Predictive Modelling of Cardiometabolic Outcomes and Treatment Discontinuity in ADHD Patients

Given the substantial comorbidity and shared genetics between ADHD and cardiometabolic diseases, appropriate medication management of adults with ADHD is a crucial component of clinical managment of co-occurring cardiometabolic disease. We hypothesize that the discountinuation of the ADHD medication is associated with increased risks and worsened clinical outcomes of the cardiometabolic diseases in adults with ADHD. Many factors, socio-demograhic, behavioral and biological/clinical variables, can modify the patients’ adhearance to ADHD treatment and their subsequent risks of cardiometabilic conditions. Our planned study will utilize state-of-the-art multivariate machine learning methods to clarify the predictive relationship of these factors and the treatment discountinuation and cardiometabolic outcomes in the subjects with ADHD. Specificaly, we will use health-related data and measurements obtained from the Lifeline cohort to predict: 1) the adherence to ADHD treatment and the risk of  discountinuation in ADHD subjects; 2) and their risk of developing cardiometabolic conditions and adverse clinical outcomes later in life.

We are aware that study 2 can be addressed in the lifelines data while study 1 needs to be done within the CBS environment. We are currently exploring the latter possibilities with CBS. We will start with project 2 and proceed with project 1 If feasible.