Abstract
INTRODUCTION: Specific health guidance programs provide motivational support to individuals in the precontemplation stage of the Transtheoretical Model (TTM) to encourage behavioral changes. However, this stage may encompass heterogeneous attitudes beyond simple lack of health interest. This study aimed to develop a novel segmentation classification incorporating behavioral economics concepts to better characterize the precontemplation population and validate targeted interventions. METHODS: We conducted a two-phase study among Japanese adults aged 40-59 years eligible for specific health guidance. Phase 1 (n = 1,125) involved systematic questionnaire development based on behavioral economics literature and expert interviews, followed by factor analysis and k-means clustering to create a segmentation algorithm. Phase 2 (n = 4,900) validated targeted motivational messages for each identified segment through randomized message testing. Primary outcomes included changes in behavioral stage and health behavior intentions. RESULTS: Seven distinct segments were identified: 'High health awareness and practices' (18.7%), 'High risk perception and busy' (12.0%), 'High health threat' (21.1%), 'High health anxiety' (11.8%), 'Procrastination and improvement interest' (12.7%), 'Busy and no future image' (19.2%), and 'Procrastination and improvement resistance' (4.5%). A simplified 9-item algorithm achieved 50.7% classification accuracy (κ = 0.42, 95% CI: 0.39-0.45) compared to the original clustering. Message testing revealed significant segment effects (F = 7.48, p < 0.001, ηp (2) = 0.015). 'High risk perception and busy' and 'High health threat' segments showed greatest responsiveness to targeted motivational messages (Cohen's d = 0.28-0.35), while the 'Procrastination and improvement resistance' segment showed consistent low responsiveness except to gain-framed messages focusing on specific actions. DISCUSSION: This novel segmentation approach provides deeper understanding of the precontemplation stage by identifying distinct subgroups with unique behavioral economic characteristics including present bias and reference point orientations. Findings suggest that tailored interventions considering these behavioral economic factors may be more effective than traditional stage-based approaches for engaging resistant populations in health behavior change.