Abstract Summary
Spine kinematics are often recorded in vivo by tracking motion of multi-segment regions (e.g. T12 to S1). Many models of the spine must distribute T12/S1 motion across segmental levels, commonly by assuming linear proportions. This raises concern when modelling non-sagittal plane movements as 3-D angles are only distributive if the assumptions for small-angle approximation are met. Here we demonstrate that modelled spine orientations, based on distributed motion, do not coincide with recorded data. Systematic errors were observed in all three Carden angles with maximum errors occurring in twist. This is concerning as intervertebral discs are sensitive to twist damage and injury.