Clinical assessments in amyotrophic lateral sclerosis (ALS) are typically collected during clinic visits and may miss domain-specific changes in daily function. Digital health technologies (DHTs) can enable remote monitoring, but passive free-living measures often need long wear times and are affected by non-motor factors. This study evaluated whether brief, standardized at-home lower-limb exercises recorded with ankle-worn accelerometers yield objective, interpretable metrics that reflect lower-limb disease progression in ALS.
The analysis used data from 349 participants enrolled in the decentralized ALS Research Collaborative (ARC) Study. Participants completed repeated self-entry ALS Functional Rating Scale–Revised assessments (ALSFRS-RSE) and wore bilateral ankle accelerometers during monitoring periods conducted between September 2014 and January 2023. During each monitoring period, participants performed brief seated knee flexion-extension exercises at home.
A previously developed signal-processing pipeline converted accelerometer recordings from the seated knee flexion-extension task into four exercise-derived metrics: count, duration, intensity, and similarity. These metrics aim to capture different aspects of movement performance: the number of repetitions (count), total time engaged (duration), vigor or power of movements (intensity), and consistency or reproducibility of the movement waveform across repetitions (similarity). The pipeline provided standardized, objective measures intended to be interpretable in the context of lower-limb motor function.
At baseline, the exercise-derived metrics demonstrated associations with ALSFRS-RSE total and subdomain scores. Metrics of intensity and similarity showed the strongest relationships with the gross motor subdomain, suggesting these movement-quality measures reflect functional capacity for lower-limb tasks. The analysis identified links between exercise metrics and clinical measures without reporting additional numerical correlation coefficients in the source.
Longitudinal change in exercise metrics was examined using linear mixed-effects models. Over time, duration increased while intensity and similarity decreased, with all reported changes meeting statistical significance (p < 0.001). These patterns were interpreted as consistent with progressive slowing, reduced movement vigor, and reduced movement consistency in lower limbs as ALS progressed. The count metric did not show a statistically significant change over time. The source did not provide detailed model coefficients or effect sizes in the text provided here.
Worsening responses to specific gross motor ALSFRS-RSE items — namely turning in bed, walking, and stair climbing — were consistently associated with exercise-derived signals indicating fewer, slower, less vigorous, and less consistent lower-limb repetitions. Additionally, baseline intensity and similarity were lower in participants whose disease onset involved the lower limb on the corresponding side, indicating anatomical specificity of the exercise metrics for unilateral disease onset.
The study compared exercise-derived metrics with passive free-living gait measures. Intensity derived from the seated exercise demonstrated model fit comparable to the strongest free-living gait metrics while requiring substantially less observation time. The source emphasizes that short, standardized exercises may offer efficient, scalable assessment options relative to longer-duration passive monitoring approaches.
The authors conclude that short at-home lower-limb exercises recorded using ankle-worn accelerometers provide scalable, objective, and interpretable measures of ALS-related functional decline. Movement-quality metrics, particularly intensity and similarity, may complement passive free-living monitoring and support remote digital clinical outcome assessment in ALS research. Trial registration is reported as NCT06885918. Data may be shared upon request and after review by the data owners via the ALS TDI ARC Data Commons; the source indicates a data request pathway but does not include raw datasets within the article text presented here.
The study adhered to the Declaration of Helsinki and received institutional review board approval from ADVARRA Center for IRB Intelligence (CIRBI). All participants provided informed consent; there was no participant compensation. The authors disclosed a range of competing interests in the source: several authors reported paid consultancies, research support, or advisory roles with industry or nonprofit organizations, and ALS TDI under one author’s leadership had received research support. Specific competing-interest details are reported in the original source text.