Diabetic kidney disease (DKD), a devastating microvascular complication of diabetes mellitus, arises from intricate crosstalk between metabolic disorders and immune dysregulation. Metabolomics has emerged as a powerful tool to unravel the immunometabolic mechanisms underlying DKD, facilitating early disease diagnosis, mechanistic pathway interpretation, and therapeutic biomarker discovery. This study aimed to systematically delineate the global research landscape and evolutionary trends of DKD metabolomics, with a key emphasis on immunometabolic interactions.

Relevant publications on DKD metabolomics published between 2015 and 2025 were comprehensively retrieved from the Web of Science Core Collection and PubMed databases. Multiple bibliometric and visualisation tools, including CiteSpace and the online bibliometric platform (https://bibliometric.com/), were utilised for data processing, visual mapping, and integrated quantitative and qualitative analysis.

A total of 1,410 eligible articles were included in the final analysis, among which 1,110 were sourced from the Web of Science and 299 from PubMed, demonstrating a sustained annual growth in publication volume. China contributed the largest number of publications (663 articles), followed by the United States (207 articles). The University of Michigan and Shandong University were the most productive research institutions. Li Ping and Sharma Kumar were identified as the leading productive and highly cited authors, respectively, with Kidney International recognised as the flagship journal in this research field. Keyword co-occurrence and co-citation analyses confirmed that DKD pathogenesis is predominantly governed by immunometabolic crosstalk. Specifically, renal lipotoxicity, oxidative stress, and insulin resistance trigger persistent renal inflammatory responses, while aberrant glucose metabolism, amino acid dysfunction, and gut microbiota disturbance disrupt renal immune homeostasis via the gut-kidney axis. Branched-chain amino acids and gut microbiota-derived metabolites serve as pivotal immunometabolic biomarkers. Clinical trial data from PubMed further validate the potential applications of these biomarkers, alongside the functional roles of immune regulatory molecules and their correlations with pathological alterations in DKD.

This bibliometric study systematically profiles the global research panorama of DKD metabolomics with a focus on immunometabolic regulation. It consolidates well-established research domains covering lipid- and oxidative stress-induced immune disorders, and further identifies amino acid metabolism-related immunomodulation as a burgeoning research frontier. These findings establish a refined research roadmap for future mechanistic investigations and the development of targeted immunometabolic therapeutic strategies for DKD.

While exercise remains an essential part of managing type 2 diabetes mellitus (T2DM), evidence remains limited regarding which exercise modalities and doses provide the greatest benefit for glycemic control in older adults. Accordingly, we assessed the effects of exercise modality and dose on glycated hemoglobin (HbA1c), fasting blood glucose (FBG), and 2-hour postprandial glucose (2hPG) in older adults with T2DM.

We searched seven databases from inception to 18 August 2025. Eligible studies were RCTs involving adults with T2DM in which the mean participant age was at least 60 years. Under a Bayesian framework, we conducted a random-effects network meta-analysis combined with dose-response modeling, using MET-min/week as the standardized metric for exercise dose, to assess the effects of exercise on glycemic control in older adults.

Thirty-four RCTs involving 2461 participants were ultimately included. The intervention network comprised continuous aerobic exercise (CAE), combined aerobic and resistance exercise (CE), traditional Chinese sports (TCS), resistance exercise (RE), high-intensity interval training (HIIT), and control group (CG). For HbA1c, CE (MD = -1.05%; 95% CrI: -1.43, -0.64), TCS (MD = -0.80%; 95% CrI: -1.09, -0.52), HIIT (MD = -0.63%; 95% CrI: -1.07, -0.19), CAE (MD = -0.48%; 95% CrI: -0.74, -0.24) and RE (MD = -0.44%; 95% CrI: -0.80, -0.08) showed greater reductions than CG. The SUCRA ranking probabilities favored CE for HbA1c (SUCRA = 94.86%). For FBG, significant reductions versus CG were observed for CE (MD = -1.44 mmol/L; 95% CrI: -2.16, -0.66), CAE (MD = -0.98 mmol/L; 95% CrI: -1.52, -0.41), and TCS (MD = -0.77 mmol/L; 95% CrI: -1.30, -0.23). Ranking probabilities favored CE (SUCRA = 88.03%). In the overall dose-response analyses, significant improvements began at approximately 520 MET-min/week for HbA1c and 500 MET-min/week for FBG. For 2hPG, evidence was limited. Only CE showed a significant reduction versus CG (MD = -3.15 mmol/L; 95% CrI: -5.42, -0.83), and no dose-response analysis was performed because of sparse evidence.

Exercise interventions improved glycemic control in older adults with T2DM, and CE appeared to show a favorable overall profile across glycemic outcomes. Total exercise volume showed nonlinear dose-response relationships with HbA1c and FBG. Further well-designed studies are needed to confirm these dose-response patterns, refine dose ranges for specific exercise modalities, and support more individualized exercise strategies.

https://www.crd.york.ac.uk/PROSPERO/view/CRD420251127345, identifier CRD420251127345.

Diabetic foot ulcers (DFUs) are a serious diabetes-related complication characterized by high rates of amputation and mortality. Emerging evidence suggests that DFUs are not simply the result of infection, but also involve microbiome dysbiosis, which impairs healing. Systemically, disturbances to the gut microbiota via the gut-skin axis promote systemic inflammation and metabolic dysfunction. Locally, skin microbial diversity is significantly reduced, allowing opportunistic pathogens such as Staphylococcus aureus and Pseudomonas aeruginosa to form resilient biofilms. These biofilms resist antibiotics and host immunity, while microbial virulence factors exacerbate tissue damage and disrupt the healing cascade. This synergy between host pathology and dysbiosis perpetuates chronic ulceration. Novel therapeutic strategies therefore aim to modulate this aberrant ecology by shifting from broad-spectrum eradication to targeted restoration. Promising approaches include probiotics, phage therapy, traditional Chinese medicine, and faecal microbiota transplantation, which seek to recalibrate the microbiome and promote healing. However, translation into clinical practice requires more robust evidence from large-scale trials. Future perspectives point towards personalized microbial medicine, integrating multi-omics data and artificial intelligence to match interventions with specific microbial ecotypes, which may reduce the global burden of DFUs.

Diabetic retinopathy (DR) is a leading cause of preventable visual impairment worldwide, and its precise severity grading is critical for optimizing clinical management. Conventional frameworks, notably the International Clinical Diabetic Retinopathy (ICDR) scale, are often hindered by substantial inter-observer variability and high dependency on specialist expertise. While deep learning (DL) has recently emerged as a transformative approach for automated stratification, a comprehensive synthesis of evidence regarding its diagnostic performance and clinical application remains lacking.

This systematic review and meta-analysis aimed to comprehensively assess the diagnostic accuracy of fundus image-based deep learning models in the grading of diabetic retinopathy.

PubMed, Embase, Web of Science, and the Cochrane Library were systematically searched for relevant studies published up to October 28, 2025. Diagnostic accuracy studies utilizing DL algorithms alongside ICDR criteria for diabetic retinopathy grading were included. Literature screening and data extraction were performed independently by two researchers, and the risk of bias was assessed using the QUADAS-2 tool.

A total of 41 studies were included, encompassing various DL architectures and multiple public and private fundus image datasets. In the five-class classification task based on ICDR criteria, the pooled sensitivities of DL-based models varied significantly across severity levels: 95.19% (95% CI: 93.00%-97.00%) for no DR (stage 0), 72.06% (95% CI: 62.06%-81.09%) for mild NPDR (stage 1), 84.33% (95% CI: 78.90%-89.10%) for moderate NPDR (stage 2), 75.84% (95% CI: 68.42%-82.57%) for severe NPDR (stage 3), and 78.82% (95% CI: 71.76%-85.13%) for PDR (stage 4). In the simplified four-class classification task, sensitivities markedly improved across all grades: 96.85% (95% CI: 90.18%-99.93%) for stage 0, 92.94% (95% CI: 79.50%-99.72%) for stage 1, 92.75% (95% CI: 79.31%-99.61%) for stage 2, and 88.19% (95% CI: 68.99%-98.93%) for stage 3.

DL exhibits high sensitivity and substantial potential for DR grading, particularly in screening for no DR and vision-threatening DR. Nevertheless, precisely differentiating between adjacent non-proliferative stages remains a clinical challenge. The observed heterogeneity underscores the imperative for methodological standardization, rigorous external validation, and multimodal data integration. Future research should prioritize enhancing clinical utility and generalizability to facilitate their translation into real-world clinical practice.

https://www.crd.york.ac.uk/PROSPERO/, identifier CRD420261338867.

Diabetic ketoacidosis (DKA) is a life-threatening acute complication of diabetes mellitus, most commonly precipitated by infection. Patients with diabetes are at a significantly elevated risk for severe infections and sepsis. Odontogenic infections represent a prevalent but potentially overlooked source of sepsis in this population, given the established bidirectional relationship between diabetes and periodontitis.

A 57-year-old female with poorly controlled type 2 diabetes presented with a one-week history of left jaw pain and one day of severe dyspnea and lethargy. Examination revealed left submandibular swelling, Kussmaul respirations, and lethargy. Laboratory findings confirmed severe DKA (pH 7.04, β-hydroxybutyrate 7.19 mmol/L) and sepsis (leukopenia, procalcitonin 2.34 ng/mL). Echocardiography demonstrated severe left ventricular systolic dysfunction (ejection fraction 36%) with normal cardiac biomarkers, consistent with sepsis-induced cardiomyopathy (SICM). Imaging identified a left submandibular abscess. Management involved prompt sepsis bundle implementation, including early empiric antibiotics (amoxicillin-clavulanate and tinidazole), aggressive fluid resuscitation, and concurrent intravenous insulin infusion for DKA, alongside proactive electrolyte repletion. Multidisciplinary source control was achieved. Outcomes: The patient showed rapid clinical improvement. Metabolic acidosis and ketosis resolved. Infection markers normalized. Remarkably, follow-up echocardiography on day 14 demonstrated complete recovery of cardiac function (ejection fraction 72%). She was discharged after 15 days and remained well at one-month follow-up.

This case underscores that odontogenic infection can be a potent trigger for life-threatening DKA and septic shock with multi-organ dysfunction in diabetic patients. It highlights the critical importance of thorough physical examination, including oral cavity inspection, in the evaluation of sepsis or DKA. The presentation and complete reversibility of SICM reinforce that treatment must focus on correcting the underlying septic and metabolic insults. Early multidisciplinary management is essential for optimal outcomes.

A 14-year-old female with type 1 diabetes for over six years presented with abdominal pain for two days and vomiting for three hours. Her regimen includes NovoRapid before meals and glargine at bedtime, but she has irregular meal patterns, inconsistent blood glucose monitoring, and poor insulin adherence. She has a history of recurrent diabetic ketoacidosis (DKA) and related hospitalizations.

The patient is alert but mildly weak, with adequate responsiveness. Breathing is slightly increased, deep, and regular. Extremities are cool. Radial pulses are strong, but dorsalis pedis pulses are diminished bilaterally. There is scattered right abdominal tenderness. The liver is palpable 5 cm below the costal margin, with medium texture and blunt borders, and percussion tenderness is positive.

Venous blood glucose was 21.86 mmol/L, and HbA1c was 10%. Arterial blood gas showed pH 6.933, pCO2 31.4 mmHg, and BEb -25.3 mmol/L. Insulin was below 1.39 pmol/L, and C-peptide below 0.003 pmol/L. Urinalysis revealed glucose 4+ and ketones 1+, leading to a diagnosis of type 1 DKA. Liver function was abnormal, with hepatomegaly and blood lactate at 9.85 mmol/L. Metabolic, infectious, and immune causes were excluded. Liver biopsy confirmed glycogenic hepatopathy, establishing a diagnosis of Mauriac syndrome with lactic acidosis. Serum uric acid ranged from 434 to 545 mmol/L, indicating hyperuricemia. Whole-exome sequencing identified a homozygous HFE mutation, suggesting hereditary hemochromatosis, though ferritin was normal and no iron deposition was seen on pathology.

Continuous intravenous insulin therapy was initiated, along with diabetes education, dietary guidance, exercise, glucose monitoring, and symptomatic care. An insulin pump was subsequently used for continuous subcutaneous insulin infusion (with Gansulin R before meals). Later, the pump was discontinued and switched to a four-injection subcutaneous regimen, consisting of NovoRapid before meals and insulin glargine at bedtime. Bicyclol Tablets was given for hepatoprotection.

The patient was hospitalized for 13 days, remained afebrile, and reported no discomfort. Blood glucose ranged from 3.1 to 15.2 mmol/L. Liver size slightly decreased, but tenderness persisted. Liver function and lipids were mostly normal at discharge. Two months later, she was readmitted for DKA; liver function was normal, but hepatomegaly with tenderness persisted.

Coronary heart disease (CHD) remains a leading cause of morbidity and mortality worldwide. Early identification of high-risk hypertensive patients is crucial for preventing cardiovascular events. While traditional risk scores rely on static clinical measurements, 24-h ambulatory blood pressure monitoring (ABPM)-derived time in target range (TTR) captures dynamic blood pressure control patterns that may improve risk stratification. Machine learning methods, particularly deep neural networks, offer an enhanced capability to model complex non-linear relationships in high-dimensional clinical data, compared with conventional statistical approaches.

This single-center retrospective cohort study included 1,026 patients admitted between January 2023 and December 2024, with 718 patients allocated to model development and 308 to internal validation. A deep neural network model with three hidden layers was developed and compared against eight conventional machine learning algorithms (logistic regression, naïve Bayes, k-nearest neighbors, random forest, support vector machine, XGBoost, LightGBM, and CatBoost). Thirty-two variables spanning demographics, clinical data, laboratory results, echocardiographic measures, and blood pressure indices were evaluated. Continuous variables were discretized into quartile-based categories to enhance clinical interpretability. Feature selection employed a two-step process combining the Boruta algorithm and least absolute shrinkage and selection operator (LASSO) regression, with variance inflation factor analysis confirming the absence of collinearity. Model selection prioritized balanced performance across discrimination (AUC), calibration (Brier score), and clinical utility (decision curve analysis) in the independent validation cohort. Interpretability was evaluated using SHAP (SHapley Additive exPlanations) values.

The deep neural network model achieved optimal balanced performance with an AUC of 0.822 (95% CI: 0.793-0.850) in the training cohort and 0.796 (95% CI: 0.749-0.846) in the validation cohort, accompanied by the lowest Brier score (0.172), indicating superior calibration. Nine predictors were retained: diabetes mellitus, mean systolic blood pressure, time in target range of systolic blood pressure, left atrial diameter, left ventricular end-systolic diameter, left ventricular ejection fraction, use of antihypertensive medications, calcium channel blockers, and β-blockers. SHAP analysis identified TTR and blood pressure control parameters as the primary drivers of model predictions.

The developed deep neural network model enables early identification of high-risk CHD patients with hypertension through interpretable, routinely available clinical variables. Prospective multicenter external validation is warranted to confirm its generalizability across diverse populations and clinical settings.

Type 2 diabetes mellitus is a major contributor to cognitive dysfunction and neurodegeneration, driven by complex metabolic, vascular, and inflammatory disturbances. Although conventional antidiabetic therapies primarily focus on glycemic control, few effectively preserve the integrity of the neurovascular unit (NVU), a critical determinant of brain health. This review examines the neuroprotective potential of dipeptidyl peptidase-4 (DPP-4) inhibitors, highlighting their unique ability to link metabolic regulation with neural and vascular preservation. A literature search was conducted in PubMed and Google Scholar for English-language articles published up to December 2025, using keywords related to cognitive dysfunction, DPP-4 inhibitors, incretins, glucagon-like peptide-1, and the nervous system. Eligible studies included original research, randomized trials, meta-analyses, animal studies, reviews, and mechanistic investigations addressing the effects of DPP-4 inhibitors on NVU stability. Editorials and studies lacking relevance to diabetes-related cognitive impairment or clear biological mechanisms were excluded. Evidence indicates that DPP-4 inhibitors exert dual neuroprotective actions by enhancing incretin signaling (GLP-1/GIP), which supports synaptic plasticity and attenuates neuroinflammation, and by preserving stromal cell-derived factor-1α, thereby activating the CXCR4 pathway to promote endothelial repair. Additional benefits include modulation of the Nrf2/GPX4 axis, reducing oxidative stress and ferroptosis in neural tissue. Comparative analyses suggest potential advantages over other antidiabetic classes, although clinical data on dementia risk remain heterogeneous. Overall, DPP-4 inhibitors emerge as promising agents for stabilizing the NVU beyond glucose lowering, supporting their repositioning in the management of diabetes-associated cognitive dysfunction.

The lack of a standardised definition of large for gestational age (LGA) complicates clinical assessments and the interpretation of research.

The objective of this study is to assess how different definitions of LGA affect the interpretation of outcomes in a stepped wedge cluster randomised controlled trial (SW-CRT).

Secondary analysis of the Changing Diagnostic Criteria for Gestational Diabetes (CDC4G, ISRCTN41918550) study population (n = 47,080 pregnancies), including eight delivery units in Sweden during 2018. The outcome measures were the prevalence of LGA and severe LGA before and after switching from the previous Swedish diagnostic criteria for gestational diabetes (SWE-GDM) to the World Health Organisation (WHO-2013) criteria. LGA and severe LGA were defined as birthweight > 90th and > 97th percentile, respectively, using the following reference growth charts: The new Swedish reference ranges for foetal weight (SWE-REF), the Gestation Related Optimal Weight (GROW) and the International Foetal and Newborn Growth Consortium for the 21st Century (Intergrowth-21st). Analyses were performed on a modified intention-to-treat (mITT) population and on the subgroup discordant for GDM diagnosis between new and former criteria, i.e., glucose levels between the WHO-2013 and SWE-GDM diagnostic thresholds.

In the mITT population, the prevalence of LGA decreased from 19.4% to 13.1% after the switch to the WHO-2013 GDM criteria, as defined by SWE-REF (adjusted relative risk [RR], 0.92; 95% confidence interval [CI] 0.86, 0.98). There was no change in severe LGA using any of the reference charts. In the subgroup, LGA defined by SWE-REF (RR 0.68, 95% CI 0.50, 0.91) and GROW (RR 0.78, 95% CI 0.63, 0.95) was reduced, but not when defined by Intergrowth-21st (RR 0.88, 95% CI 0.74, 1.06). Severe LGA was reduced only when defined by Intergrowth-21st (RR 0.73, 95% CI 0.62, 0.85).

The choice of LGA definition affected the interpretation of the results in the CDC4G trial.

To highlight retinal ischemia as an important component of diabetic retinopathy (DR) pathogenesis, a significant prognostic factor, and a driver of various DR complications, including neovascularization. We aim to describe the main features of ischemic areas that can be identified through fundoscopy, color fundus photography, optical coherence tomography (OCT), and OCT angiography and provide clinicians with a step-by-step approach to identifying prognostically unfavorable markers of DR and accurately staging DR to improve management and prevent severe proliferative DR.

Cases of DR with subtle but significant retinal findings that are not readily identifiable unless a systematic multimodal imaging approach is used. We emphasize the presence of retinal ischemia in each case and show how to identify retinal vascular nonperfusion and its complications.

Advanced stages of DR may lack prominent fundus findings and may be easily misdiagnosed as milder DR without a comprehensive multimodal imaging diagnostic approach.