A Machine Learning Model for Predicting the Risk of Developing Diabetes - T2DM Using Real-World Data from Kilifi, Kenya

Type 2 Diabetes Mellitus (T2DM) is a growing public health concern in low-resource settings, where early detection remains limited due to infrastructural and diagnostic constraints. This study presents a machine learning-based risk prediction model developed using real-world data from Kilifi County Referral Hospital in Kenya, aiming to identify individuals at risk of developing T2DM before clinical onset. The study applied the CRISP-DM framework to guide the end-to-end process, from data collection to model deployment. A dataset comprising 2,500 anonymized electronic health records was used, incorporating a diverse range of features including clinical, behavioral, demographic, and socioeconomic variables. Feature selection was conducted using both statistical (Chi-square test) and algorithm-based methods (Random Forest, Recursive Feature Elimination, and XGBoost importance), resulting in two candidate feature sets (14-feature and 7-feature subsets). Four supervised learning algorithms; Logistic Regression, Support Vector Machine (SVM), Random Forest, and XGBoost were trained and evaluated using 5-fold cross-validation. Among them, the XGBoost model achieved the best performance, with a test set accuracy of 91.33%, F1-score of 88.66%, and an AUC-ROC of 96.24%, outperforming other models across all metrics. This study demonstrates that integrating multi-domain features with machine learning can enhance early risk stratification for T2DM in under-resourced environments. The final model’s ability to categorize individuals into low, medium, and high-risk groups offers a practical tool for targeted screening and preventive healthcare interventions in Kenyan public health systems.

Type 2 Diabetes Mellitus, Machine Learning, Risk Prediction, XGBoost, CRISP-DM, Kenya

1. Bhargava, S., & Zafar, S. (2019). Socioeconomic and behavioral predictors in diabetes risk: An ML-based population health study in Pakistan. Journal of Public Health Research, 8(3), 164–170. https://doi.org/10.4081/jphr.2019.164 [Google Scholar] [Crossref]

2. Chen, H., et al. (2021). Using real-world data from rural China to predict diabetes risk via ensemble learning. BMC Endocrine Disorders, 21, 198. https://doi.org/10.1186/s12902-021-00870-w [Google Scholar] [Crossref]

3. Deberneh, H. M., & Kim, I. (2021). Prediction of type 2 diabetes based on machine learning algorithm. International Journal of Environmental Research and Public Health, 18(6), 3317. https://doi.org/10.3390/ijerph18063317 [Google Scholar] [Crossref]

4. Farran, B., et al. (2022). An explainable machine learning approach to early T2DM prediction in Qatar’s primary care. BMC Medical Informatics and Decision Making, 22, 183. https://doi.org/10.1186/s12911-022-01948-6 [Google Scholar] [Crossref]

5. International Diabetes Federation. (2021). IDF Diabetes Atlas (10th ed.). Brussels, Belgium: International Diabetes Federation. https://diabetesatlas.org/ [Google Scholar] [Crossref]

6. Islam, S. M. S., et al. (2022). Development of a non-invasive diabetes prediction tool using behavioral and anthropometric data in rural Bangladesh. Scientific Reports, 12, 14378. https://doi.org/10.1038/s41598-022-18022-6 [Google Scholar] [Crossref]

7. Kavakiotis, I., Tsave, O., Salifoglou, A., Maglaveras, N., Vlahavas, I., & Chouvarda, I. (2017). Machine learning and data mining methods in diabetes research. Computational and Structural Biotechnology Journal, 15, 104–116. https://doi.org/10.1016/j.csbj.2016.12.005 [Google Scholar] [Crossref]

8. Lee, S., et al. (2021). Diabetes risk classification with explainable ML: Application in underserved Korean population. PLoS ONE, 16(6), e0253312. https://doi.org/10.1371/journal.pone.0253312 [Google Scholar] [Crossref]

9. Mohan, V., et al. (2019). A deep learning model for diabetes prediction using Indian rural cohort. Diabetes Technology & Therapeutics, 21(10), 562–569. https://doi.org/10.1089/dia.2019.0172 [Google Scholar] [Crossref]

10. Nguyen, Q. C., et al. (2020). Leveraging social determinants and EHR data to predict diabetes risk in underserved populations. International Journal of Medical Informatics, 141, 104241. https://doi.org/10.1016/j.ijmedinf.2020.104241 [Google Scholar] [Crossref]

11. Rahman, M. M., et al. (2020). T2DM risk assessment using random forest and decision tree in community health datasets. Informatics in Medicine Unlocked, 21, 100461. https://doi.org/10.1016/j.imu.2020.100461 [Google Scholar] [Crossref]

12. Wang, F., & Hu, J. (2019). Predicting chronic disease risk using machine learning on health survey data: A case study on diabetes. IEEE Journal of Biomedical and Health Informatics, 23(6), 2548–2556. https://doi.org/10.1109/JBHI.2018.2887383 [Google Scholar] [Crossref]