It is not completely clear which organs are responsible for glucagon elimination in humans, and disturbances in the elimination of glucagon could contribute to the hyperglucagonemia observed in chronic liver disease and chronic kidney disease (CKD). Here, we evaluated kinetics and metabolic effects of exogenous glucagon in individuals with stage 4 CKD (n = 16), individuals with Child-Pugh A–C cirrhosis (n = 16), and matched control individuals (n = 16), before, during, and after a 60-min glucagon infusion (4 ng/kg/min). Individuals with CKD exhibited a significantly lower mean metabolic clearance rate of glucagon (14.0 [95% CI 12.2;15.7] mL/kg/min) compared with both individuals with cirrhosis (19.7 [18.1;21.3] mL/kg/min, P < 0.001) and control individuals (20.4 [18.1;22.7] mL/kg/min, P < 0.001). Glucagon half-life was significantly prolonged in the CKD group (7.5 [6.9;8.2] min) compared with individuals with cirrhosis (5.7 [5.2;6.3] min, P = 0.002) and control individuals (5.7 [5.2;6.3] min, P < 0.001). No difference in the effects of exogenous glucagon on plasma glucose, amino acids, or triglycerides was observed between groups. In conclusion, CKD, but not liver cirrhosis, leads to a significant reduction in glucagon clearance, supporting the kidneys as a primary site for human glucagon elimination.
This study was undertaken to gain insights into the elimination of glucagon in humans, as this is an important, yet not fully uncovered, part of glucagon physiology.
We investigated whether individuals with chronic kidney disease or liver cirrhosis were characterized by an altered elimination of glucagon.
We found that individuals with stage 4 chronic kidney disease exhibit a significantly reduced elimination of glucagon, while patients with liver cirrhosis have a preserved glucagon elimination.
Our findings support the kidneys as the prevailing site of glucagon elimination in humans and shed light on glucagon-related pathophysiology in liver and kidney diseases.
Introduction
Elevated levels of glucagon, hyperglucagonemia, can increase the risk of detrimental hyperglycemia and are reported to play a role in several diseases, including obesity, liver diseases, chronic kidney disease (CKD), and type 2 diabetes (1–6).
The causes of hyperglucagonemia are not fully elucidated. Most research has focused on hyperglucagonemia as a consequence of disturbances in glucagon secretion, whereas less attention has been paid to potentially altered elimination of glucagon (7,8). Glucagon is thought mainly to be eliminated in the kidneys and the liver; however, the relative role of each of these organs has not been fully determined, and the existence of interspecies differences in glucagon handling and conflicting evidence in humans have left human glucagon metabolism unresolved (9–11). In humans, hyperglucagonemia is frequently observed in both acute and chronic kidney failure, and fasting glucagon levels are inversely correlated with kidney function, suggesting the kidneys as a likely site of glucagon clearance (3,4). In early- and late-stage liver diseases, hyperglucagonemia has also been demonstrated, but controversies regarding the hepatic handling of glucagon exist (12,13). Hyperglucagonemia in various disease states has also been at least partly attributed to cross-reactivity of glucagon assays with other products of proglucagon processing (e.g., proglucagon [1–61] and glicentin), especially in individuals with CKD where such moieties may accumulate (14). Recent development of sandwich ELISA-based glucagon analyses with limited cross-reactivity has decreased the risk of unspecific interference (14).
Here, we conducted a glucagon infusion study in individuals with either CKD or liver cirrhosis and in matched healthy control individuals, measuring glucagon by ELISA, to describe the importance of the liver and the kidneys for the clearance of glucagon.
Research Design and Methods
The study was approved by the Danish Scientific Ethics Committees for the Capital Region of Denmark (ID: H–16043802), registered with ClinicalTrials.gov (ID: NCT05056584), and performed in accordance with the Declaration of Helsinki.
Participants
The study included three groups of 16 participants: individuals with CKD, individuals with liver cirrhosis, and healthy control individuals. Following informed consent, eligibility was evaluated including assessment of kidney and liver function by routine biochemistry (Table 1) and transient elastography (FibroScan). Body composition was estimated by bioelectrical impedance analysis (mBCA 515; SECA). Inclusion criteria for the CKD group were stage 4 CKD with an eGFR between 15 and 30 mL/min/1.73 m2, normal liver function, and normal transient elastography. Inclusion criteria for the cirrhosis group were verified liver cirrhosis (clinically and/or histologically) without hepatic encephalopathy, and normal kidney function. Inclusion criteria for the control group were normal kidney function and liver function and normal transient elastography. Exclusion criteria for all participants included any form of diabetes, severe anemia, and treatment with drugs known to affect glucagon physiology. The participants were recruited to obtain an equal distribution of sex, age, body weight, and BMI across the three groups.
. | Cirrhosis group . | CKD group . | Control group . |
---|---|---|---|
n (female) | 16 (6) | 16 (6) | 16 (6) |
Age (years) | 64.8 [60.5;69.1] | 59.8 [53.5;65.8] | 61.2 [56.3;66.1] |
Weight (kg) | 80.2 [69.8;90.9] | 87.6 [82.5;92.8] | 82.9 [76.7;89.1] |
BMI (kg/m2) | 25.6 [23.3;27.9] | 28.5 [26.9;30.1] | 27.1 [25.9;28.4] |
Lean body mass (kg) | 50.5 [43.8;57.3] | 57.1 [51.2;63.1] | 55.9 [51.0;60.8] |
Skeletal muscle mass (kg) | 22.3 [18.6;26] | 25.1 [20.2;30] | 24.5 [20.3;28.8] |
Fat mass (kg) | 28.5 [23.2;33.8] | 31.3 [27.5;35.3] | 27.1 [24.2;30.0] |
ALT (units/L) | 29.5 [22.2;39.2]** | 18.9 [16.7;21.3]** | 22.5 [18.9;26.9] |
AST (units/L) | 48.1 [31.8;64.3]**§§ | 21.5 [18.7;24.3]** | 24.1 [21.1;27.0]§§ |
eGFR (creatinine, mL/kg/min) | 96.0 [88.3;103]*** | 26.3 [24.3;28.3]***### | 86.8 [80.6;92.9]### |
eGFR (cystatin C–creatinine, mL/kg/min) | 83.0 [75.1;91.7]*** | 27.9 [26.2;29.7]***### | 93.3 [86.2;101]### |
Fibrosis 4 score1 | 3.06 [2.17;4.32]***§§§ | 1.24 [0.94;1.64]*** | 1.21 [1.00;1.46]§§§ |
HbA1c (%) | 5.3 [5.1;5.5] | 5.4 [5.2;5.6] | 5.2 [5.1;5.4] |
HbA1c (mmol/mol) | 34.3 [32.1;36.4] | 35.9 [33.7;38.2] | 33.8 [32.0;35.6] |
Plasma glucose (mmol/L) | 5.4 [5.2;5.7] | 5.2 [5.0;5.4] | 5.2 [5.0;5.4] |
Plasma insulin (pmol/L)1 | 79.4 [55.6;113]§ | 52.5 [43.09;63.92] | 38.9 [30.6;49.5]§ |
Plasma C-peptide (pmol/L) | 853 [633;1.07 × 103] | 996 [874;1.12 × 103]### | 513 [417;609]### |
HOMA-IR | 3.6 [2.2;4.93]*§ | 1.9 [1.5;2.2]* | 1.5 [1.1;1.9]§ |
UACR2 | 4 [0;12.3]*** | 309 [31.2;928]*** | 2.5 [0;3.05]### |
CAP (dB/m) | 272 [235;309] | 241 [215;268] | 233 [204;263] |
TE (kPa) | 20 [14;30]***§§§ | 3.6 [3.3;4.6]*** | 4.1 [3.7;4.7]§§§ |
. | Cirrhosis group . | CKD group . | Control group . |
---|---|---|---|
n (female) | 16 (6) | 16 (6) | 16 (6) |
Age (years) | 64.8 [60.5;69.1] | 59.8 [53.5;65.8] | 61.2 [56.3;66.1] |
Weight (kg) | 80.2 [69.8;90.9] | 87.6 [82.5;92.8] | 82.9 [76.7;89.1] |
BMI (kg/m2) | 25.6 [23.3;27.9] | 28.5 [26.9;30.1] | 27.1 [25.9;28.4] |
Lean body mass (kg) | 50.5 [43.8;57.3] | 57.1 [51.2;63.1] | 55.9 [51.0;60.8] |
Skeletal muscle mass (kg) | 22.3 [18.6;26] | 25.1 [20.2;30] | 24.5 [20.3;28.8] |
Fat mass (kg) | 28.5 [23.2;33.8] | 31.3 [27.5;35.3] | 27.1 [24.2;30.0] |
ALT (units/L) | 29.5 [22.2;39.2]** | 18.9 [16.7;21.3]** | 22.5 [18.9;26.9] |
AST (units/L) | 48.1 [31.8;64.3]**§§ | 21.5 [18.7;24.3]** | 24.1 [21.1;27.0]§§ |
eGFR (creatinine, mL/kg/min) | 96.0 [88.3;103]*** | 26.3 [24.3;28.3]***### | 86.8 [80.6;92.9]### |
eGFR (cystatin C–creatinine, mL/kg/min) | 83.0 [75.1;91.7]*** | 27.9 [26.2;29.7]***### | 93.3 [86.2;101]### |
Fibrosis 4 score1 | 3.06 [2.17;4.32]***§§§ | 1.24 [0.94;1.64]*** | 1.21 [1.00;1.46]§§§ |
HbA1c (%) | 5.3 [5.1;5.5] | 5.4 [5.2;5.6] | 5.2 [5.1;5.4] |
HbA1c (mmol/mol) | 34.3 [32.1;36.4] | 35.9 [33.7;38.2] | 33.8 [32.0;35.6] |
Plasma glucose (mmol/L) | 5.4 [5.2;5.7] | 5.2 [5.0;5.4] | 5.2 [5.0;5.4] |
Plasma insulin (pmol/L)1 | 79.4 [55.6;113]§ | 52.5 [43.09;63.92] | 38.9 [30.6;49.5]§ |
Plasma C-peptide (pmol/L) | 853 [633;1.07 × 103] | 996 [874;1.12 × 103]### | 513 [417;609]### |
HOMA-IR | 3.6 [2.2;4.93]*§ | 1.9 [1.5;2.2]* | 1.5 [1.1;1.9]§ |
UACR2 | 4 [0;12.3]*** | 309 [31.2;928]*** | 2.5 [0;3.05]### |
CAP (dB/m) | 272 [235;309] | 241 [215;268] | 233 [204;263] |
TE (kPa) | 20 [14;30]***§§§ | 3.6 [3.3;4.6]*** | 4.1 [3.7;4.7]§§§ |
Data are mean with 95% CI in brackets unless otherwise specified. Baseline characteristics of participants in the CKD group (n = 16), the cirrhosis group (n = 16), and the healthy control group (n = 16) were obtained in the fasting state. Causes of CKD were hypertension (7 of 16), systemic lupus erythematosus (1 of 16), previous malignancy (2 of 16), amyloidosis (2 of 16), sarcoidosis (1 of 16), hyperparathyroidism (1 of 16), chronic granulomatous disease (1 of 16), and hereditary kidney disease (1 of 16). Causes of cirrhosis were alcohol-induced cirrhosis (13 of 16), metabolic dysfunction–associated steatotic liver disease with increased alcohol intake (2 of 16), and hemochromatosis-induced cirrhosis (1 of 16). The individuals in the cirrhosis group comprised 14 individuals with Child-Pugh grade A (10 of whom had experienced previous episodes of decompensation), 1 subject with Child-Pugh grade B, and 1 subject with Child-Pugh grade C. Significant differences are highlighted with the following symbols: *between the CKD group and the cirrhosis group, #between the CKD group and the control group, and §between the cirrhosis group and the control group. A single symbol denotes P < 0.05; a double symbol denotes P < 0.01; a triple symbol denotes P < 0.001. CAP, controlled attenuation parameter; HOMA-IR, HOMA for insulin resistance; TE, transient elastography; UACR, urine-albumin-creatinine ratio. 1Log-normalized values. 2UACR and alcohol intake are displayed as median with interquartile range in brackets.
Study Design
Participants were studied after an overnight fast (10 h) including abstinence from medicine, tobacco, and liquids. Participants were instructed not to consume alcohol or exercise vigorously in the 3 days leading up to the experiment. Intravenous cannulas were inserted into a cubital vein in both arms of the resting participants. At time −120 min, a primed infusion of isotopically stable [6,6-D2]-glucose tracer was initiated and continued for the entirety of the experiment (priming bolus dose of 26.4 µmol/kg, infusion rate of 0.6 µmol/kg/min). At time point 0 min, an infusion of recombinant human glucagon was initiated (4 ng/kg/min) (GlucaGen; Novo Nordisk A/S) and continued for 60 min followed by a 60-min washout period. From the contralateral arm, venous blood samples were frequently drawn throughout the 240 min. A heating pad (∼40°C) was applied to the forearm from which blood was sampled, to promote arterialization of the venous blood. Urine was collected at baseline and end of the study. Vital signs were assessed throughout the study (Supplementary Fig. 4).
Analyses of Blood Samples
Plasma glucagon concentrations were quantified using both a sandwich ELISA (Mercodia AB, Uppsala, Sweden) and a radioimmunoassay (RIA) directed against the C-terminal end of the glucagon molecule (in-house, antibody code 4305) (14). ELISA-based glucagon results are reported unless otherwise stated. An overview of secondary end point analyses is provided in the Supplementary Material.
Statistical Methodology
The primary end point was the metabolic clearance rate (MCR) of glucagon. Secondary end points included glucagon apparent volume of distribution (Vd), glucagon plasma half-life (t1/2), glucose rate of appearance (Ra) and disappearance (Rd), and plasma/serum levels of amino acids, insulin, and triglycerides, both fasting and in response to the glucagon infusion. The calculation of MCR, Vd, and t1/2 has been described in detail previously (15). Change from baseline in response to the glucagon infusion was evaluated as incremental area under the curve. Differences between the three groups were assessed using ANOVA. Individual comparisons between any two groups were made using t tests. Normally distributed data are presented as arithmetic mean with 95% CI in brackets unless otherwise stated. Skewed data were logarithmically transformed prior to analysis and presented as geometric mean with 95% CI in brackets unless otherwise stated. A two-sided P value <0.05 was chosen to indicate significance. The false discovery rate method by Benjamini and Hochberg was used to correct for multiple testing of the secondary end points. Uncorrected P values and full overview of calculations are presented in the Supplementary Material.
Data and Resource Availability
The data sets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request. No applicable resources were generated or analyzed during the current study.
Results
Participants
Between February 2021 and March 2023, 48 participants were enrolled, and all successfully completed the study. An equal distribution of age, sex, weight, BMI, and body composition was obtained between the three groups (Table 1). Individuals in the CKD group all had stage 4 CKD with a markedly reduced mean eGFR while exhibiting normal markers of liver function, similar to those in the control group. The cirrhosis group displayed significantly elevated liver enzymes and liver stiffness/fibrosis markers and normal markers of kidney function, not differing from the control group. No differences in FibroScan-assessed hepatic fat content, fasting plasma glucose, or HbA1c were observed between the three groups.
Fasting Glucagon
No differences in fasting glucagon concentrations were observed between groups when assessed by ELISA (cirrhosis group 6.68 [5.19;8.6] pmol/L, CKD group 4.82 [3.48;6.67] pmol/L, and control group 5.96 [3.82;9.32] pmol/L) (Table 2 and Supplementary Fig. 1). However, interestingly, the RIA analysis showed elevated glucagon concentrations in the two disease groups, reaching statistical significance between the CKD group and the control group (15.5 [11.9;19.1] vs. 6.6 [3.3;9.9] pmol/L, P = 0.004).
. | Cirrhosis group . | CKD group . | Control group . |
---|---|---|---|
Baseline ELISA (pmol/L)1 | 6.68 [5.19;8.6] | 4.82 [3.48;6.67] | 5.96 [3.82;9.32] |
Baseline RIA (pmol/L) | 12.4 [7.9;20.1] | 15.5 [11.9;19.1]## | 6.6 [3.3;9.9]## |
Steady state (pmol/L) | 56.7 [51.0;62.5]*** | 80.6 [71.6;89.7]***### | 56.9 [51.1;62.7]### |
MCR (mL/kg/min) | 19.7 [18.1;21.3]*** | 14.03 [12.2;15.7]***### | 20.4 [18.1;22.7]### |
t1/2 (min) | 5.7 [5.2;6.3]** | 7.5 [6.9;8.2]**### | 5.7 [5.2;6.3]### |
Vd (mL/kg) | 210 [168;252] | 178 [152;205] | 179 [161;197] |
. | Cirrhosis group . | CKD group . | Control group . |
---|---|---|---|
Baseline ELISA (pmol/L)1 | 6.68 [5.19;8.6] | 4.82 [3.48;6.67] | 5.96 [3.82;9.32] |
Baseline RIA (pmol/L) | 12.4 [7.9;20.1] | 15.5 [11.9;19.1]## | 6.6 [3.3;9.9]## |
Steady state (pmol/L) | 56.7 [51.0;62.5]*** | 80.6 [71.6;89.7]***### | 56.9 [51.1;62.7]### |
MCR (mL/kg/min) | 19.7 [18.1;21.3]*** | 14.03 [12.2;15.7]***### | 20.4 [18.1;22.7]### |
t1/2 (min) | 5.7 [5.2;6.3]** | 7.5 [6.9;8.2]**### | 5.7 [5.2;6.3]### |
Vd (mL/kg) | 210 [168;252] | 178 [152;205] | 179 [161;197] |
Data are mean with 95% CI in brackets. Glucagon levels in the fasting state and during steady-state conditions in response to an exogenous glucagon infusion (4 ng/kg/min) in participants with CKD and with cirrhosis and healthy matched control participants (n = 16 in each group). Steady state, MCR, t1/2, and Vd are all calculated from ELISA glucagon measurements. Significant differences are highlighted with the following symbols: *between the CKD group and the cirrhosis group, #between the CKD group and the control group, and §between the cirrhosis group and the control group. A single symbol denotes P < 0.05; a double symbol denotes P < 0.01; a triple symbol denotes P < 0.001.
1Log-normalized values.
Glucagon Levels in Response to Infusion of Glucagon
Glucagon levels quickly rose in response to the glucagon infusion (Fig. 1 and Table 2), with the CKD group reaching significantly higher glucagon steady state concentrations (80.6 [71.6;89.7] pmol/L) compared with the concentrations measured in the cirrhosis group (56.7 [51.0;62.5] pmol/L, P = 0.001) and control group (56.9 [51.1;62.7] pmol/L, P = 0.001).
Glucagon Kinetics
The MCR of glucagon was markedly lower in the CKD group (14.0 [12.2;15.7] mL/kg/min) compared with both the cirrhosis group (19.7 [18.1;21.3] mL/kg/min, P < 0.001) and the control group (20.4 [18.1;22.7] mL/kg/min, P < 0.001) (Table 2). There were no differences in glucagon Vd between the three groups. Consequently, the t1/2 of glucagon was prolonged in the CKD group (7.5 [6.9;8.2] min) compared with the t1/2 in both the cirrhosis group (5.7 [5.2;6.3] min, P = 0.002) and the control group (5.7 [5.2;6.3] min, P < 0.001).
Insulin
No differences in fasting insulin levels were observed between the CKD group and the two other groups, whereas fasting insulin levels were significantly higher in the cirrhosis group compared with the control group (Fig. 2A and Table 1). Insulin change from baseline was similar in all groups in response to the glucagon infusion (Fig. 2A and Supplementary Table 1).
Glucose
All participants had fasting plasma glucose values within the normal range, with no differences between the groups (Table 1). Plasma glucose levels rose in response to the glucagon infusion, with no differences in change from baseline between groups (mean increase 1.2 mmol/L, range 0.4–2.8 mmol/L) (Fig. 2C). No differences in change from baseline in glucose Ra and Rd were observed between the groups (Fig. 2Dand E and Supplementary Table 1).
Amino Acids
Total fasting amino acid levels did not differ among groups, but baseline differences in several individual amino acids were observed (Supplementary Table 2). Total amino acids decreased similarly and approximately 10% in all groups in response to glucagon infusion (Fig. 2F), with no significant differences between the three groups at the level of individual amino acids (Supplementary Fig. 3).
Triglycerides
No differences in triglyceride levels were observed between the three groups (Fig. 2G and Supplementary Table 1).
Discussion
Here, we evaluated kinetics and effects of exogenous glucagon in individuals with CKD, individuals with liver cirrhosis, and healthy control individuals to provide insights into human glucagon physiology and the hyperglucagonemia reportedly associated with CKD and liver cirrhosis. We show that individuals with stage 4 CKD have significantly reduced glucagon elimination, clearly demonstrated by an approximately 30% lower MCR and a markedly longer t1/2, compared with both individuals with liver cirrhosis and healthy control individuals. In contrast, the individuals with liver cirrhosis exhibited a preserved ability to eliminate glucagon. Together, the kidneys appear to be the most important organ for glucagon clearance in humans.
Our study adds novel insights both by the direct comparison between CKD and liver cirrhosis and by the use of the newer sandwich ELISA. In previous RIA-based studies (15,16), and in a recent ELISA-based study (17), glucagon MCR of healthy participants was similar to that of our control group. Of note, Laurenti et al. report slightly shorter t1/2 and smaller Vd, which may be due to technical differences between the studies (17). ELISA-based studies evaluating glucagon kinetics in CKD and liver cirrhosis are lacking, but our results agree with previous RIA-based studies, showing glucagon clearance to be decreased in end-stage renal disease (4,18), and preserved in liver cirrhosis (12,19), while contradicting others who report signs of hepatic glucagon uptake and clearance (20–22). In our study, the fasting hyperglucagonemia previously observed in CKD and liver cirrhosis could only be reproduced using RIA, and not ELISA, likely because of increased cross-reactivity between the RIA and other proglucagon derivatives (e.g., proglucagon 1–61 [23]) in the groups with advanced organ disease. Notably, the absence of fasting hyperglucagonemia in the CKD group despite a significantly reduced MCR of glucagon suggests a compensatory reduction in pancreatic glucagon secretion and questions the presence of actual hyperglucagonemia in isolated CKD.
We further investigated whether potentially altered glucagon kinetics and/or compromised organ function could result in altered effects of glucagon, as previously reported (24). We uncovered robust differences in fasting levels of several amino acids, including tyrosine, tryptophane, leucine, and serine, but observed no significant differences in the glucagon-induced changes in amino acid levels between the groups. Together with the comparable effects on plasma glucose, glucose kinetics, and triglycerides, our results suggest that the participants with liver cirrhosis and CKD have a preserved metabolic response to the glucagon infusion, contrary to our expectations. Notably, insulin levels should be considered when evaluating these effects, and the observed elevated insulin levels in the cirrhosis group—likely due to both increased insulin resistance and decreased insulin clearance—could influence our results (25).
Our study has limitations. The participants in the cirrhosis group had predominantly Child-Pugh A cirrhosis, and we cannot exclude the possibility that our results would have been different if more individuals with clinical impairment had been included. On the other hand, 10 individuals with Child-Pugh A had previous decompensation episodes, suggesting that some degree of impairment of liver function could exist. Similarly, participants in the CKD group did not have end-stage renal disease. Our relatively small sample size limits the power of our secondary statistical analyses. We did not clamp endogenous insulin and glucagon concentrations and thus cannot disentangle the endogenous effects of these on our results. Finally, we cannot deduce exact single-organ clearance from our study, as we only measured whole-body clearance.
An important strength of our study is the comparison of three groups who, outside their primary disease state, were well matched on BMI, sex, age, hepatic fat content, and body composition, limiting bias by confounding factors. To our knowledge, this is the first direct comparison of these three groups, using a highly sensitive and specific sandwich ELISA for glucagon quantification.
Taken together, we conclude that glucagon clearance is decreased in CKD, while glucagon clearance is preserved in liver cirrhosis, highlighting the kidneys as a prevailing organ for glucagon elimination in humans.
Clinical trial reg. no. NCT05056584, ClinicalTrials.gov
This article contains supplementary material online at https://doi.org/10.2337/figshare.26317105.
This article is featured in a podcast available at diabetesjournals.org/diabetes/pages/diabetesbio.
Article Information
Acknowledgments. The authors thank the participants for their time spent on this project. The authors are also grateful for the assistance of the recruiting clinicians and the laboratory expertise of Brian Jensen, Center for Clinical Metabolic Research, Copenhagen University Hospital Herlev and Gentofte, Hellerup, Denmark.
Funding. The present work was funded by the Novo Nordisk Foundation and the Helen and Ejnar Bjørnow Foundation.
The funding parties have not been involved in the design, data collection, data analysis, data interpretation, writing, or publication of the study.
Duality of Interest. A.B.L. is on the speaker’s bureau of and/or has received research support from Sanofi, Boehringer Ingelheim, Novo Nordisk, and AstraZeneca. B.H. is a cofounder of Bainan Biotech. J.I.B. has received lecture fees from Novo Nordisk. J.J.H. was on the advisory panel of, consultant for, in the speaker’s bureau of, and/or has received research support from AstraZeneca, GlaxoSmithKline, Hamni, Intarcia, Merck Sharp & Dohme, Novartis, Novo Nordisk, Sanofi, and Zealand Pharma. F.K.K. has been on the advisory panel of, consultant for, in the speaker’s bureau of, and/or received research support from 89bio, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Gubra, Novo Nordisk, Merck Sharp & Dohme, Sanofi, Structure Therapeutics, Zealand Pharma, and Zucara, and is cofounder of and minority shareholder in Antag Therapeutics and currently employed by Novo Nordisk. M.F.G.G. has received lecture fees and travel compensation from Novo Nordisk Denmark A/S, is a minority shareholder of Zealand Pharma A/S, and is cofounder of Medvaegt ApS. M.H. has received speaker and consultancy fees from AstraZeneca, Bayer, Boehringer Ingelheim, Novo Nordisk, Vifor, and GSK within the last 3 years. T.V. is on the advisory panel for, consultant for, in the speaker’s bureau of, and/or has received research support from AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Gilead, GSK, Merck Sharp & Dohme, Novartis, Novo Nordisk, Sanofi, and Sun Pharmaceuticals. No other potential conflicts of interest relevant to this article were reported.
Author Contributions. M.F.G.G., M.B.C., A.B.L., and F.K.K. conceived and designed the study. L.L.G., M.H., and T.V. contributed to the design of the study. M.T. and D.H.K. recruited participants. M.F.G.G. and A.H.L. collected the clinical data. G.v.H., S.A.J.T., T.J.G., B.H., and J.J.H. provided analyses, data acquisition, and data interpretation. M.F.G.G. performed the statistical analyses. M.P.S. and J.I.B. researched data. M.F.G.G. drafted the manuscript. All authors reviewed and approved the manuscript. M.F.G.G. and F.K.K. are the guarantors of this work and, as such, had full access to all data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.