RightRep EN · PT Open the app
BLOG

Can You Train With Sacroiliac Joint Pain? The Pelvis-Floor Door Audit: The Door Where the Pain Sits Below the Disc's Zip Code and the Internet Has No Address for It, Written for the Reader Whose Back Pain Is Low, Deep and Off-to-One-Side at the Beltline and Who Has Been Handed Either the Rest-It-It's-Your-Back Script or the SI-Joints-Don't-Really-Cause-Pain Script — Where the Epidemiology Column Prices the Sacroiliac Joint at 15–25% of Axial Low Back Pain in Cohen's Comprehensive Review and 15–30% of Pain Mainly Below L5 in the 2025 International Consensus Guidelines, Where Sembrano and Polly's 200-Patient Etiology Workup Found 14.5% Sacroiliac and 12.5% Hip Pathology Sharing the Same Waiting Room, Where Laslett's Cluster Puts Three-or-More of Six Provocation Tests at Sensitivity 94% and Specificity 78% — the Bedside Exam That Outranks an Imaging File the Consensus Itself Rates Unclear or Negative, Where the Randomized Exercise Evidence Lives in the Postpartum Pelvis (Stuge: Evening Pain 30 mm Better and Disability Halved, 85%-vs-47% Minimal Disability at Two Years — With the Honest Regression Leg Printed Beside Them), Where the Procedure Column Shows Its Decay Curve (the Brazilian Stepped-Care File: 82% Improved After Intra-Articular Steroid at Pain 8→2 cm for 25 Weeks; Radiofrequency 79% at One Month but 14% at One Year), and Where Surgery Finally Has a Real RCT (INSITE: 148 Randomized, Fusion −55.4 VAS Points at 24 Months, the Non-Surgical Arm's Disability Improvement Under 10%)

By RightRep AI Assistant · · 23 min read

Sacroiliac joint pain is the file where the gym's two default scripts — it's your back, rest it and SI joints don't really cause pain, push through — both fail the reader at the exact door they walked in through, because this reader's pain sits below the disc's zip code: low, deep, off-to-one-side at the beltline, often into the buttock or the groin, worse on one leg loaded than two — and the internet has no address for it. The rest-it script dies on the epidemiology column: the sacroiliac joint carries 15–25% of axial low back pain [1] — up to 30% of pain mainly below L5 in the 2025 international consensus [5] — and on the treatment ladder where a stabilizing-exercise randomized trial halved disability [7][8] and a stepped-care program improved 82% of confirmed cases without surgery [11]. The push-through script dies on the same file's fences: the inflammatory differential that routes new bilateral, morning-stiff sacroiliac pain to a rheumatology screen and the ankylosing spondylitis door, the red-flag column every spine door in this library carries, a procedure column whose relief decays on a visible curve [6], and a fusion trial strong enough that it deserves to be read before the consult, not after [12]. 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 back-surgery door owns the cauda equina details). Second: leg-dominant pain, numbness or true weakness in a dermatomal pattern routes through the sibling doors — the herniated disc file, the sciatica file — because a talking nerve root is a different door than a talking pelvic joint [3][5]. Third: the inflammatory pattern — new pain, both sides, under age 45, worst in the second half of the night and morning stiffness over 30 minutes — earns a rheumatology screen before a barbell program, because sacroiliitis is the one thing that wears this door's costume and answers to different medicine [5]. And fourth: this file assumes the workup is happening — the sacroiliac joint is the one major low-back pain source where imaging helps least [5], which makes the clinician's hands the diagnostic instrument and the reader's job to arrive with the right map, not a self-diagnosis [4]. With the fences standing, the evidence underneath is unusually practical for a joint the internet argues doesn't exist: it is real, it is common, it is exam-diagnosed, it answers to exercise and selective procedures, and the training question has a cleaner answer than most doors in this library — the joint's job description is load transfer.

The short answer: yes — load transfer is literally the job description

Strip the file to its spine. The joint is real and common: the sacroiliac joint accounts for 15–25% of axial low back pain in the comprehensive review [1], 15–30% of pain mainly below L5 in the 2025 multispecialty consensus [5], and when Sembrano and Polly worked up 200 non-radicular low-back-pain patients properly, 14.5% carried sacroiliac pathology and another 12.5% hip pathology — the low-back waiting room is not a monoculture [2]. You do not diagnose it off a checklist, but your clinician can diagnose it off an exam: three or more of Laslett's six provocation tests ran at sensitivity 94% and specificity 78% [4] — and the imaging file the reader keeps asking about is rated unclear-or-negative by the consensus itself [5]. The strongest randomized exercise evidence in this joint's file is postpartum: Stuge's specific stabilizing-exercise program for pelvic girdle pain beat standard physical therapy — evening pain 30 mm better on the visual analog scale and disability cut by more than half post-treatment [7] — and at two years, 85%-vs-47% minimal disability and 68%-vs-23% minimal evening pain, with the honest regression leg printed beside the numbers [8]. The procedure column works and decays: intra-articular steroid improved 82% of a Brazilian stepped-care cohort at pain 8→2 cm lasting to 25 weeks [11], and radiofrequency denervation ran 79% success at one month but only 14% persistent relief at one year [6]. And the surgery column finally has a real randomized trial: INSITE's 148 patients, fusion at −55.4 VAS points and 83.1% at ≥20-point improvement through 24 months, against a non-surgical arm whose disability improvement stayed under 10% [12]. That is the audit in one paragraph: this reader needs the costume cleared (nerve, inflammation, red flags), the exam booked, and a training program that respects load transfer in both legs — which is not a limitation on training, it is a specification.

What the joint actually is: the keystone that doesn't move much, on purpose

The mechanics in a paragraph: the sacroiliac joints are the two L-shaped synovial joints where the sacrum — the wedged bone carrying the entire spinal column — meets the iliac wings of the pelvis, and their entire physiological job is force closure: transferring every load from trunk to leg and every ground reaction from leg to trunk across a joint that moves only millimeters, stabilized less by its shape (form closure) than by the compression of ligaments and muscles screwing it shut — the gluteals, the deep rotators, the pelvic floor, the latissimus via the thoracolumbar fascia, and above all the transversus abdominis and multifidus acting like a corset pre-tensioned before every load [1][5]. That anatomy explains this door's entire symptom signature: pain at the beltline (the joint's location), worse on single-leg loading (stairs, lunges, hopping, running gait's stance phase — one leg now carries the whole transfer alone), worse with long standing or sitting in a soft chair (sustained compression without muscular pre-tension), the classic difficulty rolling in bed or getting out of a car (the transitional moment when the joint switches loading strategy), and referral that stays below the knee — groin, buttock, thigh — because the joint's referral map barely reaches the calf, unlike a talking L5 or S1 root [1][3]. And it explains why this is the one door where the scan is quietest: the joint's pain generators are capsular and ligamentous stretch and compression, not structural damage an MRI frame captures, which is exactly why the 2025 consensus rates imaging unclear-or-negative for this diagnosis and puts the weight on the provocation exam and, in selected cases, the diagnostic block [5]. The reader whose MRI "came back clean" and whose beltline still hurts is not imagining anything — they are standing in the one room of the spine house where the camera doesn't work and the hands do [4][5].

The epidemiology column: the waiting room is not a monoculture

How common is this door, really? The base rate comes from three directions and they agree. Cohen's comprehensive review prices the sacroiliac joint at 15–25% of axial low back pain — roughly one in five chronic back-pain patients walking through a pain clinic's door [1]. The 2025 multispecialty international consensus — pain physicians, surgeons, physiatrists — carries the estimate as 15–30% of patients with axial pain mainly below L5, calling the joint "an underappreciated source of chronic low back pain" [5]. And the etiology census that proves the point prospectively: Sembrano and Polly worked up 289 consecutive low-back-pain patients, excluded 77 with prior surgery, and diagnosed the remaining 200 — 82% carried spine pathology, 14.5% (29 patients) sacroiliac pathology, 12.5% hip pathology, 17.5% mixed spine-plus-hip/SI, 8% hip/SI only, 10% undefined [2]. Read that column the way a coach reads it: in any gym's low-back-pain population, roughly one in six or seven members is not a disc case at all — they are a pelvis case, and many of them are stretching their hamstrings for a problem that lives two joints away [1][2][5]. The original block-diagnosed cohort sharpened the location rule: among 43 chronic low-back-pain patients whose pain was maximal below L5 — this door's exact fingerprint — 13 (30%) obtained gratifying relief from fluoroscopically guided intra-articular sacroiliac blocks, and nothing on the arthrogram predicted who [3]. The zonal rule is the takeaway: the lower the pain's center of mass sits, the higher this joint's share of the probability — which is why the fences and the routing come before any programming [2][3].

The diagnosis column: the exam outranks the picture

Here is the section that separates this door from every other spine door in this library, and the reader needs it stated plainly: the sacroiliac joint is the one major low-back pain source where the imaging file is quietest and the bedside exam carries the diagnostic weight. The instrument with the numbers on it is Laslett's cluster: six provocation tests — distraction, compression, thigh thrust, sacral thrust, Gaenslen's, the sacroiliac stretch — each loading the joint to ask it directly whether it is the one talking; three or more positive ran at sensitivity 94% and specificity 78% against double diagnostic blocks as the reference standard, and any two of the four best tests carried an area under the curve of 0.842 [4]. Against that, the 2025 consensus prices the rest of the diagnostic stack with unusual candor: exam maneuvers have reasonable sensitivity but lower specificity — and negative tests are worth more than positive ones; intra-articular injections have diagnostic validity for intra-articular but not extra-articular pain; the evidence for imaging is unclear or negative; and the block thresholds are stated outright — 50% relief to call a diagnostic block positive, 30% for therapeutic decision-making [5]. The gym translation is not "self-diagnose off a YouTube checklist" — the cluster is a clinician's instrument, performed in sequence, interpreted against blocks [4][5]. The gym translation is two things: first, why the clean MRI doesn't clear you — the camera is the wrong instrument for this joint [5]; and second, what to ask for — if your pain fingerprint fits (low, unilateral, beltline, single-leg-loaded) and the spine workup is dry, the question to bring to the clinician is whether the provocation exam and, if needed, a diagnostic block have been done [3][4][5]. The reader who arrives at that appointment with this map gets to the answer faster than the reader who arrives with a list of exercises they found online.

The pregnancy column: the population the internet forgets trains

One column of this door's evidence deserves its own section because it is both the joint's biggest risk factor and the source of its best randomized data. Pregnancy relaxes the pelvic ligaments and adds a growing anterior load — and the sacroiliac and pubic joints pay for it. The classic cohort: 49% nine-month period prevalence of back pain in pregnancy, point prevalence 22–28% from week 12 to delivery — and crucially, the pain-drawing analysis showed the group localized to the sacroiliac areas was the group whose pain worsened as pregnancy advanced, while true dermatomal sciatica stayed rare at 1% [9]. That finding — sacroiliac-pattern pain that ramps through pregnancy — is the signature this door exists for. The Brazilian anchor puts local numbers on it: in 203 pregnant women interviewed in primary care in Paulínia, São Paulo, 79.8% reported back pain, 80.8% localized lumbar, 49.1% localized sacroiliac — half of pregnant women with back pain pointing at this joint's territory — with pain more frequent in younger women and, interestingly, not climbing with gestational progression [10]. And the postpartum tail is where the training story lands: pelvic girdle pain does not always clear at delivery, and it was exactly this population — postpartum pelvic girdle pain, sacroiliac-dominant — that provided the randomized exercise column below [7][8]. The cross-links for this reader: the training-through-pregnancy door and the after-giving-birth door own the trimester-by-trimester and return-to-training programs; this file owns the joint those programs need to respect.

The exercise column: the randomized evidence, and its honest legs

The treatment question a coach actually cares about: does specific exercise beat generic care for this joint? The best randomized answer in the file is Stuge's Oslo trial — 81 women with postpartum pelvic girdle pain, randomized to a treatment program focusing on specific stabilizing exercises (the transversus abdominis and multifidus corset, trained to fire before load, then integrated into function) or to standard physical therapy over 20 weeks: post-treatment, the exercise group's evening pain was 30 mm better on the visual analog scale and their disability fell by more than half, with negligible change in controls; at one year they still held statistically and clinically significant advantages in pain, disability and quality of life [7]. The two-year follow-up is the number column to remember: 85% of the specific-exercise group at minimal disability versus 47% of controls, and 68% at minimal evening pain versus 23% [8]. Now the honest legs, because this library prices them: the control group improved too (median change 6.0 on the function scale); when the regression adjusted for one-year scores, the between-group differences disappeared for all measures; and the worst-baseline participants recovered most regardless of arm [8]. And the candid floor beneath the whole non-interventional column: the 2025 consensus states outright that the evidence base for non-interventional therapies is indirect, extrapolated mostly from low back pain studies — the specific-exercise file for this joint is thin, postpartum-heavy, and honest about it [5]. So what does a coach do with a trial that won at two years and dissolved under regression adjustment? The same thing the field does: program the corset — because it won every unadjusted comparison, it cost nothing, and beneath it sits the wide floor of the general exercise-for-back-pain file where the Cochrane-grade effects live (the back-pain floor door prices that column) [5][7][8]. The gym translation: stabilization work here is not magic — it is the joint's own anatomy trained deliberately (the force-closure corset from the mechanics section), stacked on the ordinary progressive-strength floor every other door in this library stands on.

The procedure column: works, and shows its decay curve

This is the door where many readers will meet injections and radiofrequency before they meet a barbell program, so this file prices that column with the curve visible. The stepped-care Brazilian file is the cleanest single arc: 192 low-back-pain patients who had already failed conservative care were consecutively examined — 26% carried sacroiliac joint dysfunction (the refractory-end prevalence, consistent with the epidemiology column [1][5]); of the 50 treated, 82% reported pain and quality-of-life improvement after a single intra-articular corticosteroid block, with mean pain falling from 8 to 2 cm and holding to 25 weeks, and rescue analgesic consumption dropping; the 18% without lasting improvement at week three went to cooled radiofrequency — 7 of 9 successful; and only 4% of the whole cohort reached a surgery recommendation [11]. The randomized placebo-controlled leg on radiofrequency: Cohen's trial ran 79% success (≥50% relief) at one month, 64% at three months, 57% at six months — against a placebo arm at 14% at one month and 0% at three — but at one year, only 2 patients (14%) in the treatment arm still held their relief [6]. The consensus file prices the rest: steroid injections buy at least 4 weeks of relief in well-selected patients (the evidence slightly stronger for extra-articular than intra-articular), dextrose prolotherapy and PRP carry weak evidence for at least 3 months, and the whole interventional column earns its place inside an interdisciplinary, multimodal plan — not instead of one [5]. The gym translation for a reader scheduled around these procedures: train through the calendar, not around it — the exercise column above is the multimodal plan's base, the injection window is when symptom-limited arcs are easiest to find, and the decay curve is why the dated ledger below matters more for this door than for most: relief that fades on a schedule is a program variable you can see coming [5][6][11].

The surgery column: finally a real RCT — priced honestly both ways

Most chronic-pain doors in this library carry surgery files with marginal or negative verdicts. This one is different, and the reader deserves the difference stated cleanly. INSITE randomized 148 patients with chronic sacroiliac joint pain to minimally invasive fusion with triangular titanium implants (102) or non-surgical management (46, with crossover to surgery allowed after six months): at six months the surgical arm's VAS improvement outpaced non-surgical by 38.3 points (p<0.0001); at 24 months the fusion arm held a mean 55.4-point VAS improvement, 83.1% reached ≥20-point clinical improvement, 82.0% substantial clinical benefit, and 68.2% cleared the ≥15-point ODI disability threshold — while the non-surgical arm's disability-improvement rate stayed under 10% — with only 3 revision procedures in the fusion arm across 24 months [12]. That is a strongly positive randomized file, and this library reports it as one. The honest legs, also from the file itself: the non-surgical comparator was 2016-standard care — not the stabilizing-exercise-plus-stepped-injection ladder of the columns above [7][11]; the trial enrolled patients who had already failed that standard care, so it prices surgery late, not first; and the conflict-of-interest disclosures of device-industry funding are part of the published record a reader walking into a consult is entitled to weigh [12]. The gym translation is not "avoid surgery" and not "surgery works, stop reading" — it is that the conservative ladder gets a genuine, unhurried first turn (the stepped-care file reached 4% surgery [11]), and if fusion does happen, the return-to-training question routes through the same post-surgical door rules every fusion reader follows, with the implant-side specifics owned by the surgical team [12].

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

Trains always: the whole library, dosed symmetrically. Whole-body progressive strength two-to-three times a week on the ordinary floor (the machines-vs-free-weights question answers to preference and tolerance here, not to the joint) [5][7] — plus the corset work this joint specifically answers to: the deep-corset activation family (dead-bug and bird-dog progressions, carries) trained to fire before load, exactly the shape Stuge's program prescribed and the warm-up makes routine [7][8]. Symmetry is the joint's love language: prefer split-stance and unilateral patterns that load both sides evenly across the week (split squats, lunges, step-ups, single-leg hinges) — not because bilateral lifts are forbidden, but because single-leg work is simultaneously this joint's provocation test and its training stimulus, dosed small and grown slowly [1][5].

Ramps single-leg load like a beginner, whatever your bench press says: the joint's symptom trigger is single-leg load transfer [1][3] — so stairs-before-lunges-before-hopping-before-sprinting is a ladder, running comes back through walk-run intervals on the zone-2 pattern before speed work, and the deadlift family loads the joint honestly and is trainable here — what earns respect is the ramp, not the avoidance [5][7].

Re-doses during flares, never cancels: flare = symmetry bias (double-leg and supported patterns while the joint calms), volume and impact down, pain-free arcs only, keep moving — the same architecture every spine door in this library runs, and the direction the postpartum trial's control group confirmed improves on its own timeline [7][8]. A belt-line flare after a heavy asymmetric day is information for the ledger, not a cancelled program.

Stops and refers: the red-flag column from the intro is today-level care; the inflammatory fingerprint (new, bilateral, under 45, second-half-of-night worst, morning stiffness >30 min) is a rheumatology screen, not a program [5]; dermatomal leg patterns route to the herniated and sciatica doors; and a pain fingerprint that fits this door but hasn't had the exam earns the clinician's provocation cluster, not a self-diagnosis [3][4].

Keeps one dated ledger: sessions, pain (0–10 at rest, after sessions, and after single-leg days specifically), procedure dates if any — one timeline. With injections in the picture, the ledger is also how the decay curve gets seen rather than felt [6][11].

The Brazil anchors: the stepped-care file and the prenatal census

The pt-BR twin of this file exists for a Brazilian audience with two local anchors the English reader also benefits from knowing. The stepped-care file: 192 Brazilian patients with low back pain after failed conservative care, evaluated prospectively — 26% carried sacroiliac joint dysfunction — and the ladder that followed treated them without surgery for 96% of the cohort (82% improved on a single steroid block held to 25 weeks; 7 of 9 radiofrequency patients succeeded; 4% to surgery) [11]. The prenatal census: Paulínia, São Paulo, 203 pregnant women in primary care — 79.8% with back pain, and 49.1% of them localizing it to the sacroiliac area — a number that says half of pregnant Brazilian readers pointing at back pain are pointing at this door's territory [10]. Beside them, the São Paulo low-back-pain census anchors the general floor this door sits inside (point prevalence 9.8%, one-year 48.1%) — the pt-BR twin cross-links rather than re-audits.

So what do you actually do? The practical file

Six rules, every one sourced from the columns above:

  1. Clear the costumes before you program the joint. Red flags today; dermatomal leg pain to the herniated/sciatica doors; the inflammatory fingerprint to rheumatology [5]; and if your low, unilateral, belt-line pain fits this door and hasn't had the exam, book the provocation cluster — three-or-more of six at Se 94%/Sp 78% is the instrument that finds this joint [3][4].
  2. Stop waiting for the MRI to answer. Imaging is the wrong instrument for this joint — the consensus says unclear-or-negative, the exam and the block carry the weight [5]. A clean scan with this pain fingerprint means the camera looked at the wrong thing, not that you invented the pain.
  3. Train the corset deliberately, stack the ordinary floor under it. Deep-corset activation before load, integrated into function — the program shape that halved disability and still held 85%-vs-47% minimal disability at two years [7][8] — on top of ordinary progressive strength, because the general exercise floor is where the field admits its evidence is borrowed from [5].
  4. Ramp single-leg work like a beginner. One-leg load transfer is this joint's job and its trigger — lunges, step-ups and run-walk intervals grow on a ladder, whatever the barbell numbers say [1][3].
  5. Price the procedures with the curve visible. The steroid block's 25-week arc, radiofrequency's 14%-at-one-year honesty [6][11], and the consensus's ≥4-week / ≥3-month pricing tags [5] — the ledger around a procedure calendar is how faded relief gets caught early and re-dosed.
  6. Read INSITE before the fusion consult, not after. Surgery late prices at −55.4 VAS points over 24 months with 83.1% improving [12]; the conservative ladder treats 96% of even the refractory cohort without it [11]. Both tables belong in the room where the decision is made.

How RightRep fits into this

The SIJ file is, at its core, a symmetry-and-ramp question — how much single-leg load, loaded how evenly, at what progression — and neither is answerable from memory:

  • Symmetry made measurable. The joint flares on asymmetric loading [1][3]; camera-scored reps give both sides the same counting and the same form standard, so left-right drift shows up as data long before it shows up as a flare.
  • The ramp governed. Stuge's program was progressive by design [7] — logged sets, reps and single-leg progressions make the ladder a number instead of a feeling.
  • The flare instrument. Pain logged beside sessions — with single-leg days tagged — turns "re-dose, don't cancel" into a measurable set point: which session size the joint actually tolerates [7][8].
  • The dated ledger, built in. One timeline of sessions, pain and procedure dates is the artifact that sees the decay curve coming and makes the next consult concrete [6][11].

The bottom line

  • The epidemiology: the sacroiliac joint carries 15–25% of axial low back pain [1], 15–30% of below-L5-dominant pain [5]; of 200 worked-up LBP patients, 14.5% sacroiliac and 12.5% hip [2]; of 43 whose pain peaked below L5, 30% block-confirmed [3].
  • The diagnosis: three-or-more of six provocation tests at Se 94%/Sp 78% [4]; imaging rated unclear-or-negative; negative tests worth more than positive; 50%/30% block thresholds — the hands outrank the camera at this door [5].
  • The pregnancy column: 49% period prevalence, the sacroiliac-localized group worsening through pregnancy, sciatica 1% [9]; Brazil — 79.8% back pain in pregnancy, 49.1% sacroiliac-localized [10].
  • The exercise column: Stuge — evening pain 30 mm better, disability halved [7]; two years 85%-vs-47% minimal disability, 68%-vs-23% minimal evening pain, with the honest regression legs printed [8]; the consensus's candid floor — non-interventional evidence borrowed from the LBP literature [5].
  • The procedure column: Brazil's stepped care — 26% of refractory LBP; 82% improved on one steroid block at 8→2 cm to 25 weeks; 7-of-9 radiofrequency; 4% surgery [11]; radiofrequency's randomized decay — 79%/64%/57% at 1/3/6 months, 14% at one year [6]; consensus pricing — steroids ≥4 weeks, prolotherapy/PRP weak at ≥3 months [5].
  • The surgery column: INSITE — 148 randomized, fusion −55.4 VAS points at 24 months, 83.1% ≥20-point improvement, non-surgical disability improvement under 10%, three revisions [12] — priced as surgery late, after a conservative ladder that treats 96% without it [11].
  • The Brazil anchors: the 192-patient stepped-care file [11] and the 203-woman Paulínia prenatal census [10].

Clear the costumes, book the exam, train the corset on the ordinary floor, ramp the single-leg work like a beginner — and let the camera count the symmetric 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. Cohen SP. Sacroiliac joint pain: a comprehensive review of anatomy, diagnosis, and treatment. Anesthesia & Analgesia, 2005;101(5):1440–1453. pubmed.ncbi.nlm.nih.gov
  2. Sembrano JN, Polly DW. How often is low back pain not coming from the back? Spine, 2009;34(1):E27–E32. pubmed.ncbi.nlm.nih.gov
  3. Schwarzer AC, Aprill CN, Bogduk N. The sacroiliac joint in chronic low back pain. Spine, 1995;20(1):31–37. pubmed.ncbi.nlm.nih.gov
  4. Laslett M, Aprill CN, McDonald B, Young SB. Diagnosis of sacroiliac joint pain: validity of individual provocation tests and composites of tests. Manual Therapy, 2005;10(3):207–218. pubmed.ncbi.nlm.nih.gov
  5. McCormick ZL, et al. Consensus practice guidelines on sacroiliac joint complex pain from a multispecialty, international working group. Regional Anesthesia and Pain Medicine, 2025. pubmed.ncbi.nlm.nih.gov
  6. Cohen SP, Hurley RW, Buckenmaier CC 3rd, et al. Randomized placebo-controlled study evaluating lateral branch radiofrequency denervation for sacroiliac joint pain. Anesthesiology, 2008;109(2):279–288. pubmed.ncbi.nlm.nih.gov
  7. Stuge B, Veierød MB, Laerum E, Vøllestad N. The efficacy of a treatment program focusing on specific stabilizing exercises for pelvic girdle pain after pregnancy: a randomized controlled trial. Spine, 2004;29(4):351–359. pubmed.ncbi.nlm.nih.gov
  8. Stuge B, Veierød MB, Laerum E, Vøllestad N. The efficacy of a treatment program focusing on specific stabilizing exercises for pelvic girdle pain after pregnancy: a two-year follow-up of a randomized clinical trial. Spine, 2004;29(10):E197–E203. pubmed.ncbi.nlm.nih.gov
  9. Ostgaard HC, Andersson GB, Karlsson K. Prevalence of back pain in pregnancy. Spine, 1991;16(5):549–552. pubmed.ncbi.nlm.nih.gov
  10. Martins RF, Pinto e Silva JL. Prevalência de dor nas costas na gestação. Revista da Associação Médica Brasileira, 2005. pubmed.ncbi.nlm.nih.gov
  11. Lopes BCP, et al. Sacroiliac joint dysfunction and lumbar pain: evaluation in a Brazilian population. Acta Ortopédica Mexicana, 2023;37(4). pubmed.ncbi.nlm.nih.gov
  12. Polly DW, Swofford J, Whang PG, et al. Two-Year Outcomes from a Randomized Controlled Trial of Minimally Invasive Sacroiliac Joint Fusion vs. Non-Surgical Management for Chronic Sacroiliac Joint Pain. International Journal of Spine Surgery, 2016;10:28. pubmed.ncbi.nlm.nih.gov