Saskia Neumann

Background: Real-world gait assessment has gained momentum in populations with walking impairments, offering insights beyond standardized tests and supporting the integration of remote monitoring into clinical care. However, its potential remains limited by the lack of validated population- and context-specific digital biomarkers. Objective: The primary objective was to summarize and critically evaluate the current state of real-world gait assessment using wearable sensors in Parkinson disease (PD) and stroke, addressing methodological approaches, sensor configurations, validation strategies, and outcome reporting. The secondary objective was to quantify methodological heterogeneity by reporting pooled means with 95% confidence intervals (CIs) and prediction intervals (PIs) of commonly reported gait parameters. Methods: PubMed, Scopus, Web of Science, and Embase were searched for English-language studies published up to April 20, 2026. Eligible studies used wearable sensors to assess gait quality parameters in real-world settings in at least 5 individuals with PD or poststroke. Laboratory- or rehabilitation-only studies, non–peer-reviewed articles, abstracts, and studies before 2014 were excluded. Methodological quality was assessed with a checklist adapted from Hubble et al. Gait features available in at least 5 reports were included in a meta-analysis using a random-effects model with Hartung-Knapp-Sidik-Jonkman adjustment, deriving 95% CIs and PIs to characterize the pooled estimate and its dispersion. Results: Of 2489 records, 43 reports were included: 37 on PD (2774 participants; mean age 67.89, SD 8.68 years) and 6 on stroke (217 participants; mean age 63.83, SD 11.84 years). Methodological quality was high, but external validity was consistently the weakest domain in both populations. Sensor configurations most commonly consisted of a lower back accelerometer. In total, 13 distinct walking bout definitions were identified, whereas 17 reports provided none, indicating substantial terminological heterogeneity. In PD, 4 gait parameters met the meta-analysis threshold, each showing a PI far wider than its CI, precluding generalization to future settings. For example, gait speed pooled at 0.84 m/s (95% CI 0.77‐0.91; 95% PI 0.56‐1.12), and number of steps per day at 7202 (95% CI 3975‐10,429), the latter with a PI extending below zero, a mathematically impossible range reflecting extreme dispersion. No stroke parameter was reported consistently enough to permit meta-analysis, underscoring a critically underdeveloped evidence base. Conclusions: Unlike previous reviews focusing on single diseases or specific hardware, this review contrasts PD and stroke to highlight shared methodological challenges. Real-world gait assessment in PD is advancing but fragmented, whereas the stroke evidence base remains sparse. The observed variance is compounded by procedural and algorithmic inconsistencies that cannot currently be separated from genuine clinical differences. The methodological groundwork established in PD offers a foundation for early adoption of standardized terminology, validation procedures, and core outcomes, essential for transitioning wearable sensors from exploratory tools to reliable clinical and remote-monitoring instruments.

BACKGROUND Real-world gait assessment has gained momentum in populations with walking impairments, offering insights beyond standardized tests and supporting the integration of remote monitoring into clinical care. However, the full potential of wearable sensors remains limited by the lack of validated population- and context-specific digital biomarkers. OBJECTIVE The primary objective was to update the state of the art and summarize challenges in real-world gait assessment for PD and stroke. The secondary objective was to report pooled means and standard deviations of gait parameters. METHODS PubMed and Scopus were searched for English-language studies published up to December 31, 2024. Eligible studies included a minimum of five individuals with Parkinson’s disease (PD) or post-stroke and used wearable sensors to assess gait in real-world settings. Studies conducted solely in laboratory or rehabilitation environments, non-peer-reviewed articles, abstracts, or studies before 2014 were excluded. RESULTS Of 167 records identified, 34 studies were included, comprising 30 on PD (n=209; 812 [37%] female, 1359 [63%] male, mean age 68,59 years [SD 7,86]) and four on stroke (n=159; 77 [49%] female, 80 [51%] male, mean age 64,16 years [SD 10,51]). The meta-analysis for PD covered seven gait parameters with high heterogeneity across outcomes (I²>97%). CONCLUSIONS Wearable sensors show strong potential for real-world gait assessment, but inconsistent methods call for standardization in sensor placement, algorithm validation, and metric definitions. Stroke populations are underrepresented, highlighting the need for targeted validation. CLINICALTRIAL This rapid review and meta-analysis was registered with PROSPERO, CRD42024531665.

Successful postural control depends on the integration of visual, vestibular, and proprioceptive inputs. With age, postural control degrades, leading to impaired balance and greater fall risk. Understanding how this integration changes over the lifespan is invaluable for designing more effective interventions that enable healthy postural control in older age. Earlier studies measured visual dependence using perceptual tasks or spontaneous sway comparisons across visual conditions. This study evaluates how visual dependence differs between younger and older adults within the postural control mechanism using a Central Sensorimotor Integration (CSMI) test. Eighty healthy adults (n=40, 60-87 years, n=40, 21-52 years) were exposed to small pseudorandom visual scene movements implemented in virtual reality while standing on a compliant surface. Sway responses were measured using virtual reality trackers and interpreted using an established frequency domain balance control model. Model parameters included visual weight, proportional and derivative feedback gains, time delay, and torque feedback gain. Test–retest reliability was assessed in a subgroup (n = 40) and showed excellent intra-class correlation coefficients for visual weight, proportional and derivative feedback gains (ICC = 0.89– 0.96), and lower ICCs for time delay (ICC=0.59) and torque parameters (ICC=0.39). The main difference between age groups was visual dependence, with older adults relying 40% on vision, compared to 33% for the younger group (p = 0.042). No significant group differences were found in other model parameters. Our results provide direct evidence of an increase in visual contribution to posture control with age.

Background: Gait impairments are common in stroke survivors, negatively impacting their overall quality of life. Therefore, gait rehabilitation is often targeted during in-clinic rehabilitation. While standardized assessments are available for inpatient evaluation, the literature often reports variable results when these assessments are conducted in a home environment. Several factors, such as the presence of an observer, the environment itself, or the technology used, may contribute to these differing results. Therefore, it is relevant to establish unsupervised capacity assessments for both in-clinic use and across the continuum of care. Objective: This study aimed to investigate the effect of supervision on the outcomes of a sensor-based 10-meter walk test conducted in a clinical setting, maintaining a controlled environment and setup. Methods: In total, 21 stroke survivors (10 female, 11 male; age: mean 63.9, SD 15.5 years) were assigned alternately to one of two data collection sequences and tested over 4 consecutive days, alternating between supervised test (ST) and unsupervised test (UST) assessments. For both assessments, participants were required to walk a set distance of 10 meters as fast as possible while data were collected using a single wearable sensor (Physilog 5) attached to each shoe. After each walking assessment, the participants completed the Intrinsic Motivation Inventory. Statistical analyses were conducted to examine the mean speed, stride length, and cadence, across repeated measurements and between assessment conditions. Results: The intraclass correlation coefficient indicated good to excellent reliability for speed (ST: κ=0.93, P<.001; UST: κ=0.93, P<.001), stride length (ST: κ=0.92, P<.001; UST: κ=0.88, P<.001), and cadence (ST: κ=0.91, P<.001; UST: κ=0.95, P<.001) across repeated measurements for both ST and UST assessments. There was no significant effect of testing order (ie, sequence A vs B). Comparing ST and UST, there were no significant differences in speed (t39=–0.735, P=.47, 95% CI 0.06-0.03), stride length (z=0.835, P=.80), or cadence (t39=–0.501, P=.62, 95% CI 3.38-2.04) between the 2 assessments. The overall motivation did not show any significant differences between the ST and UST conditions (P>.05). However, the self-reported perceived competence increased during the unsupervised assessment from the first to the second measurement. Conclusions: Unsupervised gait capacity assessments offer a reliable alternative to supervised assessments in a clinical environment, showing comparable results for gait speed, stride length, and cadence, with no differences in overall motivation between the two. Future work should build upon these findings to extend unsupervised assessment of both capacity and performance in home environments. Such assessments could allow improved and more specific tracking of rehabilitation progress across the continuum of care.

BACKGROUND Gait impairments are common in stroke survivors, negatively impacting their overall quality of life. Therefore, gait rehabilitation is often targeted during in-clinic rehabilitation. While standardized assessments are available for inpatient evaluation, the literature often reports variable results when these assessments are conducted in a home environment. Several factors, such as the presence of an observer, the environment itself, or the technology used, may contribute to these differing results. Therefore, it is relevant to establish unsupervised capacity assessments for both in-clinic use and across the continuum of care. OBJECTIVE This study aimed to investigate the effect of supervision on the outcomes of a sensor-based 10-meter walk test conducted in a clinical setting, maintaining a controlled environment and setup. METHODS In total, 21 stroke survivors (10 female, 11 male; age: mean 63.9, SD 15.5 years) were assigned alternately to one of two data collection sequences and tested over 4 consecutive days, alternating between supervised test (ST) and unsupervised test (UST) assessments. For both assessments, participants were required to walk a set distance of 10 meters as fast as possible while data were collected using a single wearable sensor (Physilog 5) attached to each shoe. After each walking assessment, the participants completed the Intrinsic Motivation Inventory. Statistical analyses were conducted to examine the mean speed, stride length, and cadence, across repeated measurements and between assessment conditions. RESULTS The intraclass correlation coefficient indicated good to excellent reliability for speed (ST: κ=0.93, P<.001; UST: κ=0.93, P<.001), stride length (ST: κ=0.92, P<.001; UST: κ=0.88, P<.001), and cadence (ST: κ=0.91, P<.001; UST: κ=0.95, P<.001) across repeated measurements for both ST and UST assessments. There was no significant effect of testing order (ie, sequence A vs B). Comparing ST and UST, there were no significant differences in speed (t39=–0.735, P=.47, 95% CI 0.06-0.03), stride length (z=0.835, P=.80), or cadence (t39=–0.501, P=.62, 95% CI 3.38-2.04) between the 2 assessments. The overall motivation did not show any significant differences between the ST and UST conditions (P>.05). However, the self-reported perceived competence increased during the unsupervised assessment from the first to the second measurement. CONCLUSIONS Unsupervised gait capacity assessments offer a reliable alternative to supervised assessments in a clinical environment, showing comparable results for gait speed, stride length, and cadence, with no differences in overall motivation between the two. Future work should build upon these findings to extend unsupervised assessment of both capacity and performance in home environments. Such assessments could allow improved and more specific tracking of rehabilitation progress across the continuum of care.