Stability of dynamic trunk movement

Abstract

Study design: Nonlinear systems analyses of trunk kinematics were performed to estimate control of dynamic stability during repetitive flexion and extension movements.

Objective: Determine whether movement pace and movement direction of dynamic trunk flexion and extension influence control of local dynamic stability.

Summary of background data: Spinal stability has been previously characterized in static, but not in dynamic movements. Biomechanical models make inferences about static spinal stability, but existing analyses provide limited insight into stability of dynamic movement. Stability during dynamic movements can be estimated from Lyapunov analyses of empirical data.

Methods: There were 20 healthy subjects who performed repetitive trunk flexion and extension movements at 20 and 40 cycles per minute. Maximum Lyapunov exponents describing the expansion of the kinematic state-space were calculated from the measured trunk kinematics to estimate stability of the dynamic system.

Results: The complexity of torso movement dynamics required at least 5 embedded dimensions, which suggests that stability components of lumbar lordosis may be empirically measurable in addition to global stability of trunk dynamics. Repeated trajectories from fast paced movements diverged more quickly than slower movement, indicating that local dynamic stability is limited in fast movements. Movements in the midsagittal plane showed higher multidimensional kinematic divergence than asymmetric movements.

Conclusion: Nonlinear dynamic systems analyses were successfully applied to empirically measured data, which were used to characterize the neuromuscular control of stability during repetitive dynamic trunk movements. Movement pace and movement direction influenced the control of spinal stability. These stability assessment techniques are recommended for improved workplace design and the clinical assessment of spinal stability in patients with low back pain.

Figure 1

Experimental setup of the targeted movement task.

Figure 2

Example of a reconstructed movement trajectory with n = 3 state-space dimensions. Although the movement data were analyzed with n = 5, 3 embedding dimension is the largest that can be illustrated.

Figure 3

Typical plot of the state-space expansion with time. The dashed line represents the best-fit line between t = 0:1 cycles (with a cycle length of approximately 1.5 seconds for this trial). The slope of this best-fit line was used to represent the state-space expansion, i.e. local dynamic stability of the task.

Figure 4

λ_Max values were greater during fast paced movement trials than slow paced cyclic movement. Values were also greater during asymmetric movement tasks than during sagittal midplane movements. Larger values of λ_Max represent more chaotic, i.e. less stable, movement dynamics.