Background: Considering the thoracic, lumbar spine or whole spine as rigid segments has been the norm until recent studies highlighted the importance of more detailed modelling. A better understanding of the requirement for spine multi-segmental analysis could guide planning of future studies and avoid missing clinically-relevant information.

Research question: This study aims to assess the correlation between adjacent spine segments movement thereby evaluating segmental redundancy in both healthy and participants with low back pain (LBP).

Methods: A 3D motion capture system tracked the movement of upper and lower thoracic and lumbar spine segments in twenty healthy and twenty participants with LBP. Tasks performed included walking, sit-to-stand and lifting, repeated 3 times. 3D angular kinematics were calculated for each spine segment. Segmental redundancy was evaluated through cross-correlation (Rxy) analysis of kinematics time series and correlation of range of motion (RROM) of adjacent spine segments.

Results: The upper/lower lumbar pairing showed weak correlations in the LBP group for all tasks and anatomical planes (Rxyrange:0.02–0.36) but moderate and strong correlations during walking (Rxy _frontalplane:0.4) and lifting (Rxy _sagittalplane:0.64) in the healthy group. The lower thoracic/upper lumbar pairing had weak correlations for both groups during lifting and sit-to-stand in the frontal plane and for walking (Rxy:0.01) in the sagittal plane only. The upper/lower thoracic pairing had moderate correlations during sit-to-stand in sagittal and transverse plane in patients with LBP (Rxy _sagittalplane:0.41; Rxy _transverse plane:−0.42) but weak in healthy (Rxy _sagittalplane:0.23; Rxy _transverseplane:−0.34); the contrary was observed during lifting.

The majority of RROM values (55/72) demonstrated weak correlations.

Significance:The results suggest that multi-segmental analysis of the spine is necessary if spine movement characteristics are to be fully understood. We cannot establish a priori where redundancy occurs based on healthy data, therefore extra consideration should be made when planning studies with pathological cohorts.