Using network structure to better understand associations between adversities and child development outcomes

• In Event: 2-147 - Is there more to childhood adversity than cumulative stress? Distinguishing threat/ deprivation in large-scale studies

Integrative Statement

Exposure to adversity in childhood is common and associated with negative outcomes in childhood and adulthood. The current most common approach used to examine the consequences of adversity exposure for health and well-being is a cumulative risk model where all forms of adversity are considered equipotent and likely to impact outcomes through the same pathway. Recently, we have proposed a novel approach, the dimensional model of adversity and psychopathology (DMAP), where different dimensions of adversity are hypothesized to impact health and well-being outcomes through different pathways. Specifically, we expect deprivation or a lack of rich caregiving inputs to selectively disrupt cognitive processing whereas we expect threat or exposure to interpersonal violence to selectively disrupt emotional processing. Recent hypothesis driven approaches have confirmed that deprivation and threat have these selective impacts and that these pathways increase risk for negative health outcomes. However, this work has not answered a fundamental question: are deprivation and threat observable using data-driven approaches, or are they only observeable when a hypothesized structure is forced on the data. Here we present a network analysis in an initial study (Study 1: N = 277 adolescents aged 16–17 years; 55.1% female) and a replication (Study 2: 262 children aged 8-16 years; 45.4% female) designed to address this question. Both samples were recruited from the community but included targeted recruitment to identify at-risk children and adolescents. As a result, both studies included a significant number of children with parents who had completed high school or less than high school (Study 1: 16.7%; Study 2: 41.9%) and maltreatment exposure (Study 1: 25.1%; Study 2: 38%). In Study 1, children reported on all variables. In Study 2, children and parents served as reporters. In both samples we observed a network structure consistent with the DMAP model (Figures 1 and 2) and further demonstrate that our observed network is statistically different than a hypothetical cumulative risk network using bootstrap resampling procedures. In sum, we believe that we have good evidence that this approach reveals organization within the data consistent with DMAP and inconsistent with the cumulative risk model. Future work interested in the pathways through which adversity comes to impact well-being should take multiple dimensions of adversity into account.

Authors

• Margaret Sheridan, University of North Carolina at Chapel Hill
  Presenting Author

• Feng Bill Shi, University of North Carolina, Chapel Hill
  Non-Presenting Author

• Adam Miller, University of North Carolina at Chapel Hill
  Non-Presenting Author

• Carmel Salhi, Bouvé College of Health Sciences, Northeastern University
  Non-Presenting Author

• Katie McLaughlin, Harvard University
  Non-Presenting Author