A simple polynomial that predicts low-back compression during complex 3-D tasks

While most existing models that predict loads on the low back for occupational risk analysis are restricted to assessing moments in the sagittal plan, a few have the ability to determine spine compression from three-dimensional (3-D) loading. The objective of this work was to find a method to estimate low-back compression forces during 3-D loading tasks from a model that contains as much biological content validity as the authors could incorporate, such as the effects of muscle co-contraction, but was simple enough to be implemented into a model appropriate for industrial use. The problem that had to be solved was how to represent the anatomical reality that a given muscle force vector contributes simultaneously to all three moments, flexion or extension, lateral bend and axial twist. Simply summing the compression components of independent equivalent muscles in each plane unrealistically assumes that muscles are uncoupled (i.e. each one works independently to support only one moment). In this study, loads in the various tissues of the low back that resulted from the simultaneous generation of moments about the three orthopaedic axes during 3-D tasks were estimated by an anatomically detailed 90 muscle model. A four-dimensional regression equation was developed to predict low-back compression from the three moments generated about the three axes (R2 = 0.94). Comparison of compression estimates from an uncoupled model showed that accounting for muscle coupling reduces compression by 22% on average. The predictive equation is presented to simplify analysis of complex 3-D industrial tasks for those who would like to incorporate it into their own models that produce moments of force in three plans of the L4/L5 level of the lumbar spine.