In this study, three-dimensional nonlinear finite element models of age-specific one year old, three year old, and six year old pediatric human cervical spine (C4-C5-C6) structures were developed. Their biomechanical responses were compared with the adult human cervical spine behavior under different loadings and at all load levels. The adult human cervical spine model was constructed from close-up computed tomography sections in the axial and sagittal planes, and sequential anatomic cryomicrotome sections. The adult model was validated with experimental moment-rotation data under flexion-extension and compression by correlating bilateral strains in the vertebral body and the lateral masses, and the force-deflection responses with experiments conducted in our laboratory. The adult model was modified to create one, three and six year old pediatric spines by incorporating the local geometrical and material characteristics of the developmental anatomy.
The biomechanical force-displacement/moment-rotation responses of the pediatric structures under compression, flexion and extension forces were nonlinear. All pediatric structures were consistently more flexible than the adult spine under all loading modes and at all load levels. Further, the responses of the pediatric structures were compared to the well developed adult structure using three approaches: one, pure overall structural scaling (reduce size) of the adult model without incorporating the local component geometrical and material property changes to obtain the "representative" pediatric responses; two, using three separate age-specific pediatric models incorporating local component geometrical and material property changes; and three, using the above three pediatric models and combining the overall structural scaling effects.
The pediatric responses obtained using the pure overall structural scaling to the adult cervical spine model increased the flexibilities slightly (maximum 118%). In contrast, the inclusion of the local component hard and soft tissue geometry and their material property changes to create the three individual pediatric cervical spine models, produced significantly higher changes in the flexibilities under all loading modes and at all load levels (maximum 465% increase under compression). When overall structural scaling effects were added to the three pediatric models, the increase in the flexibilities was not considerably higher (maximum 534% under compression). The change in the flexibilities was uniform for all the three pediatric models under flexion, i.e., the flexibility was independent of moment level. In contrast, the change in flexibilities decreased with increasing levels of loading under compression and extension. While the one year old pediatric model was the most flexible followed by the three and six year old models in flexion (168%) and extension (404%), the three year old pediatric model was the most flexible under compression (534%) followed by the six and one year old models. The differing biomechanical responses among different pediatric groups were ascribed to the individual developmental anatomical features. The present findings of significant increase in biomechanical response due to local geometry and material property changes emphasize the need to consider the developmental anatomical features in the pediatric structures to better predict their biomechanical behavior.