Volume 18 · Issue 04 · 2026

Type 1 diabetes mellitus (T1DM) is an autoimmunity-mediated chronic metabolic disease characterized by the progressive failure of pancreatic β-cell function, requiring lifelong dependence on exogenous insulin to sustain life. In 2025, multiple global breakthroughs were achieved in the field of T1DM, particularly regarding the refinement of epidemiological features, early diagnosis and risk prediction, treatment innovations, and whole-life-cycle management. This review systematically integrates key research findings published in authoritative international journals in 2025, categorizing them into four main dimensions: prevalence, diagnosis and prediction, treatment, and management. Furthermore, it analyzes the current limitations of research and explores the future directions toward precision, intelligent, and functional cures, providing a reference for clinical practice and scientific exploration.

Diabetic foot is one of the most severe and complex complications of diabetes. It not only significantly impairs patients′ quality of life but also imposes a heavy economic burden on families and society, becoming a major global public health challenge. Currently, regional disparities exist in the diagnosis and treatment of diabetic foot in China, with numerous challenges remaining in its comprehensive assessment and tiered diagnosis and treatment. The Burn Surgery Branch of the Chinese Medical Association and the Yangtze River Delta Integrated Diabetic Foot Disease Alliance jointly developed the Practical guideline on the prevention and management of diabetic foot in China, which covers comprehensive medical assessment, integrated medical and surgical treatment, and the construction of diagnostic and therapeutic systems. From the perspective of endocrinologists, this article interprets the guideline regarding the comprehensive assessment, advanced technologies, and frontiers of multidisciplinary tiered diagnosis and treatment of diabetic foot, aiming to enhance clinicians′ capabilities in the comprehensive diagnosis and treatment of this disease.

To investigate the predictive value of the triglyceride-to-high-density lipoprotein cholesterol ratio (TG/HDL-C) for the risk of incident type 2 diabetes mellitus (T2DM) in adults.

This was a retrospective cohort study. Data were extracted from the NAFLD in the Gifu area, longitudinal analysis (NAGALA) project, which has been uploaded to the Dryad public database. The study population comprised adults who underwent health examinations at Murakami Memorial Hospital from 2004 to 2015. Baseline data were collected, including sex, physical activity, alcohol consumption, triglycerides (TG), and high-density lipoprotein cholesterol (HDL-C). TG/HDL-C was calculated using TG and HDL-C values. Participants were divided into five groups according to baseline TG/HDL-C: Q₁ (TG/HDL-C 0.10-0.26), Q₂ (TG/HDL-C 0.27-0.40), Q₃ (TG/HDL-C 0.41-0.62), Q₄ (TG/HDL-C 0.63-1.03), and Q₅ (TG/HDL-C 1.04-3.89). Participants were followed up 1-2 times annually, with incident T2DM as the study outcome. Cox proportional hazards regression models were used to evaluate the association between TG/HDL-C and incident T2DM, with subgroup analyses performed. Kaplan-Meier curves were plotted to examine changes in T2DM incidence over time across groups, and differences were assessed using the log-rank test.

A total of 15 453 participants were included. Over a 10-year follow-up, the overall incidence rate of T2DM among the study subjects was 2.41% (373/15 453). The cumulative incidence of T2DM in Q₁-Q₅ was 0.61% (19/3 093), 0.84% (26/3 088), 1.42% (44/3 091), 2.82% (87/3 090), and 6.37% (197/3 091), respectively. After adjustment for multiple confounders, the results of the Cox proportional hazards regression model showed that the risk of T2DM continued to increase progressively from group Q₁ to group Q₅; the risk of T2DM in Q₅ was 2.550 times that in Q₁ (95%CI 1.522-4.271). For each 1-standard-deviation increase in TG/HDL-C, the risk of incident T2DM increased by 24.3% (HR=1.243, 95%CI 1.184-1.304). Kaplan-Meier analysis showed a gradual increase in T2DM incidence from Q₁ to Q₅ (log-rank P<0.001). Subgroup analyses indicated interactions of sex, physical activity, and alcohol consumption with the association between TG/HDL-C and T2DM risk (P for interaction<0.05). The association was stronger in women (HR=1.861, 95%CI 1.648-2.102) and individuals without regular exercise (HR=1.573, 95%CI 1.492-1.659).

TG/HDL-C is an independent predictor for T2DM, and elevated TG/HDL-C levels are associated with an increased risk of incident T2DM. This association is particularly pronounced in women and individuals lacking regular physical activity.

To investigate the value of the metabolic score for insulin resistance (METS-IR) in the prognostic assessment of diabetic kidney disease (DKD) in patients with type 2 diabetes mellitus (T2DM).

This was a retrospective cohort study. Patients with T2DM complicated by chronic kidney disease (CKD) who attended the Department of Nephrology at the First Affiliated Hospital of Nanjing Medical University from January 2013 to December 2020 were selected as the subjects. Baseline clinical data and pathological indicators were collected, and the patients were categorized into three groups based on the tertiles of METS-IR levels: the METS-IR≤38.96 group (55 cases), the 38.96<METS-IR≤42.93 group (55 cases), and the METS-IR>42.93 group (55 cases). Follow-up was performed to assess the primary outcome of renal endpoint events. Logistic regression analysis was used to evaluate the effect of METS-IR on the risk of diabetic kidney disease (DKD).Cox regression model was used to identify the influencing factors of poor renal prognosis. Kaplan-Meier survival curves were used to compare cumulative renal event-free survival rates. Smoothing curve fitting and threshold effect analysis were used to analyze the relationship between METS-IR and poor renal prognosis. The receiver operating characteristic (ROC) curve was used to analyze the diagnostic value of METS-IR alone and in combination with estimated glomerular filtration rate (eGFR) and 24-hour urinary total protein in the evaluation of renal prognosis. Decision curve analysis (DCA) was used to determine the net benefit of the prediction model.

A total of 165 patients with T2DM complicated by CKD were included in this study. During a median follow-up of 32.0 (14.92, 44.47) months, 87 patients (52.73%) experienced the renal composite endpoint. Compared with the the METS-IR≤38.96 group, the the METS-IR>42.93 group had more significant 24-hour proteinuria, worse renal function, and more severe pathological changes. Multivariate logistic regression analysis identified that higher levels of METS-IR were significantly associated with an increased risk of pathologically proved DKD (P<0.05). Kaplan-Meier survival analysis showed that the cumulative survival rate without renal endpoint events was lower in METS-IR>42.93 group than that in METS-IR≤38.96 group (χ²=12.505, P<0.05). Cox regression analysis showed that METS-IR was an independent risk factor for renal endpoint events (P<0.05). There was a linear relationship between METS-IR and the relative risk of poor renal prognosis (β=1.04, 95%CI 1.02-1.07). ROC curve analysis showed that the AUC of METS-IR was 0.749 (95%CI 0.675-0.822), and the AUC of METS-IR combined with eGFR and 24-hour urinary total protein could reach 0.860 (95%CI 0.805-0.915). The DCA results showed that the clinical variable model based on METS-IR to predict renal outcome would get more net benefits than the intervention-all scheme and intervention-none scheme.

In patients with T2DM and CKD, a higher METS-IR level is associated with more severe renal function impairment and can aid in clinicaldifferential diagnosis. Furthermore, METS-IR exhibits a positive linear correlation with, and serves as an independent risk factor for, poor renal prognosis.

To compare the real-word glycemic outcomes and safety of the open-source automated insulin delivery (OS-AID) system and the commercial AID system (MiniMed^TM 780G) in patients with type 1 diabetes mellitus (T1DM).

This real-world retrospective cohort study consecutively enrolled patients with T1DM who initiated either OS-AID (Loop system) or MiniMed^TM 780G during routine outpatient follow-up between September 2021 and March 2025 at Peking Union Medical College Hospital. Glycemic outcomes were evaluated at baseline and at 1, 3, and 6 months after AID initiation. The primary outcome was the variation of time in range (TIR) between pre-AID initiation and after 6 months of AID use. Secondary outcomes included the changes in other CGM-derived metrics, including time in tight range (TITR), time below range (TBR), glucose management indicator (GMI), mean glucose (MG), and coefficient of variation (CV), as well as safety events during follow-up. Between-group comparisons were performed using independent-samples t tests, Mann-Whitney U tests, or χ² tests as appropriate. Analysis of covariance (ANCOVA) was applied to compare 6-month endpoints adjusted for baseline values and relevant covariates, and linear mixed-effects models were used to evaluate longitudinal changes and between-group differences.

A total of 55 patients with T1DM were screened, of whom 40 met the predefined criteria and were included in the final analysis (20 in the OS-AID group and 20 in the MiniMed^TM 780G group). After 6 months of AID use, TIR increased by 13.8% (95%CI 9.1%-18.5%) in the OS-AID group and by 14.6% (95%CI 9.9%-19.3%) in the MiniMed^TM 780G group, with no significant between-group difference in TIR change (t=-0.279, P=0.781). Nighttime TIR was higher in the MiniMed^TM 780G group than in the OS-AID group (F=11.313, P=0.001). Regarding secondary outcomes, no statistically significant between-group differences were observed in changes in TITR, GMI, or MG (all P>0.05). The OS-AID group exhibited a higher TBR compared with the MiniMed^TM 780G group (F=10.771, P=0.003), while no statistically significant difference was found in CV between groups (F=0.002, P=0.969). No severe hypoglycemia or diabetic ketoacidosis events were reported during follow-up.

Both the Open-source AID and the MiniMed^TM 780G systems can effectively and safely improve glycemic metrics in patients with T1DM in a real-world setting, with overall comparable efficacy. However, the MiniMed^TM 780G system exhibits a certain advantage over the OS-AID system in reducing hypoglycemia exposure.

To investigate the relationship of serum thyroid hormones (TH) levels with the shape and peak time of the glucose curve during an oral glucose tolerance test (OGTT) in patients with type 2 diabetes mellitus (T2DM).

This was a single-center cross-sectional study. Consecutive patients with T2DM who were hospitalized in the Department of Endocrinology, Hebei General Hospital, between May 2018 and March 2023 were enrolled. Data on height, body weight, free triiodothyronine (FT3), free thyroxine (FT4), total triiodothyronine (TT3), total thyroxine (TT4) were collected. Body mass index (BMI), the thyroid feedback quantile-based index (TFQI) and the FT3/FT4 ratio were calculated. Based on the OGTT glucose peak time, patients were divided into an early-peak group (glucose peak≤60 min) and a non-early-peak group (glucose peak>60 min).According to dynamic glucose changes, patients were classified into a monophasic curve group (decrease>0.25 mmol/L after glucose peak) and a sustained elevation curve group (continuous glucose elevation without decline). Between-group comparisons were performed using the two independent samples t-test, Mann-Whitney U test, or χ² test. Multivariate logistic regression analysis was used to analyze the correlation between relevant thyroid hormone levels and the OGTT glucose curve shape and peak time. Subgroup analysis was performed according to BMI (BMI≤24 kg/m² and BMI>24 kg/m²) to verify whether obesity affected the association of TH levels with OGTT peak time and curve shape.

A total of 405 patients with T2DM were finally enrolled, including 112 in the early-peak group and 293 in the non-early-peak group, and 275 in the monophasic group and 130 in the sustained elevation group. Compared with the early-peak group, FT3, FT4, TT3, and the FT3/FT4 ratio were significantly lower in the non-early-peak group. Compared with monophasic group the FT3, TT3, TT4, and the FT3/FT4 ratio were significantly lower in the sustained elevation group (all P<0.05). After adjusting for multiple confounders, multivariate logistic regression showed that FT3 (OR=2.83, 95%CI 1.72-4.67), FT4 (OR=1.17, 95%CI 1.13-1.31), and TT3 (OR=5.13, 95%CI 4.62-8.11) were associated with peak time, and TT3 (OR=5.08, 95%CI 3.41-7.44) and TT4 (OR=1.03, 95%CI 1.02-1.05) were associated with curve shape. TFQI was positively correlated with an early OGTT peak (OR=3.88, 95%CI 1.46-10.27). Subgroup analysis showed that in T2DM patients with BMI>24 kg/m², the FT3 level was positively correlated with an early-peak curve (OR=2.55, 95%CI 1.21-6.15). In T2DM patients with BMI≤24 kg/m², FT3 (OR=2.42, 95%CI 1.23-5.57), FT4 (OR=1.15, 95%CI 1.06-1.27), and TT3 (OR=9.44, 95%CI 2.75-32.23) levels were positively correlated with an early-peak curve, while FT3 (OR=1.57, 95%CI 1.16-2.63), TT3 (OR=4.15, 95%CI 1.33-12.26), and TT4 (OR=1.03, 95%CI 1.02-1.05) levels were positively correlated with a monophasic curve.

In patients with T2DM, levels of FT3, FT4, TT3, and TT4 are elevated. The OGTT peak tends to occur earlier, and the glucose response curve more frequently exhibits a monophasic pattern. Additionally, the associations between thyroid hormone levels, OGTT peak time, and curve morphology vary significantly across different BMI strata.

To investigate the efficacy and safety of a hybrid closed-loop insulin delivery system (HCLS) in Chinese adults with type 2 diabetes mellitus (T2DM).

This was a single-arm, prospective, multicenter clinical trial. From October 2023 to May 2024, 19 participants with T2DM were enrolled from six centers. Participants underwent a 2-week baseline therapy phase followed by a 12-week closed-loop therapy phase with continuous glucose monitoring (CGM).The primary outcome was the change of time in range (TIR), and the secondary outcomes included glycated hemoglobin A_1c (HbA_1c), time above range (TAR), time below range (TBR), mean sensor glucose (MSG), coefficient of variation (CV), and changes in body weight between the baseline therapy and study periods. Safety events included rates of severe hypoglycemia, severe hyperglycemia, and diabetic ketoacidosis (DKA). Paired sample t-tests or Wilcoxon signed-rank tests were used to analyze and compare changes in various indicators between the baseline and closed-loop therapy periods.

Compared with baseline therapy, HCLS significantly improved TIR in adults with T2DM. TIR (3.9-10.0 mmol/L) increased from 69.86%±16.02% to 76.17%±11.67%, with a mean increase of 6.31% (P=0.041). HbA_1c decreased from 8.09%±1.20% to 6.81%±0.62%, with a mean reduction of 1.28% (P<0.001). TAR (>10.0 mmol/L) decreased from 28.58%±17.17% to 23.29%±11.99%, with a mean reduction of 5.20%, but the difference was not statistically significant (P=0.106). There were no statistically significant differences in TBR (<3.9 mmol/L), MSG and CV (P=0.258, 0.530, 0.166, respectively). Following HCLS therapy, body weight increased by 1.10 kg, although this difference was not statistically significant (P=0.068). Regarding safety, no device-related adverse events occurred during the trial, and there were no incidents of severe hypoglycemia, severe hyperglycemia, or DKA.

HCLS can significantly improve blood glucose levels in adult patients with T2DM without increasing the risk of hypoglycemia, preliminarily verifying its efficacy and safety.

To compare the consistency and safety of the cardiac autonomic nervous function analysis software (CANFASt) versus the electrocardiograph (ECG) machine method in diagnosing Cardiac Autonomic Neuropathy (CAN) in diabetic patients.

This study was a self-paired, multi-center clinical trial. From December 2023 to July 2025, diabetic patients were recruited from the Endocrinology Departments of Shenzhen People′s Hospital, Shenzhen University General Hospital, the Diabetes Department of Shenzhen Bao′an Traditional Chinese Medicine Hospital, and the Endocrinology Department of Shenzhen Bao′an People′s Hospital. All subjects simultaneously underwent cardiac autonomic reflex tests (CARTs) and heart rate variability (HRV) analysis lasting >5 minutes using both CANFASt and the ECG machine method. The results of various indicators of CARTs (the longest RR interval during the lying-to-standing for 25-35 heart beats, the shortest RR interval during the lying-to-standing for 10-20 heartbeats, the heart rate response to lying to standing; the longest RR interval during deep breathing, the shortest RR interval during deep breathing, heart rate response to deep breathing; the longest RR interval during the Valsalva maneuver, the shortest RR interval during the Valsalva maneuver, the Valsalva ratio) were collected. Moreover, the indicators of HRV were collected [the standard deviation of normal-to-normal interval (SDNN), the root mean square of successive RR interval difference (rMSSD), standard deviation of successive difference (SDSD), the percent of adjacent RR intervals with a difference more than 50 ms (pNN50)]. The primary endpoints for consistency evaluation were positive percent agreement and negative percent agreement. Secondary endpoints for consistency included overall percent agreement, kappa coefficient for categorical variables, and intraclass correlation coefficient (ICC) for continuous variables. Safety was evaluated based on the number and incidence of adverse events. Paired t-tests were used to compare the interpretation time required by the two methods.

A total of 150 diabetic subjects were recruited. Based on the ECG machine method, 60 cases were diagnosed as CAN-positive and 90 as CAN-negative. CANFASt diagnosed 61 cases as CAN-positive and 89 as CAN-negative. The positive percent agreement between CANFASt and the ECG machine method for CAN diagnosis was 1.000 (0.940-1.000), and the negative percent agreement was 0.989 (0.940-0.998). The overall percent agreement for CAN diagnosis was 0.993 (0.963-0.999), with a kappa value of 0.986 (P<0.001). The weighted kappa value for assessing the severity of CAN was 0.991 (P<0.001). The ICC coefficients for CARTs parameters assessed by CANFASt and the ECG machine method were as follows: longest RR interval during beats 25-35 of lying-to-standing (0.990); shortest RR interval during beats 10-20 of lying-to-standing (0.988); heart rate response to lying to standing (0.993); longest RR interval during deep breathing (0.987); shortest RR interval during deep breathing (0.987); heart rate response to deep breathing (0.987); longest RR interval during Valsalva maneuver (0.991); shortest RR interval during Valsalva maneuver (0.987); Valsalva ratio (0.984) (all P<0.001). The ICC coefficients for HRV parameters assessed by both methods were: SDNN (0.997); rMSSD (0.981); SDSD (0.938); pNN50 (0.995) (all P<0.001). In this clinical trial, no adverse events such as dizziness, headache, palpitation, chest tightness, blackout, elevated blood pressure, or other conditions deemed by the physician as warranting discontinuation of the trial were observed. The interpretation time for CARTs using CANFASt and for HRV was significantly lower than that of the ECG machine method t=66.6,46.3,respectively, (both P<0.001).

CANFASt demonstrated acceptable diagnostic consistency with traditional ECG method for CARTs in diagnosing CAN, along with favorable safety.

To investigate the role and mechanism of aquaporin 1 (AQP1) in diabetic myocardial injury.

In this study, eight-week-old C57 BL/6J mice were completely randomized into a normal control (NC) group and db/db mice into a diabetic model (DM) group, 15 mice in each group. H9c2 cardiomyocytes were completely randomized into four experimental groups: normal glucose (NG) group, high glucose (HG) group, high glucose with AQP1 Inhibitor (HG+TC AQP1) group, and hypertonic solution (HS) group, with each group consisting of three replicates. Single-cell sequencing was conducted on mouse cardiac tissues, while echocardiography and Masson′s trichrome staining were utilized to evaluate myocardial function and pathological changes. The expression levels of aquaporin 1 (AQP1), the anti-apoptotic protein B-cell lymphoma 2 (Bcl-2), the apoptotic protein Bcl-2-associated X protein (Bax), and activated cleaved caspase-3 in tissues and cells were quantified using real-time quantitative PCR (qPCR) and Western blotting. Cardiomyocyte apoptosis was assessed via the TUNEL assay. Statistical comparisons between two groups were made using the independent samples t-test, while comparisons among multiple groups were analyzed using one-way analysis of variance (ANOVA). Subsequently, Tukey′s post hoc test was applied to perform pairwise comparisons between the groups.

Single-cell sequencing results indicate that compared to normal cardiac tissue, diabetic myocardial injury exhibits a significantly increased proportion of C0 cardiomyocyte subpopulations. The Aqp1 gene was highly expressed in the C0 subpopulation and significantly elevated under diabetic myocardial injury conditions (|log₂ fold change|≥1, P<0.05). Gene functional enrichment analysis shows that Aqp1 was primarily enriched in the apoptosis process. Animal experiments revealed that compared to the NC group, the DM group exhibited significantly upregulated Aqp1 gene expression (P<0.001), but no significant difference in protein levels (P>0.05). Cell experiments showed that compared to the NG group, the HG group had significantly upregulated Aqp1 gene expression (P<0.001), but significantly downregulated protein levels (P<0.05). Compared to the NG group, the HG group exhibited significant downregulation of the anti-apoptotic protein Bcl-2 and significant upregulation of the pro-apoptotic proteins Bax and cleaved caspase-3. Further downregulation of AQP1 protein expression using an inhibitor resulted in additional downregulation of the anti-apoptotic protein Bcl-2 and further upregulation of the pro-apoptotic proteins Bax and cleaved caspase-3 (P<0.05). TUNEL assays revealed significantly increased fluorescence intensity in the HG and HG+TC AQP1 1 groups compared to the NG group. Furthermore, fluorescence intensity in the HG+TC AQP1 1 group was significantly higher than that in the HG group (P<0.05).

AQP1 participates in the pathological changes of diabetic cardiomyopathy by promoting cardiomyocyte apoptosis in diabetic myocardial injury.

1 case reportedCUL7Cases of 3M syndrome with metabolic abnormalities caused by gene complex heterozygous mutations, in order to improve clinicians' understanding of the disease. The patient was a 26-year-old male, mainly due to slow growth from childhood, progressive weight gain, chronic course of disease, small birth length, low weight, slow growth from childhood, short stature, special appearance, severe obesity, and abnormal gonadal development. Auxiliary examination revealed abnormal glucose tolerance, insulin resistance, metabolic syndrome, mild valgus of knee joint, and poor physiological curvature of spine. Whole exome sequencing showed that the proband carriedCUL7Gene composite heterozygous variants: NM_014780.5: c.3201G>A (p.Trp1067) and NM_014780.5: c.3355+5G>A. Among them, NM_014780.5:c.3201G>A (p.Trp1067) variant was inherited from the mother and is a nonsense mutation, which has not been reported in the previous literature. NM_014780.5: The c.3355+5G>A variant is inherited from the father and is a splice site mutation. The final diagnosis was 3M syndrome type 1. The patient's epiphysis is currently closed, and growth hormone therapy is meaningless. Sex hormone levels are low and androgen replacement therapy is given. The patient is obese, severe insulin resistance, and abnormal glucose tolerance. The patient is instructed to eat, exercise and lose weight reasonably. After follow-up, the patient gradually lost weight and had an erection in the morning.

Diabetic nephropathy (DKD) is the main cause of end-stage renal disease, and early diagnosis is very important to delay the progression of the disease and improve the prognosis. At present, the urinary albumin/creatinine ratio (UACR) and the estimated glomerular filtration rate (eGFR) commonly used in clinical practice have certain limitations in the early recognition of DKD. In recent years, the development of omics technology provides an opportunity for discovering new biomarkers. Several studies have shown that the combined detection of multiple markers may be superior to UACR and eGFR in diagnostic efficacy. This paper systematically reviews the value of proteomics-based CKD273 classifier, PromarkerD model, V-ATPase, microRNA combination model, gut microbiome marker combination and metabolomics combination marker in early prediction and risk stratification of DKD. At the same time, from the perspectives of detection cost, technical standardization and clinical promotion feasibility, the challenges faced by the combined application of omics markers were analyzed, which provided a reference for the clinical translation of this strategy in the early diagnosis of DKD.

The persistence of chronic wounds such as diabetic foot ulcer and venous leg ulcer seriously affects the quality of life of patients, and their occurrence is closely related to dysfunction in each stage of wound repair. This paper reviews the dysfunctional mechanisms of four stages of wound repair (hemostasis, inflammation, proliferation and remodeling), focuses on the core pathological links of chronic wound persistence, and focuses on the key regulatory role of S100A9 protein and circadian rhythm in it, in order to provide new theoretical support for the development of time-dependent preparations targeting S100A9 and the optimization of clinical treatment strategies for chronic wounds such as diabetic foot ulcer.

The comorbidity of sarcopenic obesity (SO) and type 2 diabetes mellitus (T2DM) is becoming increasingly severe, driven by trends such as global population aging, insufficient physical activity and nutritional imbalance. SO and T2DM interact with each other through core mechanisms such as insulin resistance, chronic low-grade inflammation, autophagy dysfunction and intestinal flora dysbalance, which exacerbate body composition abnormalities and metabolic dysregulation. At present, nutrition and exercise therapy are the basic treatment options for SO combined with T2DM. There is currently no recognized specific drug for SO in clinic. When hypoglycemic drugs, metabolic surgery and other methods are used to reduce glucose and lipid, it is necessary to pay attention to its potential impact on muscle. This paper briefly reviews the research progress of SO and T2DM in order to provide reference for the prevention and treatment of SO and T2DM.