Work

Analysis of Participant-Level Characteristics Predicting Adherence to Long-Term EMA and Fitbit Monitoring in the 4HAIE Prospective Cohort Study

Year: 2026

Type: article

Abstract: BACKGROUND: Ecological momentary assessment (EMA) and wearable devices provide valuable real-time data on health behaviors, but their utility depends on sustained participant adherence. Limited evidence exists on which individual characteristics predict adherence in long-term, large-scale studies. PURPOSE: This study examined participant-level factors associated with adherence to a 12-month monitoring protocol that combined wearable activity tracking and repeated mobile surveys in a prospective cohort of community-dwelling adults. METHODS: A total of 1,314 adults from two regions in the Czech Republic (Moravia-Silesia and South Bohemia) wore a Fitbit activity tracker daily and completed EMA surveys during four two-week bursts across one year. Adherence was operationalized as the number of valid Fitbit monitoring days, completed time-based surveys, and completed weekly surveys. Bidirectional stepwise regression and random forest analyses identified predictors of adherence from demographic, psychological, motivational, and health-related characteristics assessed at baseline. RESULTS: Adherence was generally high, with participants providing an average of 77% valid Fitbit-days. Older age, lower stress, greater life satisfaction, higher optimism, and stronger barrier self-efficacy predicted greater adherence. Conversely, poorer physical health, higher body mass index, and higher controlled regulation predicted lower adherence. CONCLUSIONS: Sustained long-term monitoring with mobile surveys and wearables is feasible in large cohorts. Adherence was lower among participants with poorer physical health, higher BMI, and stronger controlled motivation, suggesting that protocols which reduce burden, promote autonomous rather than externally driven motivation, and provide additional support for individuals with health-related barriers may help optimize adherence and data quality.

Source: International Journal of Behavioral Medicine

DOI: 10.1007/s12529-025-10433-3