The Coronary Artery Risk Development in Young Adults (CARDIA) study is to identify the risk factors in early life that have influence on the development of clinical CVD in later life. Longitudinal study of 5115 subjects started in 1985-86 in 4 study centers in the US includes various clinical and physical measurements and in-depth questionnaires about sociodemographic background like psychosocial issues, smoking, drinking habits etc. We in our work used only basic socio-demographic information and health behaviour information for early prediction of incidence of CAC-levels using interpretative Dynamic Bayesian Networks. Our results demonstrate the superiority of behavioral features over other clinical measurements in the prediction task.