(PS1-D96) Predictors of ADHD Symptom Trajectories During Late Childhood and Early Adolescence Using the Adolescent Brain Cognitive Development (ABCD) Dataset

Attention deficit hyperactivity disorder (ADHD) is one of the most common childhood mental health disorders, affecting approximately 7% of youth (Thomas et al., 2015), and is associated with significant and serious impairment across domains (Nigg et al., 2020). While considered a chronic condition, previous research has found that ADHD symptom severity fluctuates intraindividually, with persistent, increasing, or decreasing symptoms observed over time (Murray et al., 2021), though studies have conflicting findings regarding the number and characterization of total trajectories (e.g., Brinksma et al., 2021; Sasser et al., 2016; Tandon et al., 2016). Early adolescence is a crucial period for variability in symptom expression (Shaw & Sudre, 2021), and symptom pattern is linked to later outcomes, such as more impaired social and occupational functioning when more symptoms persist (Agnew-Blais et al., 2016). Therefore, determining predictors of symptom trajectory is crucial for identifying individuals likely to follow a symptom course with negative consequences.
 

Of the few studies that have examined factors predicting differing symptom trajectories over time, most focus on demographic factors (e.g., race/ethnicity, sex, SES) that, while important, are not linked to the underlying causes of the disorder. As currently conceptualized, ADHD is a neurodevelopmental disorder, and executive functions (EFs) are theorized to underlie ADHD symptoms and impairments (Willcutt et al., 2005). Since EFs are heterogeneously impacted in ADHD yet implicated in ADHD expression, EFs are likely to be meaningful predictors of ADHD symptom trajectories. However, no study to date has examined this question.
 

The NIH Adolescent Brain Cognitive Development (ABCD) study is a large-scale (N= 11,878) representative, population based, longitudinal study. ADHD symptoms were assessed four times annually starting between ages 8 and 10, and all youth meeting DSM criteria for ADHD at any study timepoint will be included in the present analysis. Growth mixture modeling will identify ADHD symptom trajectories. Three trajectories of symptoms are predicted based on extant literature in community samples (Brinksma et al, 2021; Riglin et al., 2016), so models with two, three, and four trajectories will be fit based on current best practices of fitting one more trajectory than expected. Bayesian Information Criterion (BIC) scores for each model will be compared to examine model fit and parsimony, and Bootstrap Likelihood Ratio Tests will be used to compare the models with the lowest BIC scores. A composite score of baseline EFs (working memory, inhibition, set-shifting from the NIH Toolbox) created using Principal Component Analysis will be examined as a predictor of trajectory using multinomial logistic regression. More impaired EF is hypothesized to predict more severe symptom trajectories. As this study is being conducted as part of a thesis prospectus, analyses have yet to be run. However, all data will be analyzed and the thesis is expected to be fully defended prior to the 2023 ABCT conference. This study has the potential to identify individuals with ADHD at highest risk, allowing for efficient allocation of intervention resources to those with the greatest need.