Predicting Depression Risk Using Logistic Regression: A Case Study Based on Academic and Lifestyle Factors

Xianwen Jiang

doi:10.54097/z9j20j37

Authors

Xianwen Jiang

DOI:

https://doi.org/10.54097/z9j20j37

Keywords:

Depression Prediction, Logistic Regression, Mental Health Screening, Academic and Lifestyle Factors, Machine Learning.

Abstract

Depression has increasingly become a serious public health issue, with students and working professionals often among the most affected. Academic demands and financial challenges have been identified as key stressors, and recent global surveys show a noticeable rise in depression rates among young people, partly due to heavier workloads and reduced social support. In this study, logistic regression is applied to perform binary classification of depressive tendencies using the publicly available “Playground Series S4E11” dataset from Kaggle. This dataset contains demographic details along with academic and psychological indicators. The analysis followed a structured process—covering data cleaning, feature transformation, and training of the logistic regression model. Evaluation of the model using accuracy, recall, and Receiver Operating Characteristic-Area Under Curve (ROC-AUC) produced scores of 84%, 88%, and 0.913, respectively. Among the input variables, suicidal thoughts, academic or work pressure, and financial stress were identified as the most influential indicators. These findings indicate that even relatively simple and transparent models can serve as effective tools for the early identification of depression. When incorporated into digital systems, such approaches could help deliver timely support within educational institutions and healthcare environments.

References

[1] World Health Organization: “Depression: Key facts”, https://www.who.int/news-room/fact-sheets/detail/depression, accessed 6 August 2025.

[2] Q. Qiang, J. Hu, X. Chen, et al. “Identifying risk factors for depression and positive/negative mood changes in college students using machine learning,” Front. Public Health, 2025, 13, 1606947.

[3] C. Mimikou, C. Kokkotis, D. Tsiptsios, et al. “Explainable Machine Learning in the Prediction of Depression,” Diagnostics (Basel), 2025, 15, (11), 1412.

[4] M. A. Mamun, F. Al-Mamun, M. E. Hasan, et al. “Predictive Modeling of Comorbid Depression and Anxiety Symptoms Among Prospective University Students: A GIS-Based and Machine Learning Study,” Psychol. Rep., 2025, Advance online publication.

[5] J. Massell, M. Preisig, M. Miché, et al. “Prospective prediction of first onset of major depressive disorder in midlife using machine learning”, Soc. Psychiatry Psychiatr. Epidemiol., 2025, Advance online publication.

[6] F. T. Cruz, E. E. Flores, S. J. Quispe, “Prediction of depression status in college students using a Naive Bayes classifier based machine learning model,” Proc. Int. Conf. on Emerging Trends in Computing and Information Technology, Lima, Peru, March 2023, pp. 87–92

[7] Y. Yang, L. Wen, L. Li, “Explainable AI for time series prediction in economic mental health analysis,” Front. Med., 2025, 12, 1591793.

[8] L. Meng, Q. Zhu, F. Ma, et al. “Logistic regression analysis of risk factors for anxiety and depression in patients with coronary heart disease and subclinical hypothyroidism,” Sci. Rep., 2024, 14, (1), pp. 27781.

[9] P. Prakrankamanant, S. Watanabe, E. Chuang, “Explainable Depression Detection using Masked Hard Instance Mining,” arXiv preprint, 2025, abs/2505.24609

[10] E. Kerz, S. Zanwar, Y. Qiao, et al. “Toward explainable AI (XAI) for mental health detection based on language behavior,” Front. Psychiatry, 2023, 14, 1219479.

[11] L. C. López Steinmetz, M. Sison, R. Zhumagambetov, et al. “Machine learning models predict the emergence of depression in Argentinean college students during periods of COVID-19 quarantine,” Front. Psychiatry, 2024, 15, 1376784.