bims-glumda Biomed News
on CGM data in management of diabetes
Issue of 2026–01–18
ten papers selected by
Mott Given
  1. [Requirements for continuous glucose monitoring].
  2. Pregnancy Outcomes in Patients with Type 1 Diabetes Using Continuous Glucose Monitoring.
  3. A foundation model for continuous glucose monitoring data.
  4. Inequity in Continuous Glucose Monitor (CGM) Prescribing Behaviors in Primary Care.
  5. The Impact of a Hands-on Workshop on Continuous Glucose Monitor Prescribing in a Family Medicine Residency.
  6. Systematic Review on Deep Learning Algorithms for Blood Glucose Forecasting in Type 1 Diabetes.
  7. Teplizumab treatment for stage 2 type 1 diabetes: a real-world evaluation of metabolic and immunological outcomes.
  8. The Relationship Between HbA1c and GMI and Glucose Metrics in Children and Adolescents with Type 1 Diabetes Using AID.
  9. Assessing Time Below Range as a Predictor of Severe Hypoglycemia: Insights From Six Clinical Trials.
  10. Relationship Between Laboratory-Measured HbA1c and Continuous Glucose Monitoring-Derived Glucose Management Indicator in Adults With Cystic Fibrosis-Related Diabetes.
  1. Lakartidningen. 2026 Jan 12. pii: 25104. [Epub ahead of print]123
    [Requirements for continuous glucose monitoring].
    Peter Adolfsson, Johan Jendle.
      Most people with type 1 diabetes rely on continuous glucose monitoring (CGM) for glucose control. However, the EU Medical Device Regulation (MDR) does not sufficiently ensure quality control due to lack of a standard for CGM devices. Several cases have been reported where healthcare providers had to discard CGM systems after purchase due to poor performance. Moreover, the mean absolute relative difference (MARD) is inadequate for measuring accuracy. MARD can be influenced by study design, e.g. by including individuals with type 2 diabetes, using only data during stable glucose levels, or assessing only the most accurate days of sensor survival. Instead, accuracy should be based on methods reflecting real-world challenges in people intended to use the device. Diagnosis, age and gender should be considered. The International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) is currently developing criteria for the evaluation of CGM systems.
  2. Am J Perinatol. 2026 Jan 12.
    Pregnancy Outcomes in Patients with Type 1 Diabetes Using Continuous Glucose Monitoring.
    Shannon McCloskey, Joseph Biggio, John Morgan, Naiha Mussarat, James Toppin, Sarah Sternlieb, Nicholas Manuel, Kali Juracek, Sharon Shu, Jesse Stone, Frank B Williams.
      Objective Continuous glucose monitoring (CGM) use among patients with type 1 diabetes (T1DM) has been associated with improved glycemic control, though improvement in non-glycemic outcomes is less consistent. We hypothesize that CGM use in patients with T1DM in a real-world clinical setting is associated with both improved glycemic and clinical outcomes. Study Design This was a retrospective cohort study of patients with T1DM receiving care at a large health system from 2016 to 2023. Primary outcomes included: 1) glycemic control and 2) a composite comprised of severe maternal morbidity, preeclampsia with severe features, delivery prior to 34 weeks, and admission for diabetic ketoacidosis. Primary glycemic outcome was Hb A1c < 6% in the 2nd trimester. Patients using CGM were then evaluated by device setting, with those set to targets consistent with American Diabetes Association (ADA) recommendations compared to those with more permissive goals. Adjusted odds ratios were calculated using multivariable logistic regression to adjust for potential confounding variables. Results Among 288 patients with T1DM, there were 145 deliveries in the CGM group and 143 in the traditional capillary blood glucose monitoring group. Midtrimester on-target glycemic control was improved in the CGM group compared to traditional monitoring (40.7% vs 17.5%, aOR 2.32; 95% CI 1.21-4.12). There was no difference in composite outcome (CGM: 42.8% vs TBGM: 49.0%, aOR 0.70; 95%CI 0.40-1.22), nor was there a difference in secondary outcomes. In patients using CGM, stricter targets were associated with improved glycemic control as well as reduced preterm delivery (18.8% vs 56.9%, aOR 0.16, CI 0.05-0.48) and NICU admission (37.5% vs 60%,aOR 0.37, CI 0.14-0.96). Conclusion CGM use in T1DM is associated with improved glycemic control throughout pregnancy, however this does not uniformly translate to improved clinical outcomes. Lack of adherence to ADA blood glucose targets may contribute to these findings.
    DOI:  https://doi.org/10.1055/a-2781-7373
  3. Nature. 2026 Jan 14.
    A foundation model for continuous glucose monitoring data.
    Guy Lutsker, Gal Sapir, Smadar Shilo, Jordi Merino, Anastasia Godneva, Jerry R Greenfield, Dorit Samocha-Bonet, Raja Dhir, Francisco Gude, Shie Mannor, Eli Meirom, Eric P Xing, Gal Chechik, Hagai Rossman, Eran Segal.
      Continuous glucose monitoring (CGM) generates detailed temporal profiles of glucose dynamics, but its full potential for achieving glucose homeostasis and predicting long-term outcomes remains underutilized. Here we present GluFormer, a generative foundation model for CGM data trained with self-supervised learning on more than 10 million glucose measurements from 10,812 adults mainly without diabetes1,2. Using autoregressive prediction, the model learned representations that transferred across 19 external cohorts (n = 6,044) spanning 5 countries, 8 CGM devices and diverse pathophysiological states, including prediabetes, type 1 and type 2 diabetes, gestational diabetes and obesity. These representations provided consistent improvements over baseline blood glucose and HbA1c levels and other CGM-derived measures for forecasting glycaemic parameters3,4. In individuals with prediabetes, GluFormer stratified those likely to experience clinically significant increases in HbA1c over a 2-year period, outperforming baseline HbA1c and common CGM metrics. In a cohort of 580 adults with short-term CGM and a median follow-up of 11 years5, GluFormer identified individuals at elevated risk of diabetes and cardiovascular mortality more effectively than HbA1c. Specifically, 66% of incident diabetes cases and 69% of cardiovascular deaths occurred in the top risk quartile, compared with 7% and 0%, respectively, in the bottom quartile. In clinical trials, baseline CGM representations improved outcome prediction. A multimodal extension of the model that integrates dietary data generated plausible glucose trajectories and predicted individual glycaemic responses to food. Together, these findings indicate that GluFormer provides a generalizable framework for encoding glycaemic patterns and may inform precision medicine approaches for metabolic health.
    DOI:  https://doi.org/10.1038/s41586-025-09925-9
  4. J Gen Intern Med. 2026 Jan 13.
    Inequity in Continuous Glucose Monitor (CGM) Prescribing Behaviors in Primary Care.
    Jovan Milosavljevic, Clyde Schechter, Melissa Fazzari, Priyanka Mathias, Justin Mathew, Shivani Agarwal.
       BACKGROUND: Continuous glucose monitoring (CGM) has become a standard of care for diabetes management, yet its adoption in primary care remains limited where the majority of people with diabetes are served. Examining contributors to CGM prescriptions in primary care may identify new targets for intervention that could improve diabetes population health.
    OBJECTIVE: To determine factors associated with CGM prescription by primary care providers (PCPs) in a safety-net primary care practice network serving over 11,000 adults with insulin-treated type 2 diabetes.
    DESIGN: Retrospective cohort study.
    PATIENTS: Adults with type 2 diabetes treated with insulin with at least one primary care visit between July 31, 2020, and July 31, 2023, in a safety-net primary care clinic network in the Bronx, NY.
    MAIN MEASURES: Primary outcome was first-time CGM prescription by PCPs. Multivariable analysis was performed using cause-specific Cox regression, with death and prescriptions by endocrinology treated as censoring events.
    RESULTS: Among 11,037 insulin-treated patients (mean age 61 ± 14 years; 56% female; 44% Hispanic, 39% non-Hispanic Black), 17% received first-time CGM prescriptions from primary care, with a median monthly prescription rate of 3.1% (2.8-3.3%). Older patients, Spanish-speaking, and publicly insured had a lower likelihood of receiving CGM prescription. Conversely, patients with higher HbA1c levels, more intensive insulin regimens, and providers with more years in practice were more likely to receive CGM regardless of race-ethnicity.
    CONCLUSIONS: CGM prescribing rates in primary care remain low, with inequity that favored younger, English-speaking, commercially insured patients, and more experienced providers. We have highlighted gaps in workflow efficiency, language and age-appropriate training support, and trainee education as key next steps to address in order to increase CGM prescriptions in primary care practice. Addressing inequitable CGM prescribing patterns is a critical first step to elevating the standard of diabetes care in primary care practice and breaking the cascade of inequity in outcomes for underserved diabetes populations.
    Keywords:  continuous glucose monitoring; diabetes technology; inequity; primary care
    DOI:  https://doi.org/10.1007/s11606-025-09923-7
  5. PRiMER. 2025 ;9 60
    The Impact of a Hands-on Workshop on Continuous Glucose Monitor Prescribing in a Family Medicine Residency.
    Ashley M Runyon, Jessica B Koran-Scholl.
       Introduction: Continuous glucose monitoring (CGM) has become an essential tool in managing diabetes, offering real-time insights that improve patient engagement, decision-making, and clinical outcomes. Despite its benefits, CGM is underutilized in primary care due to perceived complexity and time constraints. This study aimed to enhance health care professionals' understanding of and confidence in prescribing CGM technology through two multidisciplinary, hands-on workshops.
    Methods: A total of 51 participants, including family medicine residents, faculty, clinical staff, and pharmacists, attended hands-on workshops and wore a CGM for 10-14 days. We analyzed prescribing patterns using EHR data to compare CGM orders placed before and after the workshops. Additionally, we evaluated pre- and postworkshop surveys' changes in participants' comfort, knowledge, and likelihood of recommending CGMs to patients.
    Results: Results showed a 31% improvement in participants' prescribing patterns of CGMs after holding the workshops. Qualitative feedback highlighted insights into device usability, blood sugar patterns, and patient counseling.
    Conclusion: This study demonstrates the value of experiential learning in improving health care providers' competence with diabetes technologies. Future research should explore the long-term impact of such training on clinical practice and patient outcomes.
    DOI:  https://doi.org/10.22454/PRiMER.2025.940150
  6. IEEE J Biomed Health Inform. 2026 Jan 14. PP
    Systematic Review on Deep Learning Algorithms for Blood Glucose Forecasting in Type 1 Diabetes.
    Andrea Calzavara, Francesco Prendin, Giacomo Cappon, Simone Del Favero, Andrea Facchinetti.
      Type 1 Diabetes (T1D) is a chronic metabolic disease characterized by elevated blood glucose (BG) concentrations, resulting from the immune-mediated destruction of insulin-producing $\beta$-cells in the pancreas. Effective management of T1D greatly benefits from constant monitoring of BG levels, achievable in real-time using minimally invasive continuous glucose monitoring (CGM) devices. These devices provide data streams that can be leveraged by forecasting algorithms to predict BG levels minutes in advance, enabling timely therapeutic interventions to prevent adverse events, such as hypo/hyperglycemia. With the increasing availability of data, deep learning (DL) algorithms have emerged as the state-of-the-art for BG forecasting, owing to their ability to autonomously learn complex nonlinear relationships, such as those underlying the glucoregulatory system. Despite a growing body of research, a comprehensive review specifically focusing on DL applications for BG prediction is still lacking. To address this gap, a systematic review was conducted following the PRISMA guidelines, involving extensive searches across PubMed, Scopus, and Web of Science databases. A total of 26 studies satisfied the inclusion criteria and were evaluated based on dataset characteristics, model inputs, training paradigm, prediction horizon, model architecture, evaluation metrics, performance, and baseline comparators. While DL models show great promise, several challenges persist-particularly in ensuring physiological fidelity and interpretability, both essential for clinical adoption. To overcome these barriers, future research should prioritize the integration of explainable AI (XAI) techniques to improve model reliability and safety, ultimately supporting the effective deployment of DL models in real-time T1D management.
    DOI:  https://doi.org/10.1109/JBHI.2025.3630214
  7. Diabetologia. 2026 Jan 15.
    Teplizumab treatment for stage 2 type 1 diabetes: a real-world evaluation of metabolic and immunological outcomes.
    Kagan E Karakus, Lexie Chesshir, Sonya Walker, Erin E Baschal, Kristen A McDaniel, Taylor M Triolo, Andrea K Steck, Brigitte I Frohnert, Peter A Gottlieb, Aaron W Michels, Kimber M Simmons.
       AIMS/HYPOTHESIS: This is the first real-world prospective observational study of teplizumab use following approval by the United States Food and Drug Administration in individuals with stage 2 type 1 diabetes. We examined whether glycaemic responses observed in controlled trials were reproduced in real-world practice and explored immunological biomarkers associated with treatment.
    METHODS: Children and adults with stage 2 type 1 diabetes were prospectively followed in the Early Type 1 Diabetes Clinic at the Barbara Davis Center. Individuals who received teplizumab between April 2023 and February 2025 (n=30) were compared with an untreated group (n=10). Key assessments included OGTTs, HbA1c measurements and continuous glucose monitoring (CGM) data, collected before and after treatment. Longitudinal metabolic data were available for up to 22 treated participants, with follow-up assessments occurring between 2 and 13 months after treatment. Changes in Epstein-Barr virus (EBV) and islet antigen-targeting T cell receptor (TCR) β chains were measured longitudinally from genomic DNA via a PCR-based targeted TCR sequencing assay.
    RESULTS: Among treated individuals followed up between 2 and 6 months after treatment, OGTT 2 h glucose improved (10.7 ± 2.1 to 8.8 ± 2.8 mmol/l, p=0.007), as did HbA1c (40 ± 4 to 39 ± 6 mmol/mol [5.8 ± 0.4% to 5.7 ± 0.5%], p=0.044). In addition, 67% had a stable or reduced CGM time ≥7.8 mmol/l. CD4 preproinsulin-specific TCRs declined after treatment, with no change in the untreated group. These reductions correlated with higher C-peptide AUC (r=-0.656, p=0.013). EBV TCR sequences were similar before and after teplizumab treatment.
    CONCLUSIONS/INTERPRETATION: Teplizumab can be safely and effectively administered in clinical practice. Early glycaemic improvements and reductions in CD4 preproinsulin-specific TCRs suggest that post-treatment immunological changes may serve as biomarkers to guide early-stage intervention.
    Keywords:  C-peptide; CGM; OGTT; Real-world; T cell receptor; Teplizumab; Type 1 diabetes
    DOI:  https://doi.org/10.1007/s00125-025-06646-6
  8. J Clin Res Pediatr Endocrinol. 2026 Jan 12.
    The Relationship Between HbA1c and GMI and Glucose Metrics in Children and Adolescents with Type 1 Diabetes Using AID.
    Emrullah Arslan, Hanife Gül Balkı, Deniz Özalp Kızılay, Günay Demir, Samim Özen, Şükran Darcan, Damla Gökşen.
       Introduction: HbA1c remains the standard biomarker for long-term glycemic control, but it lacks precision in capturing short-term glucose variability and acute excursions. This limitation is especially relevant in children with type 1 diabetes (T1D) who use continuous glucose monitoring systems (CGMS) and automated insulin delivery (AID) systems.
    Aim: To evaluate the temporal relationship between HbA1c and the glucose management indicator (GMI), and their associations with CGMS-derived glycemic parameters over 12-weekperiod in children and adolescents with T1D using AIDsystems.
    Material-methods: In this retrospective cross-sectional observational study,81 children and adolescents with T1D on the Medtronic MiniMed 780G™ system were included.CGMS data covering 12weeks prior to HbA1c measurement were analyzed in two-week intervals.Correlations between HbA1c,GMI, andCGMS metrics were assessed.
    Results: HbA1c was positively correlated with all GMI values,with the strongest correlation observed for the last six-weekGMI (r=0.728,p<0.001). The mean difference between HbA1c and last12-weekGMI was0.57% (95%CI:-1.13to 2.27).GMI demonstrated stronger correlations than HbA1c with time in range (TIR),time above range(TAR),and time below range(TBR).Notably, in individuals with similar TIR (~70%), HbA1c values varied widely (6.6-9.6% /48-81mmol/mol),while GMI remained stable (6.8-7.1%).
    Discussion: HbA1c exhibited the strongest correlation with GMI calculated over the last six weeks,suggesting that it primarily reflects recent glycemic trends rather than cumulative exposure.GMI also showed closer alignment with CGMS-derived indices such as TIR,TAR,and TBR,indicating its enhanced sensitivity in capturing day-to-day glycemic variability,especially in suboptimally controlled individuals.
    Conclusion: Given its temporal limitations,HbA1c may not reliably capture 12-weekglycemic patterns in pediatric AIDusers.GMI, as CGMS-derived metric,offers a more consistent and clinically actionable estimate of glycemic control,supporting its integration into routine care for children with T1D.
    Keywords:  AID; CGMS; GMI; T1D; sampling period
    DOI:  https://doi.org/10.4274/jcrpe.galenos.2025.2025-10-11
  9. Diabetes Care. 2026 Jan 13. pii: dc252353. [Epub ahead of print]
    Assessing Time Below Range as a Predictor of Severe Hypoglycemia: Insights From Six Clinical Trials.
    Eslam Montaser, Clarence Williams, Viral N Shah.
       OBJECTIVE: To evaluate whether baseline continuous glucose monitoring (CGM)-derived time below range (TBR) metrics-TBR level 1 (TBR1) (<70 mg/dL) and TBR level 2 (TBR2) (<54 mg/dL)-predicts severe hypoglycemia (SH) during follow-up of individuals with type 1 diabetes.
    RESEARCH DESIGN AND METHODS: Baseline CGM TBR levels and their association with SH adverse events during six clinical trials were analyzed using Wilcoxon rank-sum tests and Spearman correlations. Analyses were stratified by sex, race, and age group. Sensitivity, specificity, and false-positive rates (FPRs) were calculated for thresholds (1-5%), and receiver operating characteristic (ROC) analysis assessed discrimination.
    RESULTS: Participants (n = 1,433; median age 4-43 years; 50-62% female sex; 83-96% White race) had a baseline median TBR2 range of 0.1% to 0.7% and TBR1 from 1.2% to 4.1% across the six clinical trials. Those who developed SH had slightly higher baseline TBR2 (0.41% vs. 0.32%; P = 0.022) and TBR1 (2.58% vs. 2.23%; P = 0.044). Predictive accuracy was limited: sensitivity of TBR2 fell from 48.9% at a 1% cutoff to 18.2% at 5% (specificity 75.9-95.0%), and TBR1 sensitivity declined from 81.8% to 44.3% (specificity 29.6-77.7%), with modest discrimination by ROC analysis (area under the curve = 0.62 [95% CI 0.55-0.69] for TBR2; and 0.65 [95% CI: 0.58-0.71] for TBR1).
    CONCLUSIONS: Baseline TBR1 and TBR2 had limited predictive value for SH. No threshold achieved strong discrimination, a finding that supports the need to integrate additional clinical factors for SH risk stratification.
    DOI:  https://doi.org/10.2337/dc25-2353
  10. Diabetes Care. 2026 Jan 13. pii: dc252936. [Epub ahead of print]
    Relationship Between Laboratory-Measured HbA1c and Continuous Glucose Monitoring-Derived Glucose Management Indicator in Adults With Cystic Fibrosis-Related Diabetes.
    Nithya Kadiyala, Amanda Brennan, Alistair Lumb, Richard I G Holt, Dawn Lau, Parag Yajnik, Yee S Cheah, Shahideh Safavi, Imogen Felton, Gordon MacGregor, Andrew Clayton, Ailish Nugent, Laura Barlow, Amanda Adler, Roman Hovorka, Charlotte K Boughton.
      
    DOI:  https://doi.org/10.2337/dc25-2936
This page is being maintained by Thomas Krichel. It was last updated on 2026‒01‒18 at 0:05.