From Steps to Context: Optimizing Digital Phenotyping for Physical Activity Monitoring in Older Adults by Integrating Wearable Data and Ecological Momentary Assessment

Feb 13, 2025Sensors (Basel, Switzerland)

Improving Activity Tracking in Older Adults by Combining Wearable Devices and Real-Time Self-Reports

AI simplified

Abstract

Over two weeks, 108 older adults used wearable sensors, achieving 67.2% adherence to real-time activity prompts.

  • Mobile technologies and wearable sensors may provide a less biased method for assessing physical activity in older adults.
  • Physical activity was primarily low (51.4%) and moderate (46.2%) in intensity, with peaks in midday activity.
  • Motivation and self-efficacy were significantly associated with low-intensity physical activity in the morning.
  • Objective step counts showed no correlation with self-reported physical activity measures, indicating complementary value in data sources.
  • The integration of ecological momentary assessment, wearable sensors, and temporal frameworks may improve physical activity evaluation.

AI simplified

Key numbers

67.2%
Adherence Rate
Overall adherence rate across the follow-up period.
51.4%
Low-Intensity PA Proportion
Proportion of physical activity classified as low intensity.
46.2%
Moderate-Intensity PA Proportion
Proportion of physical activity classified as moderate intensity.

Full Text

What this is

  • This research explores how digital phenotyping can enhance physical activity (PA) assessment in older adults.
  • It integrates ecological momentary assessment (EMA) with wearable sensor data to capture real-time behaviors.
  • The study involved 108 community-dwelling older adults monitored over two weeks to evaluate adherence and activity patterns.

Essence

  • Integrating EMA with wearable sensors provides a nuanced understanding of physical activity patterns in older adults. This approach reveals that most activity occurs at low and moderate intensities, primarily in the mornings, while highlighting discrepancies between self-reported and objective measures.

Key takeaways

  • Physical activity predominantly occurs at low (51.4%) and moderate (46.2%) intensities, with a peak around midday. This distribution underscores the need for tailored interventions to promote higher intensity activities among older adults.
  • Adherence to EMA prompts averaged 67.2%, indicating feasibility for real-time data collection in older populations. However, adherence varied slightly by time of day, suggesting potential fatigue or reduced motivation in the evenings.
  • No correlation was found between self-reported MET values and objective step counts (R = -0.026, p = 0.65), emphasizing the importance of integrating both subjective and objective data for a comprehensive understanding of physical activity.

Caveats

  • The sample consisted mainly of active and digitally literate older adults, limiting generalizability to less active or less tech-savvy populations. Future research should aim for more diverse participant recruitment.
  • The short observation period of two weeks may not capture seasonal variations in physical activity. Longer monitoring durations are recommended for a broader understanding of activity patterns.

AI simplified

what lands in your inbox each week:

  • 📚7 fresh studies
  • 📝plain-language summaries
  • direct links to original studies
  • 🏅top journal indicators
  • 📅weekly delivery
  • 🧘‍♂️always free