Diabetes Care 1 fi Christiane Winkler,1,2 Manja Jolink,1 Age, HLA, and Sex De ne a Annette Knopff,1 Nana-Adjoa Kwarteng,1 fi Peter Achenbach,1,2,3 Ezio Bonifacio,4,5 and Marked Risk of Organ-Speci c Anette-G. Ziegler1,2,3 Autoimmunity in First-Degree Relatives of Patients With Type 1 Diabetes https://doi.org/10.2337/dc19-0315 EPIDEMIOLOGY/HEALTH SERVICES RESEARCH OBJECTIVE Autoimmune diseases can be diagnosed early through the detection of autoanti- bodies. The aim of this study was to determine the risk of organ-specific autoimmunity in individuals with a family history of type 1 diabetes. RESEARCH DESIGN AND METHODS The study cohort included 2,441 first-degree relatives of patients with type 1 diabetes who were prospectively followed from birth to a maximum of 29.4 years (median 13.2 years). All were tested regularly for the development of autoanti- 1Institute of Diabetes Research, Helmholtz bodies associated with type 1 diabetes (islet), celiac disease (transglutaminase), or Zentrum Munchen,¨ German Research Center for Environmental Health, Munich-Neuherberg, thyroid autoimmunity (thyroid peroxidase). The outcome was defined as an Germany autoantibody-positive status on two consecutive samples. 2Forschergruppe Diabetes e.V. at Helmholtz Zentrum Munchen,¨ German Research Center RESULTS for Environmental Health, Munich-Neuherberg, In total, 394 relatives developed one (n = 353) or more (n = 41) of the three disease- Germany 3Forschergruppe Diabetes, Technical University associated autoantibodies during follow-up. The risk by age 20 years was 8.0% (95% Munich at Klinikum rechts der Isar, Munich, CI 6.8–9.2%) for islet autoantibodies, 6.3% (5.1–7.5%) for transglutaminase auto- Germany 4 antibodies, 10.7% (8.9–12.5%) for thyroid peroxidase autoantibodies, and 21.5% Center for Regenerative Therapies Dresden, – Faculty of Medicine, Technical University Dres- (19.5 23.5%) for any of these autoantibodies. Each of the three disease-associated den, Dresden, Germany autoantibodies was defined by distinct HLA, sex, genetic, and age profiles. The risk of 5Paul Langerhans Institute Dresden of the Helm- developing any of these autoantibodies was 56.5% (40.8–72.2%) in relatives with holtz Center Munich at University Hospital Carl HLA DR3/DR3 and 44.4% (36.6–52.2%) in relatives with HLA DR3/DR4-DQ8. Gustav Carus and Faculty of Medicine, Technical University Dresden, Dresden, Germany CONCLUSIONS Corresponding author: Anette-G. Ziegler, anette-g Relatives of patients with type 1 diabetes have a very high risk of organ-specific [email protected] autoimmunity. Appropriate counseling and genetic and autoantibody testing for Received 15 February 2019 and accepted 31 May 2019 multiple autoimmune diseases may be warranted for relatives of patients with This article contains Supplementary Data online type 1 diabetes. at http://care.diabetesjournals.org/lookup/suppl/ doi:10.2337/dc19-0315/-/DC1. Prospective studies that follow children from birth have established that autoim- C.W. and M.J. share first authorship. munity can start very early in life and years before clinical symptoms appear (1–3). In © 2019 by the American Diabetes Association. type 1 diabetes, autoimmunity often appears between 6 months and 3 years of age Readers may use this article as long as the work – is properly cited, the use is educational and not (1 3), and the large majority of children who develop autoantibodies against multiple for profit, and the work is not altered. More infor- pancreatic islet antigens at this age progress to clinical disease (4). New disease mation is available at http://www.diabetesjournals paradigms have emerged from such studies, leading to extensive screening (5), the .org/content/license. Diabetes Care Publish Ahead of Print, published online June 18, 2019 2 Familial Risks of Autoantibodies Diabetes Care investigation of potential causes of the to gluten reduces the risk of developing Outcome Definition diseases (6), and consortia that attempt autoantibodies. That intervention failed IA positivity was defined as the devel- to prevent disease in asymptomatic in- to show any effect on autoantibody de- opment of persistent IAA, GADA, IA-2A, dividuals (7). velopment, and all the participants con- or zinc transporter 8, with sample The BABYDIAB and subsequent BABY- tinued with follow-up examinations values .99th percentile of control chil- DIET studies were the first of these pro- according to the natural history protocol dren. Persistence was defined as auto- spective from-birth studies of type 1 (8). The studies were approved by the antibody positivity in at least two diabetes (8,9). These studies have ethics committees of Bavaria, Germany consecutive samples or in the last sample been following newborn children who (Bayerische Landesarztekammer¨ no. if the participant developed diabetes have a first-degree relative with type 1 95357 and Ludwig-Maximilians Univer- before providing a follow-up sample. diabetes on the basis that they have an sity no. 329/00, respectively) and were The age at the first IA-positive sample elevated risk of developing the disease. performed in accordance with the prin- was considered the age of seroconver- Over the course of follow-up, the chil- ciples of the Declaration of Helsinki, in- sion. The children were classified as dren were also observed to be at risk for cluding the provision of written informed multiple-IA positive if in addition to developing other autoantibodies (10,11), consent from all participants or their persistent IA positivity they tested pos- similar to findings in patients with type 1 parents. itive for more than one of the IAs on at diabetes (12–15) and in their family Children were scheduled for follow- least one occasion. In children who were members (16–18). The participants in up and venous blood collection at ages positive for GADA in the absence of other these studies are now approaching 9 months, 2 years, and every 3 years IAs, GADA was also assessed for its GAD- 30 years of age and provide a unique thereafter, and for the 150 children who binding affinity (19) and positivity on an opportunity to track the course of auto- participated in the dietary intervention, ELISA (Medizym anti-GAD ELISA Kit; immunity from birth to adulthood. The every 3 months until 3 years of age and MEDIPAN, Berlin, Germany). These chil- aim of the current study was to deter- yearly thereafter. The median follow-up dren were classified as single-GADA pos- mine the risk of organ-specific autoim- period from birth to the last sample itiveiftheyhadGADAwithaffinity $109 munity in individuals with a family history was 13.2 years (interquartile range L/mol or tested positive on the radio- of type 1 diabetes. Our findings demon- [IQR] 7.4–18.2 years) and from birth to binding assay and ELISA measurements strate distinct age, sex, and genetic pro- the last contact was 16.1 years (11.2–20.0 (21). In children who were positive for files for the autoantibodies associated years). IAA in the absence of other IAs, IAAs were with the three diseases examined, which assessed for their insulin-binding affinity occur with surprisingly high frequency Autoantibody Measurements and were classified as single-IAA positive in first-degree relatives of patients with Thetype1diabetes–associated islet if they had IAA with affinity $109 L/mol type 1 diabetes. Our findings are relevant autoantibodies (IAs) against insulin (19). TGA positivity was defined as the to the causes of autoimmunity and (IAA), GAD (GADA), IA2 (IA-2A), and development of persistent antibodies provide a strong rationale and prac- zinc transporter 8 were measured at above a threshold representing the tical guidelines for autoantibody testing the Helmholtz Center, Munich, Germany, 99th percentile of control children in members of families affected by type in venous blood samples obtained at all (10,11). TPOA positivity was defined as 1 diabetes. completed visits. The celiac disease– the development of persistent antibod- associated IgA autoantibodies against ies above a threshold of 50 units/mL, as RESEARCH DESIGN AND METHODS tissue transglutaminase 2 (TGA) and the defined with a quantile-quantile plot Study Population thyroid autoimmune disease–associated analysis (11). For TPOA and TGA, persis- Data from two German birth cohorts of autoantibodies against thyroid peroxi- tence was defined as a positive result in individuals with a first-degree family dase (TPOA) were not measured in two consecutive samples or autoanti- history of type 1 diabetes born between 46 and 30 children, respectively, because body levels of .10 units/mL (TGA) 1989 and 2000 (BABYDIAB) or between of insufficient sample volume. Measure- or .100 units/mL (TPOA) in the last 2000 and 2006 (BABYDIET) were com- ments were made with radiobinding measured sample if a second sample bined and analyzed (8,9). Both studies assays (10,11,19,20). The IA assays was unavailable for testing. prospectively examined the natural his- (laboratory 121) were evaluated in the Type 1 diabetes was diagnosed accord- tory of islet autoimmunity and type 1 Diabetes Autoantibody Standardiza- ing to the criteria of the American Di- diabetes, celiac disease–associated au- tion Program, which included other abetes Association Expert Committee toimmunity, and thyroid autoimmunity. assays that gave similar performances (26). Families were asked to report The BABYDIAB study recruited 1,650 (21–23). The genetic typing at the the occurrence of diabetes symptoms. infants born to a mother or father HLA-DRB1, HLA-DQA1, and HLA-DQB1 InchildrenwithIAs,anannualoral with type 1 diabetes, and the BABYDIET loci and at single nucleotide poly- glucose tolerance test was performed. study recruited 791 infants who had a morphisms (SNPs) rs689 (INS gene), Type 1 diabetes onset was defined as mother, father, or sibling with type 1 rs2476601 (PTPN22 gene), rs1990760 unequivocal hyperglycemia with acute diabetes. A subgroup of 150 children (IFIH1 gene), and rs3184504 (SH2B3 metabolic decompensation, the obser- participated in the BABYDIET gluten gene) was performed in 2,290 children vation on at least two occasions of a 2-h intervention study (Clinical trial reg.