Brigham Research Institute Poster Session Site logo-1

Grace Jacobs, PhD




Research Fellow




Postdoctoral Research Fellow




Grace R. Jacobs*, Stephanie H. Ameis, Peter Szatmari, Aristotle N. Voineskos, John D. Haltigan

Bifactor models of psychopathology using multi-informant and multi-instrument dimensional measures in the ABCD Study

I recently began as a postdoctoral research fellow in the Psychiatry Neuroimaging Laboratory with Dr. Martha Shenton. My research focuses on using multimodal neuroimaging and multivariate approaches to investigate behavioral and neurobiological risk factors involved in early psychosis symptoms and overlapping mental illnesses. Attending and participating in the Women in Medicine and Science Symposium would be a great opportunity to share my research with this new community and become more familiar with other exciting research happening at BWH. It will also help tremendously with building my research network and developing important collaborations, which are pivotal for my career aspirations.

Background: Current categorical mental illness diagnoses have limitations and there is a need for empirical investigations of continuous dimensions of psychopathology. Studies to date often focus on a subset of measures and neglect key symptoms relevant to severe illnesses. This study aimed to comprehensively examine widely-representative psychopathology structure in children by including detailed measures of psychotic-like experiences (PLEs), autism spectrum disorder symptoms (ASD), impulsivity, and sensitivity to reward and punishment.

Methods: We included ten child-report and parent-report instruments capturing diverse psychopathology in 11,185 children aged 9-10 from the ABCD Study. Using Mplus v7.11, two novel modeling approaches were employed separately for informants. Model fit was evaluated based on CFI, TLI, RMSEA, and conceptual clarity. Dimension score differences in sex, age, cognition, imaging measures, and medical service usage were assessed.

Results: All four models showed excellent fit, similar factor structure within informant, and factor score differences in external risk measures. PLEs loaded most highly onto a general p factor. In addition to internalizing and externalizing factors, measures of ASD symptoms, impulsivity, and sensitivity to reward loaded onto separate factors.

Conclusions: Findings highlight both novel distinctions and shared variance between diverse psychopathology symptoms and identify indicators of specific versus general risk in children.