Investigating Potential Risk Factors for Cardiovascular Diseases in Adult Qatari Population

Cardiovascular diseases (CVD) are one of the leading causes of mortality across the globe. In order to investigate the potential risk factors that relate to the cause of CVD in the Qatari population, this study investigated different kinds of biomarkers including (i) Vital Biomarkers, (ii) Bio-impedance, (iii) Spirometry and (iv) Vicorder readings. In order to investigate the prospective biomarkers, this study was conducted on 471 subjects comprised of 221 CVD patients and 250 normal participants forming the control group. Several machine learning models were trained using different combinations of biomarkers to distinguish the healthy subjects from the CVD subjects. Our analysis reveals the decision tree-based classifier as the best performing model, with high accuracy, among all the classifiers to distinguish these two groups. The outcome from the ablation study on different kinds of features reveals that bio-impedance measurements can be considered as the most influential risk factors in distinguishing the healthy subjects from CVD subjects. Furthermore, the combination of different Vitals with bio-impedance measures enhances the discriminatory power of the machine learning models.