Data Modeling
Mathematical and Computational Data Modeling
0
Citations
937
Views
3
Articles
Volume 7 Number 1 2026
RESEARCH ARTICLE (Open Access)
Early Prediction of Postprandial Glycemic Response in Gestational Diabetes Using Continuous Glucose Monitoring and Gradient Boosting Models
Akib Ullah Jafor 1*, Pronati Das Puja 2, Shawanti Biswas 3, Arko Saha 4, Kamruzzaman Mithu 1, Khondaker Abdullah-Al-Mamun 1
Data Modeling 7 (1) 1-8 https://doi.org/10.25163/data.7110742
Submitted: 26 January 2026 Revised: 10 April 2026 Accepted: 15 April 2026 Published: 18 April 2026
Abstract
There is, perhaps, an increasing realization that gestational diabetes mellitus (GDM) cannot be fully understood through static glucose measurements alone. Rather, it unfolds dynamically—particularly in response to meals—where postprandial glycemic fluctuations carry meaningful clinical implications. In this study, we aimed to develop a predictive modeling framework capable of estimating postprandial glycemic response (PPGR) using Continuous Glucose Monitoring (CGM) data integrated with meal and behavioral features. A dataset comprising 235 pregnant participants, including both GDM and healthy controls, was analyzed. CGM signals were synchronized with meal logs to derive clinically relevant outcome variables, including peak glucose (BGMax), glucose rise (BGRise), glucose at 60 minutes (BG60), and incremental area under the curve (iAUC120). Gradient boosting models were developed and evaluated using five-fold cross-validation and independent testing. The optimized model demonstrated strong predictive performance, achieving a coefficient of determination (R²) of 0.82 and a root mean squared error (RMSE) of 0.68 mmol/L for iAUC120 prediction (Table 3). Feature importance analysis revealed that prior meal composition, glycemic index, and temporal meal spacing were among the most influential predictors (Figure 3). Models incorporating contextual behavioral features consistently outperformed baseline physiological-only models. These findings suggest that integrating CGM data with contextual dietary information can substantially improve the prediction of postprandial glucose dynamics in GDM. While further validation is required, the proposed approach offers a step toward more personalized, data-driven management strategies in pregnancy-related metabolic care.
Keywords: Gestational diabetes mellitus; continuous glucose monitoring; postprandial glycemic response; machine learning; gradient boosting; iAUC120; predictive modeling; maternal health
References
Alfian, G., Syafrudin, M., Rhee, J., Anshari, M., Mustakim, M., & Fahrurrozi, I. (2020). Blood glucose prediction model for type 1 diabetes based on extreme gradient boosting. IOP Conference Series: Materials Science and Engineering, 803(1), 012012. https://doi.org/10.1088/1757-899X/803/1/012012
Gallwitz, B. (2009). Implications of postprandial glucose and weight control in people with type 2 diabetes: Understanding and implementing the International Diabetes Federation guidelines. Diabetes Care, 32(Suppl 2), S322–S325. https://doi.org/10.2337/dc09-S331
Gnanadass, I. (2020). Prediction of gestational diabetes by machine learning algorithms. IEEE Potentials, 39(6), 32–37. https://doi.org/10.1109/MPOT.2020.3015190
Oviedo, S., Vehí, J., Calm, R., & Armengol, J. (2017). A review of personalized blood glucose prediction strategies for type 1 diabetes. International Journal for Numerical Methods in Biomedical Engineering, 33(6), e2833. https://doi.org/10.1002/cnm.2833
Popova, P., Castorino, K., Grineva, E. N., & Kerr, D. (2016). Gestational diabetes mellitus diagnosis and treatment goals: Measurement and measures. Minerva Endocrinologica, 41(4), 421–432.
Pustozerov, E., Popova, P., Tkachuk, A., Bolotko, Y., Yuldashev, Z., & Grineva, E. (2018a). Development and evaluation of a mobile personalized blood glucose prediction system for patients with gestational diabetes mellitus. JMIR mHealth and uHealth, 6(1), e6. https://doi.org/10.2196/mhealth.9236
Pustozerov, E., & Popova, P. (2018b). Mobile-based decision support system for gestational diabetes mellitus. In Proceedings of the Ural Symposium on Biomedical Engineering, Radioelectronics and Information Technology (USBEREIT) (pp. 45–48). https://doi.org/10.1109/USBEREIT.2018.8384546
Pustozerov, E. A., Yuldashev, Z. M., Popova, P. V., Bolotko, Y. A., & Tkachuk, A. S. (2018c). Information support system for patients with gestational diabetes mellitus. Biomedical Engineering, 51(6), 407–410. https://doi.org/10.1007/s10527-018-9759-2
Pustozerov, E., Tkachuk, A., Vasukova, E., Dronova, A., Shilova, E., Anopova, A., Piven, F., Pervunina, T., Vasilieva, E., Grineva, E., & Popova, P. (2020). The role of glycemic index and glycemic load in the development of real-time postprandial glycemic response prediction models for patients with gestational diabetes. Nutrients, 12(2), 302. https://doi.org/10.3390/nu12020302
Popova, P., Vasilyeva, L., Tkachuk, A., Puzanov, M., Golovkin, A., Bolotko, Y., Pustozerov, E., Vasilyeva, E., Li, O., Zazerskaya, R., Dmitrieva, R., Kostareva, A., & Grineva, E. (2018). A randomized controlled study of different glycemic targets during gestational diabetes treatment: Effect on adipokines and ANGPTL4 expression. International Journal of Endocrinology, 2018, 1–8. https://doi.org/10.1155/2018/6481658
Wender-Ozegowska, E., Wróblewska, K., Zawiejska, A., Pietryga, M., Szczapa, J., & Biczysko, R. (2005). Threshold values of maternal blood glucose in early diabetic pregnancy—Prediction of fetal malformations. Acta Obstetricia et Gynecologica Scandinavica, 84(1), 17–25. https://doi.org/10.1111/j.0001-6349.2005.00606.x