The sacroiliac joint: a critical review.

Clinicians continue to focus attention on dysfunction of the sacroiliac joint (SIJ) as one cause of low back pain. Numerous assessment procedures and treatment approaches are described; however, little evidence has been presented to document their reliability, validity, or effectiveness. The SIJ is a curious and unique joint, definitely involved in a wide variety of established pathological conditions, affecting individuals across the life span. Because of its location, orientation, and morphology, the SIJ is a difficult joint to visualize clearly with radiographic procedures; that increases the problems in diagnosis of SIJ pathology. A considerable body of literature now exists on the SIJ. The purpose of this article is to review this literature on age changes and motion of the SIJ to determine the support for clinicians' current beliefs, opinions, and theories--often stated as facts--that provide the basis for SIJ assessment and treatment.

A few clinician-investigators recognize reports of very limited mobility in the SIJ and structural changes with aging.[1] Most manual therapists, however, be they physical therapists, chiropractors, or osteopaths, appear to believe that appreciable motion exists in the SIJ, that careful clinical testing can isolate SIJ dysfunction, and that specific treatment procedures can affect SIJ dysfunction. Sacroiliac joint dysfunction is variously termed subluxation, "upslip," "downslip," or posterior or, more frequently, anterior fixed innominate.[2-5] More than 15 specific tests for SIJ dysfunction are described.[2,3,5-7] These tests are basically of two types: (1) palpation of bony landmarks with or without measurement and (2) pain provocation tests. Disparity in leg lengths, functional or structural, and pelvic muscle length asymmetry are considered prime factors in detection of SIJD.[8] Treatment techniques range from use of sacral belts to localized oscillatory motions, larger range mobilization or manipulative procedures (eg, in which the leg may be used as a lever), and teaching patients to manage their back problems.[1,2,4-12] Evidence for reliability and validity of assessments and the effectiveness of treatment procedures will be discussed later in the article following a review of age-associated changes in the SIJ and mobility studies.

Interest in the SIJ has existed since the time of Hippocrates, who observed that a woman's pelvis separated in first labor and remained so thereafter.[13] Early studies were focused on the potential for motion and changes in motion during and after pregnancy.[13-18] These observations promoted interest in developmental and aging changes.[17,19-24] Studies in the last two decades have started to provide quantitative data[25-30] that were absent or inadequate in most of the early reports, and that are still frequently cited in review articles and will be in this review.[1-8,11,31]

Morphology

The basic anatomy of the SIJ is well described in standard anatomical texts[32] and will not be reviewed. The joint is part synovial and part a syndesmosis; it is not amphiarthrodial. Debate on the type of joint largely arose from early descriptive and light microscopic studies that repetitively demonstrated differences between the sacral and iliac surfaces, the iliac being more fibrous or fibrocartilaginous.[17,19,20]

Surface Composition and Thickness

Paquin et al,[33] with electron microscopy (EM) and biochemical techniques, studied tissues from five men. Although paucity of articular cartilage from the first sacral vertebra required pooling of tissues from four subjects and disallowed full study of iliac cartilage, electrophoresis demonstrated that collagen findings corresponded with Type II collagen peptides that are characteristic of hyaline cartilage. With EM, sacral cartilage resembled normal articular cartilage "in all aspects," whereas the iliac cartilage, throughout its depth, resembled only the superficial zone of typical articular cartilage. Paquin and colleagues concluded that the iliac cartilage had an atypical organization. This is the only study I was able to locate that used biochemistry to study the SIJ, and it must be remembered that it was conducted on 1-cm tissue cubes from a very small sample (N=5) with a narrow age range (49-60 years). Only one other study[34] used EM and made similar observations on the ultrastructural characteristics of the joint surfaces.

The greater thickness (depth) of the sacral cartilage compared with the iliac cartilage, with the sacral surface having a 1.5:1 to 3:1 greater thickness than that of the iliac, has been consistently reported (Tab. 1).[19-21,24,27,33-38] Practically all gross and microscopic studies have been conducted on postmortem and cadaver specimens, usually with the health histories of the subjects unknown. Paquin and colleagues'[33] small sample, stated to have no joint diseases, also showed a 2:1 greater thickness of the sacral cartilage compared with the iliac; the iliac surface had a higher cell density. [TABULAR DATA 1 OMITTED]

Differences in thickness of sacral and iliac cartilage are observed even in fetal specimens.[30,34,37] This suggests that the disparity is not a result of differential life stresses and strains on the two surfaces. The variability in composition and thickness of the joint surfaces may result in earlier and greater wear of the iliac surface. This is supported by observations in a number of studies.[30,34,37]

These observations of joint surface differences are not unique to humans. Ekman et al[39] observed similar differences in the composition and thickness of the two surfaces in horses with no known pathologies. No differences, however, were noted in dogs.[40] Both species are reported to have SIJ dysfunction.

Few investigators of cartilage depth and surface contours adequately reported their measurement techniques to permit replication. The basis for reported average thickness values is generally unclear, and it is unknown whether measurements were taken at maximum width or averaged from several points across the entire surface, or from a small sample of the total surface. Several studies[30,34,37] have demonstrated variability in SIJ cartilage thickness across the entire joint surfaces (Fig. 1).

Stability

The SIJ is not crossed by any muscle. All adjacent muscles (ie, the quadratus lumborum, erector spinae, gluteus maximus, gluteus minimus, piriformis, and iliacus muscles and even the more distantly located latissimus dorsi muscle) have fibrous expansions that blend with the anterior and posterior SIJ ligaments and contribute to the strength of the joint capsule and ligaments, and thus to the joint's stability. This fascial reinforcement is greater posteriorly, as more muscles are adjacent to the joint on that aspect. This gives support to clinicians' theories[1-3,11] that the joint is weaker anteriorly; however, the presence of the strong posterior interosseous ligament brings into question the assumption of anterior capsule weakness. Additionally, the tissues derived from muscle expansions may be placed in tension when the muscle bellies contract. Muscle activity, therefore, is likely to increase any symptoms arising from SIJ pathology.

The interosseous SIJ ligament is often described as the strongest ligament in the body, principally responsible for the stability of the joint.[23,41] Scholten et al,[42] from a computer-simulated biomechanical model that used the displacement data reported by Miller et al,[27] concluded that "in no case did ligaments seem to have a major role in pelvic mechanics." Scholten and colleagues agreed with Slocumb and Terry[41] that the SIJ ligaments limit flexion of the sacrum relative to the ilium. Authors of review articles often describe detailed roles for both muscles and ligaments in pelvic and SIJ stability while citing authors who provided no original data to support these opinions.[2,3,11]

The few investigators[27,38,42,43] who have studied loading and load displacement behavior of the SIJ have used cadaver material stripped of muscle that can only approximate in vivo behavior. Even when cadaver material is "fresh," tissue changes occur that decrease tissue resilience and pliability. Vukicevic and associates[43] used a holographic method to study static vertical loading. They subjected the SIJs to loads (pressure) between 15 and 200 kp in comparison with the computer model of Scholten et al,[42] which applied forces (torque) of 1,000 N and moments of <25 N.m. Both groups of investigators reached similar conclusions that the sacrotuberous and sacrospinous ligaments had no effect in preventing pelvic displacement after the maximal movement had occurred.

With significant trauma, such as that of high-impact injuries, there is a greater tendency for diastial (through the joint) fractures to occur than displacement of the SIJ surfaces.[36,44] Miller et al,[27] in their static load displacement studies, applied loads of up to 294 N and moments of up to 42 N.m. They demonstrated ligamentous failure in torsion tests but failure of bone medial to the SIJ in other motion tests in which the ilia were fixed. Eight single joints resisted loads of up to 1,440 N and 160 N.m without overt failure. Although their sample of eight fresh cadavers was scanned with computed tomography (CT) to eliminate pathology, the sample was small, with a narrow and older age range (59-74 years).

Despite such studies, manual therapists believe that by using manual pressure they can effect a "reduction" of what is variously termed an "upslip" or fixed innominate; reduction has been described as being evidenced by an "audible clunk."[2,3] Claims for the presence and then reduction of a subluxation require verification, such as pretreatment and posttreatment radiographs or the use of other imaging techniques.

Joint Surface Morphology and Orientation

The SIJ surfaces in fetal samples are smooth and flat.[19,21,22,27,34,37] Weisl[22] conducted a cartographic study of 62 recent postmortem and cadaver specimens, including 11 full-term fetuses. He observed that in the fetuses the joint surfaces were reciprocally shaped but not congruent. Elevations on the sacral surface varied in "average maximal" height from 2 mm at term to 11 mm at over 50 years of age. These elevations, located mainly cranially and caudally, were easily identifiable in only 21 specimens (Fig. 2). Weisl concluded that, with aging, interlocking between surface elevations and depressions could occur, with ankylosis commencing after 30 years of age. Kissling et al,[45] using more precise moire photogrammetry, displayed similar surface irregularities as Weisl; however, only 4 specimens were studied.

Few reports exist on the SIJ surfaces of children or adolescents. Vleeming et al,[46] however, examined both surfaces and stated they observed ridges and depressions in one 12-year-old boy. There is general agreement that joint surface irregularities increase with age. Vogler et al[26] reported a CT study of 45 subjects who had no SIJ symptoms. All subjects were examined for pelvic disease, and, in 10 subjects under the age of 30 years, no detectable asymmetry in the "joint space, subchondral sclerosis, and overall appearance"[26(p434)] was observed. Asymmetry, however, was present in more than 77% of those over age 30 years (30->59 years). The authors concluded that asymmetry in subjects over age 30 years may indicate abnormality.

With Vernier calipers, Weisl[22] measured the height of the surface irregularities of articular cartilage of both surfaces of the SIJ. His findings support descriptive observations made in many studies. The frequency with which elevations are observed on the SIJ surfaces, as well as splitting of the cartilage,[17,19-21,24,30,34] has led several investigators to conclude that these changes are normal age-related changes, reflecting the stresses and strains to which the joint is exposed, and are not pathological.[24,46,47]

Vleeming et al[46] emphasized that intra-articular ridges and depressions may be misinterpreted on roentgenographs as osteophytes. These investigators, grossly and microscopically, examined 47 specimens and determined the coefficient of friction (COF) on bone-cartilage samples taken from both joint surfaces. Measurements were made within 4 hours of autopsy. They concluded that surface irregularities had greater influence on the COF than surface texture. The highest COFs were shown to be associated with samples displaying coarse surface texture and surface irregularities. These results give no support to the theorizing by "muscle energy" proponents that fibrocartilage on the iliac surface facilitates motion between the irregular surfaces (Meadows J, unpublished course notes). Vleeming and colleagues[46,47] theorized that because SIJs in men are exposed to greater torques than SIJs in women, the complementary intra-articular elevations and depressions may enhance SIJ stability in men. Because torques may be greater in larger persons, not just in males, there is a need to consider torque in relation to the joint surface area squared.

Another factor contributing to the unique stability of the SIJ is the variability and complexity of the orientation of the joint surfaces.[23,48] Solonen[23] described the joint surfaces as asymmetrical in size, shape, and direction and reported that the surfaces lie on numerous planes. He argued that this variability contributed to misleading information when the usual views were obtained for radiographs. Dijkstra et al[48] used a variety of radiographic methods in vitro and in vivo to determine the best method for visualizing the joint and improving diagnosis of SIJ disease. They described the surface orientation as complex and sinusoidal. To clearly show both surfaces required tomography in different projections, with an oblique projection being the method of choice for the middle and ventral portions.

Tables 2 and 3 demonstrate the great variability in sacral surface orientation between sides and at different sacral vertebral levels. This variability in surface orientation should significantly contribute to the stability of the joint because it more resembles a jigsaw puzzle. Vertical load bearing is facilitated but motion is limited by the overall inverted wedge shape of the sacral surface, with the middle third of the sacral surface more vertically inclined and the lower third exhibiting an outward flare.[23] Surface orientation variability at different levels of the sacrum, together with surface ridges and depressions, has profound implications about the ability of manual therapists to perform their described joint procedures. [TABULAR DATA 2 OMITTED]

Table 3. Variation in the Orientation of the Sacral Articular Surfaces Between
Sides (in Degrees)
               Dijkstra et al, 1989    Solonen, 1957
Sacral Level   (n=10)[48]              (n=26)[23]
S1              20-50                   0-64
S2              10-60                   -6-62
S2              -30-15                  -45-37

It is difficult to conceive how in the nonhypermobile joint adequate separation of the surfaces can be achieved manually, let alone how a thrust or overpressure exerted in one direction could affect supposed malalignment. This is particularly true given the variable orientation of different areas of the joint's surfaces. Indeed, in reports of techniques used to affect SIJ alignment, whether by producing motion or play (by therapists who recognize that little motion, if any, exists), initial attempts to distract the joint surfaces are not described.[2] Also "slippage," if it occurs, cannot take place solely in one direction, as implied by the use of terms such as "upslip." "Roundabout"(*) slippage only seems feasible.

Accessory or "Axial" Sacroiliac Joints

A number of investigators have reported the presence of "supernumerary articular facets"[19] or "axial"[49] or accessory sacroiliac joints (ASIJs), which may contribute to, or be responsible for, SIJ dysfunction (Tab. 4).[49-54] These joints have been observed in 8% to 40% of samples, sometimes two or three at one SIJ. Trotter's[50] Comprehensive study of 958 pelvic articulations demonstrated a higher prevalence of ASIJs in males than in females (40% and 21%, respectively) and more in whites than in blacks (50% and 21%, respectively). Her sample, however, had twice as many blacks than whites between 15 and 44 years of age and half as many over 55 years of age. The frequency increased with age, but no statistical analysis was reported. [TABULAR DATA 4 OMITTED]

Accessory joints are located at the level of the sacral crest, at the first and second posterior foramina, on the ilium at the medial surface of the posterior superior iliac spine (PSIS), and on the tuberosity (Fig. 3). Trotter's[50] data did not support Seligmann's theory (cited by Trotter[50(p298)]) that prevalence of ASIJs was related to the distance between PSISs.

Hadley[51,52] described a "superficial" joint at the level of the S2 posterior foramina between the PSIS and the lateral crest of the sacrum and a "deep" articulation at the level of the S1 posterior foramina. The joint plane was variable, at any angle up to 60 degrees relative to the auricular surface.

Bakland and Hansen[53] are the only investigators to distinguish between "axial" and accessory joints. They described, from a small sample of 27 anatomical specimens, the "axial" joint as extracapsular and occurring regularly in contrast to accessory joints, which were stated, by citing other authors, as "true synovial joints" occurring sporadically. Bakland and Hansen's "axial" joint appears similarly located to Hadley's[51,52] deep ASIJ: slightly caudal to the S2 posterior foramina with the concavity on the sacral surface. Histological study of 22 specimens showed most surfaces to be fibrocartilage, thus supporting Schunke's[19] and Trotter's[50] observations; however, Trotter did not report any histological study. The "axial" joint was considered a syndesmosis on the basis of the surface composition and presence of loose connective tissue in the joint space.[53]

Ehara et al[54] termed the ASIJ a "common variant" and concurred with others in its location, that the shape of the joint surfaces was variable but that the surfaces were generally congruent. Hadley[51.52] observed more pathological changes (arthritis, ankylosis) in his sample than did Ehara et al. Although some investigators may have failed to detect these joints, others, using CT scans, make no mention of either axial or accessory joints.[55,56] Dijkstra et al,[48] in a discussion of analyzing radiographic films, stressed the need for detailed knowledge of the configuration of the SIJ, including the so-called "axial joint," which may encourage radiologists to note the presence or absence of ASIJs.

I found no reports of the presence of either "axial" or accessory joints in fetuses or children, giving no support to the statement by Ehara et al[54] that the ASIJ is present at birth. The higher frequencies observed in adult samples with increasing age supports the theory that the ASIJ may be acquired as a result of the stress of weight bearing. Stewart[57] commented that the ASIJ was more frequent in great apes than in humans. Great apes tend to be quadrupedal and lack hip extension. Could the ASIJ be more frequent in long-term wheelchair users or older humans who tend to have some degree of hip flexion contracture, sit for long periods, and often have low back pain? Given the observed frequency of these joints and their demonstrated pathology, it is reasonable that, when present, these joints could be a source of localized low back pain. No reports of the innervation of the ASIJ were located; pain referral patterns are therefore unknown. Whether ASIJ play or motion can be detected clinically is also unknown. Because the ASIJ is shown to be superficially located, it is possible that play may be detected by experienced manual therapists. Careful examination of radiographs for presence of such joints appears to be indicated. When present, it seems possible that motion at an ASIJ could be mistaken for SIJ motion and treatment intervention may be mistakenly applied.

Changes Through the Life Span

Fetal

In the small samples studied (N= 5-36), complete cavitation (separation of joint surfaces) has not been demonstrated before 34 weeks of gestational age,[19,30,34,37] unlike most human synovial joints, which show complete cavitation by about 12 weeks of fetal life.[58] Schunke's[19] observation of incomplete cavitation in a 40-mm, 10-week-old fetus may be related to his use of paraffin as the embedding medium because paraffin subjects the tissues to marked temperature changes.

Although cavitation usually commences in the middle of a joint and proceeds outward, no distinct directionality of SIJ cavitation has been reported. Rather, several small cavities appear along the entire length of the joint, with fibrous septa or bands connecting the surfaces; this has been observed even in fetuses at term (about 40 weeks of age) (Fig. 4).[19,30,34] Such bands were not observed by MacDonald and Hunt[21] in two full-term fetuses. Whether the fibrous bands disappear, perhaps because of greater motion or simply because of postnatal stress, or whether these bands remain and contribute to the hypomobile joints observed by clinicians cannot be discerned because healthy specimens in the first to third decade of life have not been studied.

Aging and Ankylosis

Frequent observations in gross morphology and light microscopy studies of adult specimens have been the presence of focal adhesions between the joint surfaces, the presence of debris or loose connective tissue in the joint space, and ankylosis (Tab. 5).[17,19,20,30,34,36,37,59] Ankylosis tends to be more partial than complete; may be peripheral or central; and may be bony, fibrous, or chondroid (Fig. 1). Again, it must be noted that the material studied was derived from postmortem and cadaver specimens, and health histories of the subjects often were unavailable. These specimens, therefore, may not be representative of the general population. Further, not all investigators have distinguished between extra-articular ankylosis (osteophytes) and intra-articular ankylosis. Extra-articular ankylosis has been more frequently reported in in vivo studies.[36] [TABULAR DATA 5 OMITTED]

An examination by Resnick et al[36] of a sample of 46 joints from a veterans hospital's postmortem specimens revealed para-articular bony ankylosis resulting from bridging osteophytes and focal cartilaginous fusions. They believed that true intra-articular ankylosis was only present in cases of ankylosing spondylitis. Their observation that intra-articular fusions, as distinct from extra-articular osteophytes, increased after 50 years of age supported Brooke's[17] earlier findings. In Brooke's study, which was mainly descriptive, ankylosis was present in 37% of the male joints and in none of the female joints (Tab. 5). Sashin's[20] findings were similar with regard to age; however, he reported more osteoarthritic changes in male specimens (91%) than in female specimens (53%). Bony ankylosis occurred in both sexes but was less likely in female specimens.[17]

Sex Differences and Pregnancy Changes

All morphological and histological studies have demonstrated higher frequencies of joint surface irregularities and apparent pathological changes in male specimens. Changes appear earlier, about the fourth decade, and are more extensive in men than in women. As noted previously, Carter and Loewi[24] and Vleeming et al[46] stated that these changes may reflect a normal response to greater imposed forces in the SIJs of men compared with women. The sex differences further support the prime function of the SIJ in women, perhaps in all mammals, of facilitating vaginal delivery by increasing the pelvic diameter during labor.[13,16,60-63]

Hormonal influence. From the time of Hippocrates, it has been theorized that an increase occurs in pelvic mobility in pregnancy, with Duncan, in 1867, credited with the first observation of "nutation of the sacrum" (form of flexion/extension or upward/downward motion) during vaginal delivery.[13] Goldthwait and Osgood,[16] in 1905 in a morphological study of 20 anatomical pelvic specimens, used manual pressure to detect motion; motion was detected in 19 specimens. Although early investigators suspected a hormonal influence, only recently has the role of relaxin been clearly delineated.[62]

Relaxin, a polypeptide hormone, is produced by the corpus luteum and decidua of pregnancy in humans, with postulated target organs being the pubic symphysis, myometrium, cervix, placenta, breasts, and skin fibroblasts.[62] Relaxin is thought to

... activate the collagenolytic system,

regulates new collagen formation and

alters the ground substance by decreasing

the viscosity and increasing the

water content.[62(p9)]

Relaxin's action decreases the intrinsic strength and rigidity of collagen. MacLennan[62] reported a positive correlation between patients who were "incapacitated clinically" and high levels of relaxin during pregnancy, which suggests a relation among pain, pelvic instability, and relaxation of pelvic joints. Relaxin levels may increase during menstruation and provoke renewal of pelvic mobility symptoms in the parous woman.[62]

Pregnancy-related mobility.

Change in the width of the pubic symphysis can indicate concomitant increased mobility in the SIJ, with both events attributable to hormonal influences during pregnancy. A number of investigators have reported symphysial width between 2 and 7 mm in studies of pregnant and postpartum women; however, none adequately described their measurement methods or their nonrandomly selected samples.[13,60,61,63] Similar values (average=4.9 mm), however, were observed in a study of 200 nonpregnant women in which some detail was given of the sample and measurement method on radiographs.64 Lynch commented that "widening of sacrosciatic spaces was almost a constant phenomenon."[13(p580)] Hagen[60] considered symphysial separation of > 10 mm and vertical mobility of > 5 mm as pathologic. He noted symptoms (localized pain and tenderness, pain provoked by active movement, walking problems) tended to commence from the 28th to the 35th week of pregnancy; 16 of 23 cases relapsed in subsequent pregnancies.[60]

Garagiola et al[65] have demonstrated, with CT scans, increased symphysial width in 42% of 14 women immediately after uncomplicated vaginal delivery. Also in 42% of the women, the presence of gas, which is considered an abnormal finding, was observed in the SIJs, sometimes bilaterally. These observations were considered normal anatomical changes with pregnancy; none were observed in 15 controls of similar age, some parous.

A radiological study66 did not reveal a relationship between a thigh lever stress test in the supine position and the commonly used Chamberlain technique of standing on one leg as indicative of pelvic instability or SIJ pathology. In nonpregnant women, pubic vertical motion is between 0 and 1.5 mm; in men, it is less (0-0.5 mm).[18,66] Chamberlain[18] stated that there is a need to study the entire pelvic brim to detect asymmetry and gaping of the pubic symphysis. Walheim and Selvik[67] recommended implantation of tantalum balls (steel pellets, 0.8 mm in diameter) into bone with stereophotogrammetry for comprehensive studies of pubic symphysis mobility. This technique, which has high precision, was only used on two volunteers; values for rotations were up to 3 degrees and for translations up to 2 mm, values similar to those of other reports.[67]

Pelvic girdle relaxation. Dietricks and Kogstad[68] have suggested use of the terms "physiological pelvic girdle relaxation" for normal ligament relaxation during pregnancy, "symptom-giving pelvic girdle relaxation" for ligament relaxation that results in pain or pelvic instability, and "pelvic joint syndrome" when pain in one or more pelvic joints is experienced outside pregnancy and the puerperium. The latter term reflects their findings that pain, and not mechanical joint dysfunction, was the common symptom. In a random sample of 1,045 Norwegians, 25% claimed to have pelvic joint syndrome.[69] Kogstad and Birnstad[69] reported an incidence of pelvic joint syndrome of between 1.5% and 16%, with a point prevalence of 26%, 4 months postpartum.

The SIJs are links in the pelvic bony ring; therefore, any increased mobility must affect the SIJs. There is clinical and radiological evidence to support the presence of increased SIJ mobility in the end phase of pregnancy. Gait dysfunction, pain, and tenderness of the SIJs during this phase of pregnancy have been reported.[13,60,69,70] All of the women (N=23) studied by Hagen[60] reported pain on changing positions and stair climbing. For obvious reasons, no radiological studies of SIJ width changes or motion have been reported on pregnant women. Because there are reports of relapses during menstruation, clinicians treating patients who have SIJ dysfunction should document, in case studies, menstrual and birth histories as well as the sex ratio of patients with SIJ dysfunction. Such data would help to substantiate statements that SIJ dysfunction is more frequent in women than in men. Theoretically, some multiparous women may exhibit the hypermobile SIJ observed by clinicians. Because symptoms appear to be related to hormonal influences, therapy objectives should be to improve stability and relieve pain. Use of sacroiliac belts appears to be justified in theory, but outcome data are not yet available.[1,2]

Sacroiliac Joint Mobility

The premise that the SIJ is a locus of low back pain rests on the assumption that the SIJ is capable of motion. As the SIJ is a typical synovial joint in its anterior half and involves the auricular surfaces of the ilium and sacrum, this is a reasonable assumption that is supported by a growing number of studies. Studies of SIJ mobility have been conducted, in vivo and in vitro, using a variety of methods: gross examination with or without bony landmark measurements or manual pressure[15,17,20,71,72]; roentgenography[25,28,73,74]; tomography[75]; use of kinematic systems such as the WATSMART[TM] (Waterloo Spatial Motion Analysis Recording Technique) system[dagger][76]; stereophotogrammetry[5,63,77] stress radiology[18,29,78,79]; use of Kirschner wires[80]; holography[43]; loading[27,45]; and, more recently, computer-generated biomechanical model simulations.[42,45] Alderink,[11] in 1991, published a review article that explored in detail the literature on the osteokinematics of the SIJ relative to functional movements. He incorporated both studies in which data were reported and theories of SIJ motion derived from published studies but not supported by original data.[1,2,81,82] It is not the purpose of this article to repeat Alderink's review but to attempt to address the issue of SIJ mobility by focusing on the quality of reported studies and to examine potential consensus of their results.

Investigators of SIJ mobility generally have focused on two main questions: What is the extent of movement? and What is(are) the axis(es) of motion? Motion must occur both to create SIJ dysfunction and, more importantly, to substantiate manual therapy procedures designed to relieve symptoms and restore function. Regardless of whether measurement reliability or validity was reported in the motion studies conducted, while not discounting the importance of these factors, it clearly should be accepted that some motion does occur in the SIJs. Anatomical studies suggest strongly that available motion may be greatest in women, especially late in pregnancy and the early postpartum period, and that motion decreases, possibly significantly, with natural aging.[17,19,20,30,34,36,37,59]

There is no agreement in the literature for a single model of SIJ motion, or on a single fixed axis of motion, but rather agreement for variability in axes and motion, with an instantaneous axis. Sacroiliac joint motion is a simultaneous combination of translation, median-plane motion (flexion/extension, upward/downward, or "nutation"), and rotation. Motion values, therefore, are reported both in degrees and in millimeters.[double dagger] Only in the child may the joint surfaces be flat and smooth.[19,30,34,37] No other human joint, healthy or pathological, is characterized by the intra-articular elevations and depressions described previously. Theoretically, therefore, translatory (gliding) motion should be greatest in younger subjects and least in older subjects. Although statements that "traumatic subluxation" is frequent in adolescents[2] Suggest support for greater motion in younger subjects, no substantiation for the incidence of subluxation in adolescents was located in the literature.

Is there an age change? Only three studies of SIJ mobility[25,28,37] included younger subjects (ie, under 40 years of age). Sturesson et al[25] noted no decrease in mobility in a sample between 19 and 45 years of age and reported 0.08 mm for translatory motion. Egund et al[28] reported translatory motion of 2 mm from only four subjects between 25 and 45 years of age. Other investigators, such as Schunke,[19] Sashin,[20] and Bowen and Cassidy,[37] described a decrease in motion with age. Bowen and Cassidy[37] described the motion in younger specimens as gliding, but in older specimens they described the motion as nodding or rotatory gliding. None of these investigators, however, gave any detail of how motion was assessed. Age either was not given or was over 50 years in most studies on postmortem or cadaver specimens; yet, it is from such studies that statements on decreased motion with age are derived. To date, motion studies have not clarified an age effect on SIJ mobility.

An important question appears to be, Is the motion present adequate in total range to be detected by observation and manual palpation, as extensively described by several clinicians?[1-3,7,9,83-86]

Reported motion. With few exceptions, reported values for SIJ rotatory motion (in degrees) (Tab. 6) and for SIJ translatory motion (in millimeters) (Tab. 7), whether derived from in vivo or in vitro studies (including computer modeling), were reasonably consistent in size of the average or maximal values observed. Rotatory motion generally averaged <4 degrees, with maximal values of up to 20 degrees reported from studies in which pelvic-sacral angles were measured either with an inclinometer[79] or with radiography[78] and with no reported reliability of measurements. Clayson et al[78], took measurements from full lumbar flexion to full extension in standing subjects. The angle formed between a line connecting the anterior superior iliac spine and the PSIS and the superior surface of the sacrum was considered to reflect SIJ motion. Selection of slender body builds in their sample must have influenced their results. Although, from a sample of 26 subjects, the maximum value was 20 degrees, only four values exceeded 11 degrees. Even given unstated errors of measurement, these data give support to earlier investigators who used manual pressure on cadaver material to ascertain motion and generally concluded that no marked rotation could occur.[19,20,23] [TABULAR DATA 6 AND 7 OMITTED]

Reported values for translatory motions (glides) in the "anterior-posterior" plane ranged between 0.5 and 7 mm, with a mode of about 3 mm (Tab. 7). Two studies are exceptions to this general trend and, because of their frequent citation, are worthy of further comment here. Frigerio et al[77] utilized stereoradiography and mathematical formulas and reported motion of the iliac crests relative to the sacrum of up to 26 mm in one subject. These investigators assessed the effect of moving one or both legs in the side-lying position. From their description, the landmarks used are not reproducible. Possibly because they have reported the highest value for motion in the literature, this study is frequently cited. Authors, however, appear not to have noted that Sherlock[87] stated that an error in calculations made by Frigerio et al rendered their study "completely invalid." The authors repeated their vector analysis calculations and concluded that the original results were correct and that their system was valid.[88] To date, however, there has been only one report[5] of similar values. Because Frigerio et al studied only one bony pelvis, one cadaver, and one live subject and because their methodology is unreproducible, their results should be viewed with great caution.

Using stereophotogrammetry (three-dimensional) with skin markers over the PSISs and other sites, Grieve[5] reported combined rotatory and translatory motions of between 1 and 16 mm. Motion on each side was assessed in one-legged stance with alternate hip and knee flexion in 21 subjects. When the data from 6 subjects with pain were eliminated, the range decreased to between 5 and 1 1 mm on any one axis. These results do not agree with those from more rigorous studies. Because of potential soft tissue displacement, the validity of her method is highly questionable and the conclusions of the study should be considered only speculative.

Grieve[5] attempted to measure a motion used clinically to assess the SIJ. In studies in which functional movements have been assessed, once again regardless of sample size (usually small and never randomized) or methodology, there is agreement that maximal motion of the sacrum occurs in rising from a supine to a standing or long-sitting position.[25,29,78,80] Sturesson et al[25] reported lower values than did others,[5,29,80] with a maximum value of 2.5 degrees in moving from a supine to a long-sitting position, in contrast to Weisl,[29] who reported maximal motion (5.6[degrees]) in rising from a sitting to a standing position.

Sturesson et al[25] inserted tantalum balls percutaneously over the ilium and sacrum and used roentgen stereophotogrammetry to measure "physiological end range" in 26 patients in various physiological positions. Small mean errors of measurement were reported (ie, 0.1[degrees]-0.2[degrees] for rotation and 0.1 mm for translation). Walheim and Selvik[67] reported similar precision with this technique but described tantalum balls inserted into the pubic bones, not just percutaneously as did Sturesson et al. This technique appears to be more controlled than those of most other landmark studies.[5,29,73,74,77-79] It therefore is of interest that the motion values reported by Sturesson et al were minimal (Tabs. 6, 7) and that the authors were unable to distinguish between subjects with and without SIJ symptoms. They concluded that assessment of mobility under physiological loads could not be used successfully to identify SIJ dysfunction.

Axes of motion. Although many authors have attempted to describe SIJ motion using conventional planes of motion (see Alderink's review[11]), various studies cited previously on the orientation of the articular surfaces and their morphology give no support to the idea that motion occurs in the classical planes about a single axis. Wilder et al[75] used topography and theoretical modeling with best-fit axes of rotation (AORs) for each contour to calculate "optimal axes of rotation." Data were obtained from 11 hemipelvises. They concluded that motion could not occur exclusively around axes proposed by Weisl's AOR,[29] at Bonnaire's tubercle, or by Fababeufs AOR (cited by Wilder et al[75]) or in optimized AORs in the median or frontal planes.[75] Wilder et al considered that a translatory motion could occur about a "rough axis" (Fig. 5) if some separation of the surfaces was present. Theorized required separation was a mean of 7.25 mm (SD=3.7) in the median plane and 3.4 mm (SD=1.7) in the frontal plane.[75]

Weisl,[29] on the basis of his radiographic and cartographic studies, concluded that sacral motion is not rotary and is not determined by a single factor and that previous reports of AORs based on surface morphology were "erroneous." He postulated that surface morphology, degree of congruence between sacra and ilia, and compressibility of the articular cartilage accounted for variation in position of the axes and the "occasional sliding movement of the sacrum observed in the living."[22(p11)] Reports such as these have not lessened attempts by several authors to explain the SIJ motion using conventional planes, AORs, and movement terminology.

Assessment and Treatment

The minimal range of motion present in probably most of the population casts doubt on whether therapists can detect 1 to 3 degrees or 1 to 3 mm of motion occurring specifically at the SIJ. Perhaps the term "play" should be used when referring to the SIJ, as "motion" connotes the idea of a quantity of motion similar to other synovial joints, which does not appear to be the case. Although it is not the purpose of this article to critically examine assessment tests or treatment techniques, when results of reliability studies are added to the morphological and mobility studies, current clinician attention to the SIJ and claims for efficacy are further in doubt. Results from two studies[89,90] suggest that experienced therapists do not possess higher intrarater or intertester reliability in SIJ assessment. Are experienced clinicians firmer in their belief that real motion and thus dysfunction exist and overdiagnosing, perhaps overtreating? Despite studies such as that by Sturesson et al[25] in which a distinction could not be made between symptomatic and asymptomatic individuals and reliability studies such as that of Potter and Rothstein[89] in which only 2 of 13 tests had intertester reliability greater than 70%, therapists continue to advocate clinical evaluation to detect SIJ dysfunction.[91,92] This indicates a "hang the torpedoes, full speed ahead!" type of view in which we may be disregarding information because it does not fit with our preconceptions and theories.

Most testing uses bony landmarks such as the PSIS. Fundamental to assessment and treatment using localized manual pressure is reliable and valid palpation of bony landmarks. Janos's[93] and Simmonds's[94] recent studies demonstrate high errors in palpation of bony landmarks. Simmonds studied 20 physical therapists and reported a systematic error of 7 to 14 mm for intrarater and 12 to 24.5 mm for interrater consecutive tests in which invisible ink was used. Four bony points were studied; the more deeply located landmarks (L4 spinous and transverse processes) were associated with higher errors. For the PSIS, an intrarater mean of 8 mm (SD=5.1) and an interrater mean of 20.4 mm (SD = 13.2) were reported. Several researchers[4,12,71,72] who studied the efficacy of intervention reported measurement reliability; however, critical in the light of the above studies, they did not report measurement reliability of landmark palpation. Such landmark identification also is the foundation for determination of leg-length disparity, which is considered indicative of SIJ pathology.[8,95]

Although Grieve[96] stressed that asymmetry does not necessarily mean abnormality, several clinicians appear to assume this even in the absence of Symptoms.[8,12,95] Statements, however, that the PSIS is frequently asymmetrical[94] are not supported by any published study of PSIS location variability. Given the region's suspectibility for congenital anomalies, such as sacralization of lumbar vertebra or lumbarization of S1, symmetry perhaps should not be assumed as indicating normality.

The implications from all types of studies suggest use of extreme caution in claims to isolate motion to the SIJ and give support to use of pain provocation tests and a treatment approach that emphasizes the patient's involvement in managing his or her back problem. There appears to be little support for highly specific tests and procedures for the SIJ to be taught in entry-level physical therapy programs. Clinicians attending continuing education courses need to look and listen for adequate supporting statements and to question the basis for detailed specificity of procedures. Those who advocate specific approaches should provide a basis in at least clinical case studies but preferably in randomized clinical trials.

Conclusions

Compared with other human synovial joints, the SIJ is unique and fascinating. Use of sophisticated methods of investigation such as moire photogrammetry, computerized biomechanical modeling, and stereoroentgenographic photogrammetry add to our knowledge of this joint but provide no support for consistent patterns of motion in conventional planes and single axes of motion, What range of motion exists, particularly in the adult, appears to be very small indeed, of a few degrees or millimeters at best. Normal variability, especially in women because of hormonal influences, implies that although most individuals in the population will have very limited motion, some will be "hypomobile" and others will be "hypermobile."

It is plausible that slight shifting of one articular surface in relation to the other may minimally disturb the alignment of the interlocking surfaces. This may constitute SIJ dysfunction and may respond to manual therapy procedures. Although the ASIJ does not appear to be common, consideration should be given to its potential presence, possible source of symptoms, and site of treatment intervention. Unless reliability and validity of assessments and effectiveness of treatment procedures can be demonstrated, clinicians should temper their claims for measurement of, and direct effects on, the SIJ. A battery of reliable functional tests may better serve to document a patient's status and progress. It is essential that clinicians document and publish treatment intervention with case-study series. (*) Author's term meaning motion as permitted, no one single direction.

References

[1] Porterfield JA, DeRosa CP. Mechanical Low Back Pain: Perspectives in Functional Anatomy. Philadelphia, Pa: WB Saunders Co; 1991. [2] Lee D. The Pelvic Girdle. New York, NY: Churchill Livingstone Inc; 1989. [3] DonTigny RL. Function and pathomechanics of the sacroiliac joint: a review. Phys Ther. 1985;65:35-44. [4] Nyberg R. Clinical studies of sacroiliac movement. In: International Federation of Orthopaedic manipulative Therapists Fifth International Conference, Vail, Colorado, June 1992. 1992:A90. Abstract. [5] Grieve EFM. Mechanical dysfunction of the sacro-iliac joint. Int Rehabil Med. 1983;5:46-52. [6] Corrigan B, Maitland GD. Practical Orthopaedic Medicine. London, England: Butterworth & Co (Publishers) Ltd; 1983:324-331. [7] Grieve GP: The sacro-iliac joint. Physiotherapy. 1976;62:383-400. [8] Cibulka MT, Koldehoff RM. Leg length disparity and its effect on sacroiliac joint dysfunction. Clinical Management. 1986;6(5): 10-11. [9] Cyriax JH. Textbook of Orthopaedic Medicine, Volume I: Diagnosis of Soft Tissue Lesions. 7th ed. London, England: Balliere Tindall; 1978. [10] Laslett M. The reliability of tests for sacroiliac joint pathology. In: International Federation of Orthopaedic Manipulative Therapists Fifth International Conference, Vail, Colorado, June 1992. 1992:A105. Abstract. [11] Alderink GJ. The sacroiliac joint: review of anatomy, mechanics and function. Journal of Orthopaedic and Sports Physical Therapy 1991;13:71-84. [12] Smith RL, Sebastian BA, Gajosik RL. Effect of sacroiliac joint mobilization on the standing position of the pelvis in healthy men. Journal of Orthopaedic and Sports Physical Therapy. 1988;10:77-84. [13] Lynch FW. The pelvic articulations during pregnancy, labor and puerperium: an x-ray study. Surg Gynecol Obstet. 1920;30:575-580. [14] Cantin L. Cited by: Lynch FW. The pelvic articulations during pregnancy, labor and puerperium: an x-ray study. Surg Gynecol Obstet. 1920;30:575-580. [15] Goldthwait JE. The pelvic articulations: a consideration of their anatomic, physiologic, obstetric and general surgical importance. JAMA 1907;49:768-774. [16] Goldthwait JE, Osgood RB. A consideration of the pelvic articulations from an anatomical, pathological, and clinical standpoint. Boston Med Surg J 1905;152:593-601. [17] Brooke R. The sacroiliac joint. J Anat. 1924;58:299-305. [18] Chamberlain WE. The symphysis pubis in the roentgen examination of the sacroiliac joint. AJR. 1930;24:621-625. [19] Schunke GB. The anatomy and development of the sacroiliac joint in man. Anat Rec. 1938;72:313-331. [20] Sashin D. A critical analysis of the anatomy and the pathological changes of the sacroiliac joints. J Bone Joint Surg. 1930;12:891-910. [21] MacDonald GR, Hunt TE. Sacroiliac joints: observations on the gross and histological changes in the various age groups. Can Med Assoc J. 1952;66:157-162. [22] Weisl H. The articular surfaces of the sacroiliac joint and their relation to the movements of the sacrum. Acta Anat (Basel). 1954; 22:1-14. [23] Solonen VA. The sacroiliac joint in the light of anatomical, roentgenological and clinical studies. Acta Orthop Scand Suppl. 1957;27: 1-127. [24] Carter ME, Loewi G. Anatomical changes in normal sacroiliac joints during childhood and comparison with the changes in Still's disease. Ann Rheum Dis. 1962;21:121-143. [25] Sturesson B, Selvik G, Uden A. Movements of the sacroiliac joints: a roentgen stereophotogrammetric analysis. Spine. 1989;14:162-165. [26] Vogler JB, Brown WH, Helms CA, Genant HK. The normal sacroiliac joint: a CT study of asymptomatic patients. Radiology. 1984;115: 433-437. [27] Miller JAA, Schultz AB, Andersson GBJ. Load-displacement behaviour of sacroiliac joints. J Orthop Res. 1987;5:92-101, [28] Egund N, Olsson TH, Schmid H, Selvik G. Movements in the sacroiliac joints demonstrated with roentgen stereophotogrammetry. Acta Radiol. 1978;19:833-846, [29] Weisl H. The movements of the sacroiliac joint. Acta Anat (Basel). 1955;23:80-91. [30] Walker JM. Age-related differences in the human sacroiliac joint: a histological study--implications for therapy. Journal of Orthopaedic and Sports Physical Therapy. 1986;7: 325-331. [31] Bellamy N, Park W, Rooney PJ. What do we know about the sacroiliac joint? Semin Arthritis Rheum. 1983;12:282-313. [32] Warwick R, Williams PL. Gray's Anatomy. 35th ed. London, England: Longman Group UK Ltd; 1973:351-357, 442-446. [33] Paquin JD, Rest M, Marie PJ, et al. Biochemical and morphologic studies of cartilage from the adult human sacroiliac joint. Arthritis Rheum. 1983;26:887-895. [34] Ishimino T. Histopathological study of the aging process in the human sacroiliac joint [in Japanese]. Nippon Seikeigeka Gakkai Zasshi. 1989;63:1074-1084. [35] Albee FH. A study of the anatomy and clinical importance of the sacroiliac joint. JAMA. 1909;53:1273-1275. [36] Resnick D, Niwayama G, Georgen TG. Degenerative disease of the sacroiliac joint. Invest Radiol. 1975;10:608-621. [37] Bowen V, Cassidy JD. Macroscopic and microscopic anatomy of the sacroiliac joint from embryonic life until the eighth decade. Spine. 1981;6:620-628. [38] Miura H. Biomechanical properties of the sacroiliac joint [in Japanese]. Nippon Seikeigeka Gakkai Zaahi 1987;61:1093-1105. [39] Ekman S, Dalin G, Olsson S-E, Jeffcott LB. Sacroiliac joint of the horse, 3: histological appearance. Anat Histol Embryol. 1986;15: 97-107. [40] Gregory CR, Cullen JM, Pool R, Vasseur PB. The canine sacroiliac joint. Spine. 1986;11: 1044-1048. [41] Slocumb L, Terry RJ. Influence of the sacrotuberous and sacrospinous ligaments in limiting movements at the sacroiliac joint. JAMA. 1926;87:307-309. [42] Scholten PJM, Schultz AB, Luchico CW, Miller JAA. Motions and loads within the human pelvis: a biomechanical model study. J Orthop Res. 1988;6:840-850. [43] Vukicevic S, Plitz W, Vukicevic D, et al. Holographic study of the stresses in the normal pelvis with particular reference to the movement of the sacrum. In: Huiskes R, Van Campen D, DeWijn J, eds. Biomechanics: Principles and Applications. Dordrecht, the Netherlands: Martinus Nijhoff/Dr W Junk Publishers; 1982:233-239. [44] Edeiken-Monroe BS, Browner BD, Jackson H. The role of standard roentgenograms in the evaluation of instability of pelvic ring disruption. Clin Orthop. 1989;50:63-76. [45] Kissling R, Brunner C, Jacob HA. Mobility of the sacroiliac joint in vitro [in German]. Z Orthop. 1990;128:282-288. [46] Vleeming A, Stoeckart R, Volkers ACW, Snijders CJ. Relation between form and function in the sacroiliac joint, part I: clinical anatomical aspects. Spine. 1990;13:130-132. [47] Vleeming A, Volkers ACW, Snijders CJ, Stoeckart R. Relation between form and function in the sacroiliac joint, part II: biomechanical aspects. Spine. 1990;13:133-135. [48] Dijkstra PF, Vleeming A, Stoeckart R. Complex motion tomography of the sacroiliac joint: an anatomical and roentgenological study. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 1989;150:635-642. [49] Derry DE. Note on accessory articular facets between the sacrum and ilium, and their significance. J Anat Physiol. 1911;45:202-211. [50] Trotter M. A common anatomical variation in the sacroiliac region. J Bone Joint Surg. 1940;22:293-299. [51] Hadley LA. Accessory sacroiliac articulations. J Bone Joint Surg [Am]. 1952;34:149-155. [52] Hadley LA. Accessory sacroiliac articulations with arthritic changes. Radiology. 1950; 55:403-409. [53] Bakland OD, Hansen JH. The "axial sacroiliac joint." Anat Clin. 1984;6:29-36. [54] Ehara S, El-Koury GY, Bergman RA. The accessory sacroiliac joint: a common anatomic variant. AJR. 1988;150:857-859. [55] Whelan MA, Gold RP. Computed tomography of the sacrum, 1: normal anatomy. AJR. 1982; 139:1183-1190. [56] Lawson TL, Foley WD, Carrera GF, Berland U. The sacroiliac joints: anatomic, plain roentgenographic, and computed tomographic analysis. J Comput Assist Tomogr. 1982;6:307-314. [57] Stewart TD. Accessory sacroiliac articulations in the higher primates and their significance. Am J Phys Anthropol. 1938;24:43-59. [58] O'Rahilly R, Gardner E. The embryology of moveable joints. In: Sokoloff L, ed. The Joints and Synovial Fluid. New York, NY: Academic Press Inc; 1978;1:49-97. [59] Stewart TD, Pathologic changes in aging sacroiliac joints: a study of dissecting-room skeletons. Clin Orthop. 1984;183:188-196. [60] Hagen R. Pelvic girdle relaxation from an orthopaedic point of view. Acta Orthop Scand. 1974;45:550-563. [61] Joseph J. The joints of the pelvis and their relation to posture in labour. Midwives Chronicle and Nursing Notes. 1988;101(1202):63-64. [62] MacLennan AH. The role of the hormone relaxin in human reproduction and pelvic girdle relaxation. Scand J Rheumatol. 1991; 20(suppl 88):7-15. [63] Borell U, Fernstrom I, The movements of the sacro-iliac joints and their importance to changes on pelvic dimensions during parturition. Acta Obstet Gynecol Scand. 1957;36: 42-57. [64] Vix VA, Ryu CY. Adult symphysis pubis: normal and abnormal. Am J Roentgenol Radium Ther Nucl Med. 1971;112:517-527. [65] Garagiola DM, Tarver RD, Gibson L, et al. Anatomic changes in the pelvis after uncomplicated vaginal delivery: a CT study of 14 women. AJR. 1989;153:1239-1241. [66] Death AB, Kirby RL, MacMillan CL. Pelvic ring mobility: assessment by stress radiography. Arch Phys Med Rehabil. 1982;63:204-206. [67] Walheim GG, Selvik G. Mobility of the pubis symphysis. Clin Orthop. 1984:191:129-135. [68] Dietricks E, Kogstad O. "Pelvic girdle relaxation": suggested new nomenclature. Scand J Rheumatol. 1991;20(suppl 88):3. [69] Kogstad D, Birnstad N. Pelvic girdle relaxation: pathogenesis, etiology, definition, epidemiology [in Norwegian]. Tidsskr Nor Laegeforen. 1990;110:2209-2211. English abstract. [70] Meisenbach RO. Sacroiliac relaxation: with analysis of 84 cases. Surg Gynecol Obstet. 1911; 12:411-434. [71] Cibulka MT, Delitto A, Koldehoff RM. Changes in innominate tilt after manipulation of the sacroiliac joint in patients with low back pain: an experimental study. Phys Ther, 1988; 68:1359-1363, [72] DonTigny RL. Dialogue on sacroiliac joint. Phys Ther. 1989;69:164-165. Letter to the editor. [73] Lavignolle B, Vital JM, Senegas J, et al. An approach to the functional anatomy of the sacroiliac joints in vivo. Anat Clin. 1983;5:169-176. [74] Reynolds HH. Three-dimensional kinematics in the pelvic girdle. J Am Osteopath Assoc. 1980;80:277-280. [75] Wilder DG, Pope MH, Frymoyer JW. The functional topography of the sacroiliac joint. Spine. 1980;5:575-579. [76] Pierrynowski MR, Schroeder BC, Garrity CB, et al. Three-dimensional sacroiliac motion during locomotion in asymptomatic male and female subjects. In: Cotton CE, Lamontagne M, Robertson DGE, Stothart JP, eds. Proceedings of the Fifth Biennial Conference and Human Locomotion Symposium of the Canadian Society for Biomechanics, London, Ontario. 1988: 132-133. [77] Frigerio NA, Stowe RR, Howe JW. Movement of the sacroiliac joint. Clin Orthop. 1974; 100:370-377. (Also see Correction. Clin Orthop. 1981;155:293-297.) [78] Clayson SJ, Newton IM, Debeucc DF, et al. Evaluation of mobility of hip and lumber vertebrae in normal young women. Arch Phys Med Rehabil. 1962;43:1-8. [79] Pitkin HC, Pheasant HC. Sacrathrogenetic telalgia, II: a study of sacral mobility. J Bone Joint Surg. 1936;18:365-374. [80] Colachis SC, Warden RE, Bechtol CO, Strohm BR. Movement of the sacroiliac joint in the adult male: a preliminary report. Arch Phys Med Rehabil. 1963;44:490-498. [81] Kapandji IA. The Physiology of the Joints. New York, NY: Churchill Livingstone Inc; 1982; 3:63-71. [82] Mitchell FL, Moran PS, Pruzzo NA. An Evaluation and Treatment Manual of Osteopathic Muscle Energy Procedures. Manchester, Mo: Mitchell, Moran and Pruzzo, Associates; 1979. [83] Maitland GD. Vertebral Manipulation. 4th ed. London, England: Butterworth & Co (Publishers) Inc; 1977:66. [84] Stoddard A. Conditions of the sacro-iliac joint and their treatment. Physiotherapy. 1958; 44:97-101. [85] Woerman AL. Evaluation and treatment of dysfunction in the lumbar-pelvic-hip complex. In: Donatelli R, Wooden MJ, eds. Orthopaedic Physical Therapy. New York, NY: Churchill livingstone Inc; 1989:403-483. [86] Mennell JB. The Science and Art of Joint Manipulation, Volume 2: The Spinal Column. Philadelphia, Pa: Blakiston Co; 1952. [87] Sherlock RA. A criticism of the article, "Movement of the Sacroiliac Joint." Clin Orthop. 1981;155:293-297. [88] Stowe RS. Reply to Dr Sherlock. Clin Orthop. 1981;155:296-297. Letter to the editor. [89] Potter NA, Rothstein JM. Intertester reliability for selected clinical tests of the sacroiliac joint. Phys Ther. 1985;65:1671-1675. [90] Herzog W, Read LJ, Conway PJW, et al. Reliability of motion palpation procedures to detect sacroiliac joint fixations. J Manipulative Physiol Ther. 1989;12:86-92. [91] Lee D. Intra-articular versus extra-articular dysfunction of the sacroiliac joint: a method of differentiation. In: International Federation of Orthopaedic Manipulative Therapists Fifth International Conference, Vail, Colorado, June 1992. 1992:A37. Abstract. [92] DonTigny RL. Pathomechanics and treatment of sacroiliac joint dysfunction. In: International Federation of Orthopaedic Manipulative Therapists Fifth International Conference, Vail, Colorado, June 1992. 1992:A40. Abstract. [93] Janos SC. Palpation of selected bony landmarks in the lumbo-pelvic region. In: International Federation of Orthopaedic Manipulative Therapists Fifth International Conference, Vail, Colorado, June 1992. 1992:A111. Abstract. [94] Simmonds MJ. The reliability of palpation skills in the therapeutic professions. In: Kumar S, ed. Proceedings of the Annual International Industrial Ergonomics and Safety Conference, Denver, June 1992. London, England: Taylor & Francis Ltd; 1992:665-671. [95] Schuit D, McPoil TG, Mulesa P. Incidence of sacroiliac joint malalignment in leg length discrepancies. J Am Podiatr Med Assoc. 1989; 79:380-383 [96] Grieve GP. Mobilization of the Spine. London, England: Churchill livingstone; 1984: 79-85.