A Comparison of Methods to Quantify Control of the Spine

Abstract

Low back pain (LBP) affects many individuals worldwide. The established association between LBP and spine motor control has led to the development of many control assessment techniques. To understand the association between motor control and LBP, it is essential to understand the relationship between separate assessment techniques. Systems identification (SI) and local dynamic stability (LDS) are two methods commonly used to quantify spine control. SI provides a detailed description of control but uses linear assumptions, whereas LDS provides a “black box” non-linear assessment and can be quantified during dynamic movements. Although both SI and LDS techniques aim to measure the control of the spine, each employs different experimental setups and data processing strategies. Therefore, the purpose of this thesis was to compare the motor behaviour outcomes of SI and LDS quantification techniques. To do this, 15 participants completed two tasks (SI and LDS) in a random order. For the SI task, participants were seated and ventrally perturbed at the level of the 10th thoracic vertebrae (T10). They completed this task under instructions to resist the perturbations (resist condition) or relax and remain upright (relax condition). Admittance was represented using frequency response functions, and a validated neuromuscular control model quantified lumbar stiffness, damping and muscle spindle feedback gains. The LDS task involved participants completing three repetitive movement blocks consisting of flexion/extension, axial rotation, and complex movements. In each block, the maximum finite-time Lyapunov exponent (λmax) was estimated. A stepwise linear regression determined that λmax during the rotation task was best predicted by SI outcomes in the relax condition (adjusted R square = 0.65). Many conditions demonstrated no significant relationship between λmax and SI outcomes. These findings outline the importance of a consistent framework for the assessment of spine control. This could improve clinical assessment efficiency as well as the understanding of the association between LBP and motor control.