SI Joint Pain: Real Problem, Over-Used Label — What the Evidence Actually Shows

Key Takeaways: SI Joint Pain

• SI joint pain is a real, recognised source of low back pain. Current consensus estimates it at 15–30% of patients with chronic axial pain located mainly below L5 — a selected group, not everyone with back pain.
• About 90% of low back pain seen in primary care is classified as nonspecific, meaning no single anatomical source can be reliably identified.
• Pain beside the spine rather than over it is common to several pain sources. In one study, this paramidline pattern was reported by 95.0% of patients with facet joint pain and 96.0% of patients with SI joint pain.
• In a diagnostic-block study of 85 patients, no single history feature or examination test reliably identified the joint as the pain source.
• “SI joint pain” may arise from the joint itself or from extra-articular structures around it. A 2025 international consensus groups these as “SI joint complex pain”.
• Positional “dysfunction” tests can be positive in people with no pain at all, and in a small stereophotogrammetric study of 10 patients, manipulation normalised those tests without changing the joint’s measured position.
• In a 2023 meta-analysis, a small direct comparison found no statistically significant difference in pain between exercise and manual therapy, while low-quality evidence favoured exercise for disability.
• In my clinical experience, the tender area in suspected SI joint pain is most often the gluteus medius region, and many of these patients do static work, with pain provoked by prolonged sitting or standing.
• In my experience, manipulation may bring short-term relief but rarely a lasting solution; the most lasting benefit has usually come from a proper active physiotherapy programme built around strength exercises.
• When pain clearly radiates down the leg, in my experience disc-related sciatica is an important possibility, and gluteal muscle tenderness may then be secondary.
• Minimally invasive SI joint fusion has weak or very weak supporting evidence in carefully selected patients; I have not seen any of my own patients undergo SI joint surgery.

Introduction: SI Joint Pain — Real, but Over-Diagnosed

In my clinical experience, SI joint pain is a diagnosis that is often given quickly and with confidence. Pain on one side of the lower back, near the dimple above the buttock, is often attributed to the sacroiliac (SI) joint. That label often comes with a second idea attached: that the joint has become “stuck”, “rotated” or “out of place”, and that a skilled pair of hands can put it back.

The SI joint can genuinely be a source of pain [2]. The more useful question is how often the joint is actually the source, how anyone could tell, and whether the mechanical story that usually accompanies the diagnosis holds up.

This matters because most low back pain does not have an identifiable structural source. A widely cited primary-care review estimates that nonspecific low back pain — where no specific pathoanatomical diagnosis is possible — accounts for about 90% of low back pain seen in primary care, and notes that the tests used to identify anatomical sources of nonspecific low back pain have unacceptably low validity, so such diagnoses are nominal [1].

A large share of the patients I see come to my clinic because their lower back hurts on one side. Many arrive with an explanation already in mind. Some tell me they have SI joint pain, others suspect the piriformis, and some describe it as sciatica. Often they have heard the explanation from someone else, or have been told by a chiropractor or another manual therapy provider that their low back pain comes from SI joint dysfunction. Part of a doctor’s job is to differentiate between these explanations and work out which one is genuine.

In this article, I look at what the research says about how common true SI joint pain is, why one-sided low back pain is so easily attributed to the joint, how often the pain may come from the surrounding ligaments, muscles or other structures rather than from inside the joint, and what the evidence shows about “dysfunction” and manipulation.

Is SI Joint Pain Real? What the Numbers Actually Show

Yes — but in a narrower group than the label suggests.

The most recent large consensus statement, developed by a multispecialty international working group and endorsed by 21 organisations, estimates that SI joint complex pain affects between 15% and 30% of patients with chronic axial pain located predominantly below L5 [2]. That is a meaningful proportion, but it is a proportion of a pre-selected population: people with chronic axial pain concentrated predominantly below L5. It is not a statement that a quarter of all back pain comes from the SI joint.

The early landmark study shows the same pattern. Schwarzer and colleagues performed image-guided SI joint blocks in 43 consecutive patients with chronic low back pain maximal below L5–S1; 13 of them (30%) obtained what the authors described as gratifying relief [3]. The population was again highly selected.

The consensus panel itself also notes that many of its answers were limited by low-quality evidence [2]. It is also worth noting that the guideline was initiated by two pain medicine societies, the American Academy of Pain Medicine and the American Society of Regional Anesthesia & Pain Medicine, and that much of its content concerns interventional procedures such as injections, radiofrequency ablation and fusion [2]. That does not make it wrong, but it is relevant context for a reader weighing how common the condition is.

Evidence summary: Diagnostic injection studies suggest that the SI joint contributes to pain in a subset of selected patients with chronic, low-situated back pain, although these tests have limitations. The figures come from selected specialist populations and do not describe low back pain in general.

Why One-Sided Low Back Pain Gets Labelled as SI Joint Pain

The core problem is that pain location alone cannot reliably establish the pain source.

Pain low in the back, to the side of the spine or in the upper buttock, can arise from several structures. In a study of 170 cases whose pain source had been established with diagnostic procedures, paramidline pain (to the side of the spine rather than directly over it) was reported by 95.0% of patients with facet joint pain and 96.0% of patients with SI joint pain [4]. Midline pain made the SI joint a less likely source — only 12.9% of SI joint cases reported it — but paramidline pain was similarly common in SI joint and facet joint cases, so it did not distinguish between them [4].

The SI joint’s referral pattern can also overlap with territory where other structures produce pain. Injecting the joint in 10 pain-free volunteers produced a zone of altered sensation extending roughly 10 cm downward and 3 cm outward from the posterior superior iliac spine, corresponding to where the injection pain was felt [5]. In 50 patients who responded to an SI joint injection, buttock pain was described by 94.0% and lower lumbar pain by 72.0%, but 50.0% also described pain extending into the lower limb and 14.0% described groin pain; the authors concluded that pain from the SI joint is not limited to the low back and buttock [6]. In other words, the area where SI joint pain is felt may overlap with areas where pain from the lumbar spine or the hip is also felt.

The clinical history does not solve this either. In a study of 85 patients who underwent SI joint blocks, two examiners assessed history and 12 commonly recommended SI joint tests. No historical feature, no single test and no combination of the 12 tests had worthwhile diagnostic value [7]. The only referral pattern Schwarzer found to be associated with a response to the block was groin pain [3] ; buttock pain was not identified as such a discriminator.

Pain near the SI joint may also arise from structures outside the spine altogether. In a series of 200 patients presenting with low back pain to a spine surgeon’s clinic, 14.5% had SI joint pathology and 12.5% had hip pathology, while 10% still had no defined pain source after work-up [8]. The authors concluded that clinicians need to be aware of nonspinal pain generators such as the hip and SI joint [8].

One further differential diagnosis concerns older patients, who may already have hip osteoarthritis. In my experience, when hip osteoarthritis is suspected, a hip X-ray can add useful information. In my experience, patients whose pain comes from the hip present somewhat differently: their pain may at times be felt more in the groin, and hip rotations may be more painful. Even so, imaging findings alone do not establish whether the hip is the source of pain.

Other possible sources of one-sided low back or buttock pain that may enter the differential diagnosis include facet joint pain, gluteal tendinopathy, deep gluteal pain such as piriformis syndrome, lumbar disc problems with referred or radicular pain, and hip joint disease.

In my clinical experience, the area that is most often tender in these patients is the gluteus medius region. Sometimes the pain also sits deeper in the buttock, a presentation often discussed under the label deep gluteal syndrome. When I perform Patrick’s test, however, the SI joint itself frequently does not hurt. When Patrick’s test does reproduce pain near the SI joint, I consider the broader differential diagnosis, including inflammatory sacroiliitis in patients with rheumatic disease when other clinical features raise suspicion. However, the test alone cannot establish the anatomical source or distinguish inflammatory from mechanical pain.

Patients whose pain radiates down the leg form a separate group. In my experience, when there is clear radiating pain in the leg, it has most often turned out to be genuine sciatica, frequently radiculopathy, with the disc as the most common source. However tender the gluteal muscles may be, it can be tempting to conclude that the problem is muscular, but in my experience muscle pain and muscle tightness can also be secondary to an underlying disc problem. A small group may have sciatic nerve irritation related to the piriformis or the deep gluteal region, but how that should actually be diagnosed remains unclear. These presentations are diagnostically difficult because no single investigation necessarily reveals the cause. In selected patients, lumbar MRI can help evaluate disc herniation, nerve root compression and other structural abnormalities when the findings are likely to change management. Imaging findings must still be interpreted alongside the clinical presentation, and a normal MRI does not, by itself, establish a muscular source of pain. Once the necessary exclusions have been made, the pain is often classified as nonspecific, and attributing it to the muscles or static postures is largely a clinical judgement, and how that diagnosis is reached often depends partly on the clinician’s experience.

Joint or Surrounding Tissue? The “SI Joint Complex”

Even when pain clearly comes from the SI region, it does not necessarily come from inside the joint.

The SI joint is held together by a dense network of ligaments, and the region is surrounded by gluteal muscles, fascia and nerves. The 2025 consensus statement therefore uses the term SI joint complex pain, explicitly covering pain from both intra-articular and extra-articular components [2]. The panel concluded that the prevalence rates of intra-articular and extra-articular pathology are comparable — a conclusion based on limited evidence — that an intra-articular anaesthetic injection is a valid diagnostic tool only for pain arising inside the joint, and that a battery of physical examination tests has reasonable sensitivity, but lower specificity, for intra-articular — not extra-articular — pain [2]. The full guideline adds that the prevalence of dorsal ligament-mediated pain is unknown, although comparisons of injection responses suggest that extra-articular pain appears to be at least as common as exclusively intra-articular pain [2].

That has a direct implication: part of what is labelled as an SI joint problem may be pain from extra-articular structures around the joint, for which, according to the consensus, examination tests and intra-articular injections have not been shown to be valid diagnostic tools [2].

A small prospective study, whose authors noted that SI joint pain may originate from both the joint capsule and the posterior ligamentous tissues, illustrates this. Of 25 patients thought to have pain from the SI joint who were given an injection inside the joint, 9 responded; of 25 given an injection into the tissues around the back of the joint, all responded. When the 16 non-responders to the intra-articular injection subsequently received a periarticular injection, it was effective in all of them [9]. Patients were allocated consecutively rather than randomised, so the study cannot settle the question. Its findings are consistent with a possible contribution from periarticular structures, but injection response does not establish the precise anatomical pain source. The consensus also rated the evidence for short-term pain relief as slightly stronger for extra-articular corticosteroid injections than for intra-articular ones [2].

This is the point where the everyday “SI joint” label becomes misleading in two directions. It can be too broad, attributing pain from the hip, spine or gluteal tendons to the joint. It can also be too narrow, implying a single faulty joint when the painful tissue may be the ligamentous and other periarticular structures around it.

In my view, some of what patients experience as SI joint pain may be a pain state in the gluteal muscles, related to static postures held through the working day, rather than a problem inside the joint itself, although the precise pain source often remains uncertain.

In my experience, most patients who have one-sided low back pain and suspect, or have been told about, SI joint pain do static work, and their pain is provoked by static working postures, whether they sit or stand for long periods. In those who stand, the pain may be felt more at the side of the pelvis, in the gluteus medius region, which could perhaps speak more in favour of a muscle-related problem. Usually, though, this has to be interpreted as nonspecific pain, because the exact source cannot be shown. In my experience, the most lasting benefit for most of these patients has usually come only with a proper active physiotherapy programme.

“Dysfunction” and “Out of Place”: Why Positional Findings Mislead

A large part of the SI joint narrative rests on tests claimed to detect a joint that moves asymmetrically or sits in the wrong position. The studies reviewed here do not support these tests.

In a single-blinded study of 101 people with no back pain, 20% had a positive result on at least one of three commonly used SI joint “dysfunction” screening tests (standing flexion, seated flexion and Gillet tests). The authors concluded that apparent asymmetry in SI joint motion can occur in joints that are not painful, and that these tests should not be relied on alone to diagnose symptomatic dysfunction [10].

The positional tests have also been checked against direct measurement. Tullberg and colleagues studied 10 patients whose symptoms and test results indicated one-sided SI joint dysfunction, using roentgen stereophotogrammetric analysis — a high-precision radiographic measurement method — before and after manipulation. The positional tests changed from positive to normal after treatment, but in none of the 10 patients did manipulation change the position of the sacrum relative to the ilium [11]. The authors noted that this neither proves nor disproves a clinical benefit from manipulation [11]; what it does show is that the clinical test results changed without any measurable change in the joint’s position between the pre-treatment and post-treatment assessments.

The concept of an SI joint that is out of place — and the biomechanics, palpation reliability and imaging questions around it — is covered in detail in a separate article, SI Joint Out of Place?. The short version: a change in test results, posture or pelvic symmetry after treatment does not show that a joint was displaced and then repositioned.

Many patients who suspect their SI joint have also visited a chiropractor and been told that the joint is out of place and needs to be put back. Some come to believe that their SI joint is permanently dysfunctional. I explain to them that, as I understand the evidence, manual therapy has not been shown to correct a true displacement of the SI joint, and that the diagnosis itself is open to question.

Can Manipulation Fix SI Joint Pain?

Manipulation may make some people feel better. The evidence does not support the idea that it does so by correcting the joint’s position [11], and it does not show that manipulation outperforms an exercise-based approach [12].

The stereophotogrammetric findings above already separate the two questions: patients’ clinical test results changed, but no measurable change in SI joint position was detected between the pre-treatment and post-treatment assessments [11]. Any benefit, where it occurs, is therefore unlikely to be explained by “putting the joint back”, and the actual mechanism remains uncertain.

On outcomes, a 2023 systematic review and meta-analysis of 8 randomised studies with 422 participants with sacroiliac pain or dysfunction compared therapeutic exercise with other interventions, including manual therapy. However, only three trials with 132 participants compared exercise directly with manual therapy. For pain intensity, no statistically significant difference was detected (mean difference −0.25; 95% CI −2.60 to 2.09; I² = 92%), while low-quality evidence from the pooled disability analysis favoured exercise (mean difference −8.74; 95% CI −10.92 to −6.67) [12]. The overall evidence was rated low to very low quality [12].

The broader evidence base is thin. The 2025 consensus states that evidence for non-interventional treatments for SI joint complex pain is indirect and mostly extrapolated from studies of low back pain in general [2]. For nonspecific low back pain in general, a primary-care review advises that patients be offered information on the nature of low back pain, reassurance about the likely low risk of serious underlying disease, and advice on evidence-based self-management [1].

This matches what I see in practice. Manual therapy, including SI joint manipulation, can bring short-term relief, but in my experience it usually does not resolve the problem in the long term. Some patients are surprised when their symptoms do not ease straight away; others get only very brief relief before the pain returns. Patients are of course free to use manipulation or massage as well; in my experience, these tend to work better alongside active treatment than as a replacement for it.

In my practice, the primary treatment is usually active physiotherapy, typically built around muscle loading, isometric and other strength exercises. In my experience, this has usually given patients the most lasting benefit. Pain medication, including muscle relaxants, is often prescribed for these presentations, and in my experience some patients find it helpful. Corticosteroid injections into the gluteal muscles are sometimes given, and sometimes they work, but in my experience they are not among the most effective options and are generally not first-line. These muscle injections are distinct from the periarticular SI joint injections discussed above. For many patients with low back pain, we also often suggest seeing an occupational physiotherapist. Together they assess movements that can ease the situation, but an equally important part of the work is reviewing and adjusting workplace ergonomics.

In my view, even if a subluxated or dysfunctional SI joint were suspected, I would still favour active rehabilitation over repeated attempts at presumed repositioning. In my experience, manual treatment has not resolved such presentations permanently, and actively maintaining muscle strength remains central here too.

Surgery does exist as an option. The 2025 consensus concluded that, for carefully selected patients with suspected intra-articular SI joint pain supported by controlled diagnostic blocks and an inadequate response to conservative care, there is weak or very weak evidence that minimally invasive SI joint fusion can provide benefit for at least one year [2]. Even so, I have never seen any of my own patients undergo SI joint surgery, which in my view suggests that surgery is not a typical route for these patients in the kind of practice I work in. In my experience, for nonspecific mechanical pain without an underlying disc problem or disc prolapse, conservative care often includes many of the same elements whatever the presumed source, although treatment may need adapting when a more specific diagnosis is established.

For a wider view of passive treatments, see manual therapy for athletes and chiropractic for athletes. For more on one form of active loading, see isometric exercise for pain.

When SI Joint Pain Is More Likely to Be the Real Answer

Scepticism about the label should not turn into dismissal. Some situations make genuine SI region pain more plausible.

Pregnancy and the postpartum period. Pelvic girdle pain — defined as pain between the posterior iliac crest and the gluteal fold, particularly near the SI joints — is common in pregnancy, with European guidelines estimating a point prevalence of about 20% among pregnant women [13]. This is a recognised clinical entity with its own guidelines, distinct from the general “out of place” narrative. Pelvic belts and braces for SI joint and pelvic girdle pain are reviewed separately in best hip braces.

Inflammatory sacroiliitis. Sacroiliitis in axial spondyloarthritis is a different condition from mechanical pain in the SI region. The SI joint out of place article linked above discusses how bedside testing and imaging fit into that work-up.

Pain that behaves differently. Groin pain was the only referral pattern linked to a response to SI joint block in early research [3], while pain directly over the midline of the spine makes the SI joint a less likely source [4]. Leg pain in a spinal nerve distribution, especially with sensory loss, reduced reflexes or weakness, may point toward radicular pain or radiculopathy [1] and is usually worth having assessed by a doctor.

In uncertain cases, the consensus notes that a negative cluster of examination tests is more informative than a positive one, and that imaging evidence for diagnosing SI joint complex pain is unclear or negative [2].

SI Joint Pain in Athletes

Athletes are not exempt from any of the above. They also present with one-sided low back and buttock pain, and the same diagnostic overlap applies: facet joints, gluteal tendons, deep gluteal structures, the lumbar discs and the hip all sit in the same territory. The research above was carried out largely in non-athletic clinical populations, and the studies reviewed here do not establish whether athletes have a higher or lower proportion of pain truly arising from the SI joint than comparable non-athletes.

A training-related trigger can make a mechanical explanation feel obvious. It is worth remembering that a specific event before the pain started does not identify which structure is painful. For lifting-related back pain specifically, see squat and back pain.

Conclusion: SI Joint Pain

SI joint pain is real. In selected patients with chronic pain concentrated below L5, it accounts for an estimated 15–30% of cases [2]. In my experience, however, the label is often applied more broadly than the evidence can support. One-sided low back and buttock pain can arise from the facet joints, the hip, the gluteal tendons and deep gluteal structures, and history and examination have a limited ability to identify the SI joint as the pain source [2][7].

Even when the SI region is the source, the pain may come from extra-articular structures around the joint rather than from inside it [2]. And the idea that the joint is “dysfunctional” or out of place — and can be fixed by manipulation — is not supported by the studies reviewed here: positional tests can be positive in pain-free people [10], and manipulation normalised them without any measurable change in joint position [11].

In my clinical experience, many of these patients do static work, and gluteal tenderness with symptoms aggravated by prolonged sitting or standing is common, but these findings do not establish a muscular pain source, and the pain usually has to be classified as nonspecific. When pain radiates clearly down the leg, I keep a disc-related cause in mind even when the buttock muscles are tender, and in older patients hip osteoarthritis is worth considering. Whatever the label, I have found that the most lasting benefit usually comes from active physiotherapy, with manual therapy, massage, medication or occasional injections playing a supporting rather than a central role, and with occupational physiotherapy and ergonomics helping many working patients.

A more accurate explanation may also be a less frightening one: the evidence does not establish that the pelvis repeatedly slips out of alignment, or that recovery depends on someone putting it back.

References

1. https://doi.org/10.1503/cmaj.170527
2. https://doi.org/10.1136/rapm-2025-107387
3. https://pubmed.ncbi.nlm.nih.gov/7709277/
4. https://pubmed.ncbi.nlm.nih.gov/21257131/
5. https://pubmed.ncbi.nlm.nih.gov/7939978/
6. https://pubmed.ncbi.nlm.nih.gov/10724079/
7. https://pubmed.ncbi.nlm.nih.gov/8961447/
8. https://doi.org/10.1097/BRS.0b013e31818b8882
9. https://pubmed.ncbi.nlm.nih.gov/17530380/
10. https://pubmed.ncbi.nlm.nih.gov/8059269/
11. https://pubmed.ncbi.nlm.nih.gov/9615363/
12. https://doi.org/10.20986/resed.2023.4034/2022
13. https://pmc.ncbi.nlm.nih.gov/articles/PMC2518998/