Family Related Variables’ Influences on Adolescents’ Health Based on Health Behaviour in School-Aged Children Database, an AI-Assisted Scoping Review, and Narrative Synthesis

Objects: Health Behaviours in School-aged Children (HBSC) is an international survey programme aiming to investigate adolescents’ health behaviours, subjective perception of health status, wellbeing, and the related contextual information. Our scoping review aimed to synthesise the evidence from HBSC about the relationship between family environmental contributors and adolescents’ health-related outcomes. Methods: We searched previous studies from six electronic databases. Two researchers identified the qualified publications independently by abstract and full-text screening with the assistance of an NLP-based AI instrument, ASReview. Publications were included if they were based on HBSC data and investigated the effects of family environment on adolescents’ health outcomes. Researches addressed family-related factors as mediators or moderators were also included. Results: A total of 241 articles were included. Family environmental contributors could be mapped into six categories: (1) Demographic backgrounds (N = 177); (2) General family’s psycho-socio functions (N = 44); (3) Parenting behaviours (N = 100); (4) Parental health behaviours (N = 7); (5) Family activities (N = 24); and (6) Siblings (N = 7). Except for 75 papers that assessed family variables as moderators (N = 70) and mediators (N = 7), the others suggested family environment was an independent variable. Only five studies employed the data-driven approach. Conclusion: Our results suggest most research studies focussed on the influences of family demographic backgrounds on adolescents’ health. The researches related to parental health behaviours and siblings are most inadequate. Besides, we recommend further research studies to focus on the mediator/moderator roles of the family, for exploring the deep mechanism of the family’s impacts. Also, it would be valuable to consider data-driven analysis more in the future, as HBSC has mass variables and data.