Development of Autoantibodies in the TrialNet Natural History Study

  1. for the TrialNet Natural History Study Group
  1. 1University of South Florida, Pediatrics Epidemiology Center, Tampa, Florida
  2. 2University of Western Ontario, Division of Endocrinology and Metabolism, London, Ontario, Canada
  3. 3University of Florida, College of Medicine, Gainesville, Florida
  4. 4University of Miami, Division of Endocrinology, Miami, Florida
  1. Corresponding author: Kendra Vehik, kendra.vehik{at}


OBJECTIVE Understanding the relationship between age and islet autoantibody (Ab) seroconversion can establish the optimal screening interval(s) to assess risk for type 1 diabetes, identify subjects who can participate in prevention trials, and determine associated costs. This study assessed the rates of seroconversion to glutamic acid decarboxylase positive (GAD65+), insulin positive (mIAA+), and insulinoma-associated protein 2 positive (ICA512+) in a large cohort of relatives of type 1 diabetes probands undergoing Ab rescreening in the TrialNet Natural History Study.

RESEARCH DESIGN AND METHODS Of 32,845 children aged <18 years screened for Abs, 1,287 (3.9%) were GAD65+, 778 (2.4%) were mIAA+, 677 (2.1%) were ICA512+, and 31,038 were Ab-negative. Ab-negative children were offered annual rescreening up to 18 years of age. Cox regression was used to estimate the risk for GAD65, mIAA, and ICA512 seroconversion.

RESULTS There were 205 children who seroconverted to GAD65+, 155 who seroconverted to mIAA+, and 53 who seroconverted to ICA512+ over 5.8 years of follow-up. The risk of mIAA (hazard ratio 0.89 [95% CI 0.85–0.92]) and GAD65 (0.96 [0.93–0.99]) seroconversion significantly decreased with increasing age (i.e., for each 1-year increase in age, the risk of seroconversion decreased by 11% [P < 0.0001] for mIAA and 4% [P = 0.04] for GAD65) across all ages. The cumulative Ab seroconversion was 2% for those <10 years of age versus 0.7% for those ≥10 years of age.

CONCLUSIONS The risk of development of islet Abs declines with increasing age in type 1 diabetes relatives. These data support annual screening for children <10 years of age and one additional screening in adolescence.

  • Received March 22, 2011.
  • Accepted May 30, 2011.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. See for details.

This Article

  1. Diabetes Care
  1. All Versions of this Article:
    1. dc11-0560v1
    2. 34/9/1897 most recent