The roles of muscles in arm swing and thoracic rotation during walking

Abstract

The functional role of arm swing during walking has been hypothesized to reduce net angular momentum and moment about the vertical axis, and to minimize metabolic costs of locomotion. Arm swing follows a specific timing pattern, with peak flexion occurring at the midpoint of stride. Recent studies have proposed that arm swing arises passively from driving forces imparted on the arm by thorax rotation and derived from energy input from the lower body. However, the passively driven hypothesis does not account for changes in arm swing timing that would occur with changes in stride frequency or inertia properties of the segments, or for measured muscle activity in the shoulder and trunk. Here I present a passively driven forward dynamic model of the upper body with rotational springs in the shoulder joints and in the trunk. To account for the constant timing of arm swing across different speeds and the presence of muscle activity in the upper body I propose an Actively Tuned Hypothesis: The dynamics of arm swing and thorax rotation are derived from the lower body and act like a passively driven spring-mass-pendular system, with muscles in the shoulder and trunk actively tuning the spring parameters to phase-lock arm swing in-phase with thorax rotation and anti-phase with ipsilateral heel strike. Eight male subjects participated in three walking experiments testing the ability of the model to reproduce in vivo kinematics and the actively tuned hypothesis. A Speed Experiment varied stride frequency, an Arm Inertia Experiment added inertia to both arms, and a Thorax Inertia Experiment added inertia to the thorax. The passively driven model was able to reproduce the experimentally measured kinematics of the subjects. Spring stiffness at the shoulder and trunk varied with experimental manipulations, and the result was maintenance of the natural frequency of the arms below the driving frequency of the stride. Tuning of the spring parameters cause phase-locking between thoracic rotation and arm swing in an in-phase manner, and thoracic rotation and ipsilateral heel strike in an anti-phase manner. Muscle activity in the posterior deltoid and erector spinae group showed timing similar to a spring. The tuning mechanism in the upper body was proposed to be active muscular contraction. Through active tuning of the system the timing of arm swing is controlled in order to produce the hypothesized functions during gait.