Related Experiment Video
Updated: Nov 5, 2025
10:52
Simulation of Human-induced Vibrations Based on the Characterized In-field Pedestrian Behavior
Published on: April 13, 2016
9.0K
Characterizing and predicting person-specific, day-to-day, fluctuations in walking behavior
Guillaume Chevance1,2,3, Dario Baretta4, Matti Heino5
1ISGlobal, Barcelona Institute for Global Health, Barcelona, Spain.
Plos One
|May 14, 2021
Summary
Physical activity, like daily steps, changes suddenly, not gradually. Early warning signals can predict these changes, aiding in developing timely interventions for better health outcomes.
Area of Science:
- Behavioral science
- Public health
- Data science
Background:
- Physical activity is crucial for health, yet many are insufficiently active.
- Previous research often overlooks the day-to-day fluctuations in physical activity.
- Understanding temporal dynamics is key for effective health interventions.
Purpose of the Study:
- To analyze high-resolution, day-to-day changes in walking behavior.
- To identify sudden gains and losses in daily step counts.
- To explore early warning signals for behavioral shifts.
Main Methods:
- Utilized accelerometer data from 151 young adults with overweight/obesity over ~226 days.
- Employed recursive partitioning to detect sudden gains/losses (≥30% change for ≥7 days).
- Applied dynamic complexity analysis to identify critical fluctuations as early warning signals.
Main Results:
- Walking behavior changes occur discontinuously, characterized by sudden gains and losses.
- Participants experienced an average of six sudden gains or losses during the study.
- Critical fluctuations in step count significantly predicted subsequent sudden behavioral losses.
Conclusions:
- Walking behavior is best understood through a dynamic, rather than aggregate, lens.
- Early warning signals can predict significant shifts in physical activity.
- Findings support "just-in-time adaptive" interventions for promoting physical activity.
