T29. CHILDHOOD PSYCHOPATHOLOGY ACROSS 12 YEARS PREDICTS ADULT PSYCHOTIC-LIKE EXPERIENCES: A PARALLEL TWO-PART PIECEWISE LATENT GROWTH CURVE MODEL

Abstract

Background: The prevalence of childhood psychopathology fluctuates across the lifespan, yet studies often adopt linear growth curve models (LGM) of estimation that assumes constant linear growth and do not account for the comorbidity between internalizing and externalizing behaviors in predicting later psychosis. This study tests whether internalizing and externalizing behaviors in early childhood (4-9 years) and adolescence (11-16 years) are best modelled by a two-part parallel piecewise growth model (2-PGM) or a single LGM (4-16 years) and whether specific developmental periods better predict psychotic-like experiences (PLEs) in adulthood (18 years). Methods: Parent-rated child's psychopathology on the Strengths and Difficulties Questionnaire (Goodman, 2006) at ages 4, 6, 8, 9, 11, 13, and 16 years from the Avon Longitudinal Study of Parents and Children were first modelled by a parallel LGM, then a 2-PGM, to predict clinician-rated adult PLEs. Models were re-run controlling for confounds assessed prior to age 4 years (i.e., child's gender, verbal IQ, socioeconomic status, maternal education, prior diagnosis of mental health issues, and stressful life experiences at 42 months). Results: Considering internalizing and externalizing problem behaviors in tandem, a 2-PGM fit the data better than a LGM (CFI/TLI = .97/.96, 2(129) = 781.63, p < .001, RMSEA = .033, 90%CI[.031-.035], WRMR = 1.32, N = 4717). Controlling for confounds, internalizing symptoms at baseline (b = .130, p = .004) and changes (b = .196, p