Development of a school-based health promotion intervention in Ecuadorian adolescents and its cluster randomised-controlled evaluation design

Driven by economic and societal changes, obesity levels and chronic diseases have surged in many low- and middle-income countries (LMICs) in recent years. An estimated 25% (51.8 million) of Latin American children and adolescents are considered overweight or obese. Poor dietary and physical activity (PA) behaviours fuel these developments. Considering their impact on overall health, well-being and quality of life, preventive actions and informed policies are needed to tackle chronic diseases effectively. Preventive interventions promoting a healthy diet and an active lifestyle are promising means of improving child and adolescent health, and curbing the rise of this disease burden. Most of the available evidence, however, originates from high-income countries (HICs). In addition, a number of key conceptual and methodological concerns hamper their potential to alleviate the disease burden of chronic diseases and obesity. This research targets school-going adolescents in Ecuador, a Latin American country with a high incidence of obesity and chronic diseases. The purpose of this research was to provide evidence at different levels of the systematic and stepwise planning and development processes of preventive interventions. In this context, a preventive school-based health promotion intervention, aiming to improve dietary and PA behaviours among 11 to 15-year old adolescents was developed as a case study. To facilitate the development of an intervention, existing evidence on the effectiveness of school-based obesity prevention interventions was evaluated using a systematic review. This review focused on interventions targeting dietary and/or PA behaviour in children and adolescents aged 6 to 18 years in LMICs. Most of the identified interventions (82%) had a positive effect on dietary and PA behaviour but few decreased Body Mass Index (BMI). Effective interventions targeted both diet and PA, involved multiple stakeholders, and integrated educational activities into the school curriculum. However, to reach their full potential, the current interventions need to address important methodological and conceptual issues. In general, more and better quality research is needed that examines contextual influences of PA and eating behaviours and uses theory, rigorous evaluation designs, valid evaluation tools and methods that are not prone to reporting bias. To address the latter, the validity and reliability of a PA record as a tool for PA measurement was evaluated in a sample of 302 adolescents. The PA record provided acceptable estimates for reliability and validity within a group and showed an overall fair measurement agreement for validity. There was modest reliability for assessing PA in each intensity level. Sex and setting were associated with better validity, whereas perceived difficulty in filling out the record was associated with better reliability estimates for low PA. Additional insights are needed into factors determining young adolescents’ food and PA choices to develop the intervention. Focus groups with adolescents, parents and school staff were conducted for this purpose. Financial autonomy, perceived food safety, lack of self-control, habit strength, changes in both the socio-cultural (increased workload, changed food patterns, new transport modes) and the built environment (traffic and crime perceptions, distances) were identified as culture-specific factors. Measures that recognise environmental changes are thus needed to complement health education activities. Interestingly, various factors differed between settings and socio-economic groups. As a consequence, two composite conceptual frameworks were proposed. Both conceptualised adolescent eating or PA behaviour as a function of the identified individual and environmental influences. The usefulness of a multi-level, interactive framework for understanding and explaining the drivers of dietary behaviour was also evaluated. The framework is a comprehensive and valid model that specified the inter-relationships of individual and environmental factors and their influence on key components of adolescents’ eating behaviour (i.e., sugary drink intake, breakfast intake, unhealthy snacking, and fruit and vegetable intake). The framework confirmed that factors varied with reference to the different components of eating behaviour and socio-economic status (SES) and provides valuable entry points for developing future interventions and contributes to the evidence-base of theory development in LMICs. Based on the above and using theory, local evidence and a participatory approach, a school-based health promotion intervention was developed. It involved a comprehensive, culturally-appropriate intervention package, with intervention strategies consisting of an individual classroom-based and environment-based (school and family level) component. A pair-matched cluster randomised-controlled trial was conducted in 20 schools (1430 adolescents) to assess effectiveness. An in-depth process evaluation was also carried out. Primary outcomes assessed were dietary and PA behaviour and their influencing factors. Secondary outcomes were anthropometric measurements including BMI and the prevalence of overweight and obesity in the adolescents. The transparent and stepwise planning provides clear and detailed insights into the processes of intervention development in LMICs. This will help identify both effective and ineffective intervention strategies, and subsequently allow for the replication, adoption and/or dissemination of the identified strategies in LMICs. In conclusion, this PhD research adds to the available evidence on health promotion interventions in the wider context of obesity and chronic diseases in LMICs. This work has resulted in further insights into the current evidence-base for preventive measures in schools in LMICs, validation studies for PA measures, and the identification and conceptualisation of influencing factors of adolescents’ dietary and PA behaviours. It also delivers detailed information on the systematic processes associated with intervention development through the use of local evidence, theory, and participation. Lastly, it provides a niche exploration of further implications in the progress of understanding and implementing school-based interventions in LMICs.