T1D, genes and environment – The TEDDY study

Lessons Learned from The Environmental Determinants of Diabetes in the Young (TEDDY) Study–Insights Into Early Autoimmune Type 1 Diabetes

Presented by:
Jeffrey Krischer, PhD
Health Informatics Institute, Morsani College of Medicine, University of South Florida, Tampa, FL; USA
Jill M. Norris, PhD
Department of Epidemiology, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, CO; USA
Kristian F. Lynch, PhD
Health Informatics Institute, Morsani College of Medicine, University of South Florida, Tampa, FL; USA
Maria Lönnrot, MD, PhD
Faculty of Medicine and Life Sciences, University of Tampere, Tampere, Finland
Ritta Veijola, MD, PhD
Department of Pediatrics, PEDEGO Research Unit, MRC Oulu, Oulu University Hospital and University of Oulu, Oulu, Finland

One of the goals of the Environmental Determinants of Diabetes in the Young (TEDDY) study was to identify environmental factors and gene-environment interactions that cause islet autoimmunity and type 1 diabetes. The study explored differences in the environmental determinants of type 1 diabetes across diverse populations and ethnic groups, and in children with and without first-degree relatives with type 1 diabetes.

Two diabetes-related endotypes exist defined by insulin autoantibodies (IAA) or GADA (antibodies against the enzyme glutamic acid decarboxylase) as the first appearing autoantibody. Early in life there is a peak in seroconversion in IAAs which appears to decline overtime. However, data are limited in older age children as fewer children have reached the higher age range (13 years) to date. Findings have also shown that the appearance of specific autoantibodies is different, with GADA appearing first and others developing later, especially after the 8-year-old mark. The changeover from a single antibody to multiple antibodies conversion often occurs within 12 months. From that point, many patients continue to develop antibodies which may lead to the onset of type 1 diabetes.

When reviewing both genetic and environmental risk, variability exists in the development of type 1 diabetes as not all patients with genetic mutations will develop type 1 diabetes. As a result, the predictability of patients who will develop type 1 diabetes associated with IAA or GADA is modest at best. Metabolomics, genomics, proteomics and microbiome data that is increasingly becoming available may improve prediction.

The fetal environment may play a role in the appearance of islet autoimmunity (IA), which typically precedes the clinical onset of type 1 diabetes. Infants and children exposed to vitamin D and omega-3 fatty acids have shown mixed results regarding a potential inverse association with the risk of IA, which could be linked to their role in immune regulation and inflammation. Studies completed in Norway showed a reduction in the risk of type 1 diabetes with supplements containing vitamin D, while other studies conducted in the U.S. found no association.1

The TEDDY study took specific interest in examining the association between maternal vitamin D and omega-3 fatty acid supplements during pregnancy, exploring the development of persistent IA in infants. The researchers found no association between omega-3 fatty acid and vitamin D intake in the presence of persistent IA-2A, IAA-first, or GADA-first development.

These findings from a large cohort (over 8,000 at-risk children from Sweden, Finland, Germany, and the U.S.) show that maternal vitamin D and omega-3 fatty acid supplement use is not associated with the risk of IA overall, nor IAA and GADA as first appearing autoantibodies in their offspring.2

Another aim of the TEDDY study was to examine whether gestational infections were associated with a risk of either IAA or GADA as the first appearing IA. The researchers hypothesized that gestational infections could interact with HLA and non-HLA genes of the child. If these were associated with age-related first appearing islet autoantibodies, it would create support for the role of the intrauterine environment and prenatal exposures of the risk of islet autoantibodies and type 1 diabetes.

Infections were categorized into upper respiratory tract infections, lower respiratory tract infections, gastroenteritis or diarrhea, and other infection-related illness (e.g. kidney, bladder, or urinary tract infection). Upper respiratory infections accounted for ~50% of all infections across countries. Findings demonstrated that gestational infections were not associated with IAA-only or GADA-only as first appearing IA.

Gestational infections were not associated with first appearing islet autoantibodies overall. Complex interactions on first appearing IA were noted between gestational respiratory infections and both HLA-DR-DQ genotypes and rs231774 in CTLA-4. Gestational respiratory infection was consistently associated with a lower risk of IAA-first in each country among HLA-DR4/8 children, while there was an increased risk of IAA-first if no respiratory infection was reported by the mother. Irrespective of HLA, if a mother reported a respiratory infection, children with CTLA-4 (AG, GG) allele polymorphism, were less likely to develop IAA-first and more likely to develop GADA-first.

These findings suggest that gestational events (e.g., respiratory infections) interact with CTLA-4, a well-known T cell regulatory protein, to influence how DR4-DQ8 or DR3-DQ2 react to a hypothetical trigger occurring during the first years of life. Future considerations will be to evaluate the influence of gestational infections on disease progression and determine whether the risk of childhood type 1 diabetes is reduced, or diagnosis is delayed by respiratory events during pregnancy.3,4

In this portion of the TEDDY study, the goal was to examine if gastrointestinal infectious episodes (GIEs) in up to age 48 months predicted the onset of islet autoantibodies, in particular with IAA vs. GADA.

According to the results of this study, GIEs did not predict the onset of IA overall. However, GIEs during a 12-month period were positively associated with GADA seroconversion. When reviewing the 12-month period in quarters, GIEs occurring at the 0-3 and 3-6-month periods were associated with GADA seroconversion. In addition, seroconversion was inversely associated with IAA seroconversion. These findings were consistent by age and countries, suggesting that gastrointestinal virus infections seem to modulate the risk of islet autoantibodies in genetically predisposed children.

GIEs predicted GADA only when reported with an ICD10 code of infective gastroenteritis, while no association was observed if only symptoms of gastroenteritis were reported. When compared with children who had respiratory infections, there was a higher correlation of GADA-first seroconversion in children who had more respiratory conditions before 6 months of age. While GIE prior to GADA-seroconversion was noted, it did not predict the progression to multiple autoantibodies or type 1 diabetes.

A cross sectional review analyzed whether a family history of diabetes (type 1, type 2, and gestational diabetes) and autoimmune diseases among first- and second-degree relatives had an effect on the development of islet autoimmunity, progression to type 1 diabetes, and determination of clinical characteristics at the time of diagnosis.

Patients with a first degree relative with type 1 diabetes significantly predicted autoimmunity, specifically in a father or sibling with type 1 diabetes (p <0.001 for both), but not in a mother with type 1 diabetes (p = 0.108). Patients with a second degree relative with type 1 diabetes significantly predicted IA. An interesting and unexpected finding was that a second-degree relative with type 2 diabetes was significantly protective for progression to type 1 diabetes in a TEDDY child.

Key messages

  • There appears to be 2 diabetes-related endotypes defined by IAA or GADA as the first appearing autoantibody.
  • Genetic background, environmental risk factors and type 1 diabetes risk differ when comparing these endotypes.
  • Maternal vitamin D and omega-3 fatty acid supplement use is not associated with the risk of islet autoimmunity overall.
  • Gestational infections were not associated with first appearing islet autoantibodies overall.
  • GIE in any 6-month period predicts IA onset with GADA by the next visit.
  • GIE predicts GADA only when reported with an ICD10 code of infective gastroenteritis.
  • Association tends to be stronger if respiratory infections are reported early in life.
  • The presence of first degree relative with T1D increases the risk of islet autoimmunity.
  • Type 2 diabetes in second-degree relatives significantly delayed progression from islet autoimmunity to type 1 diabetes, compared to children without such relatives.


Present disclosure: All of the presenters reported that they had no disclosures related to the TEDDY study.

Written by: Debbie Anderson, PhD

Reviewed by: Marco Gallo, MD


Diabetic kidney disease & glycemic control

Pathophysiology of DKD and Glycemic Goals in Patients with Low GFR