Factors affecting continuous participation in follow-up evaluations during a lifestyle intervention programme for type 2 diabetes prevention: The Feel4Diabetes-study.

Abstract

Community- and school-based lifestyle interventions are an efficient method of preventing type 2 diabetes in vulnerable populations. Many participants, however, fail to complete the necessary follow-ups. We investigated factors affecting the continuous participation in follow-up evaluations during the Feel4Diabetes-study, a multilevel intervention programme implemented across Europe. Socioeconomic, sociodemographic and clinical factors were assessed for 2702 participants within six participating countries: Bulgaria and Hungary (low-to-middle-income countries, LMIC), Belgium and Finland (high-income countries, HIC) and Greece and Spain (high-income countries under austerity measures, HICAM). Statistically significant differences were detected with respect to sex, control group, education level, employment status, BMI and blood pressure measurements (systolic and diastolic blood pressure). Post hoc analysis revealed significant differences within socioeconomic regions. Higher levels of education were associated with significantly lower attrition in HIC (p < 0.05) and HICAM (p < 0.001), higher employment status was associated with lower attrition in HICAM (p < 0.001) and being female was associated with lower attrition in LMIC (p < 0.001). Surprisingly, the intervention group exhibited higher-than-expected attrition in HIC (p < 0.001) and HICAM (p = 0.003), and lower attrition in LMIC (p = 0.007). When tested together in the same multivariable predictive model, all sociodemographic and socioeconomic variables along with higher BMI retained their statistical significance, while systolic and diastolic blood pressure failed to remain significant. Key socioeconomic and sociodemographic factors along with BMI play a significant role in determining continuous participation in follow-up evaluations during school- and community-based intervention programmes.