A novel approach for the analysis of longitudinal profiles reveals delayed progression to type 1 diabetes in a subgroup of multiple-islet-autoantibody-positive children

Diabetologia. 2016 Oct;59(10):2172-80. doi: 10.1007/s00125-016-4050-0. Epub 2016 Jul 11.


Aims/hypothesis: Progression to type 1 diabetes in children and adolescents is not uniform. Based on individual genetic background and environment, islet autoimmunity may develop at variable age, exhibit different autoantibody profiles and progress to clinical diabetes at variable rates. Here, we aimed to quantify the qualitative dynamics of sequential islet autoantibody profiles in order to identify longitudinal patterns that stratify progression rates to type 1 diabetes in multiple-autoantibody-positive children.

Methods: Qualitative changes in antibody status on follow-up and progression rate to diabetes were analysed in 88 children followed from birth in the prospective BABYDIAB study who developed multiple autoantibodies against insulin (IAA), GAD (GADA), insulinoma-associated antigen-2 (IA-2A) and/or zinc transporter 8 (ZnT8A). An algorithm was developed to define similarities in sequential autoantibody profiles and hierarchical clustering was performed to group children with similar profiles.

Results: We defined nine clusters that distinguished children with respect to their sequential profiles of IAA, GADA, IA-2A and ZnT8A. Progression from first autoantibody appearance to clinical diabetes between clusters ranged from 6% (95% CI [0, 16.4]) to 73% (28.4, 89.6) within 5 years. Delayed progression was observed in children who were positive for only two autoantibodies, and for a cluster of 12 children who developed three or four autoantibodies but were IAA-negative in their last samples, nine of whom lost IAA positivity during follow-up. Among all children who first seroconverted to IAA positivity and developed at least two other autoantibodies (n = 57), the 10 year risk of diabetes was 23% (0, 42.9) in those who became IAA-negative during follow-up compared with 76% (58.7, 85.6) in those who remained IAA-positive (p = 0.004).

Conclusions/interpretation: The novel clustering approach provides a tool for stratification of islet autoantibody-positive individuals that has prognostic relevance, and new opportunities in elucidating disease mechanisms. Our data suggest that losing IAA reactivity is associated with delayed progression to type 1 diabetes in multiple-islet-autoantibody-positive children.

Keywords: Clustering; Islet autoantibody; Longitudinal; Profiles; Progression; Type 1 diabetes.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms
  • Autoantibodies / immunology*
  • Cation Transport Proteins / metabolism
  • Cluster Analysis
  • Diabetes Mellitus, Type 1 / immunology*
  • Diabetes Mellitus, Type 1 / pathology*
  • Disease Progression
  • Female
  • Glutamate Decarboxylase / metabolism
  • Humans
  • Insulin / metabolism
  • Islets of Langerhans / metabolism
  • Male
  • Prospective Studies
  • Zinc Transporter 8


  • Autoantibodies
  • Cation Transport Proteins
  • Insulin
  • SLC30A8 protein, human
  • Zinc Transporter 8
  • Glutamate Decarboxylase