Daily Endocrinology Research Analysis

BACKGROUND: Improving the accessibility of screening diabetic kidney disease (DKD) and differentiating isolated diabetic nephropathy from non-diabetic kidney disease (NDKD) are two major challenges in the field of diabetes care. We aimed to develop and validate an artificial intelligence (AI) deep learning system to detect DKD and isolated diabetic nephropathy from retinal fundus images. METHODS: In this population-based study, we developed a retinal image-based AI-deep learning system, DeepDKD, pretrained using 734 084 retinal fundus images. First, for DKD detection, we used 486 312 retinal images from 121 578 participants in the Shanghai Integrated Diabetes Prevention and Care System for development and internal validation, and ten multi-ethnic datasets from China, Singapore, Malaysia, Australia, and the UK (65 406 participants) for external validation. Second, to differentiate isolated diabetic nephropathy from NDKD, we used 1068 retinal images from 267 participants for development and internal validation, and three multi-ethnic datasets from China, Malaysia, and the UK (244 participants) for external validation. Finally, we conducted two proof-of-concept studies: a prospective real-world study with 3 months' follow-up to evaluate the effectiveness of DeepDKD in screening DKD; and a longitudinal analysis of the effectiveness of DeepDKD in differentiating isolated diabetic nephropathy from NDKD on renal function changes with 4·6 years' follow-up. FINDINGS: For detecting DKD, DeepDKD achieved an area under the receiver operating characteristic curve (AUC) of 0·842 (95% CI 0·838-0·846) on the internal validation dataset and AUCs of 0·791-0·826 across external validation datasets. For differentiating isolated diabetic nephropathy from NDKD, DeepDKD achieved an AUC of 0·906 (0·825-0·966) on the internal validation dataset and AUCs of 0·733-0·844 across external validation datasets. In the prospective study, compared with the metadata model, DeepDKD could detect DKD with higher sensitivity (89·8% vs 66·3%, p<0·0001). In the longitudinal study, participants with isolated diabetic nephropathy and participants with NDKD identified by DeepDKD had a significant difference in renal function outcomes (proportion of estimated glomerular filtration rate decline: 27·45% vs 52·56%, p=0·0010). INTERPRETATION: Among diverse multi-ethnic populations with diabetes, a retinal image-based AI-deep learning system showed its potential for detecting DKD and differentiating isolated diabetic nephropathy from NDKD in clinical practice. FUNDING: National Key R & D Program of China, National Natural Science Foundation of China, Beijing Natural Science Foundation, Shanghai Municipal Key Clinical Specialty, Shanghai Research Centre for Endocrine and Metabolic Diseases, Innovative research team of high-level local universities in Shanghai, Noncommunicable Chronic Diseases-National Science and Technology Major Project, Clinical Special Program of Shanghai Municipal Health Commission, and the three-year action plan to strengthen the construction of public health system in Shanghai.

BACKGROUND: Limited data exist on the association of glucagon-like peptide 1 receptor agonists (GLP-1RAs) with the risk of venous thromboembolism. This meta-analysis aimed to investigate the association between GLP-1RAs and the risk of venous thromboembolism including deep vein thrombosis (DVT) and pulmonary embolism. METHODS AND RESULTS: A systematic search of PubMed, Web of Science, EMBASE, and Cochrane library was conducted from inception until July 3, 2024, to identify randomized controlled trials comparing GLP-1RAs with placebo or other anti-iabetic drugs, with reported data on DVT and pulmonary embolism. The primary outcome was venous thromboembolism, and secondary outcomes included DVT and pulmonary embolism. Pooled odds ratios (ORs) were calculated using fixed-effects models with Mantel-Haenszel method and treatment arm continuity correction for zero-event trials. A total of 39 randomized controlled trials involving 70 499 participants were included. A nonsignificant upward trend in the risk of venous thromboembolism was observed among participants using GLP-1RAs (OR, 1.19 [95% CI, 0.94-1.50]). GLP-1RAs were significantly associated with an increased risk of DVT (OR, 1.64 [95% CI, 1.14-2.36]); risk difference 25 (5-52) more events per 10 000 person-years). Subgroup analyses revealed that increased risk of DVT was particularly prominent in randomized controlled trials with treatment duration >1.5 years (OR, 2.32 [95% CI, 1.49-3.60]) and in cardiovascular outcome trials (OR, 2.18 [95% CI, 1.36-3.49]). No significant association was observed between GLP-1RAs and risk of pulmonary embolism. CONCLUSIONS: GLP-1RAs might increase the risk of DVT, especially for long-term use of GLP-1RAs. Clinicians should be aware of this potential risk when prescribing GLP-1RAs.

BACKGROUND: In the US Diabetes Prevention Program (DPP), a 3-year randomised clinical trial in 3234 adults with prediabetes, type 2 diabetes incidence was reduced by 58% with intensive lifestyle intervention (ILS) and by 31% with metformin, compared with placebo. We sought to assess the long-term effects and potential heterogeneity of treatment effects over approximately 21 years of follow-up. METHODS: The DPP trial was continued with protocol modifications as the DPP Outcomes Study (DPPOS). In the DPPOS, placebo was discontinued, metformin (850 mg twice a day as tolerated) was continued after unmasking, and group-based booster intervention classes were offered to the ILS group twice a year; additionally, all participants were offered group-based lifestyle intervention four times a year. The prespecified primary outcome during DPP and DPPOS was diabetes incidence defined by American Diabetes Association criteria. The DPPOS protocol specified continued diabetes incidence as an outcome; Feb 23, 2020, was chosen as the closing date for the present analysis, as a date prior to the COVID-19 pandemic, which caused major disruptions in clinic visits and complicated longitudinal data analyses. We assessed long-term persistence of intervention effects on diabetes incidence, and heterogeneity of effects in subgroups defined by baseline diabetes risk factors. Follow-up is reported for the combined study from July 31, 1996, to Feb 23, 2020, and analysis was by intention to treat. The trial is registered with ClinicalTrials.gov, NCT00004992 (DPP) and NCT00038727 (DPPOS); follow-up is ongoing but the trial is closed to enrolment except for previous DPP participants. FINDINGS: 3195 participants originally enrolled in the DPP were included in the present analyses. This population comprised 2171 (67·9%) female participants and 1024 (32·1%) male participants, with a mean baseline age of 50·6 years (SD 10·7). Individual follow-up times ranged from 0·2 to 23·2 years (median 8·0 years [IQR 3·0 to 18·0]); remaining numbers at risk decreased sharply after 21 years because of administrative censoring and thus follow-up was considered to represent a 21-year period. During follow-up, compared with placebo, diabetes incidence rate was reduced in the original ILS group (hazard ratio [HR] 0·76 [95% CI 0·68 to 0·85], rate difference [RD] -1·59 cases [95% CI -2·25 to -0·93] per 100 person-years) and in the original metformin group (HR 0·83 [0·74 to 0·93], RD -1·17 [-1·85 to -0·49]), with corresponding increases in median diabetes-free survival of 3·5 years and 2·5 years, and mean diabetes-free survival of 2·0 years (95% CI 1·2 to 2·8) and 1·2 years (0·4 to 2·0), respectively. The diabetes cumulative incidence curves separated early, especially in the first 3 years, with lower incidence rates in the metformin and ILS groups than in the placebo group. The metformin and ILS curves progressively converged with longer follow-up. The overall treatment effects appeared to result from large early effects during the DPP. Absolute intervention effects, measured as RDs versus placebo, were greater with ILS in participants with higher values for baseline fasting glucose, HbA INTERPRETATION: The large initial intervention effects seen in the DPP trial were followed by sustained reductions in cumulative diabetes incidence for 21 years. Intervention effects were heterogeneous according to some baseline variables. These findings could guide precision interventions to help address the current type 2 diabetes epidemic. FUNDING: US National Institute of Diabetes and Digestive and Kidney Diseases and other agencies.