Horizon Academic Research Program

My Horizon research project examines the relationship between chronic digestive symptoms and ADHD, anxiety, and depression in children and adolescents. The work has evolved from an IBD–ADHD question into a broader, more carefully defined study of symptom co-occurrence.

The revised proposal centers on pooled National Survey of Children’s Health public-use data from 2016–2023 for children aged 6–17. It compares reported digestive symptoms with parent-reported neurodevelopmental and mental-health outcomes. The dataset does not isolate diagnosed inflammatory bowel disease, so IBD is the motivation for the question rather than a diagnosis the study can claim to measure.

The analysis plan compares broad digestive-symptom exposure with a more persistent or impairing symptom definition, subject to the available survey variables. Proposed covariates include demographic, socioeconomic, household, healthcare-access, and other health factors; each outcome is modeled separately to avoid defining exposure through the outcome itself.

I am developing a framework that compares interpretable logistic and elastic-net baselines with random forests and gradient-boosted trees. Planned evaluation includes cross-validation, a later-year holdout, survey-weighted association estimates, AUROC and precision–recall measures, and probability calibration. SHAP would help explain model behavior, alongside checks across age, sex, and socioeconomic groups.

The central question is how robust the observed associations are across symptom definitions and modeling choices. Because the data are cross-sectional and parent-reported, the study is framed around association rather than causation or future disease prediction. These are the proposed methods, not a claim of completed results or publication.