ASSESSING SYMPTOM CHANGE OF NEURODEVELOPMENTAL AND PSYCHOPATHOLOGY FACTORS USING LATENT GROWTH CURVE ANALYSIS IN THE ABCD STUDY

Diagnostic models such as the hierarchical taxonomy of psychopathology (HiTOP) are gaining notice amongst the psychopathology research community due to issues with traditional classification methods (Kotov et al., 2021). Yet, the placement of neurodevelopmental symptoms in HiTOP remains unclear, partly due to a lack of longitudinal research on their stability, and of their distinction from externalising symptoms. Using data from the Adolescent Brain Cognitive Development (ABCD) study, we conducted latent growth curve analysis to assess the comparable change of neurodevelopmental symptoms and other latent factors across pre-adolescence, using factor structure determined by Michelini and colleagues (2019). We predicted that neurodevelopmental symptoms would show the lowest improvement of model fit when constrained to represent change across time. We expected that this would also be reflected in greater associations with birth weight, due to functional developmental differences, in line with previous findings in this sample (Dooley et al., 2022). Findings indicated that while the specification of a slope (constrained change of symptom severity) for the neurodevelopmental factor fit the data less well than a slope for the internalising and externalising factors, these differences were only moderate. After controlling for covariates, we found no relationship to be significant between birth weight and any psychopathology domain, with pre-term birth appearing to be a strong and unique predictor of neurodevelopmental symptoms, and household income being the strongest predictor of internalising and externalising symptoms. These findings implicate both the distinctiveness of neurodevelopmental symptoms and the effect of the p factor, which represents co-occurrence of symptoms in general. Future research should continue to investigate the placement of neurodevelopmental symptoms within psychopathology factor models.