A group-derived activity count cutpoint of 141 counts per 5 seconds demonstrated strong classification performance (AUC = 0.903) for distinguishing moderate-to-vigorous from sedentary-to-light physical activity in toddlers with Down syndrome during a home-based mobility intervention.
Key Findings
Results
The optimal group-derived activity count cutpoint for distinguishing MVPA from SLPA in toddlers with Down syndrome was 141 activity counts per 5-second epoch.
Cutpoint was derived using receiver operating characteristic (ROC) curve analysis
Accelerometer data were binned in 5-second epochs synchronized with behaviorally coded video recordings
The accelerometer was integrated into a body-weight support harness system worn during the intervention
Video recordings were behaviorally coded for MVPA and SLPA and served as the reference standard
Results
The group-derived cutpoint of 141 activity counts per 5 seconds demonstrated strong classification performance.
Area under the curve (AUC) = 0.903, indicating strong discriminative ability
Sensitivity of 94.4% for detecting MVPA
Specificity of 70.2% for detecting SLPA
Overall classification accuracy of 79.2%
Results
Participant-specific cutpoints resulted in only minimal improvements in classification performance compared to the group-derived cutpoint.
Both group-derived and participant-specific cutpoints were determined using ROC curve analysis
The marginal improvement from individualized cutpoints suggests a standardized group cutpoint may be sufficient
Authors note that using individualized cutpoints may not be necessary for classifying MVPA in this population
Replication in a larger sample was noted as warranted before drawing firm conclusions
Background
Behavioral video coding, while used as the reference standard, was identified as time and labor-intensive, motivating the development of an objective accelerometer-based classification approach.
Video recordings were behaviorally coded for MVPA and SLPA and synchronized with accelerometer data
The study was conducted in the context of a home-based mobility intervention for toddlers with DS
Quantifying movement during early mobility interventions is described as important for evaluating intervention participation and progression
A wearable accelerometer integrated into a body-weight support harness system was used to capture activity counts
Conclusions
A standardized activity count cutpoint may provide a practical approach for objectively monitoring physical activity during home-based mobility interventions in toddlers with Down syndrome.
The group-derived cutpoint of 141 counts per 5 seconds showed high sensitivity (94.4%), prioritizing detection of MVPA epochs
Specificity was lower at 70.2%, meaning some SLPA epochs may be misclassified as MVPA
Authors describe the standardized cutpoint as a 'practical approach for objectively monitoring activity'
The study population was toddlers with Down syndrome, a group for whom early mobility intervention is particularly relevant
What This Means
This research suggests that a wearable accelerometer built into a body-weight support harness can reliably distinguish between more intense physical activity (moderate-to-vigorous, or MVPA) and lower-intensity or resting activity (sedentary-to-light, or SLPA) in toddlers with Down syndrome during home-based movement therapy. The researchers found that a single cutoff value of 141 activity counts per 5 seconds worked well for the whole group, correctly identifying active periods about 94% of the time and inactive periods about 70% of the time, with an overall accuracy of roughly 79%. This cutpoint was validated against detailed human video coding of the same sessions, which is considered the gold standard but is very time-consuming.
Importantly, the study found that creating customized thresholds for each individual child offered only marginal improvements over the one-size-fits-all group cutpoint. This is practically significant because it means researchers and clinicians may not need to go through the extra effort of calibrating the sensor separately for each child — a single standard value may be sufficient for monitoring how active children are during these sessions.
This research matters because toddlers with Down syndrome often experience delays in developing mobility skills, making early intervention critical. Having an objective, wearable tool to measure how much a child is actively participating in therapy sessions — without requiring hours of video review — could make it much easier for families and clinicians to track progress and adjust interventions over time. The authors caution that these findings should be replicated in a larger group of children before the cutpoint is widely adopted.
Check Your Own Numbers
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Athreya V, Mishra S, Collimore-Doherty A, Sutter E, Iverson J. (2026). Quantifying Physical Activity in Toddlers with Down Syndrome with Wearable Sensors During a Home-Based Mobility Intervention.. Sensors (Basel, Switzerland). https://doi.org/10.3390/s26196002