Abstract
Wearable fitness devices are increasingly common for health monitoring in adults, yet longitudinal data regarding their use in healthy adolescents remains scarce. This pilot case study documents the successes and challenges of a five-year (60-month) wearable device deployment involving a teenage girl and her father. The study identifies significant behavioral shifts, including a 60% decline in the teenager’s activity levels, and highlights the technical necessity for algorithms to account for physiological growth. The findings suggest that with periodic adjustments for growth and improved automation, wearables can provide accurate, long-term health metrics for the pediatric population.
Introduction
While wearable fitness trackers are ubiquitous in adult wellness and chronic disease management, we have very little data on how healthy adolescents interact with these devices over extended periods. Adolescence is a period of rapid physiological and behavioral change, making it a difficult target for static health-tracking algorithms. This study addresses this gap through a "citizen-science" longitudinal case study, tracking a father-daughter duo to explore the technical and behavioral dynamics of long-term pediatric wearable use.
Methods
The study spanned 60 months, from June 2013 to May 2018. The participants—a female starting at age 10 and her father starting at age 43—each utilized a Fitbit One device. Raw data, including step counts and sleep duration, was retrieved via the Fitbit API using custom R scripts in 100-day increments.
To ensure the data was contextually relevant, the researchers integrated six years of school calendars to distinguish between school days, weekends, and holiday periods. Additionally, to address the teenager's physical growth, 14 height measurements from her electronic medical records were used to fit a cubic spline, allowing for the estimation of stride length variance over the five-year duration.
Key Insights and Discussion
The "Growth Gap" in Wearable Algorithms
A significant finding of this study is the "growth gap": the discrepancy between a child’s actual physical dimensions and the static settings of a wearable device. Over the five-year period, the teenager’s stride length increased by 12.2%. If a device's software is not updated to reflect these changes, the resulting distance and caloric burn data become increasingly inaccurate. The researchers suggest that wearable manufacturers should implement "birthday prompts" or automated adjustments based on age-standardized growth charts to maintain data integrity.
Coordinated Family Behavior
The study revealed that family members are not independent variables in health data. There was a notable correlation between the father’s and daughter’s activity (Pearson r = 0.2 for steps; r = 0.23 for sleep). This correlation nearly doubled on weekends (r = 0.4), suggesting that shared family activities—such as weekend outings or synchronized household schedules—significantly influence individual health metrics. This highlights the importance of considering the family unit in pediatric health interventions.
The High School Activity Slump
The most striking behavioral trend was the dramatic decline in the teenager’s activity levels. Average daily steps fell from approximately 9,100 in Year 1 to just 3,600 in Year 5—a 60% reduction. This decline coincided with the transition into high school, where the loss of mandatory physical education and increased academic sedentary time took a toll. Statistical analysis confirmed that the teenager walked significantly more on school days than non-school days (a difference of ~2,105 steps, p = 8.17 x 10⁻³¹), emphasizing the role of structured environments in driving pediatric activity.
User Experience and Data Friction
Data collection was hampered by the manual requirements of the hardware. The Fitbit One required a manual button press to track sleep, leading to a 54.7% data loss for the teenager’s sleep records, compared to only 14.2% for steps, which were tracked passively. For pediatric use, automation is not merely a convenience but a requirement for consistent data collection.
Conclusion and Recommendations
This five-year pilot study demonstrates that wearables can be effective tools for long-term pediatric health monitoring, provided that the technology evolves alongside the user. To improve the clinical utility of these devices for adolescents, the authors recommend:
- Automated Stride Adjustments: Using growth charts to estimate height changes if manual updates are missing.
- Clinical Integration: Developing dashboards for pediatricians to monitor long-term activity trends.
- Parental Engagement: Implementing sync alerts to notify parents if a child’s device has not been updated, preventing significant data gaps.
While limited by a small sample size (N=2), these findings provide a foundational look at the longitudinal realities of pediatric wearable use.
Summary updated