Aims/hypothesis This study investigated insulin sensitivity using the hyperinsulinaemic–euglycaemic clamp technique in individuals with obesity and type 2 diabetes treated with tirzepatide at the low dose of 5 mg over a 12-week treatment period. Methods This prospective, single-arm, open-label, single-centre study was conducted in obese individuals with type 2 diabetes. Participants received tirzepatide 2.5 mg once weekly for 4 weeks; the dose was then increased to 5 mg for the remaining 8 weeks. The primary outcome was change in the glucose infusion rate. Secondary outcomes were changes in HbA1c, body weight, body composition, lipid profile, glucagon level, the HOMA2-IR and HOMA2-%β indices, and the association of these variables with the glucose infusion rate (GIR). Results Sixteen participants completed the study. The GIR increased from 3.21 to 5.16 mg min−1 kg−1 (p<0.001). HbA1c decreased from 63.4 to 43.6 mmol/mol (7.95% to 6.14%, p<0.001) and body weight decreased by 4.9 kg (5.0%, p<0.001). Muscle mass, fat mass and fat percentage significantly decreased by 1.8%, 9.1% and 4.5%, respectively. Glucagon decreased significantly from 28.8 pg/ml to 20.8 pg/ml. However, simple linear regression analysis revealed no significant relationship between changes in GIR and other clinical variables. Conclusions/interpretation Tirzepatide significantly improves insulin sensitivity within 12 weeks, indicating an early metabolic effect that is not solely attributable to weight loss. Trial registration UMIN registration number: UMIN000056862 Graphical Abstract

In economically developed countries most lean individuals presenting with insulin-dependent diabetes have autoimmune type 1 diabetes. However, in many rural areas of low- and middle-income countries (LMIC), 40–50% of individuals with a similar clinical presentation are negative for diabetes-associated autoantibodies at initial clinical presentation. The phenotype differs from the classical presentation of type 1 diabetes even in those with evidence of an autoimmune process: altered autoantibody profile; later peak age of onset; and, in those with post-pubertal clinical presentation, more marked male predominance. The incidence of insulin-dependent diabetes in LMIC is low, even when assessing those with and without autoantibodies together. A framework of possible pathophysiological mechanisms underlying the observed phenotypic differences is presented to explain how chronic undernutrition and micronutrient deficiencies might alter the presentation of insulin-dependent diabetes. Inhabitants of rural sub-Saharan Africa (SSA) depend almost entirely on staple foods grown locally in nutrient-deficient soil. The resulting chronic undernutrition, often intergenerational, affects linear growth and body morphology, and has direct immune and non-immune effects on beta cell development and function. Undernutrition directly affects thymic function, alters the autoimmune profile and is often associated with social deprivation and parasitic infection, both of which can delay and modify the (auto)immune response. Non-immune effects of undernutrition include beta cell stress, associated with apoptosis and formation of neoantigens. That environmental effects of undernutrition and social deprivation affect the altered insulin-dependent diabetes phenotype is shown by the movement back towards a classical type 1 diabetes phenotype in offspring of emigrants from SSA who are born in and develop insulin-dependent diabetes in an economically developed country. The degree of phenotype change depends on how long the parents have lived in their adopted country. It has recently been proposed that insulin-dependent diabetes in those who are mal/undernourished be called type 5 diabetes. There is need for clinician recognition of the altered phenotype(s) of insulin-dependent diabetes resulting from chronic undernutrition in rural LMIC. Additionally, changes in agricultural practice are needed to improve the nutrient content of food consumed by the rural population. Graphical Abstract

Aims/hypothesis Genetic association studies have demonstrated that partial loss of SLC30A8 Function protects against type 2 diabetes in humans. We investigated the impact of complete loss of SLC30A8 Function on type 2 diabetes risk and related phenotypes in humans. Methods The Pakistan Genome Resource (PGR), a biobank comprising whole-exome and whole-genome sequences of 145,037 participants, was analysed for phenotypic associations with SLC30A8 loss-of-function (LoF) variants. To follow up on the observations in the PGR, we conducted recall-by-genotype analyses of SLC30A8 LoF heterozygotes and homozygotes, as well as their participating family members, using OGTTs. Results We identified 18 SLC30A8 knockouts, including homozygotes for a variant enriched in South Asians (Gln174Ter), and 1024 heterozygotes for LoF variants. Type 2 diabetes risk was lower in SLC30A8 LoF heterozygotes and homozygotes relative to non-carriers, and the protective effect strengthens in a gene dose-dependent manner (ORadditive=0.62; 95% CI 0.53, 0.72; p=1.1×10–9; ORrecessive=0.34; 95% CI 0.12, 0.93; p=0.04). OGTTs in recall-by-genotype studies showed a gene dose-dependent reduction in glucose levels, coupled with elevated insulin. Conclusions/interpretation The corrected insulin response, disposition index and insulin sensitivity index in LoF heterozygotes and homozygotes indicated higher glucose-stimulated insulin secretion with preserved beta cell function that was independent of BMI. These data suggest that therapeutic inhibition of SLC30A8, up to and including complete knockout, may treat type 2 diabetes safely and effectively. Graphical Abstract

Aims/hypothesis This study aimed to assess the impact of adding dapagliflozin to insulin therapy on key hormonal determinants of glucose regulation and ketogenesis. We hypothesise that dapagliflozin increases glucagon-like peptide 1 (GLP-1), glucagon and ketone body concentrations, based on the results of a pilot study. Methods The study was designed as a randomised, placebo-controlled, open-label, crossover intervention study with two periods (dapagliflozin and placebo intake), including patients of the Department of Diabetes, Endocrinology, Clinical Nutrition & Metabolism, Inselspital, Bern University Hospital, University of Bern. Individuals with type 1 diabetes (C-peptide concentrations <0.1 nmol/l) with a duration >5 years and a BMI of 20–29 kg/m2 were included. They received 10 mg of dapagliflozin or placebo daily for 7 days throughout two independent treatment periods, separated by a 14 day washout period. Allocation was done by a computed randomisation tool (REDCap), without blinding of the participants or the investigators. On day 7 of each treatment period, hyperinsulinaemic–euglycaemic clamps (HECs) and OGTT clamps (OGTTCs) were performed to assess changes in the secretion of GLP-1, glucagon, somatostatin and total ketone bodies. The objective was to evaluate the effects of adding the sodium–glucose cotransporter 2 (SGLT2) inhibitor dapagliflozin to insulin therapy on GLP-1 during OGTTC (primary endpoint), GLP-1 secretion during HEC, and glucagon, somatostatin and ketogenesis during OGTTC and HEC (secondary endpoints). The primary endpoint was concentrations of GLP-1 during OGTTC. Secondary endpoints included GLP-1 during HEC and glucagon, somatostatin and ketone body concentrations during OGTTC and HEC. Results A total of 13 individuals with type 1 diabetes were included and randomised. All of them received dapagliflozin and placebo, finished the sequences per protocol and were analysed per protocol. GLP-1 concentrations did not differ significantly between treatments in the OGTTC (median [IQR] dapagliflozin 192.8 [129.8–257.2] pmol/l vs placebo 176.3 [138.4–227.4] pmol/l; p=0.7) or HEC (median [IQR] dapagliflozin 208.6 [133.6–294.0] pmol/l vs placebo 203.1 [150.2–291.8] pmol/l; p=0.7). Glucagon concentrations did not significantly differ between treatments in the OGTTC (median [IQR] dapagliflozin 1.54 [0.84–3.68] ng/l vs placebo 1.54 [0.82–4.64] ng/l; p=0.8) or HEC (median [IQR] dapagliflozin 1.59 [0.87–3.54] ng/l vs placebo 1.63 [0.91–3.96] ng/l; p=0.3). Somatostatin concentrations remained comparable between treatments during the HEC (median [IQR] dapagliflozin 41.1 [26.8–73.8] pmol/l vs placebo 47.0 [23.0–77.6] pmol/l; p=0.2) and OGTTC (median [IQR] dapagliflozin 51.1 [31.1–77.0] pmol/l vs placebo 45.3 [30.0–70.5] pmol/l; p=0.2). Plasma ketone bodies were higher with dapagliflozin during the HEC (median [IQR] dapagliflozin 0.15 [0.04–0.47] mmol/l vs placebo 0.03 [0.01–0.12] mmol/l; p<0.001) and OGTTC (median [IQR] dapagliflozin 0.10 [0.03–0.22] mmol/l vs placebo 0.03 [0.01–0.12] mmol/l; p<0.001). Conclusions/interpretation Short-term dapagliflozin treatment in type 1 diabetes increases plasma ketone concentrations without affecting the secretion of GLP-1, glucagon or somatostatin. Higher ketone body concentrations highlight the elevated risk of diabetic ketoacidosis associated with the adjunct intake of dapagliflozin. Trial registration ClinicalTrials.gov NCT04035031. Funding Swiss National Science Foundation, project number 32003B_185019. Graphical Abstract

Aims/hypothesis The bromodomain and extra-terminal (BET) protein family acts as ‘epigenetic readers’ to identify the acetylation marks on histones that convert the acetylated lysine residues into observable phenotypes. BET proteins have gained attention due to their ability to modulate the transcription of pathology-related genes involved in cancer and autoimmune diseases, including type 1 diabetes mellitus. However, targeting BET proteins may have secondary effects on other host cells. We aimed to elucidate possible secondary effects of BET inhibition on pancreatic beta cell function. Methods We studied the effect of the small-molecule BET inhibitor I-BET151 on pancreatic beta cells in vitro, ex vivo and in vivo. GTTs, ITTs and glucose-stimulated insulin secretion assays were performed in healthy mice and a mouse model of diabetes following daily i.p. injections of I-BET151 for 2 weeks. Transcriptomic analysis was carried out on primary mouse islets, which were subjected to ex vivo I-BET151 treatment. Changes in expression were further validated in primary human islets. Results Administration of I-BET151 modestly but significantly increased glucose excursions and reduced insulin responses in both healthy mice and diabetic mice. We found that I-BET151 exposure significantly reduced the expression of Hnf4α (also known as Hnf4a; MODY1), Gck (MODY2), Hnf1α (also known as Hnf1a; MODY3), Glut2 and other genes essential for beta cell function in rat INS-1E insulinoma cells and in mouse primary islets and human islets. Global gene expression analysis in cells treated with I-BET151 showed a downregulation of the phosphoinositide-3-kinase (PI3K)–Akt pathway. Downregulation of forkhead box protein O1, a downstream transcriptional factor of the PI3K–Akt pathway, partially rescued I-BET151-driven downregulation of Gck and insulin secretion. Likewise, islets from I-BET151-treated mice showed a modest reduction in glucose-stimulated insulin secretion. Conclusions/interpretation The results presented here suggest that BET inhibition therapy should be used with caution due to possible bimodal effects at high concentrations at the detriment of pancreatic beta cell function. Graphical Abstract

Recent advances in genome-wide approaches, the availability of isolated human islets for research and the evaluation of novel incretin mimetics in large clinical trials have brought about remarkable progress in our understanding of the role of the pancreatic beta cell in type 2 diabetes. Here, we review key developments in type 2 diabetes initiation, progression and remission, focusing mostly on human studies published in the last 5 years. Progress in multi-omics technologies has enabled researchers to identify links between type 2 diabetes risk variants and gene regulatory networks in islet endocrine cells that control beta cell development, function and stress resilience. These studies support the notion that early abnormalities in insulin secretion, rather than a reduction in beta cell mass, play a fundamental and primary role in early type 2 diabetes pathogenesis. Contributing to these intrinsic beta cell defects are various pathogenic signals from other (endocrine and non-endocrine) islet cells, the exocrine pancreas, the gut and insulin-sensitive tissues. It has also become apparent that beta cells comprise a heterogeneous population that responds differently to stress situations and that sex-related differences in beta cell responses should not be underestimated. Finally, human clinical trials have clearly demonstrated that diabetes remission can be achieved using glucose-lowering therapies and particularly strategies focused on weight loss, including bariatric surgery and, more recently, the use of highly efficient new drugs targeting the incretin system. While progress in the last 5 years has been significant, much remains to be uncovered to bring these advances to the clinic and thereby alleviate the dramatic consequences of type 2 diabetes complications for the hundreds of millions of people who live with this disease. Graphical Abstract

Aims/hypothesis Pancreatic ductal adenocarcinoma-related diabetes mellitus (PDAC-DM) is a paraneoplastic syndrome with a poorly understood pathophysiology. PDAC-DM is often clinically confused with type 2 diabetes, resulting in delayed cancer detection and poorly individualised hyperglycaemia treatment. We investigated whether these forms of diabetes can be distinguished at the metabolic level. Methods Adults with either cancer treatment-naive PDAC-DM (n=28) or type 2 diabetes (n=97), and with diabetes onset within 3 years, underwent mixed-meal tolerance tests to investigate glucose metabolism. Outcomes included insulin sensitivity (Matsuda index), insulin secretion (insulinogenic index), beta cell function (oral disposition index), insulin clearance, and postprandial glucagon, glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic peptide (GIP) responses. Results Compared with type 2 diabetes, individuals with PDAC-DM showed ~2.5-fold greater insulin sensitivity, ~81% lower insulin secretion and ~40% lower beta cell function. Insulin clearance was higher in the PDAC-DM group than the type 2 diabetes group, with and without adjustment for insulin sensitivity. Glucagon and GLP-1 levels increased after a meal in both groups, but levels were higher in the PDAC-DM group. GIP levels were similar between groups. The metabolic differences between groups persisted after adjustment for age, sex and BMI. Conclusions/interpretation PDAC-DM and type 2 diabetes are metabolically distinct, with different defects responsible for hyperglycaemia. PDAC-DM is characterised predominantly by insulin deficiency and displays higher insulin sensitivity than type 2 diabetes. There are also differences in alpha cell regulation and insulin clearance compared with type 2 diabetes. These findings identify biological characteristics that may have implications for individualised treatment of PDAC-DM and guide diagnostic biomarker discovery for early PDAC diagnosis. Graphical Abstract

To provide visual guidance in concise algorithms and tables to assist with clinical decision making in the management of adults with dyslipidemia to reduce risk of ASCVD and triglyceride-induced pancr...

Aims/hypothesis The aim of this work was to investigate tissue-specific insulin resistance in South Asian and Nordic women with previous gestational diabetes mellitus (pGDM) and to evaluate potential ethnic differences contributing to type 2 diabetes risk. Methods A cross-sectional study using a two-step hyperinsulinaemic–euglycaemic clamp with a glucose tracer was conducted to assess insulin sensitivity in muscle, liver and adipose tissue in 19 South Asian and 16 Nordic women with pGDM and prediabetes (impaired glucose tolerance and/or impaired fasting glucose), along with 16 ethnicity-specific control women. We assessed inflammation and mitochondrial genes by mRNA sequencing of adipose tissue. Results Both South Asian and Nordic women with pGDM showed reduced total glucose disposal (mainly due to muscle insulin resistance) and hyperinsulinaemia compared with the control group. Endogenous glucose production (mainly due to hepatic insulin resistance) was elevated in Nordics with pGDM, while South Asians with pGDM showed pronounced adipose tissue insulin resistance (reduced suppression of glycerol during clamp). mRNA sequencing of adipose tissue indicated increased tissue inflammation in South Asian women compared with Nordic women with pGDM. Furthermore, we observed a differential response to hyperinsulinaemia in South Asians vs Nordics related to mitochondrial mRNA, such as thymidine kinase 2 (TK2). Correlations between adiposity markers and insulin sensitivity also differed by ethnicity, suggesting that the pathways leading to type 2 diabetes may vary across populations. Conclusions/interpretation South Asian and Nordic women with pGDM exhibited differences in insulin resistance profiles, with South Asians showing greater adipose tissue insulin resistance and inflammation. Graphical Abstract

Aims/hypothesis The study aimed to explore the association between maternal glucose levels in pregnancy and offspring’s metabolism and adiposity at approximately 18 years of age. Methods Pregnant women from the Hong Kong Field Centre enrolled in the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study underwent a 75 g OGTT at 24–32 gestational weeks. Offspring’s metabolic and adiposity traits were assessed at 18 years postpartum. Associations were evaluated using multiple linear regression and logistic regression. Results Among the 506 mother–child pairs followed up to 18 years, maternal fasting plasma glucose (FPG) in pregnancy was positively associated with offspring’s FPG (β = 0.06 [95% CI 0.02, 0.09]), while maternal 1 h plasma glucose (PG) showed a positive association with offspring’s FPG (β = 0.05), 30 min PG (β = 0.21) and 2 h PG (β = 0.14). All maternal glycaemic levels were associated with an increased risk of offspring being overweight/obese, particularly maternal 1 h PG (OR 1.50 [95% CI 1.17, 1.93]). Offspring of mothers with gestational diabetes mellitus showed a higher prevalence of abnormal glucose tolerance (11.86% vs 7.97%), impaired fasting glucose (1.89% vs 0.49%) and impaired glucose tolerance (10.34% vs 7.13%) than offspring of mothers with normal glucose tolerance, although these associations did not reach statistical significance in fully adjusted models, underscoring the benefit of considering maternal glucose as a continuous trait. Conclusions/interpretation Maternal glucose levels in pregnancy showed a long-term association with offspring’s metabolic health into young adulthood, with continuous associations across the full maternal glucose spectrum, suggesting a graded effect of maternal hyperglycaemia on offspring’s metabolic risk. Graphical Abstract

Aims/hypothesis We aimed to examine offspring adiposity trajectories from birth to age 9–14 years and to assess the joint associations of maternal overweight and gestational diabetes mellitus (GDM) with these trajectories. Methods This is a prospective cohort study with 564 mother–child dyads from the Hyperglycemia and Adverse Pregnancy Outcome study Hong Kong field centre. Assessments and anthropometric measurements were taken during pregnancy, at delivery and at median ages of 7 and 10 years postpartum. Offspring adiposity was primarily assessed using sum of skinfold thickness. We used linear mixed-effect models to evaluate the independent and joint associations of maternal overweight and GDM with the offspring adiposity trajectories, and applied group-based trajectory modelling to identify distinct patterns of adiposity development based on both statistical indices and clinical interpretability. Results Offspring skinfold thickness trajectories varied significantly based on maternal overweight and GDM (p<0.05). Group-based trajectory modelling identified two trajectory groups for skinfold thickness: 52.1% with slow increase and 47.9% with rapid increase. Combined maternal overweight and GDM was associated with 6.90-fold increased risk (95% CI 1.89, 33.32; p=0.006) of the rapidly increasing trajectory. Linear mixed-effect model analysis showed greater increases in skinfold thickness among offspring of mothers with either condition, with the highest trajectory observed in offspring of mothers with both conditions (β 1.62; 95% CI 0.69, 2.54; p=0.001). Conclusions/interpretation Maternal overweight and GDM are independently and jointly associated with rapidly increasing adiposity trajectories from birth to early adolescence. The findings underscore the importance of considering both maternal metabolic conditions when evaluating offspring adiposity risk. Graphical Abstract

Aims/hypothesis The loss of pancreatic beta cell mass and identity is a hallmark of diabetes. While factors such as beta cell overwork (insulin hypersecretion) and elevated intracellular calcium have been implicated, beta cell identity loss also occurs in KATP gain-of-function (KATP-GOF) mice, a model of human neonatal diabetes, even in the absence of these factors. This suggests additional underlying mechanisms. Autophagy, a key process for cellular homeostasis, is impaired in the islets and beta cells of both type 1 and type 2 diabetes, but its role in monogenic diabetes with insulin secretory deficiency remains unclear. We hypothesise that autophagy dysfunction contributes to beta cell identity loss in KATP-GOF mice, and that intermittent fasting (IF) can restore autophagic flux, thereby preserving functional beta cell mass. Methods To test this, adult tamoxifen-inducible KATP-GOF mice and littermate controls were randomly assigned to two groups: (1) chow diet ad libitum; and (2) chow diet with alternate-day IF. Results KATP-GOF mice fed ad libitum developed severe hyperglycaemia due to impaired insulin secretion. This was followed by a reduction in insulin content, disruption of beta cell autophagic flux, autophagosome accumulation and, ultimately, the loss of beta cell identity and dedifferentiation. In contrast, KATP-GOF mice subjected to alternate-day IF exhibited lower blood glucose levels, improved mitochondrial morphology, restoration of autophagic flux and reestablishment of beta cell identity. Conclusions/interpretation This study provides the first evidence of autophagy impairment in non-obese, insulin secretory-deficient, KATP-induced diabetes mice and demonstrates that IF restores both autophagic flux and beta cell identity. This finding suggests that similar mechanisms may contribute to beta cell dysfunction in other forms of diabetes. Graphical Abstract

Aims/hypothesis The incidence of type 1 diabetes increased during the pandemic in various countries. SARS-CoV-2 infections may trigger the development of type 1 diabetes, but the evidence is inconclusive. This study aimed to assess trends in type 1 diabetes incidence between 2007 and 2023, and to quantify the association between SARS-CoV-2 infections and the risk for developing type 1 diabetes. Methods The study included all individuals under 30 years old registered in Sweden. Deviations in type 1 diabetes incidence from pre-pandemic trends (2007–2019) were assessed for each pandemic year (2020–2023) using Poisson regression. The effect of SARS-CoV-2 infections was assessed using Cox proportional hazards models in a cohort of infected individuals with five control individuals from the infection date of the case, matched by birth year, sex and region. Results Compared with the predicted linear trend, type 1 diabetes incidence increased by 12% during 2021 (incidence rate ratio [IRR] 1.12; 95% CI 1.06, 1.19) and 9% during 2022 (IRR 1.09; 95% CI 1.02, 1.16), but reverted to pre-pandemic trends in 2023. Overall, the adjusted HR for developing type 1 diabetes after SARS-CoV-2 infection was 0.96 (95% CI 0.79, 1.16). Children between 5 and 10 years old were more likely to develop type 1 diabetes within the first 28 days after infection (HR 2.68; 95% CI 1.22, 5.89), although their hazard over the whole follow-up period was not increased. Conclusions/interpretation Sweden, with its non-restrictive pandemic response, saw a transient increase in type 1 diabetes incidence that was only partially associated with SARS-CoV-2 infections. Other explanations should be investigated, including environmental and lifestyle factors. Graphical Abstract

Aims/hypothesis Previous studies have indicated that 29-amino-acid glucagon (i.e. ‘pancreatic’ glucagon) circulates in totally pancreatectomised individuals and that a postprandial glucagon response can be detected. Using a glucagon receptor antagonist (GRA), we investigated the possible role of extrapancreatic glucagon on glucose, lipid and amino acid metabolism in totally pancreatectomised individuals. Method In a randomised, crossover study, nine totally pancreatectomised individuals and nine matched healthy control individuals were given, in randomised order (planned on the website www.random.org ), 300 mg GRA (LY2409021; Eli Lilly) or placebo 10 h before two 3 h OGTTs. The experiment was double-masked (i.e. both participants and investigator were masked for the type of the experimental day [day A vs day B]). The key inclusion criteria for the healthy control participants were age >18 years, normal fasting plasma glucose and HbA1c 31–44 mmol/mol (6.0–7.2%), haemoglobin >7.0 mmol/l (men) / >6.5 mmol/l (women) and informed consent. Key inclusion criteria for the pancreatectomised individuals were age >18 years, haemoglobin in the normal range and informed consent. The primary endpoint was the difference in plasma glucose excursions between study days. Results Glucagon concentrations remained unchanged from fasting concentrations during the OGTT in the totally pancreatectomised individuals on both study days and circulating glucose, lipids and amino acid levels were unaffected by treatment with LY2409021 compared with placebo. In the control group, LY2409021 resulted in relevant pharmacodynamic effects, including lower fasting plasma glucose (4.7 [0.1] vs 5.2 [0.1] mmol/l, p=0.001) and augmented concentrations of amino acids in plasma, compared with placebo. Conclusions/interpretation We conclude that inhibition of the glucagon receptor using LY2409021 during OGTT in totally pancreatectomised individuals does not produce detectable effects on glucose, lipid or amino acid metabolism, ruling out metabolic effects of extrapancreatic glucagon. Trial registration ClinicalTrials.gov (NCT02944110). Funding This study was supported by grants from the Aase and Ejnar Danielsen’s Foundation and the Novo Nordisk Foundation. Graphical Abstract

Aims/hypothesis The role of intra-pancreatic lipid and collagen in type 2 diabetes pathogenesis remains unclear. We sought to examine this in pancreases from organ donors with and without diabetes. Methods Tissue biopsies from 36 adult donor pancreases with/without type 2 diabetes were collected from 16 anatomically defined regions, with H&E, Sirius Red Fast Green and chromogranin A immunohistochemical staining and quantification performed. Intracellular lipid droplet area was quantified using transmission electron microscopy in acinar, islet endocrine, beta and alpha cells identified through ultrastructural morphology. Results Increasing adipocyte proportional area was associated with increasing pancreas donor BMI (r=0.385, p=0.021), decreased acinar area (r=−0.762, p<0.001) and increased endocrine mass (r=0.749, p<0.001). Fibrosis was not associated with BMI, acinar area or endocrine mass. Type 2 diabetes was associated with decreased islet circularity and reduced beta:alpha cell ratio but endocrine mass was not affected. Adipocyte and fibrosis proportional areas were highest in donors with diabetes but not associated with each other. Pancreases with high fat and those with high fibrosis (>40% proportional area) appeared to form two separate subgroups. All donors with insulin-treated diabetes had a high collagen proportional area. Fibrosis but not adipocytosis was associated with decreased beta:alpha cell ratio. There was an inverse relationship between pancreatic adipocytosis and intra-acinar cell lipid content (r=−0.490, p=0.003), with the lowest levels seen in type 2 diabetes. Beta cell lipid content was associated with BMI but not type 2 diabetes. Conclusions/interpretation Systematic human pancreatic analysis revealed two separate type 2 diabetes phenotypes: fatty, associated with central obesity; and fibrotic, associated with reduced beta cell mass without central obesity. This suggests distinct underlying pathogenic mechanisms and has potential for developing personalised disease-modifying therapeutics. Graphical Abstract

Aims/hypothesis Faecal microbiota transplantation (FMT) may alleviate gastrointestinal symptoms in individuals with diabetic gastroenteropathy, as demonstrated in a recent placebo-controlled trial. In most participants, symptom relief was transient, raising the need for repeated treatments. This study assessed the long-term efficacy, safety and feasibility of repeated, on-demand FMT as a maintenance treatment in this patient population. Methods All 20 participants from the randomised clinical trial were offered extended open-label treatment with FMT. Symptom assessments were conducted by telephone every 2–3 months using the Gastrointestinal Symptom Rating Scale for Irritable Bowel Syndrome (GSRS-IBS). Secondary measures included bowel movement frequency, stool consistency assessed using the Bristol Stool Scale, perceived treatment benefit on a seven-point Likert scale, and adverse events (AEs). FMT was primarily given as oral capsules, and colonoscopy was used for participants who could not swallow capsules. Results Of the original 20 participants, 17 were included in the present study and followed from September 2021 to December 2024, with a median Duration of follow-up of 33.2 months (range 14.7–39.1 months). Participants received a total of 95 FMT treatments, with a median of five per participant and a median interval of 5.3 months between treatments. FMT induced consistent symptom relief, with reduced GSRS-IBS scores across multiple treatments. At the last FMT treatment provided, the mean GSRS-IBS score had decreased from 60 (95% CI 54, 66) at baseline to 35 (95% CI 29, 40), with a mean difference of −25 (95% CI −18, −33). The occurrence of frequent bowel movements 2 weeks after treatment (> 7 per day) decreased from 19% (95% CI 10%, 28%) to 3% (95% CI 0%, 7%). Stool consistency improved after treatment, and the frequency of normal stool types (Bristol Stool Scale score 3–5) increased from 28% (95% CI 18%, 39%) to 76% (95% CI 66%, 86%). Participant satisfaction was high, with 86% reporting considerable benefits (Likert scores 5–7). Repeated FMT was generally well tolerated, with most AEs being mild and self-limiting. Fifteen serious AEs were documented, of which only one was deemed to be possibly related to FMT. Conclusions/interpretation Repeated, on-demand FMT is effective and safe for long-term treatment of individuals with type 1 diabetes and severe diabetic gastroenteropathy. Trial registration ClinicalTrials.gov NCT04749030 Funding The study was funded by a Steno Collaborative Grant (no. 0058906) Graphical Abstract