moore CG, et al. Surgery versus nonsurgical treatment of lumbar spinal stenosis: a randomized trial. Ann Intern Med, 2015;162(7):465–473. pubmed.ncbi.nlm.nih.gov
  3. Zaina F, Tomkins-Lane C, Carragee E, Negrini S. Surgical versus non-surgical treatment for lumbar spinal stenosis. Cochrane Database Syst Rev, 2016;2016(1):CD010264. pubmed.ncbi.nlm.nih.gov
  4. Whitman JM, Flynn TW, Childs JD, et al. A comparison between two physical therapy treatment programs for patients with lumbar spinal stenosis: a randomized clinical trial. Spine, 2006;31(22):2541–2549. pubmed.ncbi.nlm.nih.gov
  5. Comer C, Williamson E, McIlroy S, et al. Exercise treatments for lumbar spinal stenosis: a systematic review and intervention component analysis of randomised controlled trials. Clin Rehabil, 2024;38(3):361–374. pubmed.ncbi.nlm.nih.gov
  6. Urata R, Igawa T, Ito S, et al. Effectiveness of non-surgical treatment combined with supervised exercise for lumbar spinal stenosis: a systematic review and meta-analysis. J Back Musculoskelet Rehabil, 2023;36(4):799–813. pubmed.ncbi.nlm.nih.gov
  7. Amundsen T, Weber H, Nordal HJ, et al. Lumbar spinal stenosis: conservative or surgical management? A prospective 10-year study. Spine, 2000;25(11):1424–1436. pubmed.ncbi.nlm.nih.gov
  8. Atlas SJ, Keller RB, Wu YA, Deyo RA, Singer DE. Long-term outcomes of surgical and nonsurgical management of lumbar spinal stenosis: 8 to 10 year results from the Maine Lumbar Spine Study. Spine, 2005;30(8):936–943. pubmed.ncbi.nlm.nih.gov
  9. Försth P, Ólafsson G, Carlsson T, et al. A randomized, controlled trial of fusion surgery for lumbar spinal stenosis. N Engl J Med, 2016;374(15):1413–1423. pubmed.ncbi.nlm.nih.gov
  10. Kgomotso EL, Hellum C, Fagerland MW, et al. Decompression alone or with fusion for degenerative lumbar spondylolisthesis (Nordsten-DS): five year follow-up of a randomised, multicentre, non-inferiority trial. BMJ, 2024;386:e079771. pubmed.ncbi.nlm.nih.gov
  11. Kanbara S, Yukawa Y, Ito K, Machino M, Kato F. Dynamic changes in the dural sac of patients with lumbar canal stenosis evaluated by multidetector-row computed tomography after myelography. Eur Spine J, 2014;23(1):74–79. pubmed.ncbi.nlm.nih.gov
  12. Rodrigues LCL, Natour J. Surgical treatment for lumbar spinal stenosis: a single-blinded randomized controlled trial. Adv Rheumatol, 2021;61(1):25. pubmed.ncbi.nlm.nih.gov
  13. Garbelotti SA Jr, Lucareli PR, Ramalho A Jr, et al. An investigation of the value of tridimensional kinematic analysis in functional diagnosis of lumbar spinal stenosis. Gait Posture, 2014;40(1):150–153. pubmed.ncbi.nlm.nih.gov
  14. Jensen RK, Jensen TS, Koes B, Hartvigsen J. Prevalence of lumbar spinal stenosis in general and clinical populations: a systematic review and meta-analysis. Eur Spine J, 2020;29(9):2143–2163. pubmed.ncbi.nlm.nih.gov
  15. Deyo RA, Mirza SK, Martin BI, et al. Trends, major medical complications, and charges associated with surgery for lumbar spinal stenosis in older adults. JAMA, 2010;303(13):1259–1265. pubmed.ncbi.nlm.nih.gov