Authors' response to the letter to the Editors by Professor M. T. Cibulka: a critical interpretation of sacroiliac joint movement studies

General comment

We welcome this opportunity to respond to the editorial by Cibulka, concerning the review article ‘Sacroiliac joint: an overview of its anatomy, function and potential clinical implications’. Professor Cibulka's main theme of critique is related to the small amount of motion in the sacroiliac joints and how the joint functions. This critique is mainly based on a limited number of studies, which are generally not in agreement with the wide body of research achieved over many decades about sacroiliac joint (SIJ) function. The same critique is presented by the author in a previous editorial, reacting to an SIJ movement study using roentgen stereophotogrammetric analysis (RSA) by Sturesson et al. (2000b). In this study, Sturesson reported consistently small but relevant movement of the SIJ. In an editorial comment, Cibulka challenged these findings by suggesting that the greatest source of error when using RSA is achieving proper three-dimensional reconstruction of the tantalum balls in the sacrum and innominate bones. Sturesson responded to the critique by explaining that the RSA technique enables optimal spatial contribution of the implanted markers, resulting in highly precise measurements. Moreover, a recent article has reported that when RSA is applied correctly, the method is reproducible and can be regarded as the gold standard in measuring small joint movements in orthopedic research (Kibsgård et al. 2012).

Specific items

Concern 1: In relation to our review, Professor Cibulka suggests that the average amount of sagittal movement (2°) of the SIJ is too small to maintain a large (17.5 cm² surface area) healthy joint.

The SIJ is situated between the lumbar spine above and the femur below; therefore, it functions in load transfer between torso and legs. As such, it is exposed to large forces and, to maintain stability, its construction has to restrict motion. The anterior aspect of the joint has a taut, smooth capsule while the posterior aspect of the joint is a dense collection of interosseous and short and long sacroiliac ligaments. Both ilia are then ‘clamped’ together by a thickened band/fascia created by the fusion of the aponeuroses of the erector spinae and the superficial and deep lamina of the posterior layer of the thoracolumbar fascia (Willard et al. 2012). Further restrictions on SI joint range of motion involve its kinematic relationship with the pubic symphysis. The two SIJs act as important stress relievers in the ‘force–motion’ relationships between trunk and lower limb and insure that the pelvic girdle is not a fixed solid ring of bone that could fracture easily from trauma. In bipedal gait, the SIJ are the ‘hub’ of forces transferred from the trunk to the ground and vice versa (Aiello & Dean, 1990; Lovejoy, 2007; Vleeming, 2007).

Cibulka attempts to relate characteristic SIJ motion exclusively to the anatomy of the auricular part of the SIJ. However, this is not a realistic approach since the SIJ is a complex joint composed of a ventrally positioned auricular component surrounded by a multi-composite capsule and lined with cartilage (diarthros) and a dorsally located extracapsular axial component (synarthros) composed of an uneven fibrous patch of cartilage surrounded by the interosseous ligaments (Bakland & Hansen, 1984). The tenet that joint form follows function and adapts to biomechanical demands is highly relevant for the SIJ. In trying to fathom the relationship between the form of this joint and its movement, it should be realized that even at such an early period in life as intrauterine development, conspicuous differences are seen between the iliac and sacral part of the joint. The sacral surface is typically characterized as smooth, white, thick articular cartilage, whereas the thin, rougher and dull iliac cartilage presents a fibrocartilaginous appearance in intrauterine life and in early childhood (Bowen & Cassidy, 1981), becoming more hyaline in nature with maturation (Paquin et al. 1983; Kampen & Tillmann, 1998). After puberty the SIJ adopts its shape in relation to increased weight with a more complex sinusoidal or propeller-shaped form (Schuncke, 1938; Solonen, 1957). Likewise, symmetrical grooves and ridges, covered by hyaline or articular cartilage, are formed on the surface of the joint. Finally, the axial portion of the SI joint (accessory SI joint) also develops around the time of birth and its main ligaments, the interosseous ligaments, are the strongest ligaments of the SIJ complex (Steinke et al. 2010). These numerous adaptations in structure (Bowen & Cassidy, 1981; Vleeming et al. 1990a,b) enhance the form closure of the SIJ and contribute to the highest coefficient of friction of the diarthrodial joints, thus further limiting the joint's range of motion (Vleeming et al. 1990b). Since the axial portion of the joint plays a large role in restricting its motion, it is incorrect to analyze SI joint motion solely on the basis of the construct of the auricular SIJ portion.

Cibulka states that the large surface area of the SI joint ‘mandates’ that the (auricular) joint has to move more than the reported mean 2° in order to maintain the health of its cartilaginous surfaces. However, combining all the data from published studies employing RSA and appropriate markers, the maximum sagittal rotation of the SIJ never exceeds 3.6° and translation of the joint never exceeds 2 mm (Kibsgård et al. 2012). Amalgamating these data with the observation that osteophytosis in women, regardless of age, is rare (1.83% of females) and not very common in men (12.27%; Dar et al. 2008), it therefore appears that small SIJ movements are present, even at an advanced age (Vleeming et al. 1992). The conclusion from these observations is that small, but essential, SIJ movement can maintain normal function of the SI joint even through the advancing age of the individual.

Concern 2: Most studies on SIJ movement are not in agreement and that the movement studies in the current SIJ review do not clarify how the SIJ move.

In our review we mention numerous studies of the SIJ, using a variety of techniques, to analyze the principal displacement characteristics of the joint. Fundamentally, these studies can be separated into two groups, those using RSA and those that did not. First, numerous older studies have used various anatomical and radiographic (pre-RSA) methods, both in vivo and in vitro, to examine SIJ motion in the transitions from the supine, seated and standing positions. Interestingly, although some of these studies mentioned here use different techniques, they all report consistent, but small, movements around an average sagittal motion of 2° (Luschka & Bender, 1864; Meyer, 1878; Walcher, 1889; Forthergill, 1896; Jarcho, 1929; Von Schubert, 1929; Sashin, 1930; Weisel, 1955; Solonen, 1957; Colachis et al. 1963).

The second group of studies applied RSA to analyze SIJ movement in various positions. These studies conclude that in the SIJ, sagittal movements occur consistently up to mean of 2° (Egund et al. 1978; Selvik, 1990; Tullberg et al. 1998). Sturesson et al. (1989, 1999, 2000a,b) studied pelvic girdle pain patients in multiple RSA studies, analyzing SIJ movement in 11 different positions. Movement from supine to standing around the helical axis (describing both simultaneous rotation and translation) was 1.3°; from supine to sitting around the same axis, the movement is increased to 1.55°. The authors also calculated the differences between male and female SIJ motion. The mean mobility for men is on average 40% smaller between positions ‘supine’ and ‘standing’. Between positions ‘standing’ and ‘prone’ the difference was 30%.

A different approach for SIJ motion analysis was taken by Kissling et al. (1990), who applied Kirschner rods in both ilia and sacrum of healthy volunteers (among them his orthopedic assistants!). Measurements were made in standing, anteflexion and retroflexion of the lumbar spine. Once again the study showed an average sagittal SIJ rotation between standing erect on both feet and one-legged stance of about 2° (range 0.4–4.3°).

Interestingly, the RSA study from Lavignolle et al. (1983), as mentioned by Cibulka, recorded much higher movement values than those reported in other RSA studies of SI joint movement. The Lavignolle study analyses SIJ movement in volunteers from a supine position while extending their legs, resulting in 2° of SIJ motion and 12° in anteflexion. However, to obtain accurate measurements with RSA it is mandatory to insert many well spaced markers in the sacrum and ilium with a good spatial configuration. In the Lavignolle study, measurement was based on natural existing pelvic bone marker points; no internal markers were implanted. The reliance only on bony landmarks could basically explain the striking different outcomes of this study when compared with other RSA studies.

The study of Smidt et al. (1995), as mentioned by Cibulka and discussed in our review, examined SIJ motion in a reciprocal straddle position using an external skeletal measurement system without internal markers. They reported mean sagittal SIJ motion of 9°. The greatest error of measurement with this technique is the calculation of bony landmarks under the skin markers. Sturesson et al. (2000b) reevaluated this study with RSA using internal markers and demonstrated that reciprocal movements takes place. However, the movements were 10 times smaller compared with the Smidt et al. (1995) study. In a subsequent pelvic study (Barakatt et al. 1996), the authors found mean SIJ oblique sagittal composite motion of 22–36°. However, in this latter study the authors commented professionally and critically that their inability to locate precisely three landmarks for each of the innominate bones and sacrum precluded acquisition of three-dimensional measurements of the SIJ. In a follow-up study by Smidt et al. (1997), fresh cadavers were prepared with radiopaque internal bone markers. The cadavers were tested for SIJ mobility in the side-lying position. The extended legs were forced into hip anteflexion and extension and reciprocally extended and flexed, throughout these movements, while external pelvic motion was restricted by a band. Measurement was done by CT scans which were taken from one X-ray source and not by stereophotogrammetry. The lack of multiple, spatially separate imaging sources in this study makes it difficult to analyze correctly the arrangements of the beads. Similarly, studying SIJ movement with RSA does not require the restriction of external pelvic motion with a band. The method allows precise analysis and differentiation between external and internal pelvic motion.

An additional possible confounding factor in the Smidt studies is that a side-lying position makes proper spatial analysis of markers very difficult, even if multiple imaging sources had been used. In an effort to avoid this situation, Sturesson et al. (2000b) chose to examine the reciprocal hip position, when standing, enabling precise RSA measurement. Nevertheless, if the results of Smidt et al. are accurate, the large movements reported could be due, at least in part, to the reciprocal displacement of the legs in a supine position in the absence of axial loading of the lumbar spine. Given the protocol used in the study, it is conceivable that the SIJ was stretched beyond their anatomical boundaries. The measured results could then be regarded typical for the unique side-lying test position of the cadavers but may not be directly applicable to in vivo situations. Adding compression into the lumbosacral junction, as was performed in a similar study (Vleeming et al. 1992), resulted in a more restricted motion of the SIJ. Nonetheless, the various Smidt et al. studies (1995, 1997) were valuable contributions to stimulate new innovative research with different methodologies.

Finally, Kibsgård et al. (2012), responding to a previous letter of criticism by Cibulka (2001), tested the accuracy and precision of RSA by comparing a plastic phantom model of the pelvis to six female patients with chronic pelvic girdle pain. Measurements were performed on the model to determine accuracy and precision, while measurements were performed on the patients to determine precision of the technique. The following markers were placed in the model as well as in each of the six patients: five dorsal markers in the ilium, three frontal markers in the inferior pubic ramus, six markers in the body of the sacrum and three markers in the cranial aspect of the sacrum. The results found high accuracy and precision as well and concluded that the RSA is a valid technique for measuring small SIJ movements.

Concern 3: The pubic symphysis act as a fusion joint except in pregnancy.

Both Chamberlain (1930) and Walheim et al. (1984) used radiographic methods to evaluate movement in the pelvis. Chamberlain examined vertical pubic kinematics as a measure for abnormal pelvic motion. In this study he found consistently small vertical pubic displacements of ∼ 2 mm. After modifying the technique, Walheim measured mean displacements of 1 mm for males and 1.3 mm for females and maximum of 3.1 mm in one multiparous woman. A further study, using healthy volunteers as subjects, confirmed these previous findings by reporting symphysial movement in men of average 1.4 mm, in nulliparous woman 1.6 mm and in multiparous women 3.1 mm (Garras et al. 2008).

In his argument that the motion in the SI joints are ‘always paired’, Cibulka also refers to the pubic symphysis as being ‘fused and does nothing more than insure equal motion’ in both left and right SI joint. To confirm this statement Cibulka refers to the study of Meissner et al. (1996); however, this study, of which the main part is available only in the original German, uses the outcome data of Walheim's symphysis measurements to analyze how the minimum amount of force needed on the symphysis to evoke the movements observed in the Walheim study. Meissner et al. describe the force/movement relation of the symphysis but did not conclude in any way that the symphysis is fused. Conversely, these researchers actually confirm that the symphysis measures of Walheim et al. (1984), which were in the range of 1–3 mm, were physiologically realistic.

In conclusion, although Cibulka has brought forward several studies demonstrating a very large range of motion for the SIJ, these studies have suffered from several problems. These complicating factors include the use of manual techniques (Cibulka et al. 1988) or the use of external markers over the skin (Smidt et al. 1995) that do not allow discrimination between movements specific to the SIJ from those which involve overall pelvic displacement; the use of a single X-ray source rather than multiple sources to measure SIJ motion (Smidt et al. 1997); or the use of RSA techniques without internal markers and therefore an inability to obtain reliable measurements (Lavignolle et al. 1983).

Concern 4: Cibulka comments that the sacroiliac joints behave as one functional unit because all spinal bicondylar joints are always paired kinematically with an equal and correlative movement between left and right. Also, that it is not known if too little or too much motion of the SIJ may cause pain.

Cibulka refers to ‘all spinal joints, except the atlanto-axial articulation’ as bicondylar joints where the movement of the left joint is ‘always paired with an equal and correlative movement of the right joint’. Bicondylar joints are not defined in either the American or British version of Gray's anatomy (Clemente, 1985; Standring, 2008), however, they were defined by Weisel (1955) as joints where both surfaces move correlatively together. This definition does not establish a magnitude for amount of movement.

Cibulka appears to make the assumption that motion in the sacroiliac joint follows similar parameters to that of the ‘all the bicondylar spinal joints’. Indeed, the sacral superior articular processes and the inferior processes of L5 form the L5-S1 zygoapophyseal joints. However, the name ‘sacroiliac’ joint clearly points at the junction were the sacrum is moving relative to the ilium and not to the sacral- vertebral junction in the vertebral spine. The SIJ closely resembles an evolved costovertebral joint with the sacrum regarded as a fusion of vertebra and the ilia regarded as fused ribs. In this comparison the auricular part of the SIJ represents the costo-vertebral articulation and the axial part the costo-transverse articulation (Gracovetsky et al. 2007). The assumed correlation between motion in a zygapophyseal joint and that of the SI joint is not realistic, since the embryological composition of the zygapophyseal joints of the spine differs significantly from the composition of the sacroiliac joint (Williams, 1995).

There is a sufficient and a varied body of evidence available, using different research methodologies, that addresses the issue as to why the left and right side of the SIJ move distinctly from each other within a very small range of motion. Our review article referred to some of the older literature on this subject; however, here are some further examples:

1. It has been shown that the interindividual geometry of the auricular and axial surfaces of the left and right SI joints differ significantly (Bakland & Hansen, 1984), leading to dissimilar movement between left and right SIJ.

2. The RSA method was used to measure the displacement of the SIJ in patients with pelvic girdle pain (PGP) with the following outcome: movement around the sagittal and helixal axes normally shows different degrees of motion intra-individually when the individual is changing position from supine to standing, supine to sitting, or standing to prone with one leg alternating left and right in hyperextension (Sturesson, 2007).

3. Symphyseal studies show that when pelvic girdle pain is present, larger movements of the symphysis occur at the painful side of the joint (Chamberlain, 1930).

4. Studies on PGP patients, using the active straight leg raise test (ASLR), show that on the side of the impaired leg, symphyseal motion around the transversal axis is increased (Mens et al. 1999).

5. When Doppler imaging of vibrations (DIV) is used to measure the stiffness-laxity values of patients with diagnosed PGP, asymmetric laxity of the SIJ leads to significantly more pain compared with patients with symmetric laxity. Additional research using DIV reports that voluntary unilateral contractions of relevant related muscles to the pelvis diminish SIJ mobility specifically on the ipsilateral side (van Wingerden et al. 2004).

6. PGP patients who test unilaterally positive on the functional ASLR test and show an inability to raise the leg, can temporarily normalize the test by manual compression or the use of a pelvic belt. This suggests that increased unilateral motion of the SIJ could lead to impairment (Mens et al. 1999).

Sturesson et al. using RSA, studied the effects of an external fixation frame for severe PGP patients on SIJ motion, prior to final arthrodesis. The effect is a median reduction of SIJ motion around the helical axis in eight patients of 55% of the left and 63% on the right side. The mean reduction around the helical axis was 59 and 74% around the X-axis. When the frame was tightened, an anterior sagittal movement of the sacrum was noticed. Most patients, directly after tightening the frame, showed increased loading capability. Furthermore, evidence-based PGP tests show an immediate positive reversal of relevant test scores. This study demonstrates that, in severe PGP patients, a non-controlled SIJ movement of 2° can present too much laxity. This and other studies mentioned here imply that ‘too much motion’ of the SIJ, could be a significant factor in the onset of PGP.

The studies presented here as well as additional studies contained in our review article, show that the two SIJs can have independent motion and that comparing the anatomy and function of the SIJ to the zygoapophyseal joints is a non-realistic point of view. Clearly, we agree with Cibulka's stance that there is a need for more testing of SIJ mobility during normal daily movements. However, such a future exploration can only commence if reliable testing methods are developed to study the SIJ during normal activities such as walking.