BACKGROUND:Diabetes care requires frequent and high-stakes decisions that must be made in the setting of substantial day-to-day physiologic variability. The growing availability of continuous glucose monitoring, connected insulin delivery devices and longitudinal electronic health record data has created an opportunity for algorithm-enabled tools that can synthesise high-frequency data, reduce cognitive burden for patients and clinicians, and support safer and more consistent decision-making. AIM:In this review, artificial intelligence (AI) is used broadly to describe computational systems that generate predictions, recommendations or automation from clinical data. METHODS:We distinguish between algorithmic automation and control methods that underpin many currently deployed automated insulin delivery (AID) systems and machine learning-based models, including deep learning and large language models (LLMs), that are increasingly used for pattern recognition, risk prediction and natural language applications. RESULTS:This distinction is clinically relevant because evidence standards, safety risks and governance needs vary substantially across these categories. CONCLUSION:This narrative review summarises current and emerging applications of AI in diabetes care with an emphasis on clinical readiness, strength of evidence and implementation considerations. We highlight established applications in AID, emerging approaches that seek greater autonomy and interoperability and newer tools such as LLMs, wearables and digital twin frameworks, focusing on where evidence is strongest, where risks are highest and what safeguards are required for responsible clinical use.

As a chronic metabolic disease, diabetes mellitus necessitates ongoing self-management to control blood sugar levels and avoid complications. Although diabetes management relies heavily on technologies like insulin pumps and continuous glucose monitors (CGMs), patient use of these devices may differ depending on personality factors. This study investigated the association between diabetes types, personality types, and preferences for diabetes management technology. The 56 patients receiving intense insulin therapy completed the Myers-Briggs Type Indicator (MBTI) to evaluate personality qualities. Multinomial logistic regression was employed for statistical analysis after technology utilization data was gathered. According to the findings, introverts are more likely to utilize CGMs. Insulin pumps were preferred by extroverts, indicating their inclination for proactive, hands-on care. There were no significant relationships between diabetes type, technology use, or other aspects of personality. These results imply that personality variables affect technology preferences and could be helpful in customizing diabetes care plans. This study highlights the importance of considering psychological issues when choosing diabetes solutions.

Objectives: Pancreas transplantation provides long-term near-normal glycemic control for recipients with type 1 diabetes, but it is unknown how this control compares with an automated insulin delivery (AID) system. Methods: In this prospective study, we compared parameters from 31 consecutive pancreas-kidney transplantation recipients versus from 377 people using an AID-either MiniMedTM 780G (n = 200) or Tandem t:slim X2TM Control-IQTM (n = 177). Results: Compared with the MiniMed and Tandem AID groups, transplant recipients at 1 month (mean ± standard deviation [SD]: 36 ± 12 days) after pancreas transplantation exhibited significantly lower glycated hemoglobin (38 mmol/mol [36, 40] vs. 55 [53, 56.5] and 56 [54.7, 57.2], respectively), lower mean glycemia (6.4 mmol/L [6, 6.8] vs. 8.5 [8.3, 8.7] and 8.2 [8.0, 8.4], respectively), and spent more time in range (90% [86, 93] vs. 72% [70, 74] and 75% [73, 77], respectively). Time in hypoglycemia did not differ significantly between the groups. Conclusions: Overall, compared with AID treatment, pancreas transplantation led to significantly better diabetes control parameters, with the exception of time below range. Clinical trials registration number is Eudra CT No. 2019-002240-24.

Introduction and Objective: Automated Insulin Delivery systems (AID) improve glycemic outcomes and reduce hypoglycemia in people with type 1 diabetes (PwT1D). We assessed the effect of AID initiation (MiniMed 780G, Tandem t:slim X2, CamAPS FX hybrid closed-loop system) in PwT1D on reduction of hypoglycemia in real-world settings. Methods: CGM data from our local registry (entered once yearly at the last clinic visit of the corresponding year) were used for the analysis. All PwT1D who were initiated on any AID in our center throughout the year 2023 were included in the study. CGM data the year before and after AID initiation were compared. Results: Altogether 219 PwT1D (121/219, 55% women) with mean (SD) age 42 (15) years, diabetes duration 24 (12) years were analyzed. Of these, 168/219 (77%) were previously on non-AID pumps and 51/219 (33%) were on MDI regimen. Following AID initiation, TBR reduced (mean paired difference [95%CI]) (-2.6 [-3.4, -1.8] %, p < 0.001), level 1 hypoglycemia (-1.7 [-2.4, -1.1] %, p < 0.001) and level 2 hypoglycemia (-0.9 [-1.2, -0.6] %, p < 0.001) reduced. Reductions in TBR, level 1 and level 2 hypoglycemia occurred with all 3 AIDs: for MiniMed 780G (n = 86) (-3.6 [-5.1, -2.1] %, p < 0.001), (-2.5 [-3.9, -1.2] %, p < 0.001) and (-1.0 [-1.6, -0.5] %, p < 0.001); for Tandem t:slim X2 (n = 94) (-2.1 [-3.2, -1.0] %, p < 0.001), (-1.3 [-2.0, -0.5] %, p < 0.001) and (-0.8 [-1.4, -0.3] %, p = 0.005), and for CamAPS FX system (n = 39) (-1.9 [-3.6, -0.3] %, p < 0.001), (-1.2 [-2.3, -0.1] %, p = 0.016) and (-0.7 [-1.4, -0.1] %, p = 0.033). The difference in hypoglycemia reduction between MiniMed 780G and CamAPS FX was statistically significant for TBR (-1.7 [-3.1, -0.2] %, p = 0.016) and level 2 hypoglycemia (-0.5 [-0.9, -0.1] %, p = 0.006) in favor of MiniMed 780G. Conclusion: Initiation of all three AIDs in PwT1D statistically significantly and clinically meaningfully reduced hypoglycemia in our cohort. Highest reductions were observed with the MiniMed 780G. P. Novodvorsky: Advisory Panel; Novo Nordisk A/S. Speaker's Bureau; Novo Nordisk A/S, Sanofi, Abbott, AstraZeneca, Boehringer-Ingelheim. R. Bem: None. K. Sochorova: Consultant; Medtronic, Ypsomed AG. R. Koznarova: None. D. Vávra: None. K. Cechová: None. F. Hrubý: None. P. Girman: Speaker's Bureau; Neovii. M. Dubsky: None. M. Haluzik: Research Support; Sanofi. Advisory Panel; Eli Lilly and Company. Speaker's Bureau; Novo Nordisk, Abbott. Advisory Panel; AstraZeneca. Speaker's Bureau; GlaxoSmithKline plc, Amgen Inc. Advisory Panel; Bausch Health. CarDia (Programme EXCELES, Project No. LX22NPO5104) - Funded by the European Union - Next Generation EU Grant: G7301

Aim: To describe the demographic and clinical characteristics of patients with Charcot neuro‐osteoarthropathy (CNO) and to examine for differences between participants with Type 1 diabetes mellitus (DM) (T1DM) and Type 2 diabetes mellitus (T2DM).Materials and Methods: Multicenter observational study in eight diabetic foot clinics in six countries between January 1, 1996, and December 31, 2022. Demographic, clinical, and laboratory parameters were obtained from the medical records. Analyses were performed using parametric or nonparametric statistical tests for variables with normally or nonnormally distributed values, respectively. Comparisons of the qualitative data were performed using the chi‐square test.Results: Seven hundred seventy‐four patients with DM and CNO were included. The mean age at diagnosis of CNO was 54.5 ± 11.7 years, and the median (interquartile range (IQR)) diabetes duration at diagnosis of CNO was 15 (10–22) years. Among participants, 71.8% (n = 546) were male and 83.2% (n = 634) had T2DM. Neuropathy was present in 91.7% (n = 688), retinopathy in 60.2% (n = 452), and nephropathy in 45.2% (n = 337). Subjects with T1DM, compared to T2DM, were diagnosed with CNO at a younger age (46.9 ± 11.0 vs. 57.9 ± 10.2 years, p < 0.001), had longer diabetes duration (median value (IQR): 29.0 (21.0–38.0) vs. 14.0 (8.0–20.0) years, p < 0.001), and had more often microvascular complications (neuropathy, 95.2% in T1DM vs. 87.4% in T2DM, p = 0.006; retinopathy, 83.3% vs. 55.4%, p < 0.001; and nephropathy 67.5% vs. 40.5%, p < 0.001).Conclusions: CNO is predominant in males, occurs in long‐standing DM, and is often accompanied by microvascular complications. People with T1DM, compared to those with T2DM, are affected at a younger age, have longer diabetes duration, and have more often microvascular complications.

There is a lack of reliable in vivo models that replicate limb-threatening ischemia in humans. To fill this gap, we developed and validated two novel porcine ischemic models: ischemic limb and dorsal flap models, both with and without streptozotocin-induced hyperglycemia (N = 3 per group, 12 in total). Hind limb ischemia model was induced via different arterial ligations, with two ischemic and three control wounds per animal. In the flap model, four full-thickness flaps were created on the dorsum with silicone sheets to block reperfusion, and excisional wounds were made on the top. One non-ischemic wound served as control. Transcutaneous oxygen pressure (TcPO2), wound area, and microvascular density were measured, with TcPO2 and wound area assessed longitudinally. Data analysis focused on detailed visualization and Bayesian hierarchical modelling to account for the small sample size. Developed models exhibited stable ischemia and prolonged wound healing, with TcPO2 remaining under 30 mmHg over 28 days, and wound healing extending beyond two weeks. The flap model showed slower TcPO2 recovery and greater chronicity compared to the limb model, without reliable effect of hyperglycemia. Thus, the porcine flap model shows the highest potential as a relevant model for chronic limb-threatening ischemia.

Charcot neuropathic osteoarthropathy (CNO) is a condition that develops in the presence of neuropathy, most commonly diabetes-related neuropathy. Owing to the neuropathy, microtrauma to the bones occur without the individual feeling them. With continued walking, bone inflammation, resorption, microfractures and structural changes occur in the bones, which result in irreversible deformities. Diagnosing this condition is often difficult and requires advanced imaging techniques, such as scintigraphy or magnetic resonance imaging, as X-ray changes may not be specific. Treatment of CNO includes immobilization, offloading, recalcification (supplementation of vitamin D and calcium) and in the most advanced cases, surgical treatment. This narrative review aims to synthesize the recent research and clinical implications relating to Charcot foot to help healthcare professionals to stay up-to-date in this relevant topic.

Diabetic foot (DF) can develop in diabetic patients after organ transplantation (Tx) due to several factors including peripheral arterial disease (PAD), diabetic neuropathy and inappropriate DF prevention. Aim: To assess the occurrence of DF and associated risk factors in transplant patients. Methods: Fifty-seven diabetic patients were enrolled as part of this prospective study. All patients underwent organ Tx (01/2013-12/2015) and were followed up for minimum of 12 months up to a maximum of 50 months. Over the study period we evaluated DF incidence and identified a number of factors likely to influence DF development, including organ function, presence of late complications, PAD, history of DF, levels of physical activity before and after Tx, patient education and standards of DF prevention. Results: Active DF developed in 31.6% (18/57) of patients after organ Tx within 11 months on average (10.7 ± 8 months). The following factors significantly correlated with DF development: diabetes control (p = .0065), PAD (p<0.0001), transcutaneous oxygen pressure (TcPO2;p = .01), history of DF (p = .0031), deformities (p = .0021) and increased leisure-time physical activity (LTPA) before Tx (p = .037). However, based on logistic stepwise regression analysis, the only factors significantly associated with DF during the post-transplant period were: PAD, deformities and increased LTPA. Education was provided to patients periodically (2.6 ± 2.5 times) during the observation period. Although 94.7% of patients regularly inspected their feet (4.5 ± 2.9 times/week), only 26.3% of transplant patients used appropriate footwear. Conclusions: Incidence of DF was relatively high, affecting almost 1/3 of pancreas and kidney/pancreas recipients. The predominant risk factors were: presence of PAD, foot deformities and higher LTPA before Tx. Therefore, we recommend a programme involving more detailed vascular and physical examinations and more intensive education focusing on physical activity and DF prevention in at-risk patients before transplantation.

Introduction & Objective: Obesity is becoming a frequent comorbidity not only in patients with type 2 but also with type 1 diabetes mellitus (T1D). The aim of this study was to compare glucose control and prevalence of diabetes-associated complications in patients with T1D according to the presence of overweight or obesity. Methods: Using data from the diabetes registry of a tertiary center, patients with T1D were stratified according to BMI into normal weight (BMI<25 kg/m2: n=831, age 40 (29-52) years (yr), T1D duration 20 (11-29) yr), overweight (BMI 25-30 kg/m2, n=741, age 46 (35-60) yr, T1D duration 23 (15-33) yr) and obese (BMI ≥30 kg/m2, n=366, age 48 (39-59) yr, T1D duration 24 (16-34) yr) and compared based on glucose control parameters (HbA1c, continuous glucose sensor metrics) and prevalence of diabetic complications and metabolic comorbidities. Results: Compared with overweight and obese, patients with normal weight had a significantly lower HbA1c (56 (49-64) vs. 57 (51-65) vs. 60 (52-67), mmol/l, p<0.001), time above target range of 3.9-10 mmol/l(25 (16-39) vs. 28 (18-41) vs 31 (17-43)%, p=0.004) and average sensor glycemia (8.2 (7.4-9.3) vs. 8.4 (7.6-9.6) vs 8.7 (7.7-9.7) mmol/l, p<0.001). No significant difference was found in time in range or glycemic variability, while time below range was slightly lower in patients with obesity (3 (1-7) vs. 3 (1-7) vs. 3 (1-6), %, p=0.024). Higher BMI was associated with increased prevalence of arterial hypertension (32.7 vs. 46.4 vs. 65.6%, p=<0.001), dyslipidemia (38.4 vs. 54.1 vs. 70.7%, p=<0.001) and cardiovascular complications (6.7 vs. 10.1 vs. 12.4%, p= 0.004), as well as diabetic retinopathy (35.8 vs. 45.2 vs. 53.0%, p <0.001) and diabetic foot disease (4.6 vs. 5.5 vs. 9.7%, p= 0.003). Conclusion: Our data indicate that an increase in BMI in patients with T1D is associated with worse glucose control and higher rate of diabetes complications and metabolic comorbidities. Disclosure L. Horváth: None. M. Mraz: None. D. Vávra: None. K. Sochorova: Consultant; Medtronic, Ypsomed AG. R. Bem: Speaker's Bureau; Abbott, A.import (Dexcom, Tandem), Medtronic. Research Support; Ministry of Health - Czech republic. Speaker's Bureau; Novo Nordisk. J. Klouckova: None. M. Haluzik: Advisory Panel; Sanofi, Novo Nordisk, Eli Lilly and Company, AstraZeneca, Bayer Inc., Johnson & Johnson Medical Devices Companies. Consultant; Merck & Co., Inc., Sanofi, Novo Nordisk, Eli Lilly and Company, AstraZeneca, Bayer Inc., Boehringer-Ingelheim, Johnson & Johnson Medical Devices Companies, Novatin. Research Support; Sanofi. Speaker's Bureau; Sanofi, Novo Nordisk. Funding Supported by the project CarDia (Programme EXCELES, Project No. LX22NPO5104); Funded by the European Union; Next Generation EU and Funded by Ministry of Health, Czech Republic; Conceptual Development of Research Organization ("Institute for Clinical and Experimental Medicine – IKEM, IN 00023001").

Background: The objective of this systematic review is to summarize the available animal models of ischemic limbs, and to provide an overview of the advantages and disadvantages of each animal model and individual method of limb ischemia creation.Methods: A review of literature was conducted using the PubMed and Web of Science pages. Various types of experimental animals and surgical approaches used in creating ischemic limbs were evaluated. Other outcomes of interest were the specific characteristics of the individual experimental animals, and duration of tissue ischemia.Results: The most commonly used experimental animals were mice, followed by rabbits, rats, pigs, miniature pigs, and sheep. Single or double arterial ligation and excision of the entire femoral artery was the most often used method of ischemic limb creation. Other methods comprised single or double arterial electrocoagulation, use of ameroid constrictors, photochemically induced thrombosis, and different types of endovascular methods. The shortest duration of tissue ischemia was 7 days, the longest 90 days.Conclusions: This review shows that mice are among the most commonly used animals in limb ischemia research. Simple ligation and excision of the femoral artery is the most common method of creating an ischemic limb; nevertheless, it can result in acute rather than chronic ischemia. A two-stage sequential approach and methods using ameroid constrictors or endovascular blinded stent grafts are more suitable for creating a gradual arterial occlusion typically seen in humans. Selecting the right mouse strain or animal with artificially produced diabetes or hyperlipidaemia is crucial in chronic ischemic limb research. Moreover, the observation period following the onset of ischemia should last at least 14 days, preferably 4 weeks.