IntroductionLow back pain (LBP) is one of the most common reasonsfor primary care visits after the common cold, with approximately 90% of adultsbeing impacted by this condition at some time in their lives 1, 2. One of themost overlooked sources of LBP is the sacroiliac joint (SIJ) due to its complexnature and the fact that the pain emanating from this region can mimic otherhip and spine conditions 1, 3. However, recent studies have reported a higherprevalence of the SIJ as a source of LBP leading physicians to place greaterfocus on the treatment and consideration of SIJ dysfunction as a pain generator4. The SIJ, the largest axial joint in the body, is thearticulation of the spine with the pelvis that allows the transfer of loads topelvis and lower extremities 5, 6. Sexual dimorphism exists in pelvis suchthat compared to the male sacrum, the female sacrum is generally wider, moreuneven, less curved, and more backward tilted.
Males tend to have a relativelylong and narrow pelvis, with a longer and more conical pelvic cavity than thoseof females 7. There are different methods to measure theSIJ motion such as roentgen stereophotogrammetric, radio-stereometric, and ultrasound8-11. Using these methods, it is shown that the SIJ rotation and translationin different planes are not exceeding 2-3° and 2 mm, respectively 12, 13. To the authors’ best knowledge, there is no studywhich discusses the biomechanical differences between male and female SIJs interms of ROM, SIJ ligaments strain, stress, and load sharing across the SIJ.Cadaver studies would be technically demanding due to low motions at SIJ.
Inaddition, quantifying stresses across the joint is not feasible. Therefore, experimentallyvalidated finite element analysis approach would be the most practical tool to assessthe ROM, stresses, and strains across the joint. The objective of this studywas to quantify these parameters at the SIJ using gender specific finiteelement models of SIJ. The study was aimed to better understand thebiomechanical differences in SIJ between genders in terms of their mobility andthe possible pain sites. Material and MethodsMale Finite Element Model of theLumbar Spine-Pelvis-FemurThe previously developed and validatedfinite element lumbar spine model 14, 15 was used for the male model. The3-dimensional (3D) pelvis geometry was generated using a 1 mm slice of computertomography (CT) of a 55 year old male pelvis without any abnormalities,degeneration or deformation of the pelvis.
The 3D reconstruction of spino-pelvismodel was done using MIMICS software (Materialise Inc., Leuven, Belgium). After3D reconstruction of the bones and spinal discs, they were imported intoGeomagic Studio software (Raindrop Geomagic Inc., USA) to reducenoises, remove spikes, smooth surfaces, and create patches and grids formeshing.
Hypermesh software (Altair Engineering, Inc., USA) was used to createthe mesh structure from the 3D model. Lumbar spine and pelvis bones were modeled astrabecular cores surrounded by a cortical layer with a thickness of 1 mm 15,16.
The linear hexahedral element type was utilized for cortical andcancellous bones of vertebrae and intervertebral discs. Tetrahedral elementtype was used for the cortical and cancellous bones of the pelvis. The trusselements were employed for ligamentous tissues including the SIJ and spinal ligaments.144,360 elements were generated for the male model. Female Finite Element Model of theLumbar Spine-Pelvis-FemurComputer tomography (CT) images of a 55years old female’s spine, pelvis without any abnormalities, degeneration wereused to reconstruct the female spino-pelvis model.
MIMICS software(Materialise, Leuven, Belgium) was utilized to build the 3D geometry of thebones and then intervertebral discs were made by filling the space between eachtwo vertebrae of the CT images. Next, smoothing and meshing were carried out byGeomagic Studio software (Raindrop Geomagic Inc., USA) and the Hypermeshsoftware (Altair Engineering, Inc., USA). Figure 1 shows the male and femalespine-pelvis-femur FE models. The linear hexahedral element type was utilized forcortical bone of vertebrae and spinal discs.
Tetrahedral elements were assignedto the cancellous bone of both the vertebrae and the pelvis as well as thecortical bone of the pelvis. The truss elements were employed for ligamentoustissues. The SIJ ligaments were anterior sacroiliac ligament (ASL),interosseous ligament (ISL), long posterior sacroiliac ligament (LPSL), shortposterior sacroiliac ligament (SPSL), sacrospinous ligament (SSL), andsacrotuberous ligament (STL).
A detailed view of the pelvis ligaments is shownin the figure 2. The female model as a whole contained 463,735 elements. Material PropertiesThe material properties used in the FEmodels were extracted from previous studies 14, 17 for cortical andcancellous bones, annulus, nucleus, ligaments, and joints are summarized in table1. Similar material properties were used for both male and female models. The SIJs,spine facets, articular cartilages, and pubic symphysis were modelled asnon-linear soft contact. The femurs were kinematically coupled to the pelvis.
Mesh Convergence Study Themesh convergence analysis was done on the segregated L4-L5 motion segment ofthe female model. An initial seed size was assigned and the model was subjectedto 7.5 N.m bending moment to simulate motions in all planes and the ROM wasmeasured. The mesh refining was repeated until the difference in the ROM in allplanes was below 4%. The final element size so determined was used to mesh theother segments of the model. The simulation was run using ABAQUS 6.14 software(SIMULIA, Inc.
, Providence, RI, USA). Finite Element Model Validation The intact male model SIJ ROM was previouslyvalidated 14, 15 against study of Miller et al. 18 under the same loadingand posture conditions. Due to lack of data on female specimens SIJ ROM undertwo leg stance condition, the validation for the female model was performed underone leg stance condition. To be consistent, the validation under one leg stancecondition was done for both male and female models. To validate the SIJs ROMfor intact male and female models, loading conditions of the cadaver study doneby Lindsey et al. 19 was simulated. This experiment was carried out forintact L4 to pelvis of the male and female specimens under single leg stancecondition.
A 7.5 Nm pure moment load was applied to the top endplate of L4 tosimulate various spinal motions. The motion at the SIJ was calculated for bothright and left joints.Loading and Boundary ConditionsIn all models, a 400 N compressivefollower load was applied through wire elements which followed the curvature ofthe lumbo-pelvis segment to simulate the effect of muscle forces and weight ofthe upper trunk.
A 10 N.m bending moment was then applied at the superiorsurface of the L1 vertebrae to simulate the physiological flexion, extension,lateral bending, and axial rotation. To constrain the models, femurs were fixedin all degrees of freedom to prevent relative displacement of the legs in twoleg stance condition 14, 15. DataAnalysisThe SIJ motion was calculated using the angulardisplacements at the sacrum minus those at the ilium for right and left joints.
The maximum von Mises stresses, normal and shear loads across theSIJ for each of the models were analyzed. The average of the maximum principalstrains were calculated and compared for all ligaments of the pelvis in theintact male and female models. ResultsModel ValidationsThe predicted data for all physiological loadings fell within onestandard deviation of the experimental data, except for right lateral bendingand right axial rotation for the male data, Figure 3.
Range of MotionThe comparison of range of motion at SIJis shown in figures 4 and 5. ROM of SIJ in the female model was the greatest inextension (1.36° left SIJ, 1.33° right SIJ) followed by flexion (0.50° leftSIJ, 0.50° right SIJ), right rotation (0.44° left SIJ, 0.44° right SIJ), rightbending (0.
30° left SIJ, 0.35° right SIJ), left rotation (0.29° left SIJ, 0.33°right SIJ), and left bending (0.24° left SIJ, 0.
30° right SIJ). In the malemodel, the maximum ROM of SIJ occurred in left rotation (0.54° left SIJ, 0.58°right SIJ) followed by right rotation (0.45 left SIJ, 0.48 right SIJ),extension (0.37 left SIJ, 0.
36 right SIJ), flexion (0.28° left SIJ, 0.27° rightSIJ), left bending (0.11° left SIJ, 0.12° right SIJ), and right bending (0.12°left SIJ, 0.
10° right SIJ). It was foundthat in flexion-extension (F-E) movements, SIJ had the highest motion in femalemodel (1.86°), however, the male model had the greatest motion in axialrotation (1.07°).
The lowest motion occurred in lateral bending in both femaleand male models (0.55° vs. 0.24°). According to the predicted motion data thefemale model experienced 86% higher mobility in flexion, 264% in extension,143% in left bending, and 228% in right bending compared to the same motions inthe male model. In left and right rotation, the ROM of the male model was 78%and 9% greater than the female model, respectively. Stressesacross SI Joint Themaximum stress in the female model occurred during the left rotation, followedby flexion, right rotation, right bending, left bending, and extension, figure 6.In the male model, the greatest stress happened during the left rotation,followed by left bending, extension, flexion, right bending, and rightrotation.
The maximum stresses at the female SIJwere higher by 27% in flexion, 28% in right bending, 49% in left bending, 45%in right rotation, and 20% in left rotation compared to those of the male model,figure 6.Sacrum had higher stresses compared to theilium in both models, figure 6. Stresses at the female sacrum and ilium were upto 49%, and 29% greater than the male model.PelvisLigaments StrainsFigure 7 and 8 show the results of the SIJligament strains for female and male models, respectively.
The anteriorsacroiliac ligament (ASL) strains were the same during all motions. The longposterior sacroiliac ligament (LPSL) experienced greatest tension duringextension motion and had no strain under the other motions. The short posteriorsacroiliac ligament (SPSL) was strained maximum during extension, but hadcomparable values under the other loads. The interosseous ligament (ISL)underwent the largest tensile strains during all motions.
The sacrospinousligament (SSL) and sacrotuberous ligament (STL) ligaments were both mostlystrained during flexion, experienced similar values under other loads, and hadno strain during extension. In the female model, ASL, LPSL, SPSL andSTL underwent larger strains compared to the corresponding male modelligaments, while SSL had similar values for both genders, and ISL showedgreater strains in the male model.Load Sharing across SI JointComputed load values, tables 2 and 3represent the greater average force magnitudes on the ilium and sacrum of bothsides. Female SIJs experienced higher loads across thejoints compared to the male SIJs under the similar loading conditions.
In bothmodels, the shear loads were higher than normal forces acting on the SIJsurfaces. DiscussionAlthoughthere are many studies which have quantified the range of motion, theliterature on the biomechanical differences between male and female SIJ is rarewith only one study comparing the range of motion differences of SIJ between genders20. However, they did not provide load sharing and stress data across the SIjoint. The current study showed that SIJ had higher mobility in femalescompared to males which is in agreement with the literature. In both male andfemale models, the motion was minimum in lateral bending. The greatest differenceof the SIJ motions between male and female models occurred during extension in whichthe female model showed significantly higher motion than the male model. Theincreased mobility in the female SIJ can be attributed to a lesser pronouncedcurvature of the SIJ surfaces, a larger gap (2 mm) at the SIJ, and a greaterpubic angle (111°) compared to the male model which had 1 mm gap at SIJ andpubic angle of 76° 12, 21.
Anatomical studyby Ebraheim et al. 22 revealed that the SIJ surface area is relativelygreater in adult males than females. The smaller joint surface area in femaleSIJ can result in higher local stresses across the joint. The current studyshowed that the maximum stress values in the female model were 49% higher thanmale model. The results of current studyalso showed that greater motion at the SIJ results in higher loads leading tohigher stresses across the joint under different motions, especially, on sacrumwhich experienced higher stresses compared to the ilium in both genders. Thishigher stress in females can result in higher risk of SIJ pain and higher riskof sacral stress fracture. Obtaining more quantitativeinformation may be essential to recognize SIJ dysfunction in both genders dueto the lack of quantitative relationship between the physiological spine motionand the biomechanical factors such as ligament strain that may be associatedwith pain in the SIJ.
In the current study, the authors illustrated thatdepending on the spine motion, the ligament strains varied. The anteriorsacroiliac ligament (ASL), short posterior sacroiliac ligament (SPSL), sacrotuberousligament (STL), sacrospinous ligament (SSL), and interosseous sacroiliacligament (ISL) underwent tension to constrain the SIJ during flexion. Longposterior sacroiliac ligament (LPSL) is one of the posterior ligaments whichonly functions during extension. SSL and STL seem to have no function in extension;however, SSL serves as a main constraint in other motions. Janssen et al. have shown that by sectioningsacrospinous and sacrotuberous ligaments, SIJ stability decreased 23. Theposterior sacroiliac ligaments contributed most to the SIJ mobility, while theanterior sacroiliac ligaments had little influence 41. Resisting the nutationand counternutation of the joint were done by ISL, STL, SSL, and LPSL 43,44.
The major role in stabilizing theSIJ was due to ISL, one of the strongest ligaments in the body. Interestingly,ASL, LPSL, SPSL and STL underwent higher strains in the female model and ISLstretched more in the male model. These high strains on certain ligaments in bothmodels can be explained by these anatomical differences which females havesmaller ASL, LPSL, and SPSL and males have smaller ISL compared to each other7. Although in our models, ligaments had same properties in both models, butdepending on the gender, the strain exerted on the ligaments were different.
Toour knowledge, this is the first study which investigated the differencebetween SIJ ROM of female and male as well as stresses, load sharing, andligaments strain across the joint in different motions. The presented data canbe used to address various critical questions regarding the anatomicaldifferences between male and female SIJ. For example, compared to men, who havea more ventral center of gravity, in females the center of gravity commonlypasses in front of or through the SIJ 24, 25.
This difference implies thatmen would have a greater lever arm than women, accounting for the stronger SIjoints in males 13. This characteristic may explain why males have morerestricted mobility. One of thelimitations of this study is the use of similar bone and ligaments materialproperties for the female and male models due to the lack of experimental data.Inconclusion, this study found that the female SIJ had relatively higher range ofmotion than male model at both sides of the SI joint. Also, the female SIJexperienced higher stresses across the joint especially on the sacrum comparedto males which implies that the females are at a higher risk of stress fractureinjury. The major role in stabilizing the SIJ was performed by ISL which is oneof the strongest ligaments in the body. The shear loads are higher across the femaleSIJ, compared to male SIJ. Thesedifferences may contribute to higher incidence of LBP in females, includingduring pregnancy.
We are presenting thedetails of model formulations, both for a male and a female sample which cannow be expanded and used to study gender differences in other postures andother conditions like pregnancy.