Background <p>Adolescence is a critical developmental period characterized by significant challenges, a time of “storm and stress”. The complex system of multidimensional mental health problems may vary across early, middle, and late adolescence due to the distinctive developmental tasks. However, previous efforts to identify prevention and treatment targets mainly focused on limited aspects of such a system and lacked age-specific considerations.</p> Objective <p>This study aims to identify pivotal age-tailored treatment and prevention targets for multidimensional mental health problems through dynamic simulation.</p> Methods <p>Three stage-specific between-subject networks were constructed to model mental health problems, using data from 80,113 Chinese adolescents. In silico intervention was implemented on these networks to identify particular symptoms that, if targeted, would prove most effective in either preventing (via aggravating simulation) or treating (via alleviating simulation) mental health issues within the mental health system.</p> Results <p>Stage-specific in silico intervention analysis revealed that, during early adolescence, academic stress and interpersonal sensitivity were the most effective targets in aggravating and alleviating interventions, respectively. In middle and late adolescence, anxiety emerged as the most significant target for aggravation, while emotional disturbance demonstrated the greatest alleviating effect.</p> Conclusions and implications <p>Academic stress experienced during early adolescence and anxiety prevalent in middle to late adolescence are recommended as primary prevention targets. Interpersonal sensitivity emerges as the most effective treatment target for early adolescents, whereas emotional disturbance should be prioritized for middle to late adolescents.</p>