To avoid the adverse consequences from abrupt increases in blood glucose, diabetic inpatients should be closely monitored. Using blood glucose data from type 2 diabetes patients, we propose a deep learning model-based framework to forecast blood glucose levels. We used continuous glucose monitoring (CGM) data collected from inpatients with type 2 diabetes for a week. We adopted the Transformer model, commonly used in sequence data, to forecast the blood glucose level over time and detect hyperglycemia and hypoglycemia in advance. We expected the attention mechanism in Transformer to reveal a hint of hyperglycemia and hypoglycemia, and performed a comparative study to determine whether Transformer was effective in the classification and regression of glucose. Hyperglycemia and hypoglycemia rarely occur and this results in an imbalance in the classification. We built a data augmentation model using the generative adversarial network. Our contributions are as follows. First, we developed a deep learning framework utilizing the encoder part of Transformer to perform the regression and classification under a unified framework. Second, we adopted a data augmentation model using the generative adversarial network suitable for time-series data to solve the data imbalance problem and to improve performance. Third, we collected data for type 2 diabetic inpatients for mid-time. Finally, we incorporated transfer learning to improve the performance of regression and classification.

In this study, we propose a personalized glucose prediction model using deep learning for hospitalized patients who experience Type-2 diabetes. We aim for our model to assist the medical personnel who check the blood glucose and control the amount of insulin doses. Herein, we employed a deep learning algorithm, especially a recurrent neural network (RNN), that consists of a sequence processing layer and a classification layer for the glucose prediction. We tested a simple RNN, gated recurrent unit (GRU), and long-short term memory (LSTM) and varied the architectures to determine the one with the best performance. For that, we collected data for a week using a continuous glucose monitoring device. Type-2 inpatients are usually experiencing bad health conditions and have a high variability of glucose level. However, there are few studies on the Type-2 glucose prediction model while many studies performed on Type-1 glucose prediction. This work has a contribution in that the proposed model exhibits a comparative performance to previous works on Type-1 patients. For 20 in-hospital patients, we achieved an average root mean squared error (RMSE) of 21.5 and an Mean absolute percentage error (MAPE) of 11.1%. The GRU with a single RNN layer and two dense layers was found to be sufficient to predict the glucose level. Moreover, to build a personalized model, at most, 50% of data are required for training.

Diabetes is a chronic metabolic disorder characterized by elevated blood glucose levels, posing significant health risks such as cardiovascular disease, and nerve, kidney, and eye damage. Effective management of blood glucose is essential for individuals with diabetes to mitigate these risks. This study introduces the Glu-Ensemble, a deep learning framework designed for precise blood glucose forecasting in patients with type 2 diabetes. Unlike other predictive models, Glu-Ensemble addresses challenges related to small sample sizes, data quality issues, reliance on strict statistical assumptions, and the complexity of models. It enhances prediction accuracy and model generalizability by utilizing larger datasets and reduces bias inherent in many predictive models. The framework's unified approach, as opposed to patient-specific models, eliminates the need for initial calibration time, facilitating immediate blood glucose predictions for new patients. The obtained results indicate that Glu-Ensemble surpasses traditional methods in accuracy, as measured by root mean square error, mean absolute error, and error grid analysis. The Glu-Ensemble framework emerges as a promising tool for blood glucose level prediction in type 2 diabetes patients, warranting further investigation in clinical settings for its practical application.

Background: Dipeptidyl peptidase-4 (DPP-4) inhibitor has been reported to have kidney-protective benefits. To elucidate how antidiabetic agents prevent diabetic kidney disease progression, it is important to investigate their effect on the kidney environment in type 2 diabetes mellitus (DM) patients. Herein, we investigated the expression pattern of urinary exosome-derived microRNA (miRNA) in patients taking a combination of DPP-4 inhibitor and metformin (DPP-4 inhibitor group) and compared them with patients taking a combination of sulfonylurea and metformin (sulfonylurea group). Methods: This was a prospective study involving 57 patients with type 2 DM (DPP-4 inhibitor group, n = 34; sulfonylurea group, n = 23) and healthy volunteers (n = 7). We measured urinary exosomal miRNA using the NanoString nCounter miRNA array (NanoString Technologies) across the three groups (n = 4 per each group) and validated findings using real-time polymerase chain reaction. Results: Twenty-one differentially expressed candidate miRNAs were identified, and six (let-7c-5p, miR-23a-3p, miR-26a-3p, miR-30d, miR-205, and miR-200a) were selected for validation. Validation showed no significant difference in miRNA expression between the DPP-4 inhibitor and sulfonylurea groups. Only miR-23a-3p was significantly overexpressed in the diabetes group compared with the control group (DPP-4 inhibitor vs. control, p = 0.01; sulfonylurea vs. control, p = 0.007). This trend was consistent even after adjusting for age, sex, and body mass index. Conclusion: There was no significant difference in urine exosome miRNA expression between diabetic participants taking DPP-4 inhibitor and those taking sulfonylurea. The miR-23a levels were higher in diabetic participants than in nondiabetic controls.

Objective Type 2 diabetes mellitus has been associated with an increased risk of cognitive decline and dementia, with patients being 1.5–2 times more likely to develop these conditions. While both sodium-glucose co-transporter 2 (SGLT2) inhibitors and thiazolidinediones (TZDs) have shown potential neuroprotective effects in previous studies, their comparative effectiveness for preventing neurodegenerative outcomes has not been established. This study aimed to compare the risk of stroke, dementia and Alzheimer’s disease (AD) between patients treated with SGLT2 inhibitors and those treated with TZDs. Design Multicentre, retrospective, observational, new-user, active-comparator cohort study. Setting Electronic health record-based databases from 11 secondary and tertiary institutions in South Korea from 1 January 2014 to 31 July 2025. The study period began in 2014, following the post-marketing surveillance initiation of SGLT2 inhibitors in Korea (November 2013), to ensure adequate drug availability and clinical adoption. Participants Patients aged 40 years or older who were newly prescribed either SGLT2 inhibitors or TZDs without prior exposure. Interventions Propensity score matching (1:1) was performed using sex as the primary covariate due to data availability constraints in the Observational Medical Outcomes Partnership Common Data Model framework. The HRs with 95% CIs were measured via Cox regression analysis. Results The study analysed 24 172 matched pairs for stroke outcomes (40 483 person-years in the SGLT2 inhibitor group and 39 363 person-years in the TZD group), 25 111 matched pairs for dementia (41 924 person-years in the SGLT2 inhibitor group and 40 726 person-years in the TZD group) and 25 237 matched pairs for AD (42 139 person-years in the SGLT2 inhibitor group and 40 895 person-years in the TZD group) across 11 participating hospitals. After a 1:1 propensity score matching, the SGLT2 inhibitors showed no significant difference in stroke risk (HR 1.18, 95% CI 0.62 to 2.23, p=0.62), while having significant reductions in dementia risk (HR 0.66, 95% CI 0.45 to 0.98, p=0.04) and AD risk (HR 0.54, 95% CI 0.35 to 0.83, p=0.005). Moreover, these protective effects for neurodegenerative outcomes were shown to be consistent across multiple hospital sites. Conclusions SGLT2 inhibitors are associated with a reduced risk of dementia and AD compared with TZDs in patients aged 40 years or older with type 2 diabetes and have neutral effects on stroke risk. These findings confirm the potential selective neuroprotective benefits of SGLT2 inhibitors for neurodegenerative outcomes, which may inform therapeutic decision-making for diabetic patients at risk of cognitive decline.