Applying Natural Language Processing to Understand Motivational Profiles for Maintaining Physical Activity After a...

Area of Science:

• Behavioral Science
• Digital Health
• Public Health

Background:

• Long-term maintenance of physical activity post-intervention remains a significant challenge.
• Regular physical activity is crucial for reducing chronic illness risk.
• Mobile health interventions show promise but require strategies for sustained engagement.

Purpose of the Study:

• To describe physical activity engagement after a mobile health intervention.
• To identify motivational profiles using natural language processing (NLP) and clustering.
• To compare sociodemographic and clinical data across identified motivational groups.

Main Methods:

• Cross-sectional analysis of 203 women post-intervention.
• Utilized a mobile phone-based physical activity education study.
• Assessed physical engagement and motivational profiles via open-ended questions, analyzed with NLP and cluster analysis.

Main Results:

• Approximately half of intervention participants maintained brisk walking post-study.
• Three motivational clusters emerged: Weight Loss (younger), Illness Prevention (Caucasian majority), and Health Promotion (lower BMI).
• No significant differences in baseline moderate-to-vigorous activity levels were found among clusters.

Conclusions:

• Findings support tailoring physical activity maintenance interventions based on motivational profiles.
• NLP and cluster analysis are effective methods for analyzing free-text data to differentiate motivational profiles.
• Advancements in NLP tools will enhance behavioral research applications.