posted on 2020-05-28, 08:21authored bySophie D. Walsh, Tal Sela, Margaretha De Looze, Wendy Craig, Alina Cosma, Yossi Harel-Fisch, Meyran Boniel-Nissim, Marta Malinowska-Cieslik, Alessio Vieno, Michal Molcho, Kwok W. Ng, William Pickett
Purpose: Adolescents' mental well-being has become a growing public health concern. Adolescents' daily lives and their engagement in risks have changed dramatically in the course of the 21st century, leading to a need to update traditional models of risk to include new exposures and behaviors. To date, studies have examined the relationship between (mainly traditional) risk behaviors and adolescent mental well-being or looked at risk factors that jeopardize mental wellbeing such as lack of social support but have not combined them together to highlight the most significant risks for adolescent mental well-being today. The present study included new and traditional risk behaviors and risk factors, robustly derived an empirically based model of clusters of risk, and examined the relative association of these clusters to adolescent mental well-being. Methods: Data from the 2017e2018 Health Behaviours in School-aged Children study were used. The sample included 32,884 adolescents (51.7% girls) aged 15 years from 37 countries and regions.
The principal component analysis was used to determine the existence of clusters of risk, using 21 items related to adolescent mental well-being that included both risk behaviors (e.g., substance use) and risk factors (e.g., peer support). Analysis was conducted in both a randomly split training and test set and in gender separate models. Mixed-effects logistic regressions examined the association between clusters of risk and mental well-being indices (low life satisfaction and psychosomatic complaints
Funding
Development of a structure identification methodology for nonlinear dynamic systems