A Rapid Artificial Intelligence Assisted Umbrella Review of the Behavioural Correlates of Excess Weight in Children and Adolescents

Background: Behavioural interventions, commonly implemented to prevent and/or manage childhood obesity, are not successfully reducing obesity prevalence. This could be the result of a narrow focus on obesity-promoting behaviours, alongside wider system-level influences. This review aimed to identify behavioural correlates of excess weight in children/adolescents aged 0-19 years, and their parents/caregivers, to better inform future behavioural intervention development. Methods: Rapid umbrella review methodology was used. PubMed, MEDLINE, and PsycINFO were searched on 22 ^nd April 2025, yielding 2,065 studies to be screened. Data were extracted and verified by both ChatGPT4.0 and human researchers. Extracted variables included behavioural correlates associated with weight, measurement of behaviour, and, where available, Equality, Diversity, Inclusivity and Intersectionality (EDII) components. Results: Evidence from 157 systematic reviews (including 45 meta-analyses) published between 2008-2025 identified 103 behavioural correlates. Child/adolescent behaviours consistently associated with excess weight included frequent consumption of sugar-sweetened beverages, ultra-processed foods and energy-dense snacks; breakfast skipping; insufficient physical activity; greater sedentary behaviour; frequent screen use; and later bedtimes. Parent/caregiver behaviours included prenatal smoking (maternal and paternal); formula feeding; earlier introduction of solid foods; restrictive or controlling feeding practices; and nonresponsive parenting style. Definitions and measurements of behaviours across reviews were wide-ranging and inconsistent. Conclusions: This review highlights the myriad behaviours occurring across the life-course that may contribute to child/adolescent obesity rates, reinforcing the complexity of obesity-promoting behaviours across varying life-stages. It complements existing evidence that excess weight is shaped by interacting behavioural systems involving children/adolescents and their parents/caregivers. Multi-level approaches are required across the wider system to address child/adolescent obesity. This review provides an important synthesis of behavioural correlates of excess weight to inform the development of future behavioural interventions, highlights areas for behavioural data harmonisation to improve measurement of behaviour and calls for improvements in the consideration and reporting of EDII characteristics in research.