BACKGROUND AND HYPOTHESIS:While physical activity (PA) is associated in observational studies with cardiovascular and renal protection few interventional studies have assessed its long-term effect on renal outcomes. This study is the first to specifically evaluate the impact of high-intensity PA on the slowing of renal function decline in type 2 diabetes (T2D) patients at high risk for kidney disease. METHODS:ActiDiaNe is an open, randomized controlled trial involving patients with T2D (pT2D) and rapid renal function decline (eGFR slope < -5 ml/min/year over 6-24 months). Participants were randomized to high-intensity PA (HIPA) or standard PA counseling (STPA) for two years. The primary endpoint was the slope of decline in cystatin C-derived eGFR (cys-eGFR). RESULTS:Across 21 centers, 178 patients were screened, 122 randomized, and 103 (29 women/74 men) analyzed for the primary endpoint (59 HIPA vs. 44 STPA). At baseline, mean age was 66 ± 8 years, median cys-eGFR was 54 ml/min/1.73m² (37; 65), and median eGFR decline was -9 ml/min/year (-12; -7). The primary endpoint showed a decline of -2.05 ml/min/year (95% CI: -3.46 to -0.64) in HIPA vs. -2.77 ml/min/year (95% CI: -4.40 to -1.14) in STPA (P = 0.512). Cardiovascular events and deaths occurred in 10 (17%) HIPA vs. 6 (14%) STPA patients (P = 0.646), with 3 deaths in HIPA (none of them protocol-related). Hypoglycemia was reported in 32 HIPA vs. 26 STPA patients (P = 0.623). CONCLUSION:In pT2D with rapid renal function decline, HIPA did not significantly alter renal function decline compared to STPA.
BACKGROUND:Despite recent progress in insulin delivery, carbohydrate counting is essential for determining prandial insulin doses and is, therefore, a key component of education for patients living with type 1 diabetes (PwT1D). This study aimed to evaluate the relationship between carbohydrate knowledge and glycemic control. METHODS:Carbohydrate knowledge was assessed using GluciQuizz, a validated, self-administered questionnaire totaling 36 points submitted to participants in the SFDT1 cohort. Glycemic control was determined using a 14-day time in range (TIR, 70-180 mg/dl) RESULTS: The median age of the 635 participants was 43 [interquartile range 32-54] years, T1D duration of 25 [14-36] years. The median overall GluciQuizz score was 25 [23-28]: 24 [21-27] among the 158 participants treated with multiple daily injections (MDI), 25 [22-27] among the 255 users of continuous subcutaneous insulin infusion (CSII), and 26 [24-29] among the 222 users of automated insulin delivery (AID) systems. Median TIR (%) was: overall 68 [55-77]; MDI 60 [50-72]; CSII 62 [49-73]; AID 75 [69-82]. The total GluciQuizz score was positively associated with TIR in the overall population (β ± SE = 0.04 ± 0.01; P < 0.001). This association remained significant in CSII users (β ± SE = 0.06 ± 0.02; P < 0.001) but not in those treated with MDI (P = 0.19) or AID (P = 0.55). CONCLUSIONS:Better carbohydrate knowledge is independently associated with improved glycemic control in PwT1D using CSII but not among those using MDI or AID systems. These findings highlight the continued relevance of structured nutritional education, particularly for individuals using CSII.
Aim Patients with type 2 diabetes (T2D) and obesity are at increased risk of metabolic dysfunction-associated steatotic liver disease (MASLD) with advanced fibrosis (AF), requiring systematic non-invasive assessment using liver stiffness measurement (LSM) or the ELF test. However, access to these biomarkers remains limited in diabetology. We therefore evaluated the diagnostic performance of a new point-of-care ultrasound device, Hepatoscope, for the risk stratification of AF. Methods Participants with T2D and/or obesity and MASLD age 40-80 years, BMI <40 kg/m2 prospectively included in a multi-centre study (NCT04435054) underwent LSM using vibration controlled transient elastography (LSM-VCTE, FibroScan) and 2D-transient elastography (LSM-2DTE) using Hepatoscope version 1 and ELF. Post-acquisition reprocessing of the LSM-2DTE data using the CE approved version 2 (LSM-2DTE-2) was performed in a subgroup of n=120 participants. The area under the ROC curve (AUROC) and 95% confidence interval (CI) were assessed for the detection of intermediate to high-risk of AF, defined by LSM-VCTE ≥ 8 kPa. Results Among 294 participants (83.3% T2D, 42.5% female, 71.4% obesity), 24.8% had LSM-VCTE ≥ 8kPa. Significant correlation was observed between LSM-VCTE and LSM-2DTE-1 r: 0.37 [0.26-0.47], P < 0.001 and was higher between LSM-VCTE and LSM-2DTE-2: r: 0.61 [0.47-0.71], P < 0.001. In the subgroup of 120 participants, the AUROC (95%CI) of LSM-2DTE-2 for the detection of LSM-VCTE ≥ 8kPa was 0.81 (0.72-0.89) compared to 0.64 (0.52-0.75) for ELF. Conclusion This proof-of-concept study provides evidence supporting the role of LSM-2DTE using the Hepatoscope in the risk stratification of AF in diabetology settings. Larger studies are needed to validate its performance.
Background The impact of insulin-resistance on bone mineral density (BMD) remains unclear. Sclerostin inhibits bone formation, and increases with ageing, metabolic syndrome, and renal failure. Genetically-determined low sclerostin levels are associated with high BMD. Our aim was to determine the bone phenotype and the influence of insulin-resistance without obesity on sclerostin levels in Familial Partial Lipodystrophy type 2 (FPLD2). Methods Demographic, metabolic, anthropometric, and bone parameters were compared after adjustment on age and menopausal status across four groups of women (19 FPLD2, 12 obese with diabetes, 14 obese without diabetes, 11 lean controls). Results The FPLD2 group had significantly higher Intra-Abdominal/Total-Abdominal Fat (IAF/TAF) ratio and lower fat mass percentage compared to the other groups. Glucose parameters (Fasting blood glucose, C-peptide, HbA1c, HOMA2-IRCP), and triglycerides were significantly higher in the FPLD2 group than in controls, with a higher sclerostin level. Sclerostin levels were positively associated with age, glucose parameters, Z-score, T-score, and negatively with osteocalcin. FPLD2 was the only subgroup in which the association between sclerostin and HOMA2-IRCP was maintained. In contrast, osteocalcin levels were negatively associated with age, glucose parameters, triglycerides, IAF/TAF, and bone parameters. Neither sclerostin or osteocalcin were associated with BMI, whole-body fat, or leptin. Conclusion The association of sclerostin (positive) and osteocalcin (negative) with HOMA2-IRCP, but not with fat mass parameters, suggests a predominant role of insulin-resistance over fat mass in the regulation of osteokines. Insulin-resistance may uncouple the usual relationship between sclerostin and BMD. The sclerostin contribution in early atherosclerosis in FPLD2 remains to be studied.
OBJECTIVE:Dysregulation of bile acid (BA) profiles is closely associated with obesity and its related metabolic abnormalities. This study aims to investigate the BA alterations between metabolically healthy obesity (MHO) and metabolically unhealthy obesity (MUO) and identify specific BA species associated with MUO. METHODS:We measured serum bile acid profiles in two cross-sectional studies. The discovery cohort included 261 normal-weight, 80 MHO, and 120 MUO individuals. The validation cohort, consisting of 104 MHO and 104 MUO participants (matched for age, sex, and BMI), was used for confirmation. RESULTS:Individual primary bile acids (PBAs), including cholic acid, glycocholic acid, chenodeoxycholic acid, and total PBA concentration, were markedly increased in MUO compared with both normal-weight and MHO subjects. Total secondary bile acids were decreased in both obesity phenotypes compared with normal-weight subjects, but did not differ between MUO and MHO. Among individuals with obesity, either individual PBA or total PBA levels were positively correlated with metabolic deterioration parameters such as WHR, HOMA-IR, HbA1c, and TG. Moreover, a higher total PBA level was associated with an increased proportion of MUO. These findings were further replicated in the validation cohort. CONCLUSIONS:Serum individual PBAs and total PBAs were significantly elevated in MUO compared with MHO. Higher PBA levels were associated with adverse metabolic parameters and an increased proportion of MUO. This study characterizes the distinct BA profiles across obesity phenotypes and highlights the associations between PBAs and MUO.
AIM:Patients with type 2 diabetes (T2D) receiving dialysis have a very high risk of cardiovascular events. Evidence for the cardiovascular effectiveness of glucagon-like peptide-1 receptor agonists (GLP-1RAs) in this population remains limited. Here, we assessed this issue using a target trial emulation framework. METHODS:We conducted a cohort study using a Japanese administrative claims database. We included adults aged 20 years or older with T2D receiving maintenance dialysis who initiated a GLP-1RA or a DPP-4i between April 2015 and March 2023. The primary outcome was the 3-year risk of major adverse cardiovascular events (MACE), defined as a composite of acute myocardial infarction, stroke, and cardiovascular death. The observational analogue of the per-protocol effect was estimated using pooled logistic regression with inverse probability weighting to adjust for baseline and time-varying confounders. RESULTS:Among 4793 patients (557 GLP-1RA initiators and 4236 DPP-4i initiators), the estimated 3-year risk of MACE was 29.7% (95% CI, 22.3% to 38.3%) for GLP-1RA users and 37.6% (95% CI, 35.2% to 40.8%) for DPP-4i users, giving a risk difference of -8.0% (95% CI, -15.5% to 0.9%) and risk ratio of 0.79 (95% CI, 0.59 to 1.03). CONCLUSION:Compared with DPP-4is, sustained use of GLP-1RAs may reduce the risk of MACE among patients with T2D receiving maintenance dialysis. These findings suggest a potential cardiovascular benefit but require confirmation in randomized controlled trials before introduction into clinical practice.
AIMS:This study addressed the challenge of glucose management in inpatients receiving nutritional support, even with automated insulin delivery (AID). We evaluated whether advanced AID would improve glycaemic control compared with standard subcutaneous insulin therapy in inpatients receiving nutrition support. METHODS:We did a single-centre, open-label, randomised controlled trial in adult inpatients receiving nutrition support who required subcutaneous insulin therapy. Patients recruited from non-critical care surgical and medical wards were randomly assigned (1:1) to receive advanced AID or conventional subcutaneous insulin therapy (control group) given in accordance with local clinical practice. Patients were followed up for up to 20 days or until hospital discharge. The primary endpoint was the TIR, the proportion of time that sensor glucose concentrations were within the target range, 3.9-10.0 mmol/l. This trial is registered with ClinicalTrials.gov, number ChiCTR2300078746. RESULTS:Between February 2023 and September 2024, 32 patients were assessed for eligibility; of these, 26 were enrolled and randomly assigned to the closed-loop group (n = 13) or the control group (n = 13). TIR was 63.1 % [SD 12.8] in the closed-loop group and 31.5 % [20.1]in the control group (difference 31.7 percentage points [95 % CI 18.0-45.2; P < 0.001]). The time spent below the target range (< 3.9 mmol/l) did not differ between groups. No serious adverse event occurred in each group. No episodes of severe hypoglycaemia or hyperglycaemia with ketonaemia occurred in either study group. CONCLUSIONS:The advanced AID system improves glycaemic control without increasing the risk of hypoglycaemia in inpatients receiving nutrition support, particularly in patients with acute severe pancreatitis.
Introduction Advanced hybrid closed-loop systems (AHCL) systems are the preferred insulin delivery method for people with type 1 diabetes (PwT1D). Time in tight range (TITR) is becoming a new target for glycemic control. The ability of different AHCL systems to achieve an improvement in TITR remains unclear. Method Multicenter observational cross-sectional study to analyze the effect on TITR of the three different AHCL systems marketed in Spain: MM780G with SmartGuard (MM780G), Tandem t:slim X2 with Control-IQ (Control-IQ) and Ypsopump with CamAPS FX (CamAPS). PwT1D ≥ 18 yrs and treated with the same AHCL device for at least 3 months were recruited consecutively. The main objective was to analyze the differences in TITR between the three types of AHCL. Results A total of 189 patients (female 66.8%) were recruited. Mean age was 42.8 ± 12.8 yrs. (18-72 yrs.) and T1D duration 23.5 ± 11.8 yrs. Time on AHCL automatization was 94.6 ± 7.6%. Overall TITR was 51.7 ± 11.6 % with no between-group differences (P = 0.154). Notably, nocturnal TITR was higher with CamAPS (53.6 ± 15.4%, P = 0.029), compared to MM780G (45.9 ± 17.8%, P = 0.029) and Control-IQ (50.6 ± 15.4%, P = 0.029). However, time below range level 1 and 2 was higher among CamAPS group (3.9 ± 2.8% and 0.7 ± 1.0%, respectively; P < 0.001), compared to MM780G (1.6 ± 1.6% and 0.2 ± 0.3%, respectively; P < 0.001) and Control-IQ (2.5 ± 2.0% and 0.5 ± 0.5%, respectively; P < 0.001). HbA1c was higher in MM780G treated patients, compared to both Control-IQ and CamAPS (6.9 ± 0.7% vs. 6.5 ± 0.6%, P = 0.003). Conclusion AHCL systems are useful for achieving TITR >50%, with no differences between treatment groups. However, this study has detected differences in some aspects of glycemic control between the devices used which could have clinical implications.
BACKGROUND:Affecting nearly 10-15% of women, polycystic ovary syndrome (PCOS) is the leading cause of infertility. Various treatments are available to induce pregnancy. PCOS is significantly associated with gestational diabetes mellitus (GDM), a condition that carries major maternal and neonatal risks. However, the potential involvement of infertility treatments as a risk factor for GDM in this population remains unclear. METHODS:Our study aims to define the prevalence of GDM in this population and to study the determinants of its onset, with a focus on infertility treatments, through an observational cohort study of pregnant women between 1997 and 2024 who were followed up at the university hospital centre where their PCOS was diagnosed. Univariate and multivariate analyses and survival curves were performed. RESULTS:The prevalence of GDM was 32.5% [95% CI: 28.8-36.3] in the total population (n = 622), 32.7% [95% CI: 27.7-38.2] in untreated women (n = 321) and 32.2% [95% CI: 27.0- 37.9] in treated women (n = 301), with difference found between the subgroups after ajustement on age and BMI (P = .04). The time to initiation of insulin therapy was similar between women receiving infertility treatment and those not receiving such treatment (Log-rank test P = .83). In logistic regression, age, BMI, blood pressure, triglycerides, testosterone and LH appeared to be independent risk factors for GDM. CONCLUSION:The prevalence of GDM in women with PCOS is not similar between those treated for infertility using ART and those not treated. Other determinants of GDM within this population have been identified, which could potentially enable early detection of women at risk and the implementation of personalized care in a precision medicine approach.
This editorial argues that technological innovation in diabetes, particularly continuous glucose monitoring and hybrid closed-loop systems, must be integrated into a humane medicine centred on both metrics and persons. It was written under the pressure of observing that diabetes distress persists in a substantial proportion of patients who benefit from the most advanced technologies, including hybrid closed‑loop systems. The author describes each patient's situation as a "mental puzzle" of beliefs, emotions and social constraints and highlights the biographical ruptures caused by disease and technology. The clinical encounter has a double object-the disease and the person-held together through clinical conversation that translates innovation into livable choices for each individual. It is where technicality meets humanity.
We report two cases of normoglycemic people with HIV well-controlled by antiretroviral therapy who were switched to a dual long-acting therapy consisting of cabotegravir (CAB), an integrase strand transfer inhibitor, and rilpivirine (RIL), a non-nucleoside analog reverse transcriptase inhibitor, administred intramuscularly once a month then every two months. Both cases developed shortly acute severe hyperglycemia with ketoacidosis or insulinopenia requiring intravenous insulin therapy. Anti-pancreatic autoantibodies were absent. The hyperglycemic episode resumed after stopping CAB/RIL and/or initiating metformin. To our knowledge these are the first reported cases of severe CAB/RIL-induced hyperglycemia. Up to now, long-acting CAB/RIL has not been associated with metabolic outcomes. These cases serve as a warning to clinicians to monitor glycemia when initiating long-acting CAB/RIL therapy.
Research on the developmental origins of health and disease has traditionally centered on maternal health and in utero exposures, with comparatively limited attention to paternal preconception factors. However, emerging evidence suggests that paternal metabolic health-particularly obesity and diabetes-may contribute to offspring metabolic risk. Observational studies associate higher paternal body mass index and cardiometabolic dysfunction with increased offspring adiposity, blood pressure, and markers of insulin resistance, although effect sizes are modest and confounding by shared genetics and environment remains a key limitation. Mechanistic data from animal models support sperm-mediated epigenetic pathways linking paternal metabolic status to offspring metabolic programming, yet human epigenetic evidence is inconsistent. At the same time, the widespread use of modern obesity pharmacotherapies, including glucagon-like peptide-1 receptor agonists and dual incretin agonists, has introduced a largely unexamined dimension of paternal exposure. Current data on paternal use of these agents and long-term offspring metabolic outcomes are sparse and fragmentary. As obesity pharmacotherapy expands among men of reproductive age, a shift beyond maternal-only frameworks is warranted. Rigorous studies integrating paternal metabolic profiling, medication exposure, and long-term offspring follow-up are urgently needed to inform clinical practice and public health guidance.
Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are widely used in type 2 diabetes, yet their management after colorectal cancer diagnosis remains uncertain. A recent population-based cohort study reported lower mortality and fewer metastatic events among GLP-1 RA users than among patients receiving other glucose-lowering therapies. However, causal inference is limited by potential immortal time bias from the exposure definition, incomplete reporting of key prognostic and treatment factors (stage, surgery, systemic therapy, performance status), and unclear cohort entry and follow-up windows. In addition, reliance on routine administrative codes may misclassify outcomes. Time-varying exposure models, active-comparator new-user or landmark designs, and registry linkage/validation would strengthen the evidence.
Aims: Antibody-negative type 1 diabetes (T1D) is a provisional diagnosis with unclear etiology. This study aimed to determine the prevalence and phenotypic characteristics of monogenic diabetes within a large antibody-negative T1D cohort. Methods: A total of 482 antibody-negative T1D patients were included from a clinically diagnosed T1D cohort. Targeted sequencing using a custom gene panel covering 36 genes and the mitochondrial 3243 A > G mutation was performed. Demographic, metabolic, and human leukocyte antigen (HLA) data were analyzed. Beta-cell function was assessed in patients with monogenic diabetes amenable to precision treatment. Results: Among 482 antibody-negative T1D patients, 2.5% (12/482) had maturity-onset diabetes of the young (MODY) and 1.7% (8/482) had mitochondrial diabetes. The prevalence of MODY increased to 11.1% (6/54) among childhood-onset patients with a single susceptible HLA haplotype, while mitochondrial diabetes reached 6.2% (7/113) in adult-onset patients lacking HLA-susceptible haplotypes. Other characteristics, including gender, age at diagnosis, hemoglobin A1c, 2-hour postprandial C-peptide (2hCP), and family history of diabetes, showed no significant differences. Compared with the "truly" antibody-negative T1D group, MODY patients had an earlier onset, whereas mitochondrial diabetes patients had later onset, higher 2hCP, and fewer HLAsusceptible haplotypes (all P < 0.05). Of eight recalled monogenic diabetes patients, 62.5% (5/8) retained random C-peptide > 100 pmol/L after a median 15.7 years. Conclusions: In antibody-negative T1D, MODY and mitochondrial diabetes accounted for 2.5% and 1.7%, respectively. HLA genotype was the key distinguishing factor. Persistent C-peptide secretion in monogenic diabetes supports the need for genetic screening in antibody-negative T1D patients.
AIMS:To evaluate whether GLP-1 RA therapy is associated with an increased risk of NAION in a large, diverse real-world population. METHODS:This retrospective cohort study utilized the TriNetX U.S. Collaborative Network from 2015 to 2024. Adults (≥18 years) with type 2 diabetes mellitus (T2DM) were identified and grouped as GLP-1 RA users versus other antidiabetic users and GLP-1 RA users versus sodium-glucose cotransporter-2 inhibitors (SGLT-2 i) users. Patients with preexisting optic neuropathy or severe ocular disease were excluded. Propensity score matching (1:1) was applied to balance baseline characteristics. The primary outcome was incident NAION, defined by ICD-10 codes. Cox proportional hazards models and Kaplan-Meier analyses were used to estimate hazard ratios (HRs) and cumulative probabilities. RESULTS:After matching, 799 036 GLP-1 RA users were compared with 799 036 other antidiabetic users, and 429 985 GLP-1 RA users were compared with 429 985 SGLT-2i users. Over 9 years, GLP-1 RA users showed higher cumulative NAION incidence vs other agents (0.21% vs 0.17%; HR 1.38; 95% CI, 1.23-1.55) and vs SGLT-2i users (0.20% vs 0.20%; HR 1.30; 95% CI, 1.11-1.52). Kaplan-Meier curves demonstrated consistent early separation favoring higher risk among GLP-1 RA users. Sensitivity analyses yielded similar patterns. CONCLUSIONS:GLP-1 RA therapy was associated with a modest but statistically significant increased risk of NAION. While the absolute risk remains low, clinicians should consider ophthalmic risk assessment-particularly in patients with anatomical susceptibility or vascular risk factors-as GLP-1 RA use expands for diabetes and obesity management. Further mechanistic and prospective research is warranted.
AIM:Toe necrosis is a common complication of diabetic foot ulcer (DFU). Amputation surgery is usually performed by a surgeon in the operating room. In response to the limited availability of operating rooms, clinicians from two specialized diabetic foot units performed bedside surgery to amputate isolated toe-necrosis-complicated DFU. The aim of our study was to compare the rate of wound healing 6 months after toe amputation following bedside amputation surgery (BAS) and conventional amputation surgery (CAS). METHODS:This retrospective observational multi-center study was conducted in two French diabetic foot units. All patients with diabetes mellitus (DM) who underwent a toe amputation for isolated necrosis in an operating room (CAS) or at bedside (BAS) were included (05/2016 - 07/2023). The primary endpoint was the 6-month healing rate, defined as a complete skin epithelialization without recurrence at 6 months, without secondary amputation. RESULTS:Out of 2029 patients admitted for DFU, 189 had isolated toe necrosis requiring limited amputation (9%). Among the 171 patients who attended follow-up at 6 months: males 82.5%, type 2 DM 94.7%, average duration of DM 18.3 ± 0.9years, average HbA1c 8.6 ± 2%. BAS was performed on 106/171(62%) patients. The 6-month healing rate was not significantly different between the two groups (BAS 53.8% vs CAS 52.3%, P = 0.852). The rate of secondary surgery was not significantly different (BAS 24.5% vs CAS 16.9%, P = 0.241). CONCLUSION:BAS is a safe and efficient approach for the treatment of isolated toe necrosis, resulting in a healing rate similar to that of conventional surgery.
AIMS:Carbohydrate counting enables flexible prandial insulin dosing in type 1 diabetes but remains cognitively demanding. Concerns persist that such sustained attention to food may contribute to disordered eating behaviors. The primary aim of this study was to examine whether carbohydrate-counting knowledge is associated with disordered eating behaviors. METHODS:This cross-sectional study (NCT07021456) was conducted online. Participants completed questionnaires assessing carbohydrate-counting knowledge (Gluciquizz), disordered eating behaviors (DEPS-R), and likely eating disorders (SCOFF-F). Additional questionnaires evaluated quality of life (ADDQoL), diabetes-related distress (PAID-5), and fear of hypoglycemia (HFS-II short form). Elevated DEPS-R was defined as a score ≥ 20, and likely eating disorders as SCOFF-F ≥ 2. RESULTS:A total of 100 adults with type 1 diabetes were included. No correlation was observed between Gluciquizz and DEPS-R (ρ = -0.03, 95% CI (-0.23 to 0.17), P = 0.73). Similarly, Gluciquizz scores did not differ between participants with SCOFF-F < 2 and ≥ 2 (P = 0.745). Diabetes-related distress was significantly higher among participants with elevated DEPS-R scores (PAID-5 median 15 vs 8; P = 0.006), whereas ADDQoL and HFS-II did not differ significantly. CONCLUSION:In this selected adult population with type 1 diabetes, carbohydrate-counting knowledge was not associated with disordered eating behaviors. However, positive DEB screening was linked with higher diabetes-related distress, supporting the importance of psychosocial assessment.
Aim Monogenic diabetes is a group of disorders arising from single gene mutations with a clear pathophysiology, most of which present with impaired beta cell function rather than insulin resistance. This study aims to evaluate the ability of TyG index and polygenetic risk score (PRS) to identify multi-type beta cell monogenetic diabetes (beta-cell-MgD) in Chinese early-onset type 2 diabetes (EOD) population. Methods A prediction model for beta-cell-MgD was established by logistic regression analysis in Cohort 1 (92 beta-cell-MgD, 512 EOD). Model performance was evaluated by receiver operating characteristic curves (ROC) and validated in an independent case-control sample (Cohort 2, 35 beta-cell-MgD, 50 EOD) and a newly diagnosed drug-naive EOD cohort (Cohort 3, 7 beta-cell-MgD, 176 EOD). PRS was constructed based on Genome-wide genotyping data from participants in Cohort 3. The ability of PRS to identify beta-cell-MgD was tested by ROC. Results The TyG-MgD score based on age at diagnosis, BMI and TyG presented a good performance to distinguish beta-cell-MgD (AUC=0.769), and achieving AUCs of 0.966 and 0.754 respectively in validation cohorts. At the optimal cutoff point -16.19, the model achieved a sensitivity of 66.3% and a specificity of 75.39%, allowing one case of beta-cell-MgD identified among every three patients. -16.85 could be used as the screening threshold prioritizing 80% sensitivity (with 59% specificity). Models combining TyG-MgD with East Asian PRS and beta-cell dysfunction-high proinsulin partitioned polygenetic score showed AUCs of 0.842 and 0.834 respectively for indentifying beta-cell-MgD. Conclusion We developed a clinical prediction model as a simple screening tool for multi-type beta-cell-MgD, identifying who are most likely to benefit from next genetic sequencing in Chinese population. PRS might be helpful for further screening of MgD.
AIM:To assess the outcomes of teleophthalmology-based diabetic retinopathy (DR) screening in individuals over 70 years within the OPHDIAT network and to compare them with those of patients aged 18-69 years. METHODS:A cohort of 16,459 diabetic patients, without known DR or with mild non-proliferative DR (NPDR), screened in 2024 in 32 OPHDIAT centers, was included and divided into two groups: < 70 years (n = 13,639) and ≥ 70 years (n = 2,820). Two non-mydriatic retinal photographs per eye were analyzed by certified ophthalmologists. RESULTS:Among patients aged ≥70 years, 21.3% (95% CI: 19.8-22.8) had any DR, and 6.1% (95% CI: 5.2-6.9) were referred to an ophthalmologist for moderate NPDR or a more severe form of the disease, including suspected macular edema. These proportions did not significantly differ from those found in patients < 70 years: 21.9% (95% CI: 21.2-22.6) and 6.1% (95% CI:5.6-6.5), respectively. Severe NPDR or proliferative DR were rare in both groups (1.0%, 95% CI: 0.6-1.4% vs. 1.7%, 95% CI: 1.5-1.9%, P < 0.001). The proportion of ungradable images was higher in the group ≥70 year (14.4%, 95% CI:13.1-15.7% vs. 6.1%, 95% CI: 5.7-6.5%, P < 0.001), particularly in phakic eyes, although 80% of patients had interpretable images for both eyes. Pupil dilation significantly improved image quality in this group. Screening also allowed detecting other ocular disorders, including age-related macular degeneration and glaucoma, which were more common in the group ≥ 70 years (2.1%, 95% CI: 1.5-2.6% vs. 0.6%, 95% CI: 0.4-0.7% P < 0.001). CONCLUSION:Teleophthalmology-based DR screening appeared feasible and clinically relevant in patients aged ≥70 years, allowing identifying patients requiring ophthalmologic evaluation, while also detecting other age-related ocular diseases. Pupil dilation is recommended to optimize image quality in this population.
Objective: The true burden of diabetes is likely underestimated by not considering the full range of complications associated with diabetes. Our aim was to compare cause-specific hospitalizations in adults with vs. without diabetes. Research Design and Methods: Our denominator included all adults with and without self-reported diabetes from the 2019 Behavioral Risk Factor Surveillance System Survey, weighted to reflect the U.S. population. Our numerator, age-standardized risks of ICD-10-CM-defined inpatient hospitalizations and emergency department (ED) visits, were identified from the 2019 National Inpatient Sample and National ED Sample, respectively, weighted to be representative of U.S. hospitalizations. Each cause-specific hospitalization was classified as traditional, emerging, or other. Results: For inpatient hospitalizations, the highest absolute risk difference per classification was for sepsis (traditional; 1,680 [95%CI: 1,649-1,712] hospitalizations per 100,000), pneumonia (emerging; 225 [218-232]), and respiratory failure (other; 280 [272-289]). For ED visits, the highest absolute risk difference was for abscess, furuncle, and carbuncle (traditional; 388 [352-423] visits per 100,000), complications of cardiac devices (emerging; 111 [104-118]), and disorders of the urinary system (other; 299 [252-346]). Conclusions: The causes of excess hospitalizations associated with diabetes extend well beyond traditional complications with implications for population-level planning, resource allocation, and individual diabetes management.
