Reward-network connectivity in childhood predicts multi-domain dysregulation in adolescence

Multi-domain dysregulation in adolescence, indexed by co-occurring affective, cognitive, and behavioural difficulties, is a robust transdiagnostic risk factor. However, its developmental course and neural antecedents are poorly understood. Given heightened emotional reactivity and impulsivity in adolescence, alterations in reward-network connectivity may represent an early neural marker of risk.

Adolescents completed four assessments approximately two years apart between ages 9-13 and 15-18 years. Multi-domain dysregulation was assessed at each wave using the Youth Self-Report Dysregulation Profile (YSR-DP), computed as the sum of the anxious/depressed, aggressive behaviour, and attention problems subscales. Resting-state fMRI was acquired at baseline (M_age = 11.34 years). Piecewise linear mixed-effects models (N = 211) characterized trajectories of YSR-DP scores across adolescence. Principal component scores indexing a Latent Dysregulation Factor were used to derive residualised change in dysregulation, and regression analyses (N = 94) tested whether baseline reward-network connectivity predicted this change.

YSR-DP scores declined from late childhood to early adolescence, increased from early to mid-adolescence, and then stabilized in late adolescence. Weaker connectivity within the reward network in late childhood predicted greater increases in the latent dysregulation factor from early to mid-adolescence, above and beyond baseline dysregulation. Connectivity in seven large-scale control networks did not predict changes in dysregulation.

Multi-domain dysregulation follows a nonlinear trajectory across adolescence, and weaker reward-network connectivity in childhood prospectively predicts subsequent escalation of this phenotype. Prevention and intervention efforts may benefit from targeting reward processing and regulatory skills in late childhood and early adolescence.