Related Experiment Video
Updated: Aug 5, 2026
05:59
Visualization of Intensity Levels to Reduce the Gap Between Self-Reported and Directly Measured Physical Activity
Published on: March 7, 2019
Identifying cluster profiles based on barriers and facilitators to physical activity during COVID-19 confinement: A
Fernanda Castro Monteiro1,2, Maria Luiza C Wuillaume3, Carlos Linhares Veloso Filho1
1Psychiatry Institute, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil.
Plos One
|July 27, 2026
Summary
COVID-19 confinement reduced physical activity (PA), with depressive symptoms and time constraints emerging as key barriers. Tailored strategies are needed to support individuals facing these challenges.
Area of Science:
- Public Health
- Behavioral Science
- Epidemiology
Background:
- Social restrictions during confinement periods often lead to decreased physical activity (PA).
- Sociodemographic factors can significantly influence the barriers and facilitators to maintaining PA during such times.
- Understanding these influences is crucial for developing effective public health interventions.
Purpose of the Study:
- To identify distinct cluster profiles of Brazilian adults based on barriers and facilitators to physical activity (PA) during COVID-19 confinement.
- To analyze how sociodemographic factors correlate with these identified barriers and facilitators.
- To inform the development of targeted strategies for promoting PA.
Main Methods:
- A cross-sectional online survey was conducted with Brazilian adults.
- Data collected included demographics, PA levels, sedentary behavior (SB), and perceived barriers/facilitators to PA.
- Unsupervised machine learning (K-modes) was employed to identify clusters based on aggregated barriers, with facilitators analysis proving less conclusive.
Main Results:
- Eight distinct clusters were identified based on barriers to PA, including profiles like 'Inactive depressive women', 'Active depressive women', and 'Super active'.
- Depressive clusters reported more barriers to PA (three each).
- Significant variations in PA levels and SB were observed across the identified clusters.
Conclusions:
- Unsupervised machine learning effectively identified latent subgroups facing specific PA barriers during confinement.
- Individuals with depressive symptoms and significant time constraints represent key target groups for interventions.
- Tailored home-based and outdoor strategies are recommended to address these identified barriers and promote PA.
