Wearable Sensors Show Daily Walking Differs from Clinic Tests for SCA Patients

We don’t usually think about how we walk until walking starts to feel different. Walking actually takes more effort than we realize. In spinocerebellar ataxia (SCA) patients, changes in walking are often the first signs that something is wrong, and these usually worsen over time. Spinocerebellar ataxia (SCA) refers to a group of progressive neurodegenerative diseases characterized by impaired motor coordination, balance, and gait. Gait is the way you walk. It includes things like how fast you walk, how long your steps are, how your body moves from side to side, and whether your steps feel steady. 

In recent years, wearable technologies have been introduced to capture detailed motor data. This allows clinicians to better assess patients’ motor performance and disease severity. This study aims to analyze the use of wearable sensors in two different settings: clinical assessment and daily life monitoring. The comparison reveals their similarities and differences. This work provides more comprehensive information about walking changes in ataxia and will inform future improvements in sensor development. 

In clinical trials, the specialists evaluate gait, balance, and six other categories of motor impairment for patients. Scale for the Assessment and Rating of Ataxia (SARA) is one of the most common one for the evaluation. However, even though SARA is the most commonly used clinical tool, it also has limitations. Three of these limitations include: 

• SARA is subjective based on specialists, so different doctors may give slightly different ratings.  

• SARA has low sensitivity requiring a large population size, because SARA may not pick up very small changes in the movement.  

• • SARA requires in person meetings with specialists, which uses hospital resources and is inconvenient for patients.  

In contrast, according to a previous study, wearable sensors can objectively record motor data with high sensitivity that can distinguish between people with ataxia and people without ataxia in a small sample size. They are easy to use, can be worn during daily activities, and reduce the need for clinical supervision. Therefore, the further development of this practical technology could benefit patients by improving accessibility and efficiency in symptom monitoring. 

This study focuses on the similarities and differences of wearing sensors in clinical and daily situations. In the clinic, participants wear sensors and are asked to walk for two minutes over a distance of ten meters. After the clinical trial, participants kept using the sensors to monitor their natural walking pace for at least seven days, with a minimum of eight hours per day.  

To narrow the perspective of the broad SCA disorders, the study focuses on SCA1, SCA2, SCA3, and SCA6. There is a total of thirty-nine participants: ten SCA1 patients, nine SCA2 patients, four SCA3 patients, and three SCA6 patients, along with thirteen same-age and same-gender healthy people as controls. This sensor consists of three small units called inertial measurement units that continuously collect data from the participant’s two feet and the lower back. The same data analysis algorithms are applied in both clinical and daily walking situations. The data from the sensors are summarized into seventeen gait measures to analyze motor functions from multiple perspectives. At the same time, traditional assessments are used as references. The study included SARA and other assessments, such as the Activities-Specific Balance Confidence Scale (ABC) and the Patient-Reported Outcome Measure (PROM-Ataxia), among others. 

The comparison between clinical walking and natural daily walking reveals several key findings. First, the wearable sensor could tell the difference between participants with SCA and healthy controls based on walking data. More walking measures were different in the clinic setting (10/17) compared to the daily walking (3/17). Among this data, the double support time standard deviations (SD) and swing time SD were most prominently altered in SCA patients. This indicates the sensors were sensitive enough to pick up on walking changes,  even with a limited sample size. 

Second, walking patterns in the clinic can be very different from how patients walk in daily life, even for the same person. Each participant is monitored in both clinic and daily walking situations, and differences are found. For SCA patients, 13 out of 17 walking measures are different between clinic and daily life. Even for healthy participants, 11 out of 17 measures are different. Even though double support SD and swing time SD showed the biggest changes, walking in the clinic still looked different from walking in daily life. These differences may happen because clinic tests are done in a controlled environment. Participants focus on walking without distractions. On the other hand, in daily life, people move while doing other tasks with more variability.  

Lastly, these walking measures from the sensors match well with the results from common clinical assessments. The outcomes of double support SD and swing time SD show strong correlations with SARA, ABC, and PROM-Ataxia. This means that sensor-derived walking measures and analyses are reliable and may be used in future ataxia clinical trials. 

In summary, this project compared sensor data from two different walking conditions: clinical walking and spontaneous walking in daily life. The use of sensor-based walking data offers several advantages, including high sensitivity, ease of use, and reliable outcomes. However, sensors are more variable in daily walking due to distractions and multitasking. Despite this variability, wearable sensors have significant potential for assessing SCA patients and can serve as a supplement to clinical tests. Daily monitoring provides additional insights beyond the clinic. This type of data reflects actual functional walking in patients’ natural environments. In the future, further studies could focus on upper-body movements. This can show more detail about the whole-body situation during walking.  Also, larger sample sizes across SCA subtypes are needed to enable more comprehensive, ataxia-type-specific assessments.