Explainable Multimodal AI for Early Cardiometabolic Risk Assessment Using Continuous Glucose Monitoring and Wearable Data: A Narrative Review
Authors
Department of Computer Science and Engineering, Mar Baselios Christian College of Engineering and Technology, Peermade – 685531, Kerala (India)
Department of Computer Science and Engineering, Mar Baselios Christian College of Engineering and Technology, Peermade – 685531, Kerala (India)
Department of Computer Science and Engineering, Mar Baselios Christian College of Engineering and Technology, Peermade – 685531, Kerala (India)
Department of Computer Science and Engineering, Mar Baselios Christian College of Engineering and Technology, Peermade – 685531, Kerala (India)
Associate Professor, Department of Computer Science and Engineering, Mar Baselios Christian College of Engineering and Technology, Peermade – 685531, Kerala, India (India)
Article Information
DOI: 10.51583/IJLTEMAS.2026.151000145
Subject Category: 151J09MAS1187
Volume/Issue: 15/10 | Page No: 1789-1802
Publication Timeline
Submitted: 2026-10-31
Accepted: 2026-11-06
Published: 2026-11-20
Abstract
Cardiometabolic conditions, including type 2 diabetes mellitus (T2DM), obesity and hypertension, develop through complex interactions among metabolic, physiological and lifestyle factors. Continuous glucose monitoring (CGM) and wearable devices can capture time-resolved information such as glucose patterns, heart rate, physical activity and sleep. Artificial intelligence (AI) methods have been investigated for glucose forecasting, physiological-state estimation, disease screening, prospective risk stratification and management of established conditions. This narrative review examines seventeen primary studies and one systematic review relevant to early cardiometabolic risk assessment, with particular attention to multimodal data integration, temporal modelling, explainable AI (XAI) and personalised preventive guidance. The reviewed primary studies cover digital-twin-based diabetes management, insulin resistance, glucose prediction, multi-task disease prediction, prediabetes screening, explainable obesity assessment, wearable-based blood-pressure assessment and CGM-based long-term risk stratification. GluFormer demonstrates prospective risk stratification from CGM representations and a dietary-data extension. The remaining gap is narrower: the included evidence does not establish one externally validated framework that jointly assesses multiple cardiometabolic conditions using CGM and broader wearable/clinical modalities while providing validated explanations and evaluated personalised preventive guidance. The review discusses validation, label quality, privacy and ethical challenges. Such systems should be evaluated using participant-level data separation, external validation and uncertainty assessment, and their outputs should be communicated as non-diagnostic risk estimates rather than clinical diagnoses.
Keywords
narrative review, explainable AI, continuous glucose monitoring, wearable data, cardiometabolic risk
This article is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).
Downloads
References
1. Wang, J., Faruqui, S. H. A., Alaeddini, A., Du, Y., et al. (2026). Human-in-the-loop AI predictive digital twin to extend virtual precision diabetes care between visits. npj Health Systems, 3, Article 59. https://doi.org/10.1038/s44401-026-00118-8 [Google Scholar] [Crossref]
2. Esmailzadeh, A., Norouzkhani, N., Sezavar Dokhtfaroughi, S., Rasoulian, A., & Mazaheri Habibi, M. R. (2026). Application of artificial intelligence in the diagnosis, prediction, and management of metabolic syndrome: A systematic review. Health Science Reports, 9(6), e72636. https://doi.org/10.1002/hsr2.72636 [Google Scholar] [Crossref]
3. Metwally, A. A., Heydari, A. A., McDuff, D., Solot, A., Esmaeilpour, Z., Faranesh, A. Z., Zhou, M., Narayanswamy, G., Xu, M. A., Liu, X., Yang, Y., Savage, D. B., Malhotra, M., Heneghan, C., Patel, S., Speed, C., & Prieto, J. L. (2026). Insulin resistance prediction from wearables and routine blood biomarkers. Nature, 652, 451–461. https://doi.org/10.1038/s41586-026-10179-2 [Google Scholar] [Crossref]
4. Rafie, Z., Talab, M. S., Koor, B. E. Z., Garavand, A., Salehnasab, C., & Ghaderzadeh, M. (2025). Leveraging XGBoost and explainable AI for accurate prediction of type 2 diabetes. BMC Public Health, 25, Article 3688. https://doi.org/10.1186/s12889-025-24953-w [Google Scholar] [Crossref]
5. Tena, F., Garnica, O., Lanchares, J., & Hidalgo, J. I. (2025). An LSTM-based neural network wearable system for blood glucose prediction in people with diabetes. IEEE Journal of Biomedical and Health Informatics, 29(8), 5515–5526. Published online 2023. https://doi.org/10.1109/JBHI.2023.3300511 [Google Scholar] [Crossref]
6. Lim, M. H., Chae, H., Yoon, J., & Shin, I. (2025). A deep learning framework for virtual continuous glucose monitoring and glucose prediction based on life-log data. Scientific Reports, 15, Article 16290. https://doi.org/10.1038/s41598-025-01367-7 [Google Scholar] [Crossref]
7. Tsai, H., Yang, T.-W., Wu, T.-Y., Tu, Y.-C., Chen, C.-L., et al. (2025). Multitask learning multimodal network for chronic disease prediction. Scientific Reports, 15, Article 15468. https://doi.org/10.1038/s41598-025-99554-z [Google Scholar] [Crossref]
8. Wolff, M. K., Schaathun, H. G., Fougner, A. L., Steinert, M., & Volden, R. (2024). Mobile software development kit for real time multivariate blood glucose prediction. IEEE Access. https://doi.org/10.1109/ACCESS.2024.3349496 [Google Scholar] [Crossref]
9. Chiang, P.-H., Wong, M., & Dey, S. (2021). Using wearables and machine learning to enable personalized lifestyle recommendations to improve blood pressure. IEEE Journal of Translational Engineering in Health and Medicine, 9, Article 2700513. https://doi.org/10.1109/JTEHM.2021.3098173 [Google Scholar] [Crossref]
10. Jeong, J.-H., Lee, I.-G., Kim, S.-K., Kam, T.-E., Lee, S.-W., & Lee, E. (2024). DeepHealthNet: Adolescent obesity prediction system based on a deep learning framework. IEEE Journal of Biomedical and Health Informatics. https://doi.org/10.1109/JBHI.2024.3356580 [Google Scholar] [Crossref]
11. Lundberg, S. M., & Lee, S.-I. (2017). A unified approach to interpreting model predictions. In Advances in Neural Information Processing Systems (NeurIPS), 30. [Google Scholar] [Crossref]
12. Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 785–794). [Google Scholar] [Crossref]
13. Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735–1780. [Google Scholar] [Crossref]
14. Vaswani, A., et al. (2017). Attention is all you need. In Advances in Neural Information Processing Systems (NeurIPS), 30. [Google Scholar] [Crossref]
15. Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. [Google Scholar] [Crossref]
16. Ribeiro, M. T., Singh, S., & Guestrin, C. (2016). ‘Why should I trust you?’: Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 1135–1144). [Google Scholar] [Crossref]
17. International Diabetes Federation. (2025). IDF Diabetes Atlas, 11th edition. International Diabetes Federation. https://diabetesatlas.org/ (2024 estimates: 589 million adults globally; 89.8 million in India). [Google Scholar] [Crossref]
18. Alberti, K. G. M. M., et al. (2009). Harmonizing the metabolic syndrome: A joint interim statement of the IDF Task Force on Epidemiology and Prevention; NHLBI; AHA; WHF; IAS; and IASO. Circulation, 120(16), 1640–1645. [Google Scholar] [Crossref]
19. Matthews, D. R., Hosker, J. P., Rudenski, A. S., Naylor, B. A., Treacher, D. F., & Turner, R. C. (1985). Homeostasis model assessment: Insulin resistance and β-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia, 28(7), 412–419. [Google Scholar] [Crossref]
20. Battelino, T., et al. (2019). Clinical targets for continuous glucose monitoring data interpretation: Recommendations from the international consensus on time in range. Diabetes Care, 42(8), 1593–1603. [Google Scholar] [Crossref]
21. Chawla, N. V., Bowyer, K. W., Hall, L. O., & Kegelmeyer, W. P. (2002). SMOTE: Synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, 16, 321–357. [Google Scholar] [Crossref]
22. Lundberg, S. M., et al. (2020). From local explanations to global understanding with explainable AI for trees. Nature Machine Intelligence, 2(1), 56–67. [Google Scholar] [Crossref]
23. Doshi-Velez, F., & Kim, B. (2017). Towards a rigorous science of interpretable machine learning. arXiv:1702.08608. [Google Scholar] [Crossref]
24. Pedregosa, F., et al. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825–2830. [Google Scholar] [Crossref]
25. Kingma, D. P., & Ba, J. (2015). Adam: A method for stochastic optimization. In Proceedings of the International Conference on Learning Representations (ICLR). [Google Scholar] [Crossref]
26. Bergstra, J., & Bengio, Y. (2012). Random search for hyper-parameter optimization. Journal of Machine Learning Research, 13, 281–305. [Google Scholar] [Crossref]
27. World Health Organization. (2000). Obesity: Preventing and managing the global epidemic (WHO Technical Report Series No. 894). World Health Organization. [Google Scholar] [Crossref]
28. Rudin, C. (2019). Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 1(5), 206–215. [Google Scholar] [Crossref]
29. Lee, M. H., Jin, S., Febriana, E., Lim, M., Baig, S., Ali, S. H., Ang, I. Y. H., Loh, T. P., Nastar, A., Chia, K. S., Kong, A. P.-S., Magkos, F., Cook, A. R., & Toh, S.-A. (2026). Integration of continuous glucose monitoring with HbA1c to improve the detection of prediabetes in Asian individuals: Model development study. JMIR Diabetes, 11, e81520. https://doi.org/10.2196/81520 [Google Scholar] [Crossref]
30. Tatli, D., Papapanagiotou, V., Liakos, A., Tsapas, A., & Delopoulos, A. (2024). Prediabetes detection in unconstrained conditions using wearable sensors. Clinical Nutrition Open Science, 58, 163–174. https://doi.org/10.1016/j.nutos.2024.09.013 [Google Scholar] [Crossref]
31. Metwally, A. A., et al. (2025). Predicting type 2 diabetes metabolic phenotypes using continuous glucose monitoring and a machine learning framework. Nature Biomedical Engineering, 9, 1222–1239. https://doi.org/10.1038/s41551-024-01311-6 [Google Scholar] [Crossref]
32. Görmez, Y., Yagin, F. H., Yagin, B., Aygun, Y., Boke, H., Badicu, G., De Sousa Fernandes, M. S., Alkhateeb, A., Al-Rawi, M. B. A., & Aghaei, M. (2025). Prediction of obesity levels based on physical activity and eating habits with a machine learning model integrated with explainable artificial intelligence. Frontiers in Physiology, 16, Article 1549306. https://doi.org/10.3389/fphys.2025.1549306 [Google Scholar] [Crossref]
33. Khater, T., Tawfik, H., & Singh, B. (2024). Explainable artificial intelligence for investigating the effect of lifestyle factors on obesity. Intelligent Systems with Applications, 23, Article 200427. https://doi.org/10.1016/j.iswa.2024.200427 [Google Scholar] [Crossref]
34. Lin, H., Li, J., Hussein, R., Sui, X., Li, X., Zhu, G., Katsaggelos, A. K., Zeng, Z., & Li, Y. (2024). Longitudinal wrist PPG analysis for reliable hypertension risk screening using deep learning. arXiv:2411.11863. https://arxiv.org/abs/2411.11863 [Google Scholar] [Crossref]
35. Kishor Kumar Reddy, C., Kaza, V. S., Madana Mohana, R., Alhameed, M., Jeribi, F., Alam, S., & Shuaib, M. (2025). Detecting anomalies in smart wearables for hypertension: A deep learning mechanism. Frontiers in Public Health, 12, Article 1426168. https://doi.org/10.3389/fpubh.2024.1426168 [Google Scholar] [Crossref]
36. AI-READI Consortium. (2024). AI-READI: Rethinking AI data collection, preparation and sharing in diabetes research and beyond. Nature Metabolism, 6(12), 2210–2212.https://doi.org/10.1038/s42255-024-01165-x [Google Scholar] [Crossref]
37. [37] Lutsker, G., Sapir, G., Shilo, S., Merino, J., Godneva, A., Greenfield, J. R., Samocha-Bonet, D., Dhir, R., Gude, F., Mannor, S., Meirom, E., Xing, E. P., Chechik, G., Rossman, H., & Segal, E. (2026). A foundation model for continuous glucose monitoring data. Nature, 650(8103), 978–986. https://doi.org/10.1038/s41586-025-09925-9 [Google Scholar] [Crossref]
38. DeFronzo, R. A., & Tripathy, D. (2009). Skeletal muscle insulin resistance is the primary defect in type 2 diabetes. Diabetes Care, 32(Suppl. 2), S157–S163. https://doi.org/10.2337/dc09-S302 [Google Scholar] [Crossref]
This work is licensed under CC BY 4.0