Diabetologie und Stoffwechsel 2025; 20(S 01): S15
DOI: 10.1055/s-0045-1807384
Abstracts | DDG 2025
Freie Vorträge
Der Weg von der Gewichtszunahme zur Organkomplikation – Eine multifaktorielle Kaskade

Excluding HbA1c from cluster assignment of persons with already treated diabetes improves prediction of insulin secretion failure and diabetic retinopathy

M Schön
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
O P Zaharia
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
C Binsch
2   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Biometrics and Epidemiology, Düsseldorf, Germany
,
K Strassburger
2   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Biometrics and Epidemiology, Düsseldorf, Germany
,
T Mori
2   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Biometrics and Epidemiology, Düsseldorf, Germany
,
G J Bönhof
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
A Strom
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
K Prystupa
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
K B Bódis
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
I Yurchenko
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
D Mendez
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
Y Karusheva
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
S Kaya
3   Medical Faculty, Heinrich Heine University Düsseldorf, Department of Ophthalmology, Düsseldorf, Germany
,
R Guthoff
3   Medical Faculty, Heinrich Heine University Düsseldorf, Department of Ophthalmology, Düsseldorf, Germany
,
S Trenkamp
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
V Burkart
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
D Ziegler
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
O Kuss
2   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Biometrics and Epidemiology, Düsseldorf, Germany
,
M Roden
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
,
R Wagner
1   German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University, Düsseldorf, Institute for Clinical Diabetology, Düsseldorf, Germany
› Author Affiliations
 

Introduction: Clustering diabetes provides a practical framework to address disease heterogeneity. However, disease duration and treatment may bias cluster assignment due to the inherent reduction of HbA1c, a variable contributing to cluster definition. We evaluated whether excluding HbA1c from clustering impacts the representation of pathological characteristics by clusters and prediction of complications.

Methods: Participants of the German Diabetes Study (GDS) with diabetes diagnosed within the past 12 months (N=940) were clustered by the original method (w/HbA1c) and leaving out HbA1c but retaining original centroids for the remaining variables (w/oHbA1c). Severe autoimmune diabetes (SAID) was defined by presence of any diabetes autoantibody in w/oHbA1c. Comprehensive phenotyping to assess insulin sensitivity (hyperinsulinemic-euglycemic clamp), insulin secretion (intravenous glucose tolerance test, IVGTT), diabetic retinopathy (fundus photography), and peripheral and autonomic neuropathy (functional and clinical measures), was performed up to 16 years. Insulin secretion failure was defined using a threshold derived from severe autoimmune diabetes in the IVGTT.

Results: The prevalence of severe insulin-deficient diabetes (SIDD) increased from 3.3% up to 19.9% clustered w/oHbA1c due to reclassification of mild age-related diabetes (N=133) and mild obesity-related diabetes (N=47). Accounting for any autoantibody increased the prevalence of SAID up to 28.6%, whereas the prevalence of the severe insulin-resistant diabetes (SIRD) remained unaltered. Considering the increased prevalence of SIDD and its association with insulin secretion failure, retinopathy and neuropathy in prior studies, we further focused on these outcomes. The approach w/oHbA1c (AUC=0.69) showed superior ability to predict insulin secretion failure in SIDD compared to w/HbA1c (AUC=0.66, p=0.01). There was no difference in insulin sensitivity between the two approaches (p>0.05). The prevalence of retinopathy was the highest in SIDD using both approaches (both p<0.03), however, the discriminative power to predict diabetic retinopathy was higher with the approach w/o HbA1c, achieving an AUC of 0.74 compared to 0.66 w/ HbA1c (p=0.048). There was no difference between the approaches in diagnostic accuracy for peripheral (p=0.13) or autonomic (p=0.59) neuropathy.

Conclusion: Excluding HbA1c from the clustering in adults with SIDD enhanced the ability to predict insulin secretion failure and diabetic retinopathy. This methodological adaptation improves the classification of diabetes heterogeneity, offering practical benefits for precision medicine in diabetes management.



Publication History

Article published online:
28 May 2025

© 2025. Thieme. All rights reserved.

Georg Thieme Verlag KG
Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany