Glucose Load Following Prolonged Fasting Increases Oxidative Stress– Linked Response in Individuals With Diabetic Complications

Efforts to understand the heterogeneity of type 2 diabetes (T2D) have led to the identification of distinct diabetes subtypes, enabling the possibility for precision diabetology (1–3). The primary findings of the Action for Health in Diabetes (Look AHEAD) trial showed that intensive lifestyle intervention does not reduce cardiovascular risk in individuals with T2D (4). However, secondary analysis has identified subgroups of study participants with increased vulnerability to intensified lifestyle interventions (5). Notably, the poor-glucose-control subgroup showed a greater cardiovascular risk following caloric restriction, despite weight loss and improved glycemic control (5). The mechanisms behind such differential responses to lifestyle interventions in individuals with T2D remain unclear. Identification of T2D individuals who are susceptible to adverse effects of lifestyle interventions will therefore help to optimize diabetes management strategies.

Recent findings show that various fasting regimens have beneficial effects in metabolic disorders and neurodegenerative diseases (6–9). Debate centers on whether prolonged fasting (≥48 h) is superior to chronic caloric restriction. Previously, we have shown that repetitive prolonged fasting in T2D individuals improved temporarily microalbuminuria, insulin resistance, and dicarbonyl stress (10), without detrimental effects on somatosensory nerve function (11). However, the improvement in dicarbonyl stress was only transient and ceased after refeeding. Studies on how fasting might challenge the metabolic flexibility and antioxidative capacity after refeeding in T2D are lacking. Such research is relevant for dietary interventions in diabetes management. Metabolic inflexibility and insulin resistance in T2D might limit the adaptation during fasting and refeeding, requiring a rapid shift from catabolism to anabolism (12–14). T2D individuals exhibit increased endogenous glucose production before and after eating, leading to fasting and postprandial hyperglycemia (15). Therefore, glucose intake after prolonged fasting may exacerbate hyperglycemic stress in T2D.

Oxidative stress is harmful to cells, yet antioxidant therapies have not proven protective in large-scale clinical studies, suggesting that the relationship between oxidative stress and disease may be more complex than previously understood (16,17). Low-level mitochondrial oxidative stress can induce changes that make cells less vulnerable to future stress, a response known as mitohormesis (18,19). Increased mitochondrial stress in yeast cells, caused by prior exposure to high glucose or to low-level dicarbonyl stress, enhances cellular survival against methylglyoxal (MG) (20). MG, a glycolysis byproduct, contributes to diabetes progression and complications (21–23). Interventions linked to enhancing mitochondrial oxidative stress have extended the lifespan in various model organisms (24–27). Hormetic pathways, involving systemic mediators of mitochondrial stress, such as GDF15 and FGF21, regulate metabolic balance (19,28,29). However, assessing mitohormesis in complex organisms remains challenging, with human studies lacking.

Based on our previous findings on hormetic response against MG, and since glucose load leads to increased glycolysis and therefore to increased production of endogenous MG (20,22), we hypothesized that glucose load after prolonged fasting increases oxidative stress–linked cellular resistance against MG in glucose-tolerant individuals. We further hypothesized that this hormetic protective response is lost in individuals with T2D and further deteriorated in those with T2D and diabetes complications.

This exploratory pilot study was performed at the Clinic of Endocrinology, Diabetology, Metabolic Diseases and Clinical Chemistry at the University Hospital of Heidelberg in Germany. The study was approved by the ethics committee of the University of Heidelberg, Heidelberg, Germany (Ethic-Nr. S–459/2018) in compliance with national guidelines and the declaration of Helsinki. This study was registered at the German Clinical Trials Register (Deutsches Register Klinischer Studien DRKS-ID: DRKS00014287). All participants gave written informed consent.

Glucose-tolerant individuals (CON, n = 10, 4/6 females/males) and individuals with T2D with (T2D+, n = 10, 4/6 females/males) and without diabetes complications (T2D−, n = 10, 4/6 females/males) were included in the study. Participants were selected from the Heidelberg Study on Diabetes and Complications (HEIST-DiC), a prospective longitudinal cohort study (ethics number S–383/2016, ClinicalTrials.gov Identifier NCT03022721). T2D groups were selected based on glucose control and thorough clinical phenotyping. They were matched for diabetes duration, glycemic control, glucose-lowering medication, and medication with renin-angiotensin-aldosterone system (RAAS) inhibitors. The CON group was matched with the T2D for sex, age, and BMI. Inclusion and exclusion criteria are provided in the Supplementary Material (10). Withdrawal criteria included subject’s request and lack of increase in blood and/or urine ketone bodies after the prolonged fasting phase.

Participants underwent a standardized 75-g oral glucose tolerance test (OGTT) after an overnight fast on the day before starting prolonged fasting (visit 1) and repeated the OGTT directly after the prolonged fasting (visit 2). The participants were instructed to comply with a 5-day fasting-mimicking diet (FMD) (details given below). Peripheral blood mononuclear cells (PBMC) were isolated before and 2 h after glucose intake for further analysis. The 2-h plasma glucose could not be measured in one participant of the CON group because of analytical error and therefore was excluded from the analysis of 2-h plasma glucose levels. Anthropometric measurements, clinical examination, and blood sampling were performed on both visits. Study design is shown in Fig. 1A.

Prolonged fasting was achieved by using a 5-day FMD. This diet mimics fasting-like effects on glucose and ketone bodies. Daily energy intake in day one is 4,600 kJ (11% protein, 46% fat, and 43% carbohydrates), whereas, from days 2 to 5, it is 3,000 kJ (9% protein, 44% fat, and 47% carbohydrate) (30). As previously described (10), patients on insulin therapy were instructed to discontinue short-acting insulin and to reduce the long-acting insulin by 50% when taking FMD. Oral glucose-lowering therapy was also discontinued during FMD. During prolonged fasting, glycemic levels were self-monitored (fasting and 2-h postprandial levels, at least four measurements daily) with a capillary blood glucose monitoring system (Accu-Check Guide, Roche), and a 24-h telephone platform was available for participants to report hypo- or hyperglycemic episodes. All participants were instructed to avoid excessive physical activity during prolonged fasting. Safety was monitored by assessment of vital signs, physical examination, adverse events following the Common Terminology Criteria for Adverse Events (v4.0), and laboratory results at each visit.

The main predictor of cellular resistance was the half-maximal effective concentration (EC₅₀) for MG measured in the PMBC with the luminescent assay CellTiter-Glo. EC₅₀ value for MG represents the concentration of MG (micromolars), which reduces cell viability by 50% after 24 h of exposure to MG. EC₅₀ value for each individual and condition was calculated from nonlinear regression fit analysis of normalized data (variable slope). The primary end point was within-group and between-group comparison of the change in EC₅₀ levels (ΔEC₅₀) for MG of each individual after glucose load compared with before glucose load, before and after prolonged fasting. The secondary exploratory end points were within-group and between-group comparison of changes in circulating levels of dicarbonyls, in mRNA expression of markers of dicarbonyl detoxification (including glyoxalase-1 Glo1), mitochondrial oxidative stress (including glutathione-disulfide-reductase Gsr, Foxo family, glutathione peroxidases, peroxiredoxins and other peroxidases), mitochondrial biogenesis (including mitochondrial heat shock protein 70, HspA9), and protein expression of mitochondrial oxidative phosphorylation complexes after glucose load compared with before glucose load, before and after prolonged fasting.

This study was an exploratory pilot study. Based on previous studies in yeast cells, exposure to low MG concentration and high glucose concentration led to a 28-fold increase in cell survival (20). Therefore, the study required a sample size of 10 patients per group to detect a mean difference among groups of 28-fold change in cell survival, assuming an SD of 15% for all groups with a two-sample t test, a two-sided significance level of α = 5%, a power of at least 90%, and a dropout rate of 20%. Sample size calculation was performed in G*Power (version 3.1.9.7).

Blood samples were drawn after overnight fasting and immediately processed in the Central Laboratory of the University Hospital of Heidelberg under standardized conditions. β-Hydroxybutyrate was measured at each visit in venous blood (StatStrip Glucose/Ketone Meter System, Nova Biomedical). Homeostatic model assessment indices for β-cell function (HOMA2-B), insulin resistance (HOMA2-IR), or insulin sensitivity (HOMA2-IS) were computed with fasting glucose and C-peptide levels in the participants that were not under insulin therapy.

Directly after isolation, PBMC were seeded in a 96-well plate with RPMI medium with 10% FCS and exposed to increasing concentrations of MG (0–400 μmol/L) for 24 h at 37°C. After 24 h, a luminescent cell viability assay (CellTiter-Glo, Promega) was used, according to the manufacturer’s protocol, to assess the number of viable cells.

Detailed protocol is shown in the Supplementary Material.

Descriptive data are shown as mean ± SD for normally distributed variables and median (interquartile range) for log-normally distributed variables. The EC₅₀ value for each individual and condition was calculated from nonlinear regression fit analysis of normalized data (variable slope), and data are shown as mean ± SE of mean. Baseline characteristics (Table 1) and metabolic and anthropometric parameters before and after prolonged fasting (Table 2) were compared among all three groups using Kruskal-Wallis test. When there was a significant difference among all three groups, Mann-Whitney test was used for the comparison between two groups. For primary and secondary end points, Wilcoxon matched-pair signed rank test was used for within-group comparison. Data for primary and secondary end points are presented as mean ± SE of mean, unless otherwise stated. Because of the exploratory nature of the analyses, P value <0.05 was considered statistically significant and was not adjusted for multiplicity. All statistical data analyses and visualization were performed using GraphPad Prism 9.5.1 (GraphPad Software, San Diego, CA).

Thirty participants (18 male/12 female, age 64.6 [62.2, 67.0] years, BMI 28.6 [27.2, 30.0] kg/m²) were included. HbA_1c was higher in T2D+ (7.4% [6.9, 7.9]) and T2D− (7.0% [6.2, 7.7]) as compared with CON (5.5% [5.1, 5.8], P < 0.001), while age, sex distribution, and BMI were comparable among study groups. Participants of the CON group were slightly hypertensive, with 50% of them receiving RAAS inhibitors compared with 90% of the participants in both diabetes groups. However, blood pressure values did not differ among study groups. All participants completed the study and were nonsmokers. Baseline anthropometric and metabolic measures are shown in Table 1. Study flow diagram is shown in Supplementary Fig. 1.

The period of prolonged fasting was well tolerated, and all participants completed the diet. The self-reported grade 1 (mild) experienced symptoms were fatigue (one participant in CON), headache (two participants in CON), flu-like body pain (one participant in T2D−), vertigo (one participants in T2D−), and nausea (one participants in CON) (Supplementary Fig. 2). One participant in CON showed, after prolonged fasting, a mild hyponatremia of 128 mmol/L (grade 2), and one participant in T2D+ showed an episode of mild hypoglycemia of 58 mg/dL (grade 1) after prolonged fasting, which did not lead to study discontinuation.

All participants showed a reduction of body weight after prolonged fasting compared with baseline (CON: −3.2 ± 0.3%, P < 0.01; T2D−: −3.5 ± 0.4%, P < 0.05; T2D+: −3.5 ± 0.31%, P = 0.01) and an increase in the blood ketones (CON: 711.1 ± 217.9%, P < 0.01; T2D−: 401.1 ± 158.8%, P < 0.05; T2D+: 345.0 ± 159.8%, P < 0.05). Change in weight loss and in increase of blood ketones after prolonged fasting did not differ among study groups (respectively, P = 0.78; P = 0.12) (Table 2).

Fasting plasma glucose decreased significantly after prolonged fasting in CON (−7.5 ± 1.7%, P < 0.05), in T2D− (−17.8 ± 5.2%, P < 0.05), and in T2D+ (−12.7 ± 5.0%, P < 0.05). Plasma 2-h glucose decreased in T2D− (−19.7 ± 9.4%, P < 0.05) and in T2D+ (−12.1 ± 4.7%, P < 0.05) and did not change after prolonged fasting in CON (22.3 ± 13.6%, P = 0.426). Insulin resistance decreased in CON and T2D+ (HOMA-2IR: CON: −46.7 ± 4.9%, P < 0.01; T2D+: −25.8 ± 11.4%, P < 0.01). β-Cell function decreased in CON (HOMA-2B: −26.6 ± 5.1%, P < 0.01). Insulin sensitivity increased in CON and T2D+ (HOMA-2S: CON: 104.4 ± 18.5%, P < 0.01; T2D+: 43.3 ± 11.2%, P < 0.01). Change in fasting plasma glucose, 2-h plasma glucose, homeostatic model assessment indices for insulins resistance, β-cell function, and insulin sensitivity after prolonged fasting did not differ among the three study groups (Table 2). Albumin-to-creatinine ratio decreased in T2D+ after prolonged fasting (−36.5 ± 15.4%, P < 0.05). Lipid profile did not change after prolonged fasting in all study groups (Table 2).

We investigated the change of cellular resistance to exogeneous MG (ΔEC₅₀) of PBMC after glucose load before and after prolonged fasting. T2D+ showed a decreased cellular resistance against MG, while cellular resistance was unchanged in T2D− and improved in CON after glucose load and prolonged fasting (−19.0% vs.−1.7% vs. 12.6%; all P = 0.017) (Fig. 1B). Moreover, in T2D+, ΔEC₅₀ of MG by glucose load was reduced after prolonged fasting as compared with before prolonged fasting (−3.9% vs.−19.0%, P < 0.01). In CON, ΔEC₅₀ of MG by glucose load increased similarly before and after prolonged fasting (12.9% vs. 12.6%, P = 0.922). In T2D−, ΔEC₅₀ of MG was unchanged by glucose load before and after prolonged fasting (1.8% vs.−1.7%, P = 0.193) (Fig. 1B). Full dose-toxicity curves/analysis for MG before and after glucose and before and after prolonged fasting are shown in Supplementary Fig. 3 and Supplementary Table 3.

To investigate potential explanatory mechanisms for the decreased cellular resistance to exogenous MG in T2D+ after glucose load following prolonged fasting, we first investigated changes in plasma levels of dicarbonyls. Although absolute values of glyoxal and 3-deoxyglucosone (3DG) increased after glucose load, before and after fasting in T2D− and T2D+ as compared with CON, we found no change in glyoxal, MG, 3DG, and dimethylglyoxal among study groups after glucose load before and after prolonged fasting. Absolute values of MG increased after glucose load in T2D− and T2D+ compared with CON only before fasting. Change in 3DG after glucose load increased significantly only in CON participants after prolonged fasting compared with before fasting (−4.4% vs. 21.5%, P = 0.006) (Supplementary Table 2).

Next, we performed qPCR analysis for several markers of dicarbonyl detoxification, oxidative defenses, and mitochondrial biogenesis in the isolated PBMC. We found no significant change among the study groups after glucose load and before prolonged fasting (Fig. 2, left panel). We found an increase by 26.9% and by 26.0% in dicarbonyl detoxification (mRNA Glo1) in T2D+ before glucose load and after prolonged fasting as compared with CON and T2D−, respectively (CON: 0.94 ± 0.04 vs. T2D+: 1.19 ± 0.14, P = 0.024; T2D−: 0.95 ± 0.03 vs. T2D+, P = 0.029) (Fig. 2, middle panel, and Supplementary Fig. 4A). This increase continued even after glucose load following prolonged fasting (T2D+: 1.15 ± 0.08 vs. CON: 0.92 ± 0.04, P = 0.039) (Fig. 2, right panel, and Supplementary Fig. 4A). We observed a 27.6% increase in oxidative stress (mRNA Gsr) in T2D+ after glucose load following prolonged fasting as compared with CON (1.36 ± 0.12 vs. 1.10 ± 0.05, P = 0.006) (Fig. 2, right panel, and Supplementary Fig. 4B). Before glucose load and after prolonged fasting, we found an increase in oxidative stress (mRNA Foxo4) by 42% in T2D− as compared with CON (1.70 ± 0.20 vs. 1.20 ± 0.11, P = 0.036) (Fig. 2, middle panel, and Supplementary Fig. 4C), and mRNA Gpx2 increased in T2D− (2.63 ± 0.73) compared with CON (1.18 ± 0.17, P = 0.034) and T2D+ (1.22 ± 0.22, P = 0,04) (Fig. 2, middle panel, and Supplementary Fig. 4D). Furthermore, mRNA HspA9 was increased by 18.7% in T2D− as compared with CON before glucose and after prolonged fasting (1.12 ± 0.09 vs. 0.94 ± 0.04, P = 0.0387) and by 19.8% after glucose load following prolonged fasting (1.09 ± 0.06 vs. 0.91 ± 0.04, P = 0.0324) but did not change in T2D+ (Fig. 2, middle and right panels, and Supplementary Fig. 4E).

Citrate synthase activity did not change among study groups after glucose load, before and after prolonged fasting (Supplementary Fig. 5). Protein expression analysis of the mitochondrial complexes showed an increase in complex V after glucose load following prolonged fasting in T2D+ (11.6 ± 5.6) compared with CON (0.60 ± 0.17, P = 0.0030) and with T2D− (1.3 ± 0.3, P = 0.0029) (Supplementary Fig. 6A). No change was found among study groups for complexes II, IV, and I (Supplementary Fig. 6B–D). Complex III was not detectable in any of the samples. Representative Western blotting images are shown in Supplementary Fig. 6E. Samples that showed signs of overall protein degradation were excluded from the analysis: n = 3 (2/1 female/male) in CON, n = 1 (female) in T2D−, and n = 2 (male) in T2D (data not shown).

We analyzed serum circulating levels of the stress-responsive cytokines FGF21 and GDF15, and Hsp70 as marker for mitochondrial biogenesis. FGF21 increased in CON after prolonged fasting as compared with before prolonged fasting (7.7% vs. 115.0%, P < 0.05) (Supplementary Fig. 7A), whereas no change was observed in T2Dv and T2D+ (115.0% vs. 69.3% vs. 47.5%, P = 0.5). Moreover, we found no change in circulating GDF15 and Hsp70 after glucose load and after prolonged fasting in all study groups (Supplementary Fig. 7A). The change in FGF21 after glucose load following prolonged fasting strongly correlated with the change in β-hydroxybutyrate only in T2D+ (r = 0.87, P = 0.001) (Supplementary Fig. 7B). No correlation was found between change in circulating GDF15 and Hsp70 and change in β-hydroxybutyrate (data not shown).

This study found that 75-g glucose intake alone does not affect PBMC cellular resistance to dicarbonyl stress in either glucose-tolerant individuals or individuals with T2D. However, after prolonged fasting, T2D individuals with diabetes complications showed reduced resistance to dicarbonyl stress after glucose load, while resistance remained unchanged in those without complications and increased in glucose-tolerant individuals. Additionally, dicarbonyl detoxification and oxidative stress response increased in T2D individuals with complications but not in those without complications or in glucose-tolerant individuals. This suggests increased susceptibility to glucose-induced oxidative stress after prolonged fasting in T2D individuals with diabetes complications.

In this study, we explored the impact of glucose load on PBMC cellular resistance against ex vivo exposure to MG. PBMC, being metabolically active and responding to insulin akin to skeletal muscle, are used as a proxy tissue for studying glucose metabolism (31,32). While glucose-tolerant individuals exhibit normal glucose disposal, those with T2D face metabolic challenges during glucose load due to increased endogenous glucose production and reduced skeletal muscle glucose utilization (15). This leads to enhanced glucose availability and altered glucose transport in other peripheral tissues, including PBMC (33,34), potentially explaining the reduced PBMC cellular resistance against dicarbonyl stress after glucose load. Moreover, increased substrate flux may enhance PBMC antioxidative capacity, as has been observed in the liver of individuals with obesity, with and without steatosis, which is subsequently lost with progression to steatohepatitis (35). Cell-cell interaction between PBMC and other peripheral tissues could also influence PBMC cellular resistance. Glucose load following prolonged fasting did not affect markers of insulin resistance and sensitivity markers, suggesting involvement of insulin-independent mechanisms in the glucose-induced response after prolonged fasting.

Of note, it was only after prolonged fasting that the cellular resistance against dicarbonyl stress after glucose load differed between glucose-tolerant individuals and T2D individuals. This resistance after glucose load following prolonged fasting may represent an evolutionary adaptation in healthy individuals, facilitating a safe metabolic transition between fasting and refeeding. Caloric restriction increases lifespan in various organisms and protects against age-related diseases in rhesus monkeys and humans (36). However, in T2D individuals without complications, this protective response is lost, and those with complications experience even more harmful effects, likely due to reduced dicarbonyl detoxification (37). Previous findings have demonstrated a hormetic response against dicarbonyl stress under high-glucose conditions in yeast and mammalian cells (20). While assessing mitohormesis in more complex organisms remains challenging, the results of this study associate the loss of hormetic response with diabetes complications in T2D individuals. Notably, both glucose load and prolonged fasting were necessary metabolic challenges to discern differences in hormetic response, highlighting the complexity of hormesis as a significant disease modulator in humans (19).

The increase in Glo1, and consequently MG detoxification, following glucose load and prolonged fasting in T2D individuals with complications aligns with our previous findings of prolonged fasting in T2D individuals with nephropathy, where we observed a significant decrease in MG-derived hydroimidazolone, likely due to increased detoxification (10). Enhanced MG detoxification is also supported by increased plasma MG after glucose load but decreased levels following prolonged fasting in individuals with T2D and complications. Although the MG levels in the CON group after fasting were higher overall, statistical significance was not reached. MG can form also from acetone conversion during ketogenesis (38), offering a new detoxification pathway via acetoacetate buffering (39). MG metabolism during fasting-related pathways like gluconeogenesis and ketogenesis therefore warrants further investigation. While the mild changes in Glo1 expression are comparable to changes using Glo1 inducers reported in individuals who are overweight or obese (40), caution is needed in interpreting these results, as they cannot impact clinical decision-making.

This study primarily addressed dicarbonyl stress based on prior findings of hormetic response against MG and the link between glucose load and increased endogenous MG production. Glucose-induced changes after prolonged fasting were not observed in PBMC resistance against ex vivo exposure to H₂O₂ (Supplementary Fig. 8), suggesting specific mechanisms for MG and dicarbonyl stress.

Enhanced oxidative stress, indicated by the increase in Foxo4 and Gpx2 expression before glucose and after prolonged fasting, appears to protect T2D individuals without complications, as cellular resistance was not reduced. Moreover, increase in HspA9 expression before and after glucose following prolonged fasting in these individuals might indicate enhanced mitochondrial biogenesis. Additionally, altered Gsr in T2D individuals with complications after glucose load suggests the involvement of distinct mechanisms in glucose-induced oxidative stress following prolonged fasting. These changes cannot be explained by alterations in mitochondrial content, since citrate synthase activity was comparable among study groups.

The current study benefits from a well-controlled ex vivo experimental setting and well-matched participants. However, it has also limitations. Firstly, the study relied on previous in vitro experimental data on yeast and mammalian cells because of the absence of similar studies in mice or humans. Secondly, limited participants’ material restricted the analysis of dicarbonyl detoxification and oxidative stress response through mRNA expression, lacking validation on protein expression and function. Despite analyzing mitochondria content and oxidative phosphorylation, caution is warranted, as we only analyzed isolated PBMC and lacked functional analysis on mitochondrial respiration. Complex III was not detectable in any of the samples, as this mitochondrial complex is known to be thermally unstable. A few samples showed sign of protein degradation and were excluded from the analysis. Additionally, the exploratory study’s analysis was not corrected for multiple comparison. While PBMC are commonly used as a proxy tissue to study glucose metabolism, future research should explore glucose-induced responses in more metabolically relevant tissues, such as skeletal muscle. Acknowledging the small sample size, future studies should aim to validate findings in larger cohorts and using other fasting regimens to improve generalizability and practicability.

In conclusion, based on the participants’ material and the methods used in this study, glucose load after prolonged fasting in individuals with T2D and diabetes complications leads to 1) decreased cellular resistance against dicarbonyl stress, 2) increased dicarbonyl detoxification, and 3) increased oxidative stress response. These findings underscore the vulnerability of individuals with T2D and complications to glucose-induced cellular stress after prolonged fasting and highlight the importance of considering such effects when designing dietary interventions for diabetes management.

Acknowledgments. The authors thank all study participants and staff members of the Study Center for Diabetes and Metabolism Heidelberg involved in the conduct of the study. The authors thank Marietta Kirchner (Institute of Medical Biometry, Heidelberg, Germany) for her assistance in the statistical analysis. The authors thank Valter Longo (Longevity Institute, School of Gerontology, and Department of Biological Sciences, University of Southern California, Los Angeles, CA) for facilitating the purchase of the FMD (from L-Nutra) used in this study.

Funding. This study was funded by the Deutsche Forschungsgemeinschaft (German Research Foundation; project number 236360313–SFB 1118). E.v.R. was supported by the Friedrich-Fischer-Estate. J.Z. was supported by the Olympia Morata funding program of the Medical Faculty Heidelberg. A.S. was supported by the Deutsches Zentrum für Diabetesforschung (German Center for Diabetes Research; project number 82DZD07C2G), by the Deutsche Diabetes Gesellschaft (German Diabetes Society), and by the Olympia Morata funding program of the Medical Faculty Heidelberg.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. J.Z. and A.S. contributed to conception of work and study design. E.v.R., C.R., E.K., R.B., M.C., J.Z., and A.S. contributed to the acquisition and interpretation of data. S.K., T.F., S.H., P.P.N., and J.S. contributed to the interpretation of data. E.v.R. and A.S. drafted the manuscript. All authors contributed to the critical revision of the manuscript and approved the final version for submission. A.S. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Preliminary data of this study were presented at the German Diabetes Congress in Berlin, Germany, 17–20 May 2023, at the Annual Meeting of the European Diabetic Nephropathy Study Group in Nantes, France, 25–27 May 2023, at the Annual Meeting of the American Diabetes Association in San Diego, California, 23–26 June 2023, and at the Annual Meeting of the European Association for the Study of Diabetes in Stockholm, Sweden, 19–23 September 2022.

Handling Editors. The journal editors responsible for overseeing the review of the manuscript were Steven E. Kahn and Adrian Vella.