Volume 18 · Issue 09 · 2026

The recently published exploratory analysis of the Soli-D study was based on a 24-week, multicenter, randomized, open-label clinical trial. The study enrolled 386 Chinese adults with type 2 diabetes who had inadequate blood glucose control (HbA1c 7.5%-11.0%) on one or two oral antidiabetic drugs. Patients were stratified by baseline β-cell function to compare the efficacy and safety of insulin glargine/lixisenatide(iGlarLixi) with insulin degludec/insulin aspart (IDegAsp). The results demonstrated that regardless of baseline β-cell function, iGlarLixi was superior to IDegAsp in glycemic control, safe target attainment, hypoglycemia risk, and weight management. Notably, in the subgroup with the poorest β-cell function and the highest risk of hypoglycemia, the rate of hypoglycemic events with iGlarLixi was significantly lower than with IDegAsp, highlighting its safety advantage in populations at high risk of hypoglycemia.

The emergence of novel hypoglycemic agents has reshaped the landscape of diabetes management, necessitating a re-evaluation of metformin′s traditional role as a first-line therapeutic agent. This study systematically reviews the evolutionary positioning and clinical value of metformin in the modern diabetes treatment system. Based on domestic and international authoritative guidelines, evidence-based medical studies, and real-world data, this article comprehensively analyzes the clinical application of metformin from multiple dimensions. The traditional stepwise treatment strategy and the exclusive first-line status of metformin have been abandoned, and a comorbidity-stratified therapeutic system has been established. Novel hypoglycemic agents are preferentially recommended for diabetic patients complicated with cardiovascular diseases, heart failure, or chronic kidney disease, whereas metformin remains a first-line option for uncomplicated non-obese diabetic patients. Current clinical strategies advocate metformin-based early combination therapy with an emphasis on cardiorenal protection. Metformin exerts hypoglycemic effects through multiple signaling pathways and provides long-term metabolic benefits, with potential anti-aging and anti-tumor properties. Its combination with novel hypoglycemic agents can further improve long-term patient prognosis; however, clinical application is accompanied by potential adverse risks such as lactic acidosis and vitamin B12 deficiency. In conclusion, metformin retains crucial clinical value in the era of novel hypoglycemic drugs and has evolved from an exclusive first-line monotherapy to a fundamental synergistic agent. It is applicable to combination therapy for patients with complications, early combined hypoglycemic regimens, and basic treatment for patients with uncomplicated diabetes. Individualized and precise administration of metformin should be implemented in clinical practice based on patients′ conditions, disease risks, drug accessibility, and economic benefits.

To examine the association between a body shape index (ABSI) and cognitive impairment in patients with type 2 diabetes mellitus (T2DM), and to characterize its nonlinear relationship and potential threshold.

The present study employs a cross-sectional approach, supplemented by validation in an external longitudinal cohort. A total of 1, 485 patients with T2DM who were hospitalized in the Department of Endocrinology of Nanjing Drum Tower Hospital from January 2023 to December 2025 were enrolled. They were divided into a normal cognition group (n=683) and a cognitive impairment group (n=802) based on Montreal Cognitive Assessment (MoCA) scores. Anthropometric measurements were collected and ABSI was calculated. Multivariable logistic regression was used to examine the association between ABSI and cognitive impairment. Restricted cubic spline (RCS) analysis was applied to investigate the nonlinear relationship, and piecewise multivariable logistic regression was used to analyze threshold effects. Subgroup analyses were further conducted to assess the consistency of the association. For external validation, 502 patients with T2DM from the Health and Retirement Study (HRS) who completed three follow-up visits in 2010, 2014, and 2018 were selected as the validation cohort. The threshold derived from the primary analysis was used for stratification, and Cox proportional hazards regression was employed to validate the stability of the threshold.

The 1 485 T2DM patients had a mean age of 61.3±6.5 years, and the cohort comprised 917 males (61.8%) and 568 females (38.2%). ABSI was higher in the cognitive impairment group than in the normal cognition group (8.33±0.50 vs. 8.24±0.43, P<0.001). After adjusting for sex, age, years of education, hypertension, diabetes duration, use of glucose-lowering drugs, glycated hemoglobin, and triglycerides, multivariable logistic regression showed that ABSI was significantly associated with cognitive impairment (OR=1.339, 95%CI 1.051-1.707, P=0.018). RCS analysis revealed a significant nonlinear J-shaped relationship (P for nonlinearity=0.003). The piecewise multivariable logistic regression identified a threshold of 8.31; when ABSI ≥8.31, cognitive impairment in T2DM patients was positively correlated with ABSI (OR=2.275, 95%CI 1.395-4.390, P=0.002). Subgroup analyses showed no statistically significant interactions between ABSI and sex, age, or years of education (all P>0.05). In the external validation cohort, ABSI ≥8.31 was associated with a higher risk of cognitive impairment (HR=1.705, 95%CI 1.009-2.882, P=0.046).

ABSI was associated with cognitive impairment in patients with T2DM in a nonlinear J-shaped manner, and the association became more pronounced when ABSI ≥8.31.

To apply unsupervised cluster analysis to explore the heterogeneous subtypes of the pre-diabetes mellitus (Pre-DM) population and to analyze the association between each subtype and glomerular filtration function impairment.

In this cross-sectional study, participants with prediabetes were recruited from those who attended health check-ups at the Health Management Center of the First Affiliated Hospital of Xinjiang Medical University from January to December 2023. The age, sex, height, weight, glycated hemoglobin A_1c (HbA_1c), fasting plasma glucose (FPG), triglyceride (TG), high-density lipoprotein-cholesterol (HDL-C), alanine transaminase (ALT), and serum cystatin C (Cys) of the research subjects were collected. Calculate their body mass index (BMI), triglyceride-glucose (TyG) index, triglyceride to high-density lipoprotein cholesterol ratio(TG/HDL-C), estimated glomerular filtration rate based on cystatin C (eGFRcys), and the prevalence of glomerular filtration dysfunction [eGFRcys<90 ml·min^-1·(1.73 m²)^-1]. A multivariate linear regression model was used to analyze the influence of different clustering groups on the eGFRcys level.

A total of 19 793 patients with prediabetes were ultimately enrolled, including 12 707 males (64.20%) and 7 086 females (35.80%), with a mean age of (49.40±12.03) years. The cluster analysis identified three distinct Pre-DM subtypes: islet dysfunction type (n=9 878), relatively healthy type (n=5 316), and severe metabolic disorder type (n=4 599). There were statistically significant differences in clinical characteristics among the three groups (P<0.001). The severe metabolic disorder type exhibited the highest TyG index, TG/HDL-C, BMI and ALT levels. Conversely, the islet dysfunction subtype showed the lowest eGFRcys and the highest levels of cystatin C and prevalence of glomerular filtration dysfunction (all P<0.001). The multivariable linear regression analysis showed that after adjusting for confounding factors, compared with the insulin dysfunction type, the relatively healthy type (β=17.86, 95%CI 17.20-18.52) and the severe metabolic disorder type (β=6.88, 95%CI 6.19-7.58) had a greater influence on the eGFRcys of Pre-DM patients. These findings indicated that cluster classification is an independent influencing factor for eGFRcys.

Heterogeneous subtypes exist within the prediabetes population, with significant differences in metabolic profiles and glomerular filtration function across subtypes. The islet dysfunction subtype exhibited the most severe renal impairment. Cluster-derived subtyping was identified as an influencing factor for glomerular filtration function.

To identify the longitudinal trajectories of fasting plasma glucose (FPG) in patients with type 2 diabetes mellitus (T2DM) under community health management, and to assess their risk of developing cardiovascular disease (CVD).

The present study was a retrospective cohort study. T2DM patients who received community-based health management under the Essential Public Health Service Package and completed annual physical examinations in Hangzhou, Zhejiang Province, for four consecutive years from 2017 to 2020 were enrolled. Latent class trajectory modeling was used to identify distinct FPG longitudinal trajectory groups. Cox proportional hazards regression models were used to assess the associations between trajectory groups and the risks of CVD incidence and mortality, with hazard ratios (HRs) and 95% confidence intervals (CIs) calculated, adjusting for age, gender, lifestyle, obesity, comorbid hypertension, blood lipids, estimated glomerular filtration rate, glucose-lowering medication use, and management classes.

A total of 12 973 T2DM patients were included, with a mean age of (66.92±4.06) years, including 5 034 males (38.80%) and 7 939 females (61.20%). The mean follow-up duration was (4.21±0.59) years. The incidence of CVD was 5.80% (752/12 973), and CVD mortality was 1.23% (160/12 973). Four distinct FPG longitudinal trajectory groups were identified: the controlled group (48.38%, n=6 276), the uncontrolled-stable group (38.35%, n=4 975), the uncontrolled increasing-then-decreasing group (6.30%, n=817), and the uncontrolled decreasing-then-increasing group (6.98%, n=905). Compared with the controlled group, the uncontrolled increasing-then-decreasing group and the uncontrolled decreasing-then-increasing group had significantly higher risks of incident CVD, with HRs (95%CI) of 1.69 (1.28-2.23) and 1.67 (1.27-2.19), respectively; the CVD mortality risks were also elevated, with HRs (95%CI) of 2.32 (1.32-4.08) and 2.03 (1.14-3.60), respectively. No statistically significant differences in CVD incidence or mortality risk were observed between the uncontrolled-stable group and the controlled group (all P>0.05). After further adjustment for mean FPG and the coefficient of variation of FPG, the associations between trajectory groups and CVD risks were no longer statistically significant. Each 1 mmol/L increase in mean FPG was associated with a 16% increased risk of CVD incidence and a 23% increased risk of CVD mortality; each 1% increase in the coefficient of variation of FPG was associated with a 2% increase in both CVD incidence and mortality risks. The greater the number of comorbid cardiovascular risk factors, the higher the CVD risk within each trajectory group (all P for trend<0.001).

T2DM patients with persistently uncontrolled FPG and marked glycemic variability have substantially elevated CVD risk, which is primarily attributable to the long-term average FPG level and its variability. Maintaining long-term stable FPG control together with comprehensive management of cardiovascular risk factors may help reduce CVD risk in patients with T2DM.

To investigate the effects of dapagliflozin (DAPA) on left atrial structural remodeling and left atrial function in patients with type 2 diabetes mellitus (T2DM), and to explore the underlying molecular mechanisms by integrating multi-omics analyses and animal experiments.

This was a retrospective cohort study. A total of 121 hospitalized patients with T2DM admitted to the Department of Cardiovascular Medicine, the Affiliated Suzhou Hospital of Nanjing Medical University between June 2023 and May 2024 were enrolled. Patients were divided into a conventional hypoglycemic group (DM group, n=60) and a dapagliflozin treatment group (DAPA group, n=61) according to their hypoglycemic regimens. Four-dimensional speckle-tracking echocardiography was used to evaluate left atrial volume and strain parameters, and electrocardiographic P-wave characteristics were analyzed. Multiple linear regression models were constructed to analyze the association between DAPA treatment and left atrial structure and function in T2DM patients. Serum inflammatory and fibrotic biomarkers were measured, and untargeted metabolomics and tandem mass tag (TMT)-labeled quantitative proteomics analyses were performed on a subset of serum samples. In addition, db/db diabetic mouse models were established, and western blotting, Masson′s trichrome staining and dihydroethidium (DHE) staining were used to verify the regulatory effects of dapagliflozin on myocardial fibrosis-related pathways and oxidative stress.

Among the 121 T2DM patients, 63 were male and 58 were female, with a mean age of (66.7±8.1) years. Four-dimensional speckle-tracking echocardiography revealed that the left atrial diameter and minimal left atrial volume in the DAPA group were significantly lower than those in the DM group [(35.3±5.4) mm vs. (39.4±4.9) mm, P<0.001; (43.1±10.5) ml vs.(47.8±11.6) ml, P=0.021, respectively], while left atrial ejection fraction was markedly higher(32.8%±9.1% vs. 26.7%±7.8%, P<0.001). For left atrial longitudinal strain parameters, reservoir strain (14.6%±5.3% vs. 10.4%±3.1%, P<0.001), conduit strain(-9.1%±4.2% vs. -7.7%±2.5%, P=0.028) and booster strain (-5.4%±5.5% vs. -3.1%±4.2%, P=0.003) were all significantly elevated in the DAPA group compared with the DM group. Similarly, left atrial circumferential reservoir strain (12.6%±8.9% vs. 5.8%±3.3%, P<0.001), conduit strain (-6.8%±5.2% vs. -2.8%±3.6%, P<0.001) and booster strain(-5.7%±7.4% vs. -0.5%±4.9%, P<0.001) were also significantly higher in the DAPA group. Electrocardiographic measurements showed that the maximum P-wave duration in the DAPA group was shorter than that in the DM group(118.5±7.9 vs. 121.3±6.7, P=0.038), and the value of P-wave terminal force in lead V1 was smaller [(-52.6±5.7) mm·ms vs. (-58.3±4.1) mm·ms, P<0.001]. Serological detection demonstrated that serum levels of inflammatory factors including interleukin-6 [(236.5±18.2) ng/L vs. (245.9±23.5) ng/L, P=0.015] and tumor necrosis factor-α [(58.3±9.2) ng/L vs.(77.1±8.0) ng/L, P<0.001], as well as fibrotic biomarkers transforming growth factor-β [(150.2±20.9) μg/L vs. (215.4±16.9) μg/L, P<0.001] and galectin-3 [(15.3±2.8) μg/L vs. (16.4±2.5) μg/L, P=0.024] were all markedly reduced in the DAPA group. Multi-omics analyses indicated that differential metabolites in the DAPA group were mainly enriched in fatty acid β-oxidation and oxidative phosphorylation pathways, accompanied by upregulated levels of the antioxidant metabolite glutathione. Differentially expressed proteins were significantly enriched in cholesterol metabolism and renin-angiotensin system pathways, with elevated angiotensin-converting enzyme 2 expression. Animal experiments confirmed that the expression levels of fibrosis-related proteins including TGF-β and structural proteins (COL1a, COL3a, α-SMA) in atrial tissues of DAPA-treated mice were remarkably lower than those in the DM model group (all P<0.01). Histological staining revealed that collagen deposition and reactive oxygen species levels in atrial tissues of the DAPA model group were lower than those in the DM model group (all P<0.01).

Dapagliflozin treatment was associated with improvements in left atrial structure, mechanical function and electrophysiological properties in patients with T2DM. The potential mechanisms may involve reprogramming of myocardial energy metabolism, enhanced antioxidant capacity, and suppression of the inflammation-fibrosis cascade.

To analyze drug costs and changes in glucose-lowering treatment regimens among patients with type 2 diabetes mellitus (T2DM) and chronic kidney disease (CKD) in Beijing from 2016 to 2018.

This cross-sectional study was based on the Beijing Medical Insurance Database and included 2 853 036 outpatients with T2DM who visited clinics and received prescriptions from 2016 to 2018. Patients were divided into the T2DM alone group and the T2DM with CKD group according to whether CKD was present. Comorbidities, diabetes-related complications, glucose-lowering and non-glucose-lowering medications, treatment regimens, and annual drug costs were collected. Negative binomial regression and multivariable regression models were used to analyze CKD-related factors.

A total of 2 853 036 patients with T2DM were included, among which 2 638 390 were in the group with pure T2DM and 214 646 were in the group with T2DM combined with CKD. The prevalence of CKD among patients with T2DM was 8.18% (73 408/897 385), 7.59% (72 804/959 509), and 6.87% (68 434/996 142) in 2016, 2017, and 2018, respectively. In the T2DM with CKD group, metformin use increased from 39.11% (28 710/73 408) in 2016 to 42.57% (30 993/72 804) in 2017, and 45.13% (30 884/68 434) in 2018, and annual total drug costs decreased from (16 963.15±13 761.03) Yuan in 2016 to (15 224.48±12 876.89) Yuan in 2017, and (14 729.27±11 876.24) Yuan in 2018; the differences across the 3 years were statistically significant (all P<0.001). In 2018, compared with the T2DM alone group, the T2DM with CKD group had more medication types, glucose-lowering medication types, and non-glucose-lowering medication types, and higher annual total drug costs; the proportions of oral combination therapy and oral medication plus insulin combination therapy were higher, whereas the proportion of monotherapy was lower (all P<0.001).

From 2016 to 2018, patients with T2DM and CKD in Beijing had more complex medication regimens and higher drug cost burdens than patients with T2DM alone.

To elucidate the expression characteristics of long-chain acyl-CoA synthetase 4 (ACSL4) in the pancreatic islets from donors with type 2 diabetes mellitus (T2DM) and to analyze its association with lipid metabolism parameters.

The murine β-cell line NIT-1 and primary human islets were stimulated with palmitic acid (PA) to simulate a high-lipid environment in vitro. ACSL4 expression was assessed by quantitative real-time PCR (qPCR) and Western blotting. Subsequently, pancreatic tissues were collected from 31 organ donors [aged 50±13 years, 22 males and 9 females] at Tianjin First Central Hospital between June 2016 and June 2025. Donors were divided into non-diabetic (ND) and T2DM groups based on glycated hemoglobin levels and history of diabetes. Immunofluorescence staining was performed to detect the colocalization of ACSL4 and insulin (INS) in islets, and the proportion of ACSL4⁺INS⁺ cells was quantified. ACSL4 mRNA levels were measured by qPCR. Spearman rank correlation analysis was used to evaluate the association between the proportion of ACSL4⁺INS⁺ cells and donor lipid metabolism parameters. Multiple linear regression model analysis was performed to identify factors associated with the proportion of ACSL4⁺INS⁺ cells.

In vitro, PA-treated human islets showed significantly higher ACSL4 mRNA expression than control islets (P<0.001). Significantly higher Acsl4 mRNA and protein expression levels were detected in NIT-1 cells treated with 500 and 1 000 μmol/L PA compared with the 0 μmol/L group (both P<0.05). Moreover, treatment with 500 μmol/L PA for 48 h significantly upregulated Acsl4 mRNA and protein expression in NIT-1 cells relative to the 0 h group (P<0.01). Immunofluorescence revealed that ACSL4 was expressed in the cytoplasm of human pancreatic β-cells. The proportion of ACSL4⁺INS⁺ cells was significantly higher in the T2DM group than in the ND group (25.35%±7.14% vs. 9.12%±6.50%, P<0.001). Similarly, relative ACSL4 mRNA expression levels in islets were higher in the T2DM group compared with the ND group [(2.95±0.95) vs. (1.02±0.24), P<0.001]. Spearman rank correlation analysis showed that the proportion of ACSL4⁺INS⁺ cells was positively correlated with triglyceride (TG) levels (r=0.633, P=0.009) and negatively correlated with high-density lipoprotein cholesterol (HDL-C) levels (r=-0.574, P=0.022). Multivariable linear regression analysis further revealed that the proportion of ACSL4⁺INS⁺ cells was negatively associated with HDL-C (β=-10.510, P=0.036) and positively associated with T2DM status (β=11.280, P=0.034).

ACSL4 expression was higher in pancreatic β-cells of donors with T2DM patients compared with non-diabetic individuals, and was associated with HDL-C levels and T2DM status.

To compare the effects of semaglutide and tirzepatide on hepatic small-molecule metabolites and metabolic pathways in mice with metabolic dysfunction-associated steatotic liver disease (MASLD).

Twenty-eight male C57BL/6J mice were randomly divided using a random number table into a control group (CON, n=7) and a model group (MOD, n=21). The CON group received a normal chow diet, whereas the MOD group was fed a high-fat, high-fructose diet (HFHFD) for 7 weeks to induce MASLD. The MOD mice were then randomized into the HFHFD, HFHFD-Sema, and HFHFD-TZP groups (n=7 each) and subcutaneously administered saline, semaglutide (30 nmol/kg), or tirzepatide (10 nmol/kg), respectively, once every 3 days for 7 weeks. Body weight, food intake, liver weight, glucose and lipid metabolic parameters, and hepatic histopathological changes were assessed. Hepatic metabolites were analyzed using ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS). Differential metabolites were defined as variable importance in projection (VIP)>1, fold change <0.5 or >2, and P<0.05, followed by Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. One-way analysis of variance (ANOVA) followed by Tukey′s post hoc test was used to compare liver weight, serum biochemical parameters, and hepatic triglyceride content among groups, whereas two-way repeated-measures ANOVA with Bonferroni correction was used to compare body weight, food intake, and blood glucose levels during the intraperitoneal glucose tolerance test (IPGTT).

Compared with the HFHFD group, both treatment groups showed lower body and liver weights, serum insulin, total cholesterol, triglyceride, low-density lipoprotein cholesterol, alanine aminotransferase, aspartate aminotransferase, and hepatic triglyceride levels, as well as a lower area under the glucose curve (AUC) during the IPGTT (all P<0.05). Hepatic lipid vacuoles and red-stained lipid droplets were also reduced. The HFHFD-TZP group had lower liver weight, IPGTT AUC, and low-density lipoprotein cholesterol than the HFHFD-Sema group (all P<0.05). Metabolomic analysis identified 296 and 183 differential metabolites in the HFHFD-Sema and HFHFD-TZP groups, respectively. Both treatments reduced γ-glutamylisoleucine, adenylosuccinic acid, xanthurenic acid, 11-dehydrocorticosterone, and phosphoenolpyruvic acid levels compared with the HFHFD group (all P<0.05). Vitamin A, dethiobiotin, and chitobiose levels were higher in the HFHFD-TZP group than in the HFHFD group (all P<0.05), whereas no statistically significant differences in these metabolites were observed between the HFHFD-Sema and HFHFD groups. Differential metabolites in both treatment groups were enriched in steroid hormone biosynthesis and cortisol synthesis and secretion pathways, whereas those in the HFHFD-TZP group were additionally enriched in biotin metabolism and amino sugar and nucleotide sugar metabolism.

Both semaglutide and tirzepatide ameliorated metabolic disturbances in mice with HFHFD-induced MASLD. Tirzepatide was superior to semaglutide in reducing liver weight, improving glucose tolerance, and lowering low-density lipoprotein cholesterol levels. The two agents exhibited both shared and distinct effects on hepatic metabolites and metabolic pathways.

A case of glucokinase (GCKA family with adult-onset type 2 diabetes (MODY2) caused by a gene mutation, which also detected forkhead box protein 3 (FOXP3Genetic variation. The first case was a 13-year-old female who visited the Department of Endocrinology at the Affiliated Hospital of Chengde Medical College in August 2020 because "a physical examination revealed elevated blood sugar". The results of auxiliary tests showed that his fasting blood glucose was 7.51 mmol/L, and his postprandial blood glucose was 10.48 mmol/L. Anti-islet cell antibodies, anti-glutamate decarboxylase antibodies, and anti-protein tyrosine phosphatase antibodies were all negative. Genetic testing results showed that the subject, along with his mother, second sister, and third sister, all carried...GCKHeterozygous nonsense pathogenic variant, consistent with autosomal dominant inheritance, combined with mild fasting hyperglycemia, diagnosed asGCK-MODY, blood sugar is stably controlled through diet and exercise alone. Six women in the family (the proband's mother, maternal grandmother, eldest sister, second sister, and third sister) were detected.FOXP3Missense mutation, unclear significance, no immune abnormalities in all carriers, two mutations are independently inherited, no synergistic pathogenic effect. This article hopes to report on this co-variant family and suggest that clinical gene detection for single-gene diabetes mellitus should combine genetic patterns to distinguish between pathogenic variants and variants of unknown significance, so as to prevent overdiagnosis.

This paper reports the diagnosis and treatment of a patient with adolescent-onset type 14 diabetes mellitus (MODY14) with a familial genetic mutation. The patient was a 32-year-old young female who presented for "15 years of elevated blood glucose and 2 months of poor blood glucose control". After admission, relevant examinations revealed overweight and obesity, along with abnormal lipid profiles, elevated uric acid, and fatty liver. Peripheral blood gene sequencing revealed...Apple1Genetic variation: c.1465A>C (p.Thr489Pro) was diagnosed with MODY14 type based on his medical history and auxiliary examinations. He was given an insulin pump combined with semaglutide injection (0.25 mg, once a week) to lower his blood sugar. He was discharged from the hospital after his blood sugar improved. After discharge, multiple adjustments to the hypoglycemic regimen failed to meet the blood glucose target. Therefore, genetic testing of commonly used hypoglycemic drugs was improved. The results showed that the efficacy of sulfonylureas and some non-sulfonylurea insulin secretagogues decreased, the risk of adverse reactions of rosiglitazone increased, while the efficacy of biguanides increased at conventional doses. Based on the test results, the hypoglycemic regimen was adjusted to ecolinsulin (70 U, once a week) combined with sufficient metformin (1.5 g in the morning, 0.5 g in the afternoon, 0.5 g in the evening), dapagliflozin (10 mg, once a day) and nateglinide (0.12 g, three times a day). The patient's overall blood glucose control improved. It is hoped that the report on this patient will improve clinicians' understanding of MODY14 and provide a reference for early clinical diagnosis, precision treatment and genetic counseling.

Type 2 diabetes mellitus (T2DM) and its complications have become a major public health problem threatening population health. This article reviews the research progress on risk prediction and disease trajectory modeling of T2DM and its complications. This paper focuses on three technical levels: statistical models, machine learning, Transformer architecture, and generative learning. It compares the applicable scenarios, advantages, and limitations of these methods, and further explores the interpretability, fairness, and transformational applications of the models. The aim is to provide a reference for building a high-precision, scalable auxiliary model for precise diabetes prevention and control.

Globally, the incidence of early-onset type 2 diabetes (EOT2DM) continues to rise, with rapid disease progression, early complications, and a significantly increased risk of long-term death. This paper systematically reviews the current epidemiological status of EOT2DM and finds that the age of onset in my country continues to shift forward. In addition to the obesity-dominant type, there is also a special lean phenotype, characterized by ectopic deposition of visceral fat. Compared to late-onset diabetes, EOT2DM is centered on insulin resistance and is accompanied by unique pathogenesis such as rapid decline in pancreatic β-cell function and characteristic gut microbiota disorders. Its risk factors cover multiple dimensions, including unhealthy lifestyles, intrauterine epigenetic changes, and socioeconomic stratification. Traditional stepped hypoglycemic regimens have limited efficacy in this population. It is recommended to initiate individualized treatments such as metformin combined with glucagon-like peptide-1 receptor agonists and sodium-glucose cotransporter 2 inhibitors early, and to improve the early screening of cardiovascular, renal, and fundus complications. This article proposes to optimize screening strategies for high-risk groups and implement full-cycle comprehensive management to slow down disease progression and improve long-term prognosis.

Diabetic foot ulcer (DFU) is a chronic complication of diabetes mellitus with a high disability and mortality rate, which causes a serious economic burden to society. Smart wearable devices (SWD) can achieve comprehensive and real-time foot health monitoring, prevent the occurrence of DFU, promote DFU healing and avoid recurrence, and improve the quality of life of patients. However, the application of smart wearable devices is limited globally. Therefore, this article reviews the application progress of SWD in the monitoring, prevention, treatment and recurrence prevention of DFU physiological indicators, raises existing problems and prospects, and aims to provide a reference for future related research at home and abroad.