ESPE Abstracts (2019) 92 P1-179

Longitudinal Metabolic Control After Initiation of Insulin Pump in 5,040 Pediatric Type-1-Diabetes Subjects – Heterogeneous HbA1c Trajectories Over Three Years from the DPV Registry

Nicole Prinz1,2, Anke Schwandt1,2, Marianne Becker3, Torben Biester4, Melanie Hess5, Martin Holder6, Beate Karges7, Andrea Näke8, Simone von Sengbusch9, Martin Tauschmann10, Reinhard W. Holl1,2


1Institute of Epidemiology and Medical Biometry, ZIBMT, University of Ulm, Ulm, Germany. 2German Center for Diabetes Research (DZD), Munich-Neuherberg, Germany. 3DECCP, Clinique Pédiatrique Centre Hospitalier de Luxembourg, Luxembourg, Luxembourg. 4Diabetes-Center for Children and Adolescents, Children's Hospital "Auf der Bult", Hannover, Germany. 5Pediatric Endocrinology and Diabetology, University Children's Hospital Basel, Basel, Switzerland. 6Department of Pediatric Endocrinology and Diabetology, Hospital Stuttgart, Olgahospital, Stuttgart, Germany. 7Division of Endocrinology and Diabetes, RWTH Aachen University, Aachen, Germany. 8Children's Hospital Carl Gustav Carus, Technical University Dresden, Dresden, Germany. 9Department of Pediatrics, University Hospital of Schleswig-Holstein, Campus Luebeck, Luebeck, Germany. 10Department of Pediatrics and Adolescent Medicine, Medical University of Vienna, Vienna, Germany


Objectives: Continuous subcutaneous insulin infusion (CSII) has been associated with lower HbA1c. To explore whether CSII initiation leads to HbA1c improvement in each individual with type-1 diabetes and to identify co-variates which might influence change in HbA1c.

Methods: 5,040 pediatric type-1-diabetes subjects (δ20 y, 49% boys, median age at diabetes onset [Q1;Q3]: 5.9 [3.5;8.4] years) with diabetes duration ε3 years at CSII initiation (median initiation year: 2009) and continuously documented pump therapy over three years were selected from the DPV registry. Group-based modeling was applied to identify heterogeneous subgroups of HbA1c change after CSII initiation (SAS: PROC TRAJ; HbA1c values aggregated quarterly; patients with <7 aggregated values excluded). HbA1c change (D) was defined as HbA1c at the respective time-point minus baseline value.

Results: Four different trajectories of HbA1c change were identified (table1). Group 1 had a HbA1c reduction within the first quarter of the year which persisted thereafter, while group 4 had a dramatic HbA1c increase during the three-year observation period. Group 2 had a slight reduction and group 3 a slight increase of HbA1c. Age at diabetes onset, age and HbA1c at CSII initiation and insulin dose were all related to group membership (each P<0.001). At CSII initiation, group 1 had the highest HbA1c and was oldest (table1). Further, they were oldest at diabetes onset and had highest insulin requirements. Analyzing boys and girls separately, the same number of trajectory groups was revealed, although gender ratio differed. In girls, the largest group (47%) had a slight HbA1c increase, whereas in boys (43%) the largest trajectory revealed a slight decrease.

Table 1: Group-specific characteristics of the four different ΔHbA1c trajectories
Group 1
(Δ -2%)
Group 2
(Δ -0.25%)
Group 3
(Δ +0.8%)
Group 4
(Δ +2.5%)
N346 (7%)1,950 (40%)2,296 (44%)448 (9%)
Age at diabetes onset, years6.7[3.7;9.9]5.7[3.4;8.3]5.9[3.5;8.3]6.6[4.0;8.8]
Age at CSII initiation, years14.2[12.0;15.6]12.4[10.1;14.4]12.2 [10.2;14.2]13.1[11.5;14.6]
HbA1c at CSII initiation, %9.7[9.0;10.7]7.9[7.3;8.5]7.3[6.7;7.9]7.4[6.7;8.2]
Insulin dose after CSII initiation, IU/kg*d
1 year0.85[0.72;1.01]0.78[0.66;0.94]0.78[0.65;0.93]0.82[0.69;0.98]
2 years0.87[0.72;1.01]0.79[0.66;0.94]0.80[0.67;0.96]0.87[0.71;1.05]

Conclusions: There are different trajectories of HbA1c change after start of CSII in pediatric type-1-diabetes patients. Further analyses are needed to characterize the subgroups in order to predict which patients may be most successful in terms of HbA1c with insulin pumps.