Evaluating Hypothetical Prevention Strategies for Internalizing Symptoms in the General Population and At-Risk Children

Objective: 

Specific modifiable factors (e.g., screen time [ST], sleep duration, physical activity, or social connections) are targets for reducing depression risk in adults. However, research in adolescents lacks causal inference implementations, as prevention trials are costly and often prohibitive. Emulating randomized trials with observational data enables inference regarding hypothetical interventions on modifiable factors that reduce depression risk, in general and at-risk populations. 

Method: 

Data from the Adolescent Brain Cognitive Development (N = 8,699) and the Generation R (N = 3,739) studies were leveraged for modifiable factors (age 10) and internalizing symptoms (age 12–14). We (a) tested prospective associations of each modifiable factor with internalizing symptoms under target trial emulation (TTE) and (b) used G-formula to estimate the potential effects on internalizing symptoms if hypothetical interventions were implemented. Analyses were conducted in the full sample (universal prevention setting) and at-risk youth (selective and indicated prevention setting).

Results: 

First, under TTE, only ST associated with internalizing symptoms. Second, in the universal prevention setting, adhering to more stringent ST guidelines (0–1 hr) decreased internalizing symptoms by 0.10–0.16 SDs while following more lenient guidelines (3–4 hr) increased them by 0.07–0.09 SDs, compared to existing guidelines (2 hr). Such changes were greater in the indicated (subclinical symptom) prevention settings and robust to residual confounding. Restricting physical activity reduced internalizing risk in the indicated setting, but findings were not robust. 

Conclusions: 

With a causally informed strategy, we demonstrated the potential effects of numerous hypothetical interventions on modifiable factors for depression risk reduction, across multiple preventive settings.