SummaryThis project aims to develop psychosocially enhanced cardiovascular disease (CVD) risk prediction models using longitudinal data from the German National Cohort (NAKO). Despite strong evidence that psychosocial factors, including depression, stress, social isolation, and exhaustion, predict CVD events as powerfully as traditional biomedical risk factors, they remain entirely absent from clinical risk algorithms used in routine practice. Using NAKO data, this project will integrate psychosocial, behavioural, and biological variables into a unified prediction framework combining penalised Cox regression and machine-learning survival methods. The project duration is 36 months. The expected public health impact is improved identification of high-risk individuals currently overlooked by biomedical-only risk models, supporting more targeted and equitable cardiovascular prevention.

Keywords Cardiovascular-risk-prediction Depression machine-learning psychosocial-factor

InstitutionsHelmholtz Zentrum München