Comparing methods to measure wearable device adherence for physical activity monitoring for persons with autism: A prospective observational cohort study.

INTRODUCTION

Individuals with autism spectrum disorder (ASD) have higher rates of obesity and exhibit more sedentary behaviors than neurotypical peers. Use of wearable devices (WD) to target insufficient physical activity (PA) in this population has increased over the past decade. However, standardized methods for assessing WD adherence remain limited among this understudied population.

PURPOSE: A prospective observational cohort design was used to compare two methods for determining WD adherence for monitoring PA in adolescents and adults with ASD.

METHODS

Subjects were enrolled in an 8-week telehealth program aimed at improving PA among adolescents and adults with ASD. The Fitbit Inspire 3 WD was utilized for PA tracking and self-monitoring. Subject step activity and heart rate (HR) data were collected for an 8-week period via FITABASE management platform. Adherence was calculated using two methods. Method-A was defined as ≥10-hr daily wear-time based on HR monitoring; Method-B was defined as ≥ 500 steps/day. Observational data collected via researcher field notes documented participant-reported factors and/or behaviors that influenced WD adherence.

RESULTS

Data from twenty-seven subjects (21 males, 6 females) ages 14 to 28 years (mean: 18.85 years) were analyzed. Adherence was significantly higher when calculated using Method B (mean = 40 days, 5.4 weeks) compared to Method A (mean = 24 days, 2.8 weeks; p  < .001). Observational data identified disability-related behaviors that negatively impacted Method A derived adherence.

CONCLUSIONS

Using HR-based measures of adherence may underestimate device use in individuals with ASD, potentially setting unrealistic expectations for continuous wear. Step-based metrics provides a more inclusive and pragmatic approach for assessing WD adherence in this population, with implications for future intervention design and evaluation. This study highlights the need to tailor WD adherence definitions to the behavioral and sensory profiles of neurodiverse populations. Further study is warranted.