- Email: [email protected]
Flexural–torsional buckling initiates idiopathic scoliosis
Medical Hypotheses 77 (2011) 924–926
Contents lists available at SciVerse ScienceDirect
Medical Hypotheses journal homepage: www.elsevier.com/locate/mehy
Flexural–torsional buckling initiates idiopathic scoliosis Ioannis A. Karnezis ⇑ Back Care Network, Erythrou Stavrou 5, Maroussi 151 23, Athens, Greece
a r t i c l e
i n f o
Article history: Received 27 May 2011 Accepted 7 August 2011
a b s t r a c t Initiation of the spinal deformity in idiopathic scoliosis (IS) has been attributed to an abnormal pattern of spinal growth during development. However, recent findings suggest that the earliest observable event in the pathogenesis of IS is a change in the shape of intervertebral discs with alterations in the shape of vertebrae being considered a secondary event. Starting from the previous description of the spinal deformity in IS as ‘buckling’ of the spine a new hypothesis describing the initial spinal deformity in IS as flexural–torsional buckling, a three-dimensional type of failure of axially loaded columns, is proposed. According to the new hypothesis the initiating event (the earliest observable event) in IS is flexural–torsional buckling developing from the flexible parts (intervertebral discs and ligaments) of the affected spinal motion segments. Since flexural–torsional buckling occurs in columns with a cross-section of one axis of symmetry characterised by a much greater in-plane than out-of-plane bending stiffness the new hypothesis predicts that the initiating condition (the condition promoting the initiation) of IS is ‘flexibility anisotropy’ namely significantly higher bending stiffness in lateral bending than bending stiffness in flexion–extension of a part of the spine. The parameter of ‘flexibility anisotropy’ as a factor for initiation of IS has never been suggested or tested before. The present hypothesis has implications in the research on the pathogenesis of IS as well as in the development of new methods for its treatment. Ó 2011 Elsevier Ltd. All rights reserved.
Introduction Idiopathic scoliosis (IS) is a common three-dimensional deformity of the spine characterised by spinal deformation in all three planes, namely the frontal (lateral curvature), sagittal (lordosis) and transverse (vertebral rotation) plane (Fig. 1). As it can lead to significant morbidity it represents a major health problem. Research into the aetiology of IS has resulted in identification of a number of causative factors including genetic, biomechanical, biochemical and neuro-hormonal [1]. However, the exact aetiology of IS remains largely unclear. Extensive focus has also been placed on identifying the mechanism of progression of the spinal deformity in IS. Research into this field lends support to the ‘vicious cycle’ hypothesis. According to this an initial lateral spinal curvature produces asymmetrical loading of the skeletally immature spine, which in turn causes asymmetrical growth and progressive spinal deformity. This hypothesis is supported by significant analytical evidence [2]. Since IS develops in a spine of apparently normal original shape the pathogenesis of IS must include some form of ‘initiation phase’ (Fig. 2) which will trigger subsequent progression of the scoliotic deformity according to the ‘vicious cycle’ hypothesis described ⇑ Tel.: +30 6942275046; fax: +30 210 8087746. E-mail address: [email protected] 0306-9877/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.mehy.2011.08.013
above. Hypotheses related to the initiation phase of IS can contribute to better understanding of its pathogenesis. However, any such hypothesis must account for both (a) an initiating condition, namely a requirement or a set of requirements essential for initiation of the deformity and (b) an initiating event, namely the earliest observable event to lead to the development of the initial spinal deformity that triggers scoliosis progression. The hypothesis Background In terms of its morphology the spinal deformity in IS has been described as ‘buckling’ of the spine. Although from the engineering point of view the term is not actually correct since buckling assumes a stiff, rather homogenous structure this term reflects the fact that the spinal deformity in IS actually resembles buckling of an axially loaded column in engineering. Lucas [3] was the first to propose that IS is initiated by a buckling phenomenon as a result of rapid increase in spinal growth. Since then IS initiation has also been described as buckling caused by loss of thoracic kyphosis [4] or anterior spinal overgrowth [5]. As early as in 1966, Roaf [6] stated that the initiating mechanism of IS is ‘buckling’ resulting from unequal growth of the vertebral bodies while years later Stokes et al. [2] demonstrated the
I.A. Karnezis / Medical Hypotheses 77 (2011) 924–926
925
Fig. 3. Theoretical strategy for reduction of the spinal deformity in idiopathic scoliosis when the latter is viewed as flexural–torsional buckling: since (A) there is a higher degree of curvature on the concave (a) compared to the convex (b) side of the deformity gradual decrease of the concave-side curvature will result in gradual return of the spine towards the coronal plane. Thus the (three-dimensional) flexural–torsional buckling configuration is first converted into (single-plane) simple flexural buckling (B) and finally into a ‘straight’ spine (C). Fig. 1. Antero-posterior radiographic view of idiopathic scoliosis of the thoracolumbar spine (T9–L3 levels).
Fig. 2. Diagram of pathogenesis of idiopathic scoliosis.
effect of deformation of the vertebral bodies (as a result of bone remodelling according to the Hueter–Volkmann law) in scoliosis progression. Using a finite-element computer model Goto et al. [7] showed that the spinal deformity in idiopathic scoliosis is reproduced when an initial buckling of the spine is combined with bone modelling although the latter study did not focus on the underlying mechanisms responsible for buckling of the spine. It is now widely accepted that the initiating mechanism for IS is asymmetry of the vertebral bodies as a result of abnormal spinal growth which then leads to asymmetric loading conditions able to trigger scoliosis progression by an ‘abnormal loading – abnormal growth’ vicious cycle. Given that the spine is, from a mechanical viewpoint, a composite column comprising both rigid (vertebral bodies) and flexible (intervertebral discs) components the question arises whether the initiating event of IS occurs in the rigid components (growthderived asymmetry of the vertebral bodies) or in the flexible vcomponents (asymmetric deformation of the intervertebral discs) of the spine. It was only recently that radiographic data have shown that the early scoliotic deformity is characterised by asymmetry of the intervertebral discs (‘disc wedging’) while asymmetry of the vertebral bodies (‘vertebral wedging’) occurs later and may well be a secondary event [8]. In other words the earliest event in pathogenesis of IS appears to be a change in the intervertebral disc shape rather than a change in the shape the vertebrae. Therefore, contrary to previous beliefs it can now be assumed that the initiating ‘buckling’ phenomenon in IS pathogenesis occurs as a result of flexible (intervertebral disc) part deformation.
Observations in buckling of axially loaded columns The simplest form of buckling is flexural (Euler) buckling which is a single-plane event. Flexural buckling occurs when the moment produced by the axial load and the column’s deflection due to bending or mechanical imperfections (Fig. 3). This moment is resisted by the flexural stiffness of the column. However, it is clear that IS is not a single-plane spinal deformation since it is known to be a three-dimensional deformity as described above. Therefore, flexural buckling cannot describe the spinal deformity in IS. Another form of buckling of axially loaded columns is a phenomenon known as flexural–torsional buckling. This occurs when the in-plane configuration of the loaded column becomes unstable and the structure is trying to reach a stable out-of-plane configuration. The result is that the column deflects laterally and twists (rotates) out of the plane of loading. While flexural buckling is a single-plane event flexural–torsional buckling is a three-dimensional event since part of the column deflects and twists out of the plane of loading resulting in significant out-of-plane deformity (bending and twisting). Flexural–torsional buckling can significantly decrease the load bearing capacity of an axially loaded column. Simple observation shows striking similarities between flexural–torsional buckling of an axially loaded column and the spinal deformity in IS (Fig. 1). It is worth mentioning here that, as known from engineering, flexural–torsional buckling is a type of failure that occurs in axially loaded columns that [9] a. are uni-symmetric (their cross-section has a single axis of symmetry) and b. have a much greater in-plane bending stiffness than out-ofplane bending stiffness. This, when translated to the human spine (which can be considered a uni-symmetric column along the sagittal plane), points to the possibility that a deformity similar to flexural–torsional buckling may occur when the bending stiffness (S) of a part of the spine becomes significantly higher in the coronal plane (lateral bending
926
I.A. Karnezis / Medical Hypotheses 77 (2011) 924–926
stiffness, SLat) than in the sagittal plane (flexion–extension bending stiffness, SExt). It should be noted here that alterations in biomechanical properties such as the bending stiffness of the spinal motion segment may not be entirely attributable to the intervertebral disc but also to spinal ligaments, the facet joints or the spinal musculature. Initiation hypothesis for IS Based on the above observations a hypothesis on initiation of IS can now be proposed as follows: a. The initiating event (i.e. the earliest observable event) of the spinal deformity in IS is a flexural–torsional buckling event developing from the flexible parts (intervertebral discs and ligaments) of the spinal motion segments. b. The initiating condition (i.e. the condition promoting the initiation) of IS is ‘flexibility anisotropy’ of a part of the spine characterised by significantly higher bending stiffness in coronal plane movements (lateral bending stiffness, SLat) than bending stiffness in the sagittal plane (flexion–extension bending stiffness, SExt) or, in other words, a significant increase in the SLat/SExt ratio of the affected part of the spine. Evaluation and consequences of the hypothesis As discussed above the introduced hypothesis is supported by the morphological similarities between flexural–torsional buckling of an axially loaded column and the spinal deformity in IS (Fig. 1) as well as recent radiographic evidence that the early deformity during the pathogenesis of IS occurs as result of deformity of the intervertebral discs [8]. The hypothesis can be tested analytically by studies of the initiation of IS using finite-element numerical models of the spine. It should be noted that the parameter of ‘flexibility anisotropy’ as a factor for initiation of IS has not been systematically tested before in analytical studies of scoliosis initiation where the intervertebral disc is invariably modelled as ‘isotropic’ in terms of its bending stiffness characteristics. Plaats et al. [10] developed a finite-element model to investigate buckling along with other possible mechanisms of initiation of idiopathic scoliosis. Interestingly, the latter study appears to support the hypothesis presented here since, according to its findings, unilaterally increased spinal stiffness (supposedly caused by unilateral postponement in ligament growth) can initiate scoliosis, while lateral eccentricity of spinal loading failed by itself to reproduce an idiopathic scoliosis curve in the entire spine. Moreover, the parameter of spinal ‘flexibility anisotropy’ (difference between lateral bending stiffness and flexion–extension bending stiffness) may become important in the light of recent evidence from analytical studies suggesting that bending stiffness of the spine undergoes significant variations during adolescent growth [11]. The latter, combined with the fact that bending stiffness in different directions is obviously affected by different parameters, suggests that the ratio of lateral to flexion–extension bending stiffness of the spine may also undergo significant variations during adolescent growth. Further support for the hypothesis could be sought in clinical assessment of the bending stiffness of the thoraco-lumbar spine during spinal development. For example, clinical studies
can assess the bending stiffness of the thoraco-lumbar spine in each direction of motion in pre-adolescent children, possibly in an indirect manner using measurements of range of motion of the entire thoraco-lumbar spine. Subsequent follow-up of the studied cohort will reveal the children which subsequently developed IS. It would then be possible to use the data obtained previously to statistically compare the ratio of lateral to extension bending stiffness (SLat/SExt) of those children against the same parameter of the children that did not developed IS. The present hypothesis may also have implications in the development of new approaches for correction of IS. This might be of particular importance for rehabilitation strategies or motion-preservation (avoiding spinal fusion) surgical methods of correction of IS. One implication of the hypothesis is that progression of the scoliotic deformity may be prevented by increasing flexion–extension bending stiffness of the spine. Furthermore, when viewed as flexural–torsional buckling of an axially loaded column (a three-dimensional deformity) IS does not appear an easy deformity to correct by manipulations (exercises or surgical) on a single plane. For example bending to the opposite of the deformity side or distraction along the long axis of the spine cannot significantly reduce the torsional component of the deformity impending thus any attempts for full correction. According to the present hypothesis a theoretically improved strategy for reduction of the spinal deformity in IS is shown in Fig. 3. Since curvature is higher on the concave side compared to the convex side of the deformity then a, by any means, gradual decrease of the concave-side curvature will result in gradual return of the spine towards the coronal plane. In this manner a three-dimensional configuration (flexural–torsional buckling) may first be converted into a simpler single-plane configuration (flexural buckling) and finally into a ‘straight’ shape. Conflict of interest statement There are no conflicts of interest related to the above paper. References [1] Kouwenhoven J-WM, Castelein RM. The pathogenesis of adolescent idiopathic scoliosis. Spine 2008;33(26):2898–908. [2] Stokes IAF, Burwell RG, Dangerfield PH. Biomechanical spinal growth modulation and progressive adolescent scoliosis – a test of the ‘vicious cycle’ pathogenetic hypothesis. Scoliosis 2006;1:16. [3] Lucas DB. Mechanics of the spine. Bull Hosp Joint Dis 1970;31:115–31. [4] Dickson RA. The etiology and pathogenesis of idiopathic scoliosis. Acta Orthop Belg 1992;58:21–5. [5] Azegami H, Murachi S, Kitoh J, et al. Etiology of idiopathic scoliosis. Clin Orthop Relat Res 1998;357:229–36. [6] Roaf R. The basic anatomy of scoliosis. J Bone Joint Surg Br 1966;48B(4): 786–92. [7] Goto M, Kawakami N, Azegami H, et al. Buckling and bone modeling as factors in the development of idiopathic scoliosis. Spine 2003;28(4):364–70. [8] Will RE, Stokes IA, Qiu X, Walker MR, Sanders JO. Cobb angle progression in adolescent scoliosis begins at the intervertebral disc. Spine 2009;34(25): 2782–6. [9] Trahair NS. Flexural–torsional buckling of structures. Boca Raton, FL, USA: CRC Press Inc.; 1993. [10] Van Der Plaats A, Veldhuizen AG, Verkerke GJ. Numerical simulation of asymmetrically altered growth as initiation mechanism of scoliosis. Ann Biomed Eng 2007;35(7):1206–15. [11] Meijer GJM, Homminga J, Hekman EEG, Veldhuizen AG, Verkerke GJ. The effect of three-dimensional geometrical changes during adolescent growth on the biomechanics of a spinal motion segment. J Biomech 2010;43:1590–7.
Contents lists available at SciVerse ScienceDirect
Medical Hypotheses journal homepage: www.elsevier.com/locate/mehy
Flexural–torsional buckling initiates idiopathic scoliosis Ioannis A. Karnezis ⇑ Back Care Network, Erythrou Stavrou 5, Maroussi 151 23, Athens, Greece
a r t i c l e
i n f o
Article history: Received 27 May 2011 Accepted 7 August 2011
a b s t r a c t Initiation of the spinal deformity in idiopathic scoliosis (IS) has been attributed to an abnormal pattern of spinal growth during development. However, recent findings suggest that the earliest observable event in the pathogenesis of IS is a change in the shape of intervertebral discs with alterations in the shape of vertebrae being considered a secondary event. Starting from the previous description of the spinal deformity in IS as ‘buckling’ of the spine a new hypothesis describing the initial spinal deformity in IS as flexural–torsional buckling, a three-dimensional type of failure of axially loaded columns, is proposed. According to the new hypothesis the initiating event (the earliest observable event) in IS is flexural–torsional buckling developing from the flexible parts (intervertebral discs and ligaments) of the affected spinal motion segments. Since flexural–torsional buckling occurs in columns with a cross-section of one axis of symmetry characterised by a much greater in-plane than out-of-plane bending stiffness the new hypothesis predicts that the initiating condition (the condition promoting the initiation) of IS is ‘flexibility anisotropy’ namely significantly higher bending stiffness in lateral bending than bending stiffness in flexion–extension of a part of the spine. The parameter of ‘flexibility anisotropy’ as a factor for initiation of IS has never been suggested or tested before. The present hypothesis has implications in the research on the pathogenesis of IS as well as in the development of new methods for its treatment. Ó 2011 Elsevier Ltd. All rights reserved.
Introduction Idiopathic scoliosis (IS) is a common three-dimensional deformity of the spine characterised by spinal deformation in all three planes, namely the frontal (lateral curvature), sagittal (lordosis) and transverse (vertebral rotation) plane (Fig. 1). As it can lead to significant morbidity it represents a major health problem. Research into the aetiology of IS has resulted in identification of a number of causative factors including genetic, biomechanical, biochemical and neuro-hormonal [1]. However, the exact aetiology of IS remains largely unclear. Extensive focus has also been placed on identifying the mechanism of progression of the spinal deformity in IS. Research into this field lends support to the ‘vicious cycle’ hypothesis. According to this an initial lateral spinal curvature produces asymmetrical loading of the skeletally immature spine, which in turn causes asymmetrical growth and progressive spinal deformity. This hypothesis is supported by significant analytical evidence [2]. Since IS develops in a spine of apparently normal original shape the pathogenesis of IS must include some form of ‘initiation phase’ (Fig. 2) which will trigger subsequent progression of the scoliotic deformity according to the ‘vicious cycle’ hypothesis described ⇑ Tel.: +30 6942275046; fax: +30 210 8087746. E-mail address: [email protected] 0306-9877/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.mehy.2011.08.013
above. Hypotheses related to the initiation phase of IS can contribute to better understanding of its pathogenesis. However, any such hypothesis must account for both (a) an initiating condition, namely a requirement or a set of requirements essential for initiation of the deformity and (b) an initiating event, namely the earliest observable event to lead to the development of the initial spinal deformity that triggers scoliosis progression. The hypothesis Background In terms of its morphology the spinal deformity in IS has been described as ‘buckling’ of the spine. Although from the engineering point of view the term is not actually correct since buckling assumes a stiff, rather homogenous structure this term reflects the fact that the spinal deformity in IS actually resembles buckling of an axially loaded column in engineering. Lucas [3] was the first to propose that IS is initiated by a buckling phenomenon as a result of rapid increase in spinal growth. Since then IS initiation has also been described as buckling caused by loss of thoracic kyphosis [4] or anterior spinal overgrowth [5]. As early as in 1966, Roaf [6] stated that the initiating mechanism of IS is ‘buckling’ resulting from unequal growth of the vertebral bodies while years later Stokes et al. [2] demonstrated the
I.A. Karnezis / Medical Hypotheses 77 (2011) 924–926
925
Fig. 3. Theoretical strategy for reduction of the spinal deformity in idiopathic scoliosis when the latter is viewed as flexural–torsional buckling: since (A) there is a higher degree of curvature on the concave (a) compared to the convex (b) side of the deformity gradual decrease of the concave-side curvature will result in gradual return of the spine towards the coronal plane. Thus the (three-dimensional) flexural–torsional buckling configuration is first converted into (single-plane) simple flexural buckling (B) and finally into a ‘straight’ spine (C). Fig. 1. Antero-posterior radiographic view of idiopathic scoliosis of the thoracolumbar spine (T9–L3 levels).
Fig. 2. Diagram of pathogenesis of idiopathic scoliosis.
effect of deformation of the vertebral bodies (as a result of bone remodelling according to the Hueter–Volkmann law) in scoliosis progression. Using a finite-element computer model Goto et al. [7] showed that the spinal deformity in idiopathic scoliosis is reproduced when an initial buckling of the spine is combined with bone modelling although the latter study did not focus on the underlying mechanisms responsible for buckling of the spine. It is now widely accepted that the initiating mechanism for IS is asymmetry of the vertebral bodies as a result of abnormal spinal growth which then leads to asymmetric loading conditions able to trigger scoliosis progression by an ‘abnormal loading – abnormal growth’ vicious cycle. Given that the spine is, from a mechanical viewpoint, a composite column comprising both rigid (vertebral bodies) and flexible (intervertebral discs) components the question arises whether the initiating event of IS occurs in the rigid components (growthderived asymmetry of the vertebral bodies) or in the flexible vcomponents (asymmetric deformation of the intervertebral discs) of the spine. It was only recently that radiographic data have shown that the early scoliotic deformity is characterised by asymmetry of the intervertebral discs (‘disc wedging’) while asymmetry of the vertebral bodies (‘vertebral wedging’) occurs later and may well be a secondary event [8]. In other words the earliest event in pathogenesis of IS appears to be a change in the intervertebral disc shape rather than a change in the shape the vertebrae. Therefore, contrary to previous beliefs it can now be assumed that the initiating ‘buckling’ phenomenon in IS pathogenesis occurs as a result of flexible (intervertebral disc) part deformation.
Observations in buckling of axially loaded columns The simplest form of buckling is flexural (Euler) buckling which is a single-plane event. Flexural buckling occurs when the moment produced by the axial load and the column’s deflection due to bending or mechanical imperfections (Fig. 3). This moment is resisted by the flexural stiffness of the column. However, it is clear that IS is not a single-plane spinal deformation since it is known to be a three-dimensional deformity as described above. Therefore, flexural buckling cannot describe the spinal deformity in IS. Another form of buckling of axially loaded columns is a phenomenon known as flexural–torsional buckling. This occurs when the in-plane configuration of the loaded column becomes unstable and the structure is trying to reach a stable out-of-plane configuration. The result is that the column deflects laterally and twists (rotates) out of the plane of loading. While flexural buckling is a single-plane event flexural–torsional buckling is a three-dimensional event since part of the column deflects and twists out of the plane of loading resulting in significant out-of-plane deformity (bending and twisting). Flexural–torsional buckling can significantly decrease the load bearing capacity of an axially loaded column. Simple observation shows striking similarities between flexural–torsional buckling of an axially loaded column and the spinal deformity in IS (Fig. 1). It is worth mentioning here that, as known from engineering, flexural–torsional buckling is a type of failure that occurs in axially loaded columns that [9] a. are uni-symmetric (their cross-section has a single axis of symmetry) and b. have a much greater in-plane bending stiffness than out-ofplane bending stiffness. This, when translated to the human spine (which can be considered a uni-symmetric column along the sagittal plane), points to the possibility that a deformity similar to flexural–torsional buckling may occur when the bending stiffness (S) of a part of the spine becomes significantly higher in the coronal plane (lateral bending
926
I.A. Karnezis / Medical Hypotheses 77 (2011) 924–926
stiffness, SLat) than in the sagittal plane (flexion–extension bending stiffness, SExt). It should be noted here that alterations in biomechanical properties such as the bending stiffness of the spinal motion segment may not be entirely attributable to the intervertebral disc but also to spinal ligaments, the facet joints or the spinal musculature. Initiation hypothesis for IS Based on the above observations a hypothesis on initiation of IS can now be proposed as follows: a. The initiating event (i.e. the earliest observable event) of the spinal deformity in IS is a flexural–torsional buckling event developing from the flexible parts (intervertebral discs and ligaments) of the spinal motion segments. b. The initiating condition (i.e. the condition promoting the initiation) of IS is ‘flexibility anisotropy’ of a part of the spine characterised by significantly higher bending stiffness in coronal plane movements (lateral bending stiffness, SLat) than bending stiffness in the sagittal plane (flexion–extension bending stiffness, SExt) or, in other words, a significant increase in the SLat/SExt ratio of the affected part of the spine. Evaluation and consequences of the hypothesis As discussed above the introduced hypothesis is supported by the morphological similarities between flexural–torsional buckling of an axially loaded column and the spinal deformity in IS (Fig. 1) as well as recent radiographic evidence that the early deformity during the pathogenesis of IS occurs as result of deformity of the intervertebral discs [8]. The hypothesis can be tested analytically by studies of the initiation of IS using finite-element numerical models of the spine. It should be noted that the parameter of ‘flexibility anisotropy’ as a factor for initiation of IS has not been systematically tested before in analytical studies of scoliosis initiation where the intervertebral disc is invariably modelled as ‘isotropic’ in terms of its bending stiffness characteristics. Plaats et al. [10] developed a finite-element model to investigate buckling along with other possible mechanisms of initiation of idiopathic scoliosis. Interestingly, the latter study appears to support the hypothesis presented here since, according to its findings, unilaterally increased spinal stiffness (supposedly caused by unilateral postponement in ligament growth) can initiate scoliosis, while lateral eccentricity of spinal loading failed by itself to reproduce an idiopathic scoliosis curve in the entire spine. Moreover, the parameter of spinal ‘flexibility anisotropy’ (difference between lateral bending stiffness and flexion–extension bending stiffness) may become important in the light of recent evidence from analytical studies suggesting that bending stiffness of the spine undergoes significant variations during adolescent growth [11]. The latter, combined with the fact that bending stiffness in different directions is obviously affected by different parameters, suggests that the ratio of lateral to flexion–extension bending stiffness of the spine may also undergo significant variations during adolescent growth. Further support for the hypothesis could be sought in clinical assessment of the bending stiffness of the thoraco-lumbar spine during spinal development. For example, clinical studies
can assess the bending stiffness of the thoraco-lumbar spine in each direction of motion in pre-adolescent children, possibly in an indirect manner using measurements of range of motion of the entire thoraco-lumbar spine. Subsequent follow-up of the studied cohort will reveal the children which subsequently developed IS. It would then be possible to use the data obtained previously to statistically compare the ratio of lateral to extension bending stiffness (SLat/SExt) of those children against the same parameter of the children that did not developed IS. The present hypothesis may also have implications in the development of new approaches for correction of IS. This might be of particular importance for rehabilitation strategies or motion-preservation (avoiding spinal fusion) surgical methods of correction of IS. One implication of the hypothesis is that progression of the scoliotic deformity may be prevented by increasing flexion–extension bending stiffness of the spine. Furthermore, when viewed as flexural–torsional buckling of an axially loaded column (a three-dimensional deformity) IS does not appear an easy deformity to correct by manipulations (exercises or surgical) on a single plane. For example bending to the opposite of the deformity side or distraction along the long axis of the spine cannot significantly reduce the torsional component of the deformity impending thus any attempts for full correction. According to the present hypothesis a theoretically improved strategy for reduction of the spinal deformity in IS is shown in Fig. 3. Since curvature is higher on the concave side compared to the convex side of the deformity then a, by any means, gradual decrease of the concave-side curvature will result in gradual return of the spine towards the coronal plane. In this manner a three-dimensional configuration (flexural–torsional buckling) may first be converted into a simpler single-plane configuration (flexural buckling) and finally into a ‘straight’ shape. Conflict of interest statement There are no conflicts of interest related to the above paper. References [1] Kouwenhoven J-WM, Castelein RM. The pathogenesis of adolescent idiopathic scoliosis. Spine 2008;33(26):2898–908. [2] Stokes IAF, Burwell RG, Dangerfield PH. Biomechanical spinal growth modulation and progressive adolescent scoliosis – a test of the ‘vicious cycle’ pathogenetic hypothesis. Scoliosis 2006;1:16. [3] Lucas DB. Mechanics of the spine. Bull Hosp Joint Dis 1970;31:115–31. [4] Dickson RA. The etiology and pathogenesis of idiopathic scoliosis. Acta Orthop Belg 1992;58:21–5. [5] Azegami H, Murachi S, Kitoh J, et al. Etiology of idiopathic scoliosis. Clin Orthop Relat Res 1998;357:229–36. [6] Roaf R. The basic anatomy of scoliosis. J Bone Joint Surg Br 1966;48B(4): 786–92. [7] Goto M, Kawakami N, Azegami H, et al. Buckling and bone modeling as factors in the development of idiopathic scoliosis. Spine 2003;28(4):364–70. [8] Will RE, Stokes IA, Qiu X, Walker MR, Sanders JO. Cobb angle progression in adolescent scoliosis begins at the intervertebral disc. Spine 2009;34(25): 2782–6. [9] Trahair NS. Flexural–torsional buckling of structures. Boca Raton, FL, USA: CRC Press Inc.; 1993. [10] Van Der Plaats A, Veldhuizen AG, Verkerke GJ. Numerical simulation of asymmetrically altered growth as initiation mechanism of scoliosis. Ann Biomed Eng 2007;35(7):1206–15. [11] Meijer GJM, Homminga J, Hekman EEG, Veldhuizen AG, Verkerke GJ. The effect of three-dimensional geometrical changes during adolescent growth on the biomechanics of a spinal motion segment. J Biomech 2010;43:1590–7.