Related Experiment Video
Updated: Aug 25, 2026
Using a Real-Time Locating System to Measure Walking Activity Associated with Wandering Behaviors Among Institutionalized Older Adults
Published on: February 8, 2019
Real-world walking and activity patterns in community-dwelling older adults: a cross-sectional analysis of baseline
Tobias Eckert1,2, Katharina Gordt-Oesterwind3, Ann-Sophie Haas3,4
1Network Aging Research, Heidelberg University, Heidelberg, Germany. tobias.eckert@med.uni-heidelberg.de.
Background:
Digital mobility outcomes (DMOs) measured under real-world conditions allow for differentiated insights into mobility patterns of older adults. This study provides an advanced description of real-world walking in community-dwelling older adults and identifies correlates of walking amount, pattern and speed.
Methods:
This cross-sectional study used baseline data from the SMART-AGE intervention trial with community-dwelling older adults (age ≥ 67 years). Assessment procedures included one-week monitoring of real-world walking using a wearable device (Axivity AX6), clinical outcomes, and tablet-based questionnaires. DMOs describing the amount (steps, walking duration), pattern (e.g., number of walking bouts [WBs]) and pace (mean and 90th percentile [P90] walking speed) of real-world walking were extracted using a validated computational pipeline. Data were included with ≥ 12 h/day wear time on ≥ 3 days. Stepwise hierarchical linear mixed modelling examined associations between DMOs and sociodemographic correlates, environmental conditions, health status (Body Mass Index), locomotor capacity (Timed Up & Go [TUG]), fall-related concerns (Short FES-I) and cognitive processing speed (Symbol Search Score).
Results:
The final sample of 569 participants with a mean age of 75.0 ± 5.6 years (52% women; 56% higher education) showed a mean walking activity of 11 746 ± 5 446 steps/day, 432.3 ± 153.8 WBs / day in total, and a mean walking speed of 0.75 ± 0.13 m/s in longer (> 30s) WBs. Age was consistently related to real-world walking across all DMOs, showing non-linear trends. We found sex-related patterns of walking, as women had more WBs in total but fewer longer (> 30 s) WBs than men. Body Mass Index and locomotor capacity were consistent predictors of walking across all DMOs (β = -0.08 to -0.25, p < 0.05).
Conclusion:
This study provides detailed and novel insights into real-world walking in a highly active population of older adults, highlighting lower walking activity with increasing age and differences in activity patterns between men and women. Mobility data may serve as a reference for community-dwelling older adults and facilitate the identification of early changes in everyday mobility, thereby informing future mobility recommendations and preventive strategies.
