Harnessing Network Science to Personalize Scalable Interventions for Adolescent Depression

点击翻译按钮获取中文摘要

Project Summary/Abstract Major depression (MD) is the leading cause of disability in youth, with a global economic burden of >$210 billion annually. However, up to 70% of youth with MD do not receive services. Even among those who do access treatment, 30-65% fail to respond, demonstrating a need for more potent, accessible interventions. A challenge underlying limited treatment potency is MD's heterogeneity: An MD diagnosis reflects >1400 symptom combinations, creating a need for treatments matched to personal clinical need. Separately, low treatment accessibility stems from the structure of existing interventions. Most span many weeks and are designed for delivery by highly trained clinicians , making them difficult to scale. This proposal aims to address the need for accessible, potent youth MD interventions by integrating methods and findings from previously separate areas: single-session intervention (SSI) research and network science. In a meta-analysis of 50 randomized trials, the investigator has found that SSIs can reduce diverse youth psychiatric problems, including MD. The investigator also found that a web-based SSI teaching growth mindset (the belief that personal traits are malleable) reduced depression and anxiety in high-symptom youth across 9 months. Thus, well-targeted SSIs can yield lasting benefits—but given MD's heterogeneity, there is a need for tools that can match youth to SSIs optimized for personal symptom structures. The proposed project harnesses computational advances from the network approach to psychopathology, which views psychiatric disorders as causal interactions between symptoms, to evaluate such a tool. The first goal is to establish a new method of characterizing MD symptom structures; the second is to test parameters from these structures as predictors of response to two SSIs targeting distinct MD features (behavioral vs. cognitive symptoms). Specifically, Aim 1 is to establish guidelines for computing personalized symptom networks using experience sampling method (ESM) data from youth with MD collected 7x/day for 3 weeks (N=50, ages 11-16; 147 time-points each). This will include a comparison of two leading approaches for computing network parameters, such as outward centrality (the degree to which a symptom prospectively predicts other symptoms). Aim 2 is to test network parameters as SSI outcome predictors among youth with MD (N=180). Youth will be randomized to a behavioral activation (BA) SSI (adapted from evidence-based BA SSIs); the mindset SSI noted above; or a control SSI. Network parameters will be tested as predictors of SSI response. For instance, youth with stronger centrality on a behavioral symptom (e.g. withdrawal from pleasurable activities) may respond more favorably to the BA SSI, and youth with stronger centrality on a cognitive symptom (e.g. hopelessness) to the mindset SSI. Results may identify a novel means of matching youth to targeted MD SSIs by personal need. The project will also include the first RCT comparing two youth MD SSIs, with the longest follow-up of any SSI trial to date (2 years), gauging their relative promise to reduce youth MD.