Characterising physical activity patterns in community-dwelling older adults using digital phenotyping: a 2-week...

Area of Science:

• Gerontology and Digital Health
• Behavioral Science and Wearable Technology

Background:

• Physical activity (PA) is vital for older adults' health, but understanding its dynamic determinants in real-world settings is challenging.
• Conventional methods often fail to capture the fluctuating nature of factors influencing PA in daily life.
• Digital phenotyping (DP) offers a novel approach for continuous, real-time behavioral assessment in natural environments.

Purpose of the Study:

• To present a study protocol for the digital phenotyping of physical activity behavior in community-dwelling older adults.
• To dynamically assess individual-level determinants of physical activity in real-world contexts.
• To explore intraday and interday variability in PA patterns and their influencing factors.

Main Methods:

• A 2-week multidimensional assessment combining supervised (questionnaires, clinical assessments) and unsupervised methods (wearable monitoring, ecological momentary assessment (EMA)).
• Continuous 24/7 data collection via Garmin Vivosmart V.5 watch (PA intensity, steps, heart rate) and EMA prompts (physical/mental health, motivation, efficacy, context).
• Machine learning (unsupervised and supervised) for pattern identification and prediction of behavioral influences, aligned with the Behavior Change Wheel framework.

Main Results:

• This section is not applicable as the abstract describes a study protocol, not results.
• The protocol outlines methods for identifying PA behavior patterns and predicting influences using machine learning.
• Temporal patterns in PA and EMA responses will be analyzed for variability.

Conclusions:

• This study protocol introduces a robust digital phenotyping approach to understand the dynamic influences on physical activity in older adults.
• The findings are expected to provide novel insights into promoting active lifestyles in this demographic.
• Ethical approval obtained, and findings will be disseminated through conferences and peer-reviewed publications.