Keystone Form & Force Closure vs. Sacral Suspension

Over the past 28 years, I have done clinical and literature research in an attempt to find a pattern that integrates the biomechanics of the musculoskeletal system. When I began reading the medical literature on the sacroiliac joint, I found information on the keystone concept of form and function that was contrary to what I had experienced. As time went on, the keystone concept became increasingly popular, but made little sense to me. At first, I questioned myself. It was hard for me to believe that I was correct and well published and respected authors were wrong. But, of many mistaken concepts, several questions stuck out in my mind: “How could osteoarthritic signs within a synovial joint be considered normal?”, “How could weight transfer through a synovial joint that was basically vertical?”, and “What could be the purpose of the large mass of ligaments if not for suspension of the sacrum from the ilia?” As I studied, I found many other concepts that didn’t fit the keystone model but did fit the historically popular suspensory model. In 2007, I started writing it down and cross-referenced the literature. I was very careful to make sure that I did not cite a reference unless it was relevant and self-explanatory in the context of my manuscript.

The currently popular understanding of the biomechanics of the sacroiliac joint is based on the keystone concept of form and force closure, which was introduced around 1990. However, when one tries to associate proven or well understood concepts into this model of form and function, they often lead to dead-ends. But, because there is no underlying cohesive model, we don’t notice when things don’t fit; we just ascribe them to the mystery of the musculoskeletal system. As time goes on without a coherent narrative, the increasingly vast sea of information clouds our ability to comprehend how the parts fit together.  How is the musculoskeletal system integrated into a functioning whole? We will not solve this mystery if we continue to look in the wrong direction.

I hope to reinstate the concept of ligamentous suspension by describing normal motion based upon the commonly known, but poorly understood principle of nutation/counternutation. I also incorporate the principle of biological tensegrity which, although controversial in some circles, has been presented in numerous peer reviewed journals and is studied at the Wyss Institute at Harvard University.

Form Closure - The Keystone in a Roman Arch
A popular concept describing a self-bracing mechanism of the sacroiliac joint is explained by Vleeming et al. [1, 2], White [3], Snijders [4] and Lee [5], in which they state that form closure depends on the structural fit of the sacrum and ilia, specifically, the wedge shape of the sacrum fitting into the accommodating shape formed between the two ilia, like a keystone in a Roman arch. The innominate bones, acting as walls of the arch, are supposed to prevent lateral movement of the sacrum and ensure stability. This concept proposes that, at the end of range of motion in nutation, the sacral and iliac articular surfaces would be compressed together, and load transmission would occur directly between the sacrum and ilia by “bone contact force” [4, 6] at the articular regions. However, a number of discrepancies exist with the keystone/form closure model.

The Anterior/Posterior Shape of the Sacrum
Vleeming [7] stated that the sacrum is “wider cranially than caudally, and wider anteriorly than posteriorly. Such a configuration permits the sacrum to become ‘wedged’ cranially and dorsally into the ilia within the pelvic ring.” However, since forces occurring during load transfer force the sacrum anteriorly, there is no practical need for the sacrum to be ‘wedged’ dorsally. To speak of wedging the sacrum dorsally is misleading and only serves as a distraction from considering true anatomical form and function. Although mistaken, at best, it is understandable, if all one looks at is the shape of the sacrum and ilia, that one could imagine a wider cranial/narrower caudal shaped sacrum may wedge into the ilia in a superior to inferior direction. However, a wider anterior/narrower posterior sacral shape is antithetical to how a true keystone could function.

The anterior/posterior shape of the sacral base demonstrates a predisposition for the sacrum to drop anteriorly, away from the ilia during weight bearing, which is reinforced by the opposing movements of the sacrum and ilium during nutation. During weight bearing, as the sacrum is driven anteriorly and inferiorly, the sacrum would wedge into the ilia only if the sacrum was wider posteriorly and would sit posterior to a corresponding wedge formed by the ilia. This way, a true keystone would provide “bone to bone” transfer of weight in both an anterior/posterior and superior/inferior direction. However, as it is, notwithstanding all the evidence to the contrary, the keystone concept of form closure can theoretically provide a supposedly stable fit in the superior/inferior direction but an unstable fit in the anterior/posterior direction; this poses a basic anatomical contradiction to the form closure model. However, from this perspective, the anatomical function of the large mass of ligaments in suspending the sacrum from the ilia becomes evident, especially when bending forward at the hips, as in other animals.

Suspension in Other Animals
Observation of the skeleton of any four legged animal would find that the sacrum sits below and is suspended from the ilia by a strong complex of ligaments. For example, in discussing horse anatomy, Haussler [8] stated that “The sacroiliac ligaments support the weight of the caudal vertebral column as it attaches to the ilial wing. The weight of the caudal vertebral column is suspended from the sacroiliac ligaments…” It would seem implausible that the human frame is constructed so differently from other animals that we evolved a completely different method of weight transfer through the pelvis, especially when ligamentous suspension is more efficient.

Normal Adaptations vs. Degeneration
A fundamental principle of the keystone concept is that roughening and interlocking ridges and grooves that occur on the articular surfaces are considered normal developments in the adult [1, 9], as a means of increasing friction to counter increased weight bearing during the aging process. They considered the interlocking ridges and grooves to be non-pathologic modifications to increase friction that would supposedly be beneficial in weight transfer during lifting, gait, and shock absorption.

However, in describing these surface changes, no mention is made of the signs of progressive degeneration within the articular region of the sacroiliac joints found by other researchers, [10, 11], including tears, deep erosions, fibrotic changes, loose connective tissue strands, capsular thickening, and flaky, amorphous, yellow debris filling the joint space, which eventually form osteophytes and ankylosis.

Although they briefly discuss Sashin’s [10] and Bowen and Cassidy’s [11] findings, they dismiss whether the signs described above actually represent degeneration as a matter of a difference in opinion. One is left to wonder what normal function this flaky, amorphous debris may have.

Accordingly, of all the signs of degeneration mentioned above, their own studies only consider two of them, coarseness of the joint surfaces and interdigitating ridges and grooves, but ignore the obvious pathologic changes mentioned by Sashin or Bowen & Cassidy that would not agree with the concept of form closure. By limiting the information presented, they are able to impart a narrow view that only presents enough evidence to give a possible reason for those two signs, while ignoring, or minimizing the significance of the others.

To maintain some semblance of balance, they did note that “the possibility that the pronounced coarseness observed in the SI joints of certain older men belongs to, or finally results in, a pathological process cannot be excluded;” this statement appears to minimize the incidence to “certain older men.” However, Sashin noted that 91% of males and 77% of females in their 40s to 50s, as well as 100% of all SIJs (both male and female) at age 60 or above have these signs of degeneration, which qualifies the term “certain” to mean almost everyone.

Instead, with their discussion limited to these two signs, they simply referred to them as “cartilage changes” which, in their opinion, should be considered normal “functional adaptations” in the SIJ, even though they denote pathologic degeneration in all other joints.

It is important to consider that degeneration starts with simple surface changes in joints. The joint cannot go directly from normal to tears, deep erosions, fibrotic changes, loose connective tissue strands, capsular thickening, and flaky, amorphous, yellow debris filling the joint space. Something must degenerate for the above signs to appear, and the only structures available are the joint surfaces. It is only reasonable that surface roughening, including ridges and grooves, form the initial surface changes that develop into the more obvious signs of degeneration described above and, therefore, are signs of degeneration themselves.

Although it is evident that the roughening and development of small ridges and grooves increases friction, I propose that these developments can only occur after the ligaments, particularly the interosseous ligament, tear and motion becomes erratic, causing the joint surfaces to start rubbing. Please see SIJ Degeneration for more.

Accordingly, a significant ridge and groove would not be expected to appear during youth in a normal joint. Yet, Vleeming et al. [7, 9] show photos of a significant ridge and groove appearing on one sacroiliac joint, but not the other, in the cadaver of a twelve year old boy. Although Vleeming stated that this is a normal development, it is difficult to understand how the appearance of this significant unilateral ridge and groove on a young boy is consistent with the keystone concept, unless he matured and gained weight on one side only. Instead, I suggest that the early development and one-sidedness of a large ridge and groove suggests a compensatory degenerative reaction to a traumatic event rather than a normal development due to aging.

Vleeming had only one young specimen, which cannot be considered representative of normal. In contrast, Bowen & Cassidy [11] had 17 specimens under the age of 20 and Sashin [10] had 43 specimens under the age of 30. In both groups, only two of Sashin’s specimens had any detectable early changes in the articular cartilage and he appeared to attribute them to severe sudden trauma, which is consistent with his and Bowen & Cassidy’s statements, and inconsistent with the form closure model.

In the suspensory concept, the above changes are indicative of dysfunction which may occur after spraining of the supportive ligaments. I suggest that aberrant movement, due to ligament laxity, could induce shear and compressive forces at the joint surfaces and result in degeneration and arthrosis, as expressed in other joints. Please see Ridge and Groove Development for more.

Flat Joint or Twists like a Propeller
Central to the keystone concept is the idea that the articular surfaces of the sacroiliac joints are flat, which would allow force transfer from the sacrum to the ilium by compression at the articular regions of the sacroiliac joints [4]. In Snijders’ seminal publication, gravitational, ligamentous, and muscular forces are illustrated and evaluated by their ability to provide force vectors that are perpendicular to the joint’s surface. All joint angles and force vectors are based on one single, laterally facing facet per side. However, as noted in previous anatomical descriptions [12, 13], Snijders also discussed the fact that the SI joint has a twist resembling the shape of a propeller blade. So, instead of one flat facet on each side of the sacrum facing laterally, he notes that this provides two facets per side, the upper facing posterolaterally and the lower facing anterolaterally. Please see Facets Curl Around the Sacrum. Snijders interprets this configuration as providing enhanced stability against sacral flexion when pressed together; a weak statement that is given little attention and then not mentioned again. He does not explain how a single medially directed force vector, as clearly illustrated in his article, can be split into two separate force vectors that approach each oblique facet, facing different directions, at a perpendicular angle. In actuality, the sacrum appears to be a modified saddle joint, in which the facets curl around the edge of the sacrum, as explained in 3D Views of the Sacroiliac Joint and Facets Curl Around the Sacrum.

Dense Trabeculation
Gracovetsky [14] points out that the sacrum and ilium do not have a dense trabecular system underlying their supposedly flat joint surfaces, strongly indicating that weight-bearing is not a primary function of the articular surfaces. Rather, he notes a dense trabecular arrangement exists at the posterior periphery of the articular sacral surface, which he named the SG ridge. He suggests that weight transfer occurs at this ridge through bone-to-bone compression; however, he is describing the location of the syndesmosis, which is filled with ligaments, and where there is no intervening cartilage to absorb compressive forces.  At first, it would appear that this concept goes against the model of the suspended sacrum because conventional thought considers increasing trabecular density to result from compressive forces. However, recent studies [15, 16] have demonstrated that trabecular patterns exist in the ilium in the neonatal human. Considering this new information, the authors hypothesized that trabecular patterns in the ilium may be developed along stress lines generated by muscular and ligamentous pull.  It should be noted that the location of the SG ridge is at the attachment points of the powerful posterior sacroiliac ligaments. Consequently, the dense trabecular patterns underlying the SG ridge may actually be established during neonatal life and further developed through the strong pulling action of the ligaments of the posterior sacroiliac joint during weight bearing. More studies on trabecular pattern development would help clarify this issue.

Viscoelastic Collagen
Gracovetsky [14] also noted that posture, rather than being a static concept, is a continuous and rapid “cycling through a sequence of different but closely related postures.” He stated that, in a viscoelastic biomechanical system, pressure deforms collagen and other tissue, so almost constant movement is required to change pressure points to avoid substantial deformation. However, in a keystone, even though some mobility exists, stability would result from a steady, relatively static posture in which the sacrum fits snugly into the arch composed of the two ilia. Increased weight bearing would force the sacrum farther into the ilia and make it more difficult to shift weight to avoid deformation of the collagen. To shift weight, the body would have to lift the sacrum out of the deep set between the ilia, inducing a small degree of instability. Thus, in a supposedly ideal situation, the stability afforded by the keystone would contrast with the requirement of the tissue to maintain movement. This example of inefficient SIJ compression suggests a dysfunctional joint. However, in a suspensory system, movement is easily and readily achievable, as the muscles are free to continually alter pulling vectors to change pressure points on the tissue, thus avoiding substantial deformation and maintaining a state of controlled instability. Because a suspensory system is relatively frictionless, much less energy would be needed to initiate and propagate movement. Evolution favors efficiency.

Effect of Compression on Vessels
Within the syndesmosis, immediately posterior to the articular region, the space between the sacrum and ilium is filled with the interosseous, axial, and short posterior ligaments. Within these ligaments are nerves, blood and lymph vessels, fat, elastic and connective tissues [10, 17]. Because of the appearance of similar ridges and grooves in both regions [17], and the likelihood that they were developed in similar manners, it is unlikely that compression could occur at the articular area without also occurring at the syndesmosis. When compression is generated by body weight, especially during heavy lifting, it could conceivably decrease blood and lymph flow and damage the nerves and vessels within the syndesmosis; thus it would seem that compression is implausible in a normal sacroiliac joint. However, if the sacrum is suspended by ligaments, the vessels would not suffer compressive damage and the pressure changes occurring during movement would assist fluid flow within the vessels, enhancing nutrition inflow and waste removal. Thus, it seems that compression would occur only in a dysfunctional joint.

Effect of Traction
Any distractive force to a keystone arch, such as hanging upside down, inversion therapy, or lumbo-pelvic traction should cause significant instability and possible injury to a keystone arch pelvic model. Conversely, distraction would enhance sacroiliac motion in a suspensory model.

Holographic Analysis
Using holographic analysis, Vukicevic et al. [18] applied a wide range of loads to 12 cadaveric pelves with preserved lumbar spines, hip joints, and ligaments. Axial loading was applied at L1. It was found that removal of the sacrospinous and sacrotuberous ligaments had no significant effect on sacroiliac movement, but removal of the interosseous ligaments caused the sacrum to drop and fix between the ilia, having a profound limiting effect on sacroiliac movement. Additionally, they found that “tight contact between the articular surfaces is never reached in a wide range of the applied loads, secured by strong sacroiliac interosseous ligaments.” Therefore, it is indicated that, in normal motion, the joint surfaces do not touch, supporting the concept of ligamentous suspension, and that the presence of ridges and grooves is evidence of degeneration in a dysfunctional joint occurring after disruption of the interosseous ligament.

Additionally, using SPECT-CT, Cusi [19] found that “The utility of SPECT/CT for the diagnosis of SIJ dysfunction and more specifically the failure of load transmission across the joint is confirmed in the current study. The pattern of scintigraphic uptake and the CT scan changes reflect the pathology of the condition, where there is a primary abnormality of the supporting ligaments in the posterior aspect of the joint that fail to support the joint, leading to abnormal motion at the level of the joint itself, hence the increased uptake, sclerosis and occasional osteophyte formation as well as uptake in the damaged ligaments and the insertion on the ilium due to the altered mechanics...the vertical loads mainly go through the posterior and interosseous ligaments. The clinical profile of patients with sacroiliac disease and scan finding that mainly affect the synovial portion of the SIJ are quite different to the pattern we have defined for SIJ instability...Most research and clinical experience has concentrated on the anterior–synovial–and to a certain extent cartilaginous portions of the joint. However, from a biomechanical point of view the posterior ligamentous apparatus of the joint plays a key role in its function as a load transmitter which has resisted accurate imaging till now.” Repetition of this study by other researchers is suggested to confirm these results.

Although Vukicevic’s and Cusi’s research papers are the only scientific studies that demonstrate that the sacrum is suspended by ligaments, there are none that demonstrate that the sacrum functions as a keystone in a Roman arch.  Further studies to distinguish between suspension and compression are suggested.

Force Closure – The ligamentous, muscular, fascial component
Force closure, a basic tenet of the keystone concept, supposedly occurs through dynamic action of the ligaments, muscles, and fascia creating compression, or resistance to opening beyond normal range of motion. However, it is interesting to note that the opposing force that may function to decompress the SIJ is not mentioned in the keystone concept.

Kapandji stated that nutation winds up the interosseous and short dorsal ligaments, holding the suspended sacrum more tightly [20]. Force closure proponents interpret this action as drawing the ilia into the sacrum, thereby increasing compression at the SIJ [6]; however, this is contradicted by Kapandji, himself,[20], who stated that the sacrum is suspended, and Vukicevic [18], who’s study found that the sacral and ilial surfaces never touch (see above) when the interosseous ligament is intact.

The gluteus maximus is supposed to be one of the most important muscles in creating force closure [4]. Paired with the contralateral latissimus dorsi, its fibers cross the sacroiliac joint and provide a nutation moment, which is considered to be crucial to self-locking of the pelvis. Because weakness of the gluteus maximus was consistently found in sacroiliac joint dysfunction, it was thought that muscle weakness precipitated sacroiliac instability, due to a reduction in compressive force. Therefore, it was suggested that strengthening the gluteus maximus would be an ideal exercise to help correct SIJ dysfunction [6, 21, 22]. However, contrary results were found by Mens et al. [23], who studied the effect of exercise on patients with peripartum pelvic pain. Prone subjects raised their lower extremities, exercising their hip extensors, thereby inducing nutation. The results suggest that contracting the hip extensors may worsen the pain. The authors stated that “The most important consequence of our study is that the hypothesis about form and force closure, described by Vleeming and colleagues, needs to be revised.” I suggest that the exercise may have driven the sacroiliac joints beyond their normal range of motion into excess nutation, forcing the sacral base away from the ilia, stressing the already sprained limiting ligaments.

Supposedly, in the keystone model, strengthening the gluteus maximus and other muscles that promote nutation would drive the sacroiliac joint into nutation where bone-on-bone contact would provide stability. This approach, however, does not allow for the possibility that the joint can go too far into nutation and stress the restraining ligaments. This approach would also disallow the possibility of a ligamento-muscular reflex [24], which has been shown to be a key factor in muscular stabilization of joints after injury, throughout the body [25-49].

Conversely, in a suspensory model of a sacroiliac lesion, the already sprained ligaments would be further stressed in nutation. The resulting ligamento-muscular reflex would inhibit the muscles that promote nutation, in order to avoid further stressing the ligaments, hence the weak gluteus maximus and associated muscles. The implied difference between the two concepts is that the keystone model suggests that muscle weakness leads to sacroiliac joint dysfunction, but the suspensory model suggests that sacroiliac dysfunction leads to muscle weakness. If the keystone model were true, the sacroiliac joint would be the only joint where the ligamento-muscular reflex was known to not apply. However, the fact that the ligamento-muscular reflex has been found to exist in the sacroiliac joint [45, 47-49] lends credence to Mens’ findings, as well as to the suspensory theory.

In order for a biomechanical model to be accepted as valid, all known and verified biomechanical functions and associations must behave as predicted by the model. Clearly, there are too many discrepancies in the keystone concept for it to be recognized as valid, while suspension has appeared to be likely in every scenario.  

      -    The anterior-posterior shape of the sacral base demonstrates a
            predisposition for the sacrum to drop away from the ilia during weight
            bearing, which is reinforced by the opposing movements of the sacrum
            and ilium. The anatomical function of the large mass of ligaments
            suspending the sacrum within the ilia becomes evident, as in other
      -    The appearance of ridges and grooves at both the articular and
            syndesmosis regions demonstrates that compression exists at both
            regions. Degenerative-type changes in both regions and potential
            impingement of the vessels in the syndesmosis region suggest that these
            changes are the result of dysfunction after injury, rather than normal
      -    Holographic analysis and SPECT/CT scanning provide evidence for
            suspension of the sacrum from the ilia as the valid mechanism for weight
            bearing at the SIJ, rather than compression of the joint surfaces. No
            scientific evidence exists demonstrating the keystone form closure
      -    Central to the keystone model is the idea that the articular surfaces of
            the sacroiliac joints are flat and facing directly laterally to absorb
            perpendicular medially facing force streams. However, the SIJ surfaces
            have a twist resembling the shape of a propeller blade, discounting the
            possibility of functioning as a single flat surface.
      -    The lack of dense trabecular patterns below the SI articular surfaces
            discount the likelihood that the articular surfaces are weight bearing;
            instead, their appearance at the syndesmosis indicates that it is the
            location of load transmission through the SIJ,  through ligamentous or
            muscular pull, rather than compression.
      -    In order to prevent deformation of the viscoelastic collagen within the
            SIJ, continuous and rapid motion is required; this is difficult to achieve in
            a keystone SIJ model, yet easy to achieve in a biotensegrity model in
            which the sacrum is suspended by ligaments.
      -    The keystone concept of form closure does not allow for the possibility
            that the SI joint can go too far into nutation and stress the restraining
            ligaments, which would further disallow the likelihood of a protective
            ligamento-muscular reflex.
      -    Too many discrepancies exist in the keystone model for it to be
            considered valid, whereas the sacral suspensory model appears to
            encompass all known and proven biomechanical functions and


1.    Vleeming, A., et al., Relation between form and function in the sacroiliac joint. Part II: Biomechanical aspects. Spine, 1990. 15(2): p. 133-6.
2.    Vleeming, A., et al., The sacroiliac joint - an overview of its anatomy, function and potential clinical implications. J Anat, 2012. 221(6): p. 537-67.
3.    White, A. and M. Panjabi, Clinical Biomechanics of the Spine. 2nd ed. 1990, Philadelphia, PA: J.B. Lippincott Company.
4.    Snijders, C.J., Transfer of Lumbosacral Load to Iliac Bones and Legs: Part 1 - Biomechanics of Self-Bracing of the Sacroiliac Joints and its Significance for Treatment and Exercise. Clinical Biomechanics, 1993a. 8: p. 285-294.
5.    Lee, D., The Pelvic Girdle. 2nd ed. 1999: Churchill Livingstone.
6.    Vleeming, A., et al., The role of the sacroiliac joints in coupling between spine, pelvis, legs and arms., in Movement, Stability, and Low Back Pain, A. Vleeming, et al., Editors. 1997, Churchill Livingstone. p. 53-71.
7.    Vleeming, A. and R. Stoeckart, The role of the pelic girdle in coupling the spine and legs: a clinical-anatomical perspective on pelvic stability, in Movement, Stability, & Lumbopelvic Pain, A. Vleeming, D. Mooney, and R. stoeckart, Editors. 2007, Churchill Livingstone. p. 113-137.
8.    Haussler, K. Functional Assessment and Rehabilitation of the Equine Axial Skeleton. in ACVS Veterinary Symposium. 2012.
9.    Vleeming, A., et al., Relation between form and function in the sacroiliac joint. Part I: Clinical anatomical aspects. Spine, 1990. 15(2): p. 130-2.
10.    Sashin, D., A critical analysis of the anatomy and the pathologic changes of the sacro-iliac joints. The Journal of Bone and Joint Surgery, 1930. 12: p. 891.
11.    Bowen, V. and J.D. Cassidy, Macroscopic and microscopic anatomy of the sacroiliac joint from embryonic life until the eighth decade. Spine, 1981. 6(6): p. 620-8.
12.    Solonen, K.A., The sacroiliac joint in the light of anatomical, roentgenological and clinical studies. Acta Orthopaedica Scandinavica. Supplementum, 1957. 27(Suppl 27): p. 1-127.
13.    Dijkstra, P.F., A. Vleeming, and R. Stoeckart. Complex Motion Tomography of the Sacroiliac Joint and an Anatomical and Roentgenological Study. in First Intedisciplinary World Congress on Low Back Pain and Its Relation to the Sacroiliac Joint. 1992. San Diego, CA.
14.    Gracovetsky, S., Stability or controlled instability?, in Movement, Stability & Lumbopelvic Pain, A. Vleeming, V. Mooney, and R. Stoeckart, Editors. 2007, Churchill Livingstone Elsevier: London. p. 279-294.
15.    Cunningham, C.A. and S.M. Black, Anticipating bipedalism: trabecular organization in the newborn ilium. J Anat, 2009 A. 214(6): p. 817-29.
16.    Cunningham, C.A. and S.M. Black, Development of the fetal ilium--challenging concepts of bipedality. J Anat, 2009 B. 214(1): p. 91-9.
17.    Rosatelli, A.L., A.M. Agur, and S. Chhaya, Anatomy of the interosseous region of the sacroiliac joint. The Journal of Orthopaedic and Sports Physical Therapy, 2006. 36(4): p. 200-8.
18.    Vukicevic, S., et al., Holographic analysis of the human pelvis. Spine, 1991. 16(2): p. 209-14.
19.    Cusi, M., SPECT-CT on patients with a clinical diagnosis of failure of load transfer of the sacro-iliac joint, in 7th Interdisciplinary World Congress on Low Back & Pelvic Pain. Effective Diagnosis and Treatment of Lumbopelvic Pain. 2010: Los Angeles, CA. p. 291-296.
20.    Kapandji, I.A., The Physiology of the Joints. Vol. 3. 1977: Churchill Livingstone.
21.    Vleeming, A., et al., The posterior layer of the thoracolumbar fascia. Its function in load transfer from spine to legs. Spine, 1995. 20(7): p. 753-8.
22.    Mooney, V., et al., Exercise treatment for sacroiliac pain. Orthopedics, 2001. 24(1): p. 29-32.
23.    Mens, J.M., C.J. Snijders, and H.J. Stam, Diagonal trunk muscle exercises in peripartum pelvic pain: a randomized clinical trial. Phys Ther, 2000. 80(12): p. 1164-73.
24.    Palmer, I., Pathophysiology of the medial ligament of the knee joint. Acta Chirurgica Scandinavica, 1958. 115(4): p. 312-8.
25.    Kim, A.W., et al., Selective muscle activation following electrical stimulation of the collateral ligaments of the human knee joint. Archives of Physical Medicine and Rehabilitation, 1995. 76(8): p. 750-7.
26.    Raunest, J., M. Sager, and E. Burgener, Proprioceptive mechanisms in the cruciate ligaments: an electromyographic study on reflex activity in the thigh muscles. The Journal of Trauma, 1996. 41(3): p. 488-93.
27.    Dyhre-Poulsen, P. and M.R. Krogsgaard, Muscular reflexes elicited by electrical stimulation of the anterior cruciate ligament in humans. Journal of Applied Physiology, 2000. 89(6): p. 2191-5.
28.    Friemert, B., et al., Differentiation of hamstring short latency versus medium latency responses after tibia translation. Experimental Brain Research. Experimentelle Hirnforschung, 2005. 160(1): p. 1-9.
29.    Knatt, T., et al., The glenohumeral-biceps reflex in the feline. Clinical Orthopaedics and Related Research, 1995(314): p. 247-52.
30.    Guanche, C., et al., The synergistic action of the capsule and the shoulder muscles. The American Journal of Sports Medicine, 1995. 23(3): p. 301-6.
31.    Solomonow, M., et al., Mechanoreceptors and reflex arc in the feline shoulder. Journal of shoulder and elbow surgery / American Shoulder and Elbow Surgeons ... [et al.]. 1996. 5(2 Pt 1): p. 139-46.
32.    Voigt, M., J. Jakobsen, and T. Sinkjaer, Non-noxious stimulation of the glenohumeral joint capsule elicits strong inhibition of active shoulder muscles in conscious human subjects. Neuroscience Letters, 1998. 254(2): p. 105-8.
33.    Tsuda, E., et al., Direct evidence of the anterior cruciate ligament-hamstring reflex arc in humans. The American journal of Sports Medicine, 2001. 29(1): p. 83-7.
34.    Phillips, D., et al., Ligamentomuscular protective reflex in the elbow. The Journal of Hand Surgery. American volume., 1997. 22(3): p. 473-8.
35.    Petrie, S., et al., Mechanoreceptors in the palmar wrist ligaments. The Journal of Bone and Joint Surgery. British volume, 1997. 79(3): p. 494-6.
36.    Stubbs, M., et al., Ligamento-muscular protective reflex in the lumbar spine of the feline. J Electromyogr Kinesiol, 1998. 8(4): p. 197-204.
37.    Solomonow, M., et al., The ligamento-muscular stabilizing system of the spine. Spine, 1998. 23(23): p. 2552-62.
38.    Indahl, A., et al., Electromyographic response of the porcine multifidus musculature after nerve stimulation. Spine, 1995. 20(24): p. 2652-8.
39.    Indahl, A., et al., Interaction between the porcine lumbar intervertebral disc, zygapophysial joints, and paraspinal muscles. Spine, 1997. 22(24): p. 2834-40.
40.    Freeman, M.A. and B. Wyke, Articular reflexes at the ankle joint: an electromyographic study of normal and abnormal influences of ankle-joint mechanoreceptors upon reflex activity in the leg muscles. The British Journal of Surgery, 1967. 54(12): p. 990-1001.
41.    Bullock-Saxton, J.E., Local sensation changes and altered hip muscle function following severe ankle sprain. Physical Therapy, 1994. 74(1): p. 17-28; discussion 28-31.
42.    Hintermann, B., Biomechanics of the unstable ankle joint and clinical implications. Medicine and Science in Sports and Exercise, 1999. 31(7 Suppl): p. S459-69.
43.    Solomonow, M. and J. Lewis, Reflex from the ankle ligaments of the feline. Journal of Electromyography and Kinesiology, 2002. 12(3): p. 193-8.
44.    Clark, R.K. and B.D. Wyke, Temporomandibular arthrokinetic reflex control of the mandibular musculature. The British Journal of Oral Surgery, 1975. 13(2): p. 196-202.
45.    Indahl, A., et al., Sacroiliac joint involvement in activation of the porcine spinal and gluteal musculature. Journal of Spinal Disorders, 1999. 12(4): p. 325-30.
46.    Tichy, M., et al., Pelvic Muscles Influence the Sacroiliac Joint. Journal of Orthopaedic Medicine, 1999. 21(1).
47.    Sakamoto, N., et al., An electrophysiologic study of mechanoreceptors in the sacroiliac joint and adjacent tissues. Spine, 2001. 26(20): p. E468-71.
48.    Holm, S., A. Indahl, and M. Solomonow, Sensorimotor control of the spine. Journal of Electromyography Kinesiology, 2002. 12(3): p. 219-34.
49.    Marshall, P. and B. Murphy, The effect of sacroiliac joint manipulation on feed-forward activation times of the deep abdominal musculature. Journal of Manipulative and Physiological Therapeutics, 2006. 29(3): p. 196-202.