Risk of Development of Type 1 Diabetes and TrialNet Natural History Study

Health Professional

I have written in the past about the predictive value of knowing that there is a "possibility" of developing Type 1 diabetes. At present, the incidence of type 1 diabetes is approximately 1 in 300 worldwide. Some families wish to know everything despite the lack of a cure; others feel that since no cure is available, why worry unnecessarily as there is nothing that one can do to prevent Type 1 diabetes at this time.   There is now more definitive information about the timing of screenings in the TrialNet Natural History Study based on a recent Diabetes Care paper by Vehik, Bean, Mahon, Schatz, Haller, Sosenko, Skyler and Krischer for the TrialNet Natural History Study Group (DOI: 10.2337/dc11-0560): "Development of Autoantibodies in the TrialNet Natural History Study."

The TrialNet Natural History Study was initiated to help predict the possible future development of autoimmune Type 1 diabetes based on the presence of islet autoantibodies. In most cases, pancreatic islet cell autoantibodies precede the onset of T1DM and thus can identify those people at greatest risk. According to the authors, the most important predictive antibodies at present are:

**Frequency: high risk                    **

low risk


**Population:   T1DM                  **

Population: general

Glutamic acid decarboxylase     (GAD-65)



Insulinoma-associated protein (ICA512)



Insulin (mIAA)



This study assessed the rates of seroconversion to GAD-65, insulin positive (mIAA), and ICA512 in a group of relatives of type 1 diabetes patients undergoing screening in TrialNet.

What is TrialNet?

Trial Net is composed of multiple clinical centers (18) and 170 affiliates in nine countries. The Trial Net Natural History Study is a prospective cohort study that works to identify future participants for prevention trials and provides information in regard to pre-clinical type 1 diabetes. Screening for the above antibodies was conducted between February 2004 and April 2010 and annual re-screening was offered to antibody negative children under18 years.

What is the relevance of this study?

By understanding the relationship between age and the development of islet autoantibody seroconversion, it is possible to establish the optimal screening intervals to determine the risk for Type 1 Diabetes. Most importantly, it may help to reassure families that their children are extremely unlikely to develop Type 1 diabetes or that their children may be at a greater risk and thus be prepared for the development of diabetes related symptoms.

  1. 32,845 children aged younger than 18 years were screened for the above antibodies
  2. 1,287 (3.9%) were GAD-65 antibody (+)
  3. 778 (2.4%) were mIAA (+)
  4. 677 (2.1%) were ICA512 (+)
  5. 31,038 were antibody (-)

Antibody negative children were offered annual re-screening up till 18 years of age. The risk for the GAD65, mIAA, and ICA512 seroconversion was calculated by statistical analysis.

**What are the results?  **

During 5.8 years of follow-up:

  1. 205 children seroconverted to GAD65 (+)
  2. 155 children seroconverted to mIAA (+)
  3. 53 children seroconverted to ICA512 (+)

The risk of mIAA and GAD65 seroconversion significantly decreased with increasing age. For each one year increase in age, the risk of seroconversion decreased by 11% for mIAA (p<0.0001) and 4% for GAD65 (p=0.04) across all ages. The cumulative antibody seroconversion was 2% for those less than 10 years vs. 0.7% for those greater than or equal to 10 years of age.

What can we conclude from this study?

Antibody seroconversion declines with age and the greatest risk for the development of Type 1 diabetes is in young children. The study confirmed that autoimmunity associated with Type 1 diabetes is age-dependent. Thus, the risk of development of islet antibodies declines with increasing age in type 1 diabetes relatives. The data thus supports annual screening for children younger than 10 years of age and one additional screening in adolescence up to 18 years.

There are limitations to the study:

  1. HLA (the area on chromosome 6 that contributes risk: high/moderate/low for autoimmune diseases such as diabetes, thyroid, and celiac disease) contribution could not be assessed (due to lack of informed consent in the screening phase to collect). Thus, seroconversion rates could not be correlated to HLA typing.
  2. Lack of follow-up for non-responders could have affected the estimates of seroconversion. (The authors felt that the true estimate of seroconversion might therefore be underestimated).
  3. Zinc transporter 8 antibody (ZnT8A) was not assessed as it was recently added to the study protocol, thus contributing to the possibility of underestimation of seroconversion.

In conclusion, until we have a cure available for the treatment of diabetes, these studies serve to provide "risk information" for the future development of Type 1 diabetes. There is, however, some comfort in the fact that the statistics may be reassuring to some families demonstrating lack of autoantibody conversion indicating that the development of diabetes would be extremely unlikely. The opposite also holds true for families whose children do seroconvert.