Diabetologia

Background. Type 2 diabetes mellitus (T2D) is defined by progressive pancreatic {beta}-cell dysfunction whose molecular underpinnings remain incompletely understood. Single-cohort transcriptomic analyses of donor islets have yielded heterogeneous gene lists of limited cross-study reproducibility, constraining both mechanistic interpretation and biomarker development. Methods. We combined two complementary analytical strategies applied to four public human islet transcriptomic cohorts (GSE25724, GSE20966, GSE38642, and GSE164416; n = 7-57 donors per contrast). For the integrative arm, three microarray datasets and one bulk RNA-seq dataset were processed independently and unified through gene-level random-effects meta-analysis, hallmark pathway scoring (GSVA/MSigDB), and iterative module refinement, yielding a two-axis disease framework. For the diagnostic arm, a consensus multi-method machine learning pipeline, combining LASSO penalized logistic regression, Support Vector Machine Recursive Feature Elimination (SVM-RFE), and Random Forest importance scoring, was applied to 184 differentially expressed genes from the RNA-seq cohort, with all normalization steps performed within leave-one-out cross-validation (LOOCV) folds to prevent data leakage. Machine learning classification of the RNA-seq cohort was additionally subjected to external transportability testing in the independent bulk human islet RNA-seq cohort GSE50244 using an overlap-restricted reduced score and a threshold fixed in the discovery cohort. Results. Meta-analysis across all four cohorts identified 337 high-confidence T2D-associated genes (96.1% directional concordance in beta-cell-enriched tissue). These were distilled into two refined 14-gene modules: ImmuneStress (MICB, HLA-DRA, HLA-DPA1, IL1R2, and others) and BetaCellIdentitySecretion (RASGRP1, PPP1R1A, SLC2A2, and others), whose composite IsletDysfunctionScore provided the most stable cross-platform separation of non-diabetic from T2D islets (Hedges' g = 1.80, p = 9.83 x $10^-17$, $\text{I}^2$= 0%). Consistent with progressive disease, IsletDysfunctionScore increased monotonically from non-diabetic to impaired glucose tolerance to T2D. Separately, the machine learning pipeline derived a 10-gene diagnostic panel: GABRA2, SLC2A2, ARG2, DKK3, PRIMA1, TAFA4, HHATL, PARVG, RNU1-70P, and the novel lncRNA ENSG00000284653, that achieved perfect discrimination in LOOCV (AUC = 1.000, sensitivity = 1.000, specificity = 1.000, zero misclassifications across all 57 donors). A leakage-verification experiment confirmed that this performance reflected genuine biological signal: global quantile normalization prior to cross-validation collapsed AUC to 0.380. External testing showed that 8 of the 10 panel genes were measurable in GSE50244. The frozen 8-gene reduced score retained strong discrimination (external AUC = 0.907), with 6 of 8 genes preserving directional concordance, but the discovery-derived threshold did not transfer because the external score distribution was shifted upward and compressed, yielding complete sensitivity but zero specificity at the frozen cutoff Conclusions. Integrating pathway-level meta-analysis with machine learning classification, we present a coherent two-axis model: immune/stress activation and loss of beta-cell identity/secretory competence, together with a compact, biologically interpretable 10-gene diagnostic signature. Panel genes converge on GABA signaling, glucose transport, arginine metabolism, WNT pathway inhibition, and a novel lncRNA, providing both mechanistic hypotheses and high-priority targets for external validation. These findings offer a reproducible transcriptomic scaffold for future mechanistic, biomarker, and clinical translation studies of human islet dysfunction. They also support external transportability of the core biological signal, while indicating that absolute operating thresholds are cohort-dependent and would require recalibration before deployment in independent datasets.

Background. Cardiovascular disease accounts for most mortality in type 2 diabetes (T2D), yet treatment is anchored on glucose-derived metrics and fasting insulin is rarely measured. We tested whether fasting insulin carries cardiovascular-mortality information across the dysglycaemic spectrum and approaches non-diabetic levels with longer oral-therapy T2D duration. Methods and Findings. We analysed six NHANES cycles [2007-2018] linked to National Death Index follow-up through 2019. Baseline characteristics were described in participants with complete kidney-function data: normoglycaemic (n = 2,913), pre-diabetes (n = 3,993), and oral-therapy T2D (n = 1,381). Survey-weighted Cox models used the covariate-complete mortality sample (pre-diabetes n = 4,019; oral-therapy T2D n = 1,387) and adjusted for age, sex, race/ethnicity, BMI, smoking, physical activity, education, poverty:income ratio, and insulin assay generation. Per +1 natural-log-unit fasting insulin, all-cause hazard ratios (HRs) were 1.69 (95% CI 1.14-2.50; P = 0.008) in pre-diabetes and 0.71 (0.51-0.98; P = 0.039) in oral-therapy T2D; cardiovascular HRs were 3.09 (1.69-5.63; P < 0.001) and 0.50 (0.26-0.99; P = 0.046), respectively. Cancer mortality was not associated with fasting insulin. In oral-therapy T2D, geometric-mean fasting insulin remained 1.6- to 2.3-fold the normoglycaemic referent across duration bands. After BMI adjustment, fasting insulin declined during the first five years (-4.0%/year, P = 0.018) and was flat thereafter (-0.1%/year, P = 0.66). Conclusions. Fasting insulin predicted cardiovascular mortality in pre-diabetes. In oral-therapy T2D, the inverse association was most consistent with survivor effects and accumulated renal and vascular damage. Fasting insulin remained above normoglycaemic levels throughout treated T2D.

Hyperglycaemia is a hallmark of diabetes and a major risk factor for diabetic kidney disease (DKD). However, the molecular consequences of long-term cumulative hyperglycaemia (CH) remain unclear. As a stable epigenetic modification, DNA methylation may capture past glycaemic exposure. Here, we assessed CH-associated DNA methylation in 1,245 participants with type 1 diabetes (T1D) from Finland and the United Kingdom-Republic of Ireland cohorts. We identified 17 CH-associated CpGs, with the strongest association at cg19693031 (TXNIP). Longitudinal analyses demonstrate that these CH-associated DNA methylation levels remain stable despite short-term glycaemic fluctuations, suggesting lasting epigenetic imprints of earlier metabolic control. Integrative analyses combining genomic, epigenetic, and proteomic data characterized these CpGs and potential target proteins. Mendelian randomization suggested a causal association between cg20853880 (KLF11) and DKD, supported by chromatin accessibility and kidney KLF11 expression. Our findings suggest that epigenetic changes contribute to metabolic memory and may mediate the effects of hyperglycaemia on DKD.

Objective: To determine whether continuous glucose monitoring (CGM) identifies clinically relevant glycemic heterogeneity and subclinical end-organ alterations in adults without diabetes. Research Design and Methods: We analyzed 1,017 AI-READI Year 3 participants without diabetes (558 with normoglycemia and 459 with prediabetes by A1C). Fifty-two metrics from 10-day blinded CGM were reduced to nonredundant glycemic axes. Partial Spearman correlations between representative CGM metrics and clinical measures across 13 domains were adjusted for age, sex, and BMI and controlled for false discovery rate. CGM-derived subphenotypes were identified using unsupervised UMAP-HDBSCAN-based clustering. Results: Among 462 glycemic-clinical associations tested, 99 (21.4%) remained significant after false discovery rate correction. Hyperglycemia-related metrics, including mean glucose, time above range, and time in tight range, showed more associations than variability metrics. The strongest signals involved cardiometabolic, cardiovascular, and cognitive measures. Greater hyperglycemia and glucose excursions were associated with lower language performance, slower processing speed, and lower cognitive efficiency ({rho} {approx} -0.10 to -0.14; all P < 0.01). Clustering identified four reproducible glycemic subphenotypes: Healthy, Mild Hyperglycemia, High Variability, and Hyperglycemia. CGM phenotypes reclassified A1C-defined groups: 58.1% of participants with normoglycemia fell into dysglycemic phenotypes, whereas 18.8% of participants with prediabetes fell into more favorable phenotypes. The Hyperglycemia phenotype had the most adverse cardiometabolic profile and lower cognitive performance. Conclusions: In adults without diabetes, CGM revealed glycemic patterns associated with distinct subclinical alterations. CGM-based phenotyping may complement A1C for characterizing early dysglycemia and selecting individuals for longitudinal risk-stratification studies.

Hypoxia is a potent stressor and a major cause of {beta}-cell failure and loss after islet transplantation. Autophagy is a critical homeostatic mechanism that preserves organelle integrity and metabolic balance in cells under stress, but whether it supports {beta}-cell adaptation to sustained oxygen deprivation is unclear. Here, we used {beta}-cell-specific Atg5 knockout together with hypoxia and transplantation models, to demonstrate that autophagy is a major determinant of {beta}-cell survival during oxygen limitation and supports islet graft function. However, prolonged hypoxia suppressed autophagic flux, reduced lysosomal activity, and led to autophagosome accumulation, indicating failure of the lysosomal clearance pathway. This was accompanied by a marked reduction in transcription factor EB (TFEB) and its lysosomal target genes. Genetic and pharmacological activation of TFEB restored lysosomal gene expression and cathepsin B activity and improved {beta}-cell viability under hypoxia, implicating TFEB decline as a contributor to autophagy-lysosome dysfunction. Together, these findings outline a sequence in which autophagy initially safeguards {beta}-cells but becomes ineffective under sustained hypoxia as TFEB levels fall, identifying TFEB as a potential target to strengthen {beta}-cell resilience and survival in islet transplantation.

Abstract Objective: To identify clinical and genetic factors associated with variation in glycemic response to glucagon-like peptide-1 receptor agonist (GLP-1RA) therapy among adults with type 2 diabetes, with a focus on common GLP1R variations, polygenic risk load, and pancreas-specific regulation annotation. Research Design and Methods: We conducted a retrospective cohort study using electronic health record (EHR)-linked biobank data from the All of Us research workbench platform that included 5784 adults with type 2 diabetes who initiated GLP-1RA therapy. Baseline HbA1c was measured within 3 months before medication initiation, and follow-up HbA1c was measured after 3 months. The patients with type 2 diabetes were classified as good responders (HbA1c reduction [&ge;] 2.5 percentage point) or poor responders (HbA1c reduction <0.5 percentage point). Models adjusted for demographic characteristics, anthropometric and metabolic measures, blood pressure, body mass index (BMI), lipid profile, liver function tests, polygenic risk score, and GLP1R variant carrier status were compared between the two groups. Common GLP1R variations were further investigated for carrier frequency and associated HbA1c levels before and after medication use. Results: The cohort included 3194 good responders and 2590 poor responders. Good responders were younger than poor responders (55.2 vs. 58.6 years) and had significantly higher glycemic improvement. HbA1c levels fell from 9.2% to 6.3% in good responders and 8.4% to 8.1% in poor responders, resulting in an absolute HbA1c reduction of 2.9% and 0.3%, respectively. Good responders also showed larger decreases in fasting glucose, BMI, systolic and diastolic blood pressure, triglycerides, total cholesterol, LDL cholesterol, and liver enzymes, as well as minor improvements in HDL-C. After multivariable adjustment, Poor responders had a greater T2D polygenic risk score (0.38 vs. 0.21), more GLP1R coding variant carrier status (10.1% vs. 8.0%), and a higher overall GLP1R variant burden (22.8% vs. 19.2%). Variant-level studies revealed rs2268650 and rs2003132 enrichment among poor responders, with negative post-treatment HbA1c patterns in carriers, whereas good-response carriers showed significant HbA1c improvement. Conclusions: Response to GLP-1RA in T2D is associated with baseline clinical and metabolic status, as well as inherited genetic susceptibility, which includes common GLP1R variation and a larger polygenic risk burden. Integrating clinical and pharmacogenomic profiling may improve patient classification and provide insight into treatment failure in poor responders.

Introduction: Ethnic minority populations are disproportionately affected by type 2 diabetes (T2D). We investigated ethnic differences in the risks of diabetes-related complications and mortality in the Netherlands, and identified clinical, sociodemographic and environmental determinants associated with these differences. Methods: We included 175,112 adults with T2D from the dynamic prospective primary care cohort DIAMANT. DIAMANT data were linked to national registries from Statistics Netherlands and GECCO, a database integrating geographic, environmental and contextual exposures. Ethnic differences in complications risks were estimated using Cox proportional hazards models. Potential mediating factors were explored using machine-learning-based variable selection and association decomposition approaches. Results: At baseline, mean age was 65.4 (SD 12.3) years, 46.6% were women and median T2D duration was 11.3 [IQR 7.2; 15.8] years. Substantial heterogeneity in complication risk was observed across ethnic groups compared with Dutch-origin individuals. Retinopathy risk was consistently higher across nearly all non-Dutch groups (HRs 1.37-2.37). For macrovascular complications, elevated risks were mainly observed among Surinamese and Turkish individuals, including heart failure (HR 1.30 and 1.46, respectively). In contrast, individuals of Indonesian and Moroccan origin showed similar or lower risk for most complications. Environmental exposures (e.g. air pollution, temperature) and sociodemographic factors (e.g. main benefit, household composition) accounted for a substantial attenuation of several observed associations. Discussion: Substantial ethnic differences exist in risks of T2D complications and mortality, which showed to be heterogeneous across outcomes and populations. Our findings suggest that a considerable proportion of these disparities is attributable to differences in environmental and sociodemographic context, highlighting the importance of interventions that take into account differences in environmental and socio-demographic context.

The decline in first-phase insulin response (FPIR) during the presymptomatic period of type 1 diabetes (T1D) is well established. In-situ functional studies with pancreas tissue slices showed that {beta}-cell loss of glucose-responsiveness was independent of T-cell infiltration into islets in recent-onset T1D cases. However, the mechanisms driving {beta}-cell dysfunction before the onset of T1D remain unclear. In pancreas tissue from donors across the natural history of T1D, we utilized an in-situ, whole-islet phenotypical and transcriptomic approach to unravel novel targets in the glucose-stimulus coupled secretion pathway that are similarly impaired in T-cell infiltrated and non-infiltrated islets. Specifically, we observed that islets from autoantibody positive (single(s) or multiple(m) AAb+) donors exhibited activation of post-transcriptional gene regulation along with reduced protein translation, processing in the endoplasmic reticulum (ER), and ER stress. Disrupted mitochondrial metabolism and bioenergetics were prominent in islets from multiple AAb+ and T1D donors with disease durations [&le;]7 years. In addition, T1D islets presented reduced mitochondrial protein import, quality control, and dynamics, together with downregulated genes in insulin secretory pathways. During infiltration, these pathways remain dysregulated while immune/inflammatory transcripts were increased. These studies identified novel mechanisms of {beta}-cell dysregulation before symptomatic onset and independent of T-cell infiltration in T1D pathogenesis.

Aims Metformin remains the primary treatment for type 2 diabetes, yet over 40% of patients fail to maintain glycaemic control. We aimed to identify patients unlikely to respond to metformin prior to treatment initiation and to evaluate whether on-treatment management can improve glycaemic outcomes in suboptimal responders, informing early treatment decisions. Materials and Methods We analyzed 59,881 longitudinal HbA1c measurements from 7,105 patients with type 2 diabetes receiving metformin monotherapy using real-world electronic health records from Kaiser Permanente Northern California with up to six years of follow-up. We integrated demographic, clinical, genetic, and pharmacological factors to characterize metformin responder phenotypes and quantify the impact of adherence and weight control on time to glycaemic failure. Results Three distinct trajectory-based phenotypes were identified: good (63.6%), poor (8.9%), and non-responders (27.5%). Poor responders initially achieved glycaemic targets but lost control within 2.5 years, while non-responders showed minimal HbA1c reduction and failed within 1 year. Five baseline factors-HbA1c, age at diagnosis, body mass index, sex, and estimated glomerular filtration rate-classified phenotypes with good discrimination (area under the receiver operating characteristic curve = 0.84). Incorporating on-treatment HbA1c further enhanced identification of non-responders. Among suboptimal responders, weight control and improved adherence delayed glycaemic failure by approximately 7 months; however, eventual glycaemic failure remained likely. Conclusions We characterized three clinically relevant metformin responder phenotypes and showed that suboptimal responders can be identified early using baseline features. Poor and non-responders are unlikely to achieve durable glycaemic control with metformin alone and may require alternative treatment strategies.

Dysregulated hormone secretion and erosion of endocrine cell identity are features of type 1 and type 2 diabetes, but the transcriptional programs maintaining adult human islet identity and function remain poorly defined. The large MAF transcription factor MAFB is expressed in human - and {beta}-cells, marks their most functionally mature subpopulations, and is downregulated in diabetes, but its role in adult human islets has not been tested directly. Using shRNA-mediated MAFB knockdown (KD) in whole and CD26+ -cell-enriched human pseudoislets, we found that whole pseudoislet MAFB KD impaired glucagon synthesis and secretion while only modestly reducing insulin content and cAMP-potentiated insulin release. Single-cell profiling detected no {beta}-cell transcriptional response beyond MAFB KD itself, consistent with buffering by the related {beta}-cell-enriched MAFA transcription factor. In contrast, -cell-restricted MAFB KD unmasked a cell-autonomous requirement for MAFB in stimulus-secretion coupling. MAFB deficiency also destabilized -cell identity, downregulating canonical -cell and neuroendocrine secretory genes while ectopically inducing mesenchymal and extracellular matrix remodeling programs. In addition, MAFB-dependent downregulation of electron transport chain genes was confined to a large -cell subcluster, manifesting as impaired islet-wide mitochondrial respiration within the broader -cell population. Together, these findings identify MAFB as an essential adult human -cell maintenance factor that links diabetes-associated downregulation to impaired glucagon secretion, -cell identity erosion, and mitochondrial dysfunction. RESEARCH IN CONTEXTO_LIWhat is already known about this subject? O_LIMAFB is expressed in adult human - and {beta}-cells, marks their most functionally mature subpopulations, and is downregulated in type 1 and type 2 diabetes C_LIO_LIIn human stem cell models, MAFB is essential for generating insulin-producing {beta}-like cells, whereas glucagon-producing -like cells are reduced but still formed C_LIO_LINeither model addresses adult human islets: rodent MafB becomes -cell restricted after birth, and stem cell models capture differentiation, not maintenance C_LI C_LIO_LIWhat is the key question? O_LIIs MAFB required to maintain identity and secretory function in adult human islet cells? C_LI C_LIO_LIWhat are the new findings? O_LIMAFB knockdown in primary human pseudoislets impaired glucagon synthesis and secretion but minimally affected {beta}-cells, consistent with buffering by MAFA C_LIO_LIKnockdown in CD26+ -cell-enriched pseudoislets revealed a cell-autonomous requirement for MAFB in stimulus-secretion coupling, and destabilized -cell identity by inducing mesenchymal and extracellular matrix programs C_LIO_LIMAFB loss downregulated electron transport chain genes in the largest -cell subcluster and reduced mitochondrial respiration C_LI C_LIO_LIHow might this impact on clinical practice in the foreseeable future? O_LIPreserving MAFB activity in adult human -cells may represent a strategy to limit -cell dysfunction in diabetes C_LI C_LI

Background: Poor glycemic control is associated with cardiovascular disease (CVD) risk. However, it is unknown whether glycemic control is related to abdominal aortic calcification (AAC), a marker of subclinical CVD. We investigated the association between glycated hemoglobin (HbA1c) and moderate-to-high automated AAC among middle-aged to older adults from the general population. Methods:We included UK Biobank Imaging Study participants free of atherosclerotic CVD at baseline. HbA1c was measured at baseline (2006-2010) and categorized as normoglycemia (<39.0 mmol/mol), prediabetes (39.0-47.9 mmol/mol), undiagnosed diabetes (HbA1c [&ge;]48 mmol/mol), and diagnosed diabetes. Machine learning-derived AAC24 (ML-AAC24) scores were estimated using a validated automated algorithm applied to dual-energy X-ray absorptiometry lateral spine images (2014-2022). The associations of HbA1c with moderate-to-high ML-AAC24 (defined as a score [&ge;]2) were assessed using logistic regression adjusting for cardiovascular risk factors. Results: Of the included 48,912 participants (mean {+/-} SD age 55 {+/-} 7.6 years, 52% women), 9.7% had prediabetes (HbA1c 39.0-47.9 mmol/mol [5.7-6.4%]), 0.4% had undiagnosed diabetes, and 2.7% had diagnosed diabetes. Each 1-SD increase in log-transformed HbA1c was associated with higher odds of moderate-to-high ML-AAC24 (adjusted odds ratio [aOR] 1.12, 95% CI: 1.09-1.16). Amongst individuals with normal HbA1c, this association was consistent but somewhat weaker for each 1-SD increase in log-transformed HbA1c (aOR 1.07, 95% CI 1.03-1.10). Compared to participants with normal HbA1c, those with prediabetes (aOR 1.19, 95% CI: 1.08-1.30) or diagnosed diabetes (1.64, 95% CI: 1.39-1.94) had higher odds of moderate-to-high ML-AAC24. These associations were consistent in stratified analyses by sex, age groups, body mass index, smoking status and total cholesterol Conclusions: Linear associations between HbA1c levels and ML-AAC24 were observed in UK adults, even in those with normal HbA1c levels. These findings indicate that AAC may develop early in the dysglycemic continuum, supporting earlier cardiometabolic risk assessment even amongst people with ?normal? levels.

The NACHT, LRR and PYD domains-containing protein 3 (NLRP3) inflammasome is a protein complex that senses metabolic disturbances and in response processes IL-1beta. Prolonged activation of the NLRP3 inflammasome by metabolic stress induces chronic low-grade inflammation and contributes to the development type 2 diabetes and its comorbidities. Here, we developed a mouse model of type 2 diabetes that features impaired proinsulin processing and the ability to form islet amyloid plaques, two determinants of human type 2 diabetes pathology. We show that at advanced ages, these mice develop amyloidosis and inflammation in their insulin-producing pancreatic islets. Beta-cell mass and function were impaired in these mice and severe hyperglycemia developed within 6 months. Oral application of OLT1177, a selective inhibitor of the NLRP3 inflammasome, prevented the development of hyperglycemia. Our data show that inhibition of the NLRP3 inflammasome in a humanized mouse model of severe type 2 diabetes prevents the development of amyloid-associated hyperglycemia.

OBJECTIVE To characterize the prevalence and penetrance of maturity-onset diabetes of the young (MODY) in a multi-ancestry population using a genotype-first design. RESEARCH DESIGN AND METHODS We analyzed whole-genome sequencing and clinical data from 374,973 unrelated All of Us participants (42.0% non-European ancestry). We identified pathogenic or likely pathogenic (P/LP) variants in 10 established MODY genes and assessed carrier prevalence, diabetes penetrance, and glycemic profiles. We evaluated age-dependent diabetes risk by comparing carriers with non-carriers stratified by type 2 diabetes polygenic risk score (T2D PRS). RESULTS We identified 370 carriers of P/LP MODY gene variants (0.099%; 1 in 1,013), with similar carrier prevalence among European- and African-ancestry participants (0.105% in both groups). Diabetes penetrance was incomplete (13.4% by age 40; 43.5% by age 60) and varied by etiology: highest for GCK (56.0% by age 60), intermediate for HNF genes (HNF1A/HNF1B/HNF4A; 45.4%), and lowest for non-GCK/HNF genes (ABCC8/INS/KCNJ11/NEUROD1/PDX1/RFX6; 29.0%). In multivariable Cox models using non-carriers in the middle 80% of the T2D PRS as the reference, non-GCK/HNF gene variant carriers had modestly increased diabetes risk (HR, 1.57), similar to non-carriers in the top 10% of T2D PRS (HR, 1.64). These associations were observed in both European- and non-European-ancestry individuals. HbA1c profiles differed by etiology, with stable mild hyperglycemia in GCK variant carriers and greater variability among HNF and non-GCK/HNF gene variant carriers. CONCLUSIONS MODY gene variants showed incomplete, etiology-dependent penetrance across ancestries. Carriers of P/LP variants in lower-penetrance genes had diabetes risk comparable to that of non-carriers with high polygenic susceptibility.

Background Glucagon-like peptide-1 (GLP-1) receptor agonists and sodium-glucose cotransporter-2 (SGLT2) inhibitors are increasingly used for type 2 diabetes. Despite established metabolic, cardiovascular, and renal benefits, it remains uncertain whether GLP-1 receptor agonists are associated with longer clinically recorded Alzheimer's disease (AD)-type dementia-free survival than sulfonylureas (SU) or SGLT2 inhibitors. Methods Using All of Us electronic health records, we emulated target trials among adults aged 55 years or older with type 2 diabetes, a 12-month washout, and no prior dementia. We compared GLP-1 receptor agonists with SU and SGLT2 inhibitors. Propensity score weighting and doubly robust estimation addressed confounding. Causal survival forests estimated individualized treatment effects on 48-month RMST free from clinically recorded AD-type dementia. Findings In the GLP-1 receptor agonist versus SU comparison (6,328 individuals; 48-month NNT approximately 202), initiation was associated with a small but statistically significant increase in AD-type dementia-free survival (ATE 0.21 months; 95% CI: 0.07-0.35). The highest-benefit stratum gained 0.45 months (95% CI: 0.28-0.62). In the SGLT2 inhibitor comparison (3,070 individuals; 48-month NNT approximately 245), the average effect was not statistically significant (ATE 0.06 months; 95% CI: -0.18 to 0.31), but treatment effects were heterogeneous. The highest-benefit stratum gained 0.83 months (95% CI: 0.49-1.17). Predicted benefit was associated with older age, insulin use, lower HbA1c, and lower BMI. Interpretation GLP-1 receptor agonists may delay clinically recorded AD-type dementia compared with SU. Comparative effectiveness versus SGLT2 inhibitors may vary, supporting further study. Given the hypothesis-generating nature of these findings, diabetes treatment selection should remain guided by glycemic, cardiovascular, renal, and patient-centered considerations. Funding German Federal Ministry of Research, Technology and Space (03LWH0181B)

Background Long-horizon kidney trajectory prediction in type 2 diabetes mellitus (T2DM) is usually reported as a point estimate or event risk, although clinical decision-making also depends on whether an individual prediction is reliable. We developed an uncertainty-aware model for 48-month estimated glomerular filtration rate (eGFR) decline and tested whether conformal interval width provides a clinically structured, patient-level signal of prediction reliability. Methods We performed a secondary prognostic modeling analysis of Action to Control Cardiovascular Risk in Diabetes (ACCORD) participants with baseline and 48-month eGFR (n=6,853). The outcome was annualized eGFR change, calculated as 48-month minus baseline eGFR divided by four years. The primary baseline feature set excluded serum creatinine since eGFR is creatinine-derived, and also excluded urine biomarkers. Random forest, gradient boosting, penalized linear models, and XGBoost were compared using fixed training, calibration, and test partitions. Split and locally adaptive conformal intervals were evaluated by empirical coverage and interval width. Interval-width analyses were repeated after conditioning on baseline eGFR. Results The best primary model was random forest (R2=0.382, MAE=3.271 mL/min/1.73m2). Split 90% conformal intervals achieved empirical coverage of 0.917. Locally adaptive 90% intervals achieved empirical coverage of 0.909 with mean width 13.759 mL/min/1.73m2. In unadjusted analyses, wider intervals were associated with larger errors and more rapid decline. After interval-width quintiles were assigned within baseline-eGFR strata, wider intervals remained associated with realized prediction error (annual adjusted increase, 0.151 mL/min/1.73m2 per quintile). Beyond baseline eGFR, wider intervals were associated with younger age, female sex, higher HbA1c, higher triglycerides, and higher systolic blood pressure. Conclusions Baseline clinical variables predicted 48-month eGFR decline with good long-horizon performance in ACCORD, even after excluding serum creatinine and urine biomarkers from the primary model. Conformal prediction provided calibrated patient-specific intervals, and interval width behaved as an informative reliability phenotype rather than a random modeling artifact. These findings support a novel uncertainty-aware framing of kidney trajectory prediction in which rapid and uncertain decline can be identified from baseline clinical data.

Introduction. Diabetes is a growing global health challenge, necessitating effective management strategies. Glycosylation, a highly regulated post-translational protein modification, has emerged as a pivotal factor in diabetes pathophysiology. However, the modulation of protein glycosylation by antidiabetic treatment is still largely unknown. This study explored the longitudinal effects of four distinct antidiabetic therapies - metformin, insulin, sodium-glucose cotransporter-2 (SGLT2) inhibitors, and glucagon-like peptide-1 receptor agonists (GLP-1RA) - on plasma protein and immunoglobulin G (IgG) glycosylation in patients with type 2 diabetes (T2D). Research Design and Methods. Plasma protein and IgG N-glycans were enzymatically released, purified and chromatographically profiled in a cohort of 124 patients, examined at four time points, to assess therapy-induced glycan alterations. Linear mixed models adjusting for covariates and multiple testing (FDR<0.05) were used to investigate the associations between plasma protein and IgG N-glycosylation and antidiabetic therapy. Results. Our findings reveal that metformin, SGLT2 inhibitors, and GLP-1RA induce significant alterations in IgG glycosylation, including the increased core fucosylation and galactosylation, features associated with a reduced inflammatory IgG potential. Notably, IgG monogalactosylation, previously linked to cardioprotective effects in women, was elevated in response to GLP-1RA and SGLT2 inhibitor treatments. Plasma protein glycosylation changes were more limited, with distinct alterations observed for each therapy. Metformin and GLP-1RA similarly reduced certain fucosylated and sialylated glycans, while SGLT2 inhibitors decreased a high-mannose glycan, previously positively associated with diabetes progression. Insulin therapy had a minimal effect on protein glycosylation, with only one plasma glycan significantly altered. Conclusions. Our findings emphasise the importance of protein glycosylation as a dynamic and responsive marker in T2D treatment. The distinct glycan alterations observed in response to metformin, SGLT2 inhibitors, and GLP-1 receptor agonists provide novel insights into the molecular effects of these therapies, potentially contributing to the development of glycan-based biomarkers for personalized diabetes management.

Body mass index (BMI) alone may underestimate clinically relevant obesity because it does not capture central fat distribution. We compared obesity prevalence in Mexico using BMI-only criteria, adiposity-confirmed criteria, and the clinical obesity definition proposed by the Lancet Diabetes and Endocrinology Commission. We conducted a population-based, cross-sectional study of 13,160 adults aged 18 years or older who participated in the 2018-2019 Mexican National Health and Nutrition Survey (ENSANUT). Obesity prevalence was estimated through survey-weighted analyses that accounted for the complex sampling design. The weighted prevalence of obesity based on BMI was 34.5% (95% CI, 33.1-35.9), while 30.9% (95% CI, 29.6-32.2) met criteria for clinical obesity. One quarter of individuals with clinical obesity had a BMI under 30 kg/m2, a phenotype more common among older adults. Half of adults with a BMI under 30 kg/m2 showed elevated central adiposity. BMI alone underestimates clinically relevant obesity in Mexican adults. Adding waist-based measurements could improve the identification of individuals with excess fat and metabolic risk, both in clinical settings and population monitoring.

The insulin-like growth factor (IGF)/IGF-binding protein (IGFBP) axis has been implicated in diabetes mellitus and the associated burden of cardiovascular complications. Higher circulating levels of IGFBP-1 and IGFBP-2 have been established as markers of protection from incident type 2 diabetes, yet their associations with cardiovascular disease remain unclear. Utilising the UK Biobank (UKB) resource to integrate disease outcomes, plasma proteomics and MRI data, we examined associations of IGFBP-1 and IGFBP-2 with incident diabetes and cardiovascular disease. Approximately 50,000 UKB participants with plasma proteomic measurements for IGFBP-1 and IGFBP-2 were included. Multivariate Cox regression models revealed that participants in the highest quartiles of IGFBP-1 and IGFBP-2 had a substantially lower risk of incident diabetes (hazard ratio (HR) = 0.31 and 0.32 respectively), but, paradoxically, had increased risks of incident macrovascular disease, all-cause and cardiovascular-related mortality (HR = 1.81 and 2.39). Both proteins were negatively associated with HbA1c levels, triglyceride/HDL ratio and abdominal adiposity, yet positively associated with NT-proBNP, troponin I, cardiac chamber size and aortic dimensions. In summary, negative associations of IGFBP-1 and IGFBP-2 with incident diabetes mellitus did not translate to a reduced cardiovascular risk, suggesting potentially complex actions of IGFBP-1 and IGFBP-2 in the pathophysiology of cardiometabolic disease.

Background: Insulin resistance is a core pathophysiologic feature of metabolic disease, but its reference-standard assessment by steady-state plasma glucose (SSPG) testing is procedurally demanding and labor-intensive, limiting use in routine clinical care and large-scale research. Because OGTT glucose profiles are widely available, we aimed to develop a glucose-only metric to characterize dynamic glucose responses and estimate SSPG-measured insulin resistance. Methods: We developed the Width-Delay Index (WDI), a glucose-only OGTT metric integrating relative exposure width, delayed exposure timing, and glycemic floor. In a dataset of 32 subjects with 16-point venous OGTT profiles and paired SSPG measurements, WDI performance was assessed using leave-one-out cross-validation (LOOCV) for SSPG prediction, together with insulin-resistance discrimination and sparse-sampling robustness analyses. Results: The 15-120 min OGTT window yielded the strongest WDI performance. WDI15-120 predicted SSPG with LOOCV R2 = 0.57 (95% CI, 0.27-0.77), Pearson r = 0.77, and Spearman rho = 0.74. WDI15-120 showed higher predictive performance than standard OGTT glucose measures and insulin-derived indices, including HOMA-IR, Matsuda index, and disposition index. WDI15-120 also discriminated insulin-resistant from insulin-sensitive subjects with AUROC = 0.969. When recalculated from conventional 5-point OGTT sampling, WDI15-120 retained substantial performance, with LOOCV R2 = 0.41 and AUROC = 0.945. Conclusions: WDI provides a simple, glucose-only, physiologically interpretable approach for estimating SSPG-measured insulin resistance from OGTT glucose dynamics.

Background Anthropometric measures do not adequately capture heterogeneity in body fat distribution and corresponding cardiometabolic risk, whereas magnetic resonance imaging (MRI) enables precise differentiation and quantification of adipose tissue compartments and ectopic fat. We aimed to validate previously derived MRI-based body composition subphenotypes and their cardiometabolic risk profiles in two independent European cohorts. Methods Using deep learning-based image analysis, we quantified bone marrow, visceral, subcutaneous, cardiac, renal sinus, hepatic, skeletal muscle, and pancreatic fat in the imaging substudies of two population-based cohorts: the German National Cohort (NAKO, N=29,314, age range 19-74 years) and the UK Biobank (N=36,109, age range 40-69 years). Body composition subphenotypes, previously identified by k-means clustering, were evaluated using a rigorous statistical cluster validation framework with method-based and results-based approaches. In NAKO, cross-sectional associations between subphenotypes and estimated cardiovascular disease risk scores were examined using linear regression. In UK Biobank, longitudinal associations between subphenotypes and incident cardiometabolic outcomes, ascertained through hospital record linkage, were analysed using Cox regression. Findings All five body composition subphenotypes were robustly validated across both cohorts, and showed distinct fat distribution patterns and cardiometabolic risk profiles: I "lean", II "average adiposity", III "bone and muscle adiposity", IV "hepato-abdominal adiposity", and V "general and pancreatic adiposity". Subphenotypes I-III showed progressive adipose tissue remodelling patterns likely reflecting ageing trajectories. The "hepato-abdominal adiposity" subphenotype showed highest risk of incident diabetes, whereas the "general and pancreatic adiposity" subphenotype showed highest overall cardiovascular disease burden and metabolic impairment. Interpretation MRI-derived body composition subphenotypes represent distinct fat distribution patterns that reflect ageing- and disease-related processes, which supports the potential of body composition phenotyping for improved cardiometabolic risk stratification and targeted prevention.