Establishing the clinical utility of walk-DMC to measure motor control following pediatric traumatic brain injury

Predicting walking recovery following pediatric traumatic brain injury (TBI) is critical, as independent mobility impacts long-term health1. However, accurate prognosis is often challenging, due, in part, to the limitations of existing measurement tools; although myriad tests are used clinically, they may be insufficiently sensitive to measure motor control deficits following TBI2. Walk-DMC is a quantitative measure of motor control that has clinical utility in cerebral palsy (CP), as it has an effect on net energy consumption, gait mechanics, and independent mobility3,4. While this suggests that walk-DMC may be similarly valuable in TBI populations, it has been minimally evaluated to this end2. To what extent does walk-DMC capture changes in motor control and mobility among a pediatric TBI population? We retrospectively analyzed electromyography (EMG) and exam data, collected as part of standard of care at Gillette Children’s Hospital, from 113 children who had sustained a TBI (Sex: 77 M, Age: 11.1 ± 3.8 years) and were in the chronic stages of recovery (7.5 ± 3.7 years following hospital admission)5. EMG data were measured during overground walking from the rectus femoris, medial hamstring, gastrocnemius, and tibialis anterior on the more-affected limb. For all individuals, walk-DMC was derived from EMG data using non-negative matrix factorization6-7 and normalized to a non-disabled (ND) cohort. A subset of individuals also had measures of (1) walking speed (n = 94), (2) net energy consumption (n = 91), and (3) self-reported mobility (n = 72; Functional Assessment Questionnaire Transform (FAQt) 8). We used a previously developed causal modeling framework to evaluate the total effect of walk-DMC on each of these three variables in a TBI and age-matched CP cohort, taken from the same database3. Dependence plots were used to evaluate how Shapley values changed over the entire domain of each variable9. Walk-DMC was significantly lower in individuals with chronic TBI (85 ± 9) compared to an ND cohort (100 ± 10; paired t-test; p < 0.001). Results from the causal model indicated that walk-DMC has a large positive effect on FAQt and walking speed in TBI and CP, whereby higher walk-DMC values result in faster walking speeds and higher FAQt (Fig. 1). Similarly, lower walk-DMC imparts an energetic penalty during walking. Fig. 1 - Dependence plots for walk-DMC on gait measuresDownload : Download high-res image (73KB)Download : Download full-size image This study indicates that walk-DMC may be a valuable addition to clinical care in pediatric TBI, as it is sufficiently sensitive to capture motor control deficits and has large effects on other clinically-relevant gait measures. However, additional work is needed to understand if walk-DMC changes with recovery and injury severity. Outcomes from this work will bolster the integration of objective motor control measures into practice to inform prognosis and intervention planning for individuals following neurologic injury.