Digital Gait Measures Discriminate People with Friedreich's Ataxia from Healthy Controls

Gait and mobility impairments are early markers of disease-related decline in individuals with Friedreich's ataxia (FRDA).

This study aimed to identify the digital gait measures most discriminative of individuals with FRDA from healthy controls (HC) during an instrumented walking assessment. We also investigated if top discriminative gait measures correlated with disease severity.

We recruited 24 participants with FRDA and 24 HCs to complete a 2-min walk at a natural pace while wearing six wearable sensors on the hands, feet, sternum, and lower lumbar area. AUCs between FRDA and HC compared 51 gait measures, including standard deviations as estimates of variability, to discriminate between ataxic and normal gait. Spearman's correlation coefficients assessed the relationships between the most discriminative measures and ataxia severity from clinical outcome measures.

We identified 10 measures with strong discriminative ability for gait characteristics of FRDA. The most discriminative measure was double-support time SD (AUC = 0.98, ICC = 0.95); a measure of variability. Eight out of the 10 most discriminate measures involved excessive gait variability in the FRDA group. The top five discriminative gait measures significantly correlated with the mFARS USS (r = 0.45-0.63) and two discriminative gait measures significantly correlated with the FARS ADL (r = 0.62 and 0.5). Test-retest reliability of the top measures were high (ICC = 0.85-0.98).

Digital gait measures from wearable sensors were discriminative, reliable, and showed concurrent validity for evaluating ataxia severity during an instrumented walk test. These results suggest promising utility of digital gait outcomes for use in FRDA clinical trials.