Using predictive simulation to explore mechanistic contributors to idiopathic toe walking

Abstract

1 Introduction: The etiology of equinus gait can be difficult to ascertain, and if no clear cause is found it is diagnosed as “idiopathic toe walking” (ITW). Predictive musculoskeletal computer simulations of gait may help unravel ITW's cause-and-effect mechanisms. It has been suggested that ITW may be caused by non-neural contractures due to a short Achilles tendon [1,2], but there is also evidence that neural (i.e., hyperreflexia or increased supraspinal drive) or higher-level control factors may contribute [3–7] to the gait pattern. 2 Research question: Can ITW be explained by tendon contracture of the plantar flexors in isolation, or are neural factors (i.e., hyperreflexia or an increased supraspinal drive), and/or other higher level control factors (i.e., prevention of high normalised fibre lengths) also involved? 3 Methods: Gait data from seventeen children diagnosed with ITW was used as comparative experimental data (i.e., ground reaction forces, joint angles, moments, and powers, and muscle excitations). A previously developed forward-dynamic modeling framework that showed good agreement with healthy gait was employed [8]. This consisted of a generic musculoskeletal model controlled by reflexes and supraspinal drive, governed by a weighted cost function that minimised cost of transport, foot-ground impact, head accelerations, muscle fatigue, and knee hyperextension. Three strategies were implemented to predict ITW gait. First, Achilles tendon contracture was implemented, matching the level of contracture from ITW children. Second, either hyperreflexia or increased supraspinal drive during the whole gait cycle was imposed on the contracted musculoskeletal model. Third, the contracted model was forced to function on the ascending limb and plateau region of the muscle force-length curve for the plantar flexors. In each step, the match of modeled gait with the experimental ITW data was quantified by root mean square errors (RMSE). 4 Results: A tendon contracture in isolation did not lead to toe walking gait and resulted in RMSE of 2.24 SD (Fig. 1). The best match with experimental data when imposing hyperreflexia or an increased supraspinal drive was RMSE of 1.84 SD. When applying the high fibre length penalty to the contracted model, the predicted gait approached ITW gait best, with RMSE of 1.37 SD (Fig. 1). This toe-walking gait employed a control strategy with higher velocity- and length-based reflex gains and supraspinal drive in stance and late swing compared to healthy gait. 5 Discussion: In line with previous research [9], implementing contracture alone resulted in the plantar flexors working on the descending limb of the force-length curve, which may not be a physiologically feasible solution [10,11]. Predictive simulations provide mechanistic insights into how non-neural and neural factors could contribute to ITW, which could guide future experimental studies. Targeting the altered control strategy, for example by employing biofeedback training, could encourage the adoption of a typical heel-toe gait, thereby delaying, or even preventing surgical intervention.L20148935702022-03-22 <br />