{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T19:01:31Z","timestamp":1790017291572,"version":"4.0.1"},"reference-count":34,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR AI"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec sec-type=\"background\">\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Type 1 diabetes is characterized by absolute insulin deficiency, requiring exogenous insulin therapy to maintain blood glucose levels within safe ranges. Postprandial glucose control remains particularly challenging, and current meal-related strategies are mainly based on carbohydrate intake. However, other macronutrients, such as fats and proteins, may also influence the magnitude and timing of the glycemic response and are not usually incorporated into glucose forecasting models.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"objective\">\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>This study aims to develop and evaluate multihorizon blood glucose forecasting models to assess the impact of incorporating detailed nutritional information, with particular emphasis on the postprandial period and to analyze how the contribution of different nutrients varies across prediction horizons.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"methods\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We trained temporal fusion transformer (TFT) models on continuous glucose monitoring, insulin, and meal data with different combinations of nutritional variables from 351 adults with type 1 diabetes (15,601 meals) to predict glucose values up to 4 hours ahead. Carbohydrates were used as the baseline nutritional input, and additional configurations included fats, proteins, sugars, and complex carbohydrates. Model performance was evaluated globally and in predictions initiated at meal intake. Prediction horizons were grouped into early and late intervals, corresponding to 0-2 hours and 2-4 hours, respectively. The interpretability mechanisms of the TFT and integrated gradients method were used to analyze the relative contribution of nutritional variables.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"results\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Models incorporating additional nutritional information generally outperformed the carbohydrate-only baseline. In the global analysis, the combination of carbohydrates, fats, and proteins achieved the best performance in the early prediction interval, reducing mean root mean squared error (RMSE) and mean absolute error (MAE) from 30.48 mg\/dL and 21.02 mg\/dL to 29.23 mg\/dL and 20.15 mg\/dL. At late intervals, distinguishing between complex carbohydrates and sugars, together with fats and proteins, provided the best performance, reducing mean RMSE and MAE from 45.34 mg\/dL and 34.73 mg\/dL to 43.24 mg\/dL and 32.65 mg\/dL. A similar temporal dependency pattern was observed in the postprandial evaluation and was partially supported by the interpretability analysis.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"conclusions\">\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>These findings suggest that incorporating a more comprehensive representation of meal composition may modestly improve postprandial glucose forecasting. Horizon-specific attribution patterns indicated that the models used nutritional inputs differently across the forecast horizon; however, these findings should not be interpreted as causal or physiological effects of individual nutrients. Further external and prospective validation is required before the potential clinical utility of these models can be assessed.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.2196\/105049","type":"journal-article","created":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T18:14:25Z","timestamp":1790014465000},"page":"e105049-e105049","source":"Crossref","is-referenced-by-count":0,"title":["Interpretable Multihorizon Glucose Forecasting for Assessing Nutritional Information Impact in Type 1 Diabetes Management: Model Development and Validation Study"],"prefix":"10.2196","volume":"5","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-0741-4436","authenticated-orcid":false,"given":"Carlos","family":"Gallardo-Garc\u00eda","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6182-313X","authenticated-orcid":false,"given":"M Elena","family":"Hernando","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3589-7842","authenticated-orcid":false,"given":"David","family":"Sub\u00edas","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6691-4871","authenticated-orcid":false,"given":"Mercedes","family":"Rigla","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8396-8185","authenticated-orcid":false,"given":"Gema","family":"Garc\u00eda-S\u00e1ez","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,9,21]]},"reference":[{"issue":"10138","key":"R1","doi-asserted-by":"publisher","first-page":"2449","DOI":"10.1016\/S0140-6736(18)31320-5","article-title":"Type 1 diabetes","volume":"391","author":"DiMeglio","journal-title":"Lancet"},{"issue":"4","key":"R2","doi-asserted-by":"publisher","first-page":"569","DOI":"10.1007\/s13300-025-01703-0","article-title":"Proposed practical guidelines to improve glycaemic management by reducing glycaemic variability in people with type 1 diabetes mellitus","volume":"16","author":"de Torres-S\u00e1nchez","journal-title":"Diabetes Ther"},{"issue":"6","key":"R3","doi-asserted-by":"publisher","first-page":"1008","DOI":"10.2337\/dc15-0100","article-title":"Impact of fat, protein, and glycemic index on postprandial glucose control in type 1 diabetes: implications for intensive diabetes management in the continuous glucose monitoring era","volume":"38","author":"Bell","journal-title":"Diabetes Care"},{"key":"R4","doi-asserted-by":"publisher","DOI":"10.1016\/j.diabres.2022.110089","article-title":"Postprandial glucose variability in type 1 diabetes: the individual matters beyond the meal","volume":"192","author":"Bozzetto","journal-title":"Diabetes Res Clin Pract"},{"key":"R5","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2023.3244712","article-title":"Impact of nutritional factors in blood glucose prediction in type 1 diabetes through machine learning","volume":"11","author":"Annuzzi","journal-title":"IEEE Access"},{"issue":"21","key":"R6","doi-asserted-by":"publisher","DOI":"10.3390\/nu16213639","article-title":"Insights in nutrition to optimize type 1 diabetes therapy","volume":"16","author":"Cadario","journal-title":"Nutrients"},{"issue":"19","key":"R7","doi-asserted-by":"publisher","DOI":"10.3390\/nu15194279","article-title":"Type 1 and type 2 diabetes mellitus: commonalities, differences and the importance of exercise and nutrition","volume":"15","author":"Krause","journal-title":"Nutrients"},{"issue":"7","key":"R8","doi-asserted-by":"publisher","first-page":"2744","DOI":"10.1109\/JBHI.2020.3040225","article-title":"Deep learning for diabetes: a systematic review","volume":"25","author":"Zhu","journal-title":"IEEE J Biomed Health Inform"},{"key":"R9","unstructured":"Bahdanau D Cho K Bengio Y . Neural machine translation by jointly learning to align and translate. arXiv. Preprint posted online on May 19, 2014. [doi: 10.48550\/arXiv.1409.0473F]"},{"key":"R10","unstructured":"Vaswani A Shazeer N Parmar N et al. Attention is all you need. arXiv. Preprint posted online on Aug 2, 2017. [doi: 10.48550\/arXiv.1706.03762]"},{"issue":"1","key":"R11","doi-asserted-by":"publisher","DOI":"10.1186\/s12911-024-02761-3","article-title":"A prior-knowledge-guided dynamic attention mechanism to predict nocturnal hypoglycemic events in type 1 diabetes","volume":"24","author":"Yu","journal-title":"BMC Med Inform Decis Mak"},{"issue":"24","key":"R12","doi-asserted-by":"publisher","first-page":"245","DOI":"10.1016\/j.ifacol.2024.11.044","article-title":"Attention networks for personalized mealtime insulin dosing in people with type 1 diabetes","volume":"58","author":"Fathi","journal-title":"IFAC-PapersOnLine"},{"key":"R13","doi-asserted-by":"crossref","unstructured":"Zhu T Chen T Kuang L Zeng J Li K Georgiou P . Edge-based temporal fusion transformer for multi-horizon blood glucose prediction. Presented at: 2023 IEEE International Symposium on Circuits and Systems (ISCAS); May 21-25, 2023; Monterey, CA. [doi: 10.1109\/ISCAS46773.2023.10181448]","DOI":"10.1109\/ISCAS46773.2023.10181448"},{"issue":"2","key":"R14","doi-asserted-by":"publisher","first-page":"236","DOI":"10.1109\/TBCAS.2023.3348844","article-title":"Population-specific glucose prediction in diabetes care with transformer-based deep learning on the edge","volume":"18","author":"Zhu","journal-title":"IEEE Trans Biomed Circuits Syst"},{"issue":"8","key":"R15","doi-asserted-by":"publisher","first-page":"5424","DOI":"10.1109\/JBHI.2024.3428921","article-title":"Multi-horizon glucose prediction across populations with deep domain generalization","volume":"29","author":"Zhu","journal-title":"IEEE J Biomed Health Inform"},{"issue":"5","key":"R16","doi-asserted-by":"publisher","first-page":"3654","DOI":"10.1109\/JBHI.2025.3530461","article-title":"Multivariate glucose forecasting using deep multihead attention layers inside neural basis expansion networks","volume":"29","author":"Kalita","journal-title":"IEEE J Biomed Health Inform"},{"issue":"6","key":"R17","doi-asserted-by":"publisher","first-page":"584","DOI":"10.3390\/diabetology5060042","article-title":"Exploration of foundational models for blood glucose forecasting in type-1 diabetes pediatric patients","volume":"5","author":"Rancati","journal-title":"Diabetology"},{"issue":"4","key":"R18","doi-asserted-by":"publisher","first-page":"1748","DOI":"10.1016\/j.ijforecast.2021.03.012","article-title":"Temporal fusion transformers for interpretable multi-horizon time series forecasting","volume":"37","author":"Lim","journal-title":"Int J Forecast"},{"issue":"2","key":"R19","doi-asserted-by":"publisher","first-page":"155","DOI":"10.1016\/j.ifacol.2025.06.027","article-title":"A comparative study of transformer-based models for multi-horizon blood glucose prediction","volume":"59","author":"Karagoz","journal-title":"IFAC-PapersOnLine"},{"issue":"15","key":"R20","doi-asserted-by":"publisher","DOI":"10.3390\/s25154647","article-title":"Incorporating uncertainty estimation and interpretability in personalized glucose prediction using the temporal fusion transformer","volume":"25","author":"Rodriguez-Almeida","journal-title":"Sensors (Basel)"},{"issue":"5","key":"R21","doi-asserted-by":"publisher","first-page":"3123","DOI":"10.1109\/JBHI.2023.3348334","article-title":"Exploring nutritional influence on blood glucose forecasting for type 1 diabetes using explainable AI","volume":"28","author":"Annuzzi","journal-title":"IEEE J Biomed Health Inform"},{"key":"R22","doi-asserted-by":"crossref","unstructured":"Shapley LS . 17. A value for n-person games. In: Kuhn HW Tucker AW , editors. Contributions to the Theory of Games. Vol 3. Princeton University Press; 2016:307-318. [doi: 10.1515\/9781400881970-018]","DOI":"10.1515\/9781400881970-018"},{"key":"R23","unstructured":"Lundberg S Lee SI . A unified approach to interpreting model predictions. Presented at: 31st Conference on Neural Information Processing Systems (NIPS 2017); Dec 4-9, 2017:4765-4774; Long Beach, CA. [doi: 10.5555\/3295222.3295230]"},{"key":"R24","doi-asserted-by":"publisher","DOI":"10.3389\/fendo.2024.1423303","article-title":"Prediction of personalised postprandial glycaemic response in type 1 diabetes mellitus","volume":"15","author":"Xiong","journal-title":"Front Endocrinol"},{"issue":"9","key":"R25","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pdig.0000996","article-title":"Explainable cluster-based learning for prediction of postprandial glycemic events and insulin dose optimization in type 1 diabetes","volume":"4","author":"Rehman","journal-title":"PLOS Digit Health"},{"issue":"4","key":"R26","doi-asserted-by":"publisher","first-page":"704","DOI":"10.2337\/dc22-1721","article-title":"Examining the acute glycemic effects of different types of structured exercise sessions in type 1 diabetes in a real-world setting: the type 1 diabetes and exercise initiative (T1DEXI)","volume":"46","author":"Riddell","journal-title":"Diabetes Care"},{"issue":"2","key":"R27","doi-asserted-by":"publisher","first-page":"85","DOI":"10.1089\/dia.2020.0357","article-title":"Assessing mealtime macronutrient content: patient perceptions versus expert analyses via a novel phone app","volume":"23","author":"Gillingham","journal-title":"Diabetes Technol Ther"},{"issue":"4","key":"R28","doi-asserted-by":"publisher","first-page":"891","DOI":"10.1038\/oby.2011.344","article-title":"Validity of the remote food photography method (RFPM) for estimating energy and nutrient intake in near real-time","volume":"20","author":"Martin","journal-title":"Obesity (Silver Spring)"},{"issue":"12","key":"R29","doi-asserted-by":"publisher","first-page":"883","DOI":"10.1089\/dia.2020.0061","article-title":"Multi-hour blood glucose prediction in type 1 diabetes: a patient-specific approach using shallow neural network models","volume":"22","author":"Kushner","journal-title":"Diabetes Technol Ther"},{"issue":"11","key":"R30","doi-asserted-by":"publisher","first-page":"2695","DOI":"10.1111\/dom.13445","article-title":"Impact of macronutrient content of meals on postprandial glucose control in the context of closed\u2010loop insulin delivery: a randomized cross\u2010over study","volume":"20","author":"Gingras","journal-title":"Diabetes Obes Metab"},{"issue":"5","key":"R31","doi-asserted-by":"publisher","first-page":"776","DOI":"10.1089\/dia.2005.7.776","article-title":"The original Clarke error grid analysis (EGA)","volume":"7","author":"Clarke","journal-title":"Diabetes Technol Ther"},{"key":"R32","unstructured":"Vivli digital object identifier information sheet: data package DOI. Vivli. URL: https:\/\/search.vivli.org\/doiLanding\/dataPackages\/PR00008428.1 [Accessed 02-09-2026]"},{"issue":"5","key":"R33","doi-asserted-by":"publisher","first-page":"3819","DOI":"10.1109\/JBHI.2025.3630214","article-title":"Systematic review on deep learning algorithms for blood glucose forecasting in type 1 diabetes","volume":"30","author":"Calzavara","journal-title":"IEEE J Biomed Health Inform"},{"key":"R34","unstructured":"Request data from Vivli. Vivli. URL: https:\/\/vivli.org\/resources\/requestdata\/ [Accessed 02-09-2026]"}],"container-title":["JMIR AI"],"original-title":[],"language":"en","deposited":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T18:14:29Z","timestamp":1790014469000},"score":1,"resource":{"primary":{"URL":"https:\/\/ai.jmir.org\/2026\/1\/e105049"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,21]]},"references-count":34,"URL":"https:\/\/doi.org\/10.2196\/105049","relation":{},"ISSN":["2817-1705"],"issn-type":[{"value":"2817-1705","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,21]]},"article-number":"v5i10138e105049"}}