Machine Learning for the Identification of Biomarker and Risk Factors associated with Depression in Adult Population: Preliminary Results on a Small Cohort

In this study we identify some risk factors for depression from the sociodemographic factors and inflammatory biomarkers from a sample of depressed subjects versus a matched group of healthy subjects. The multivariate logistic regression and random forest are used to identify the most significant factors associated with depression. We found significant differences demographic variables corresponding to education, higher levels of education seem to provide protection. Cognitive impairment (SCIP) positive results were also strongly associated with depression. The study also confirmed that married subjects suffer less of depression than separated or divorced subjects. Regarding inflammatory biomarkers, significant differences were obtained in IL6, TNFa, CAT, BDNF, and GSH Total, which indicate a strong association of depression and systemic inflamatory processes.