 This study explores the potential of Bluetooth device count, NBDC, data to predict depressive symptom severity using linear mixed effects and hierarchical Bayesian linear regression models. The results indicate that NBDC data can reflect changes in individuals' behaviors and statuses concurrent with changes in their depressive state and have a significant value in predicting depressive symptom severity. This article was authored by Yuizhou Zhang, Amos A. Fullerin, Shaoxiong Sun, and others.