br Experimental Procedures br Author Contributions br Acknow
Introduction In addition to the HLA region numerous other loci have shown association with type 1 diabetes risk , and susceptibility CHIR-124 synthesis have been identified in many of these regions . How these genes exert their function is, however, in most cases still poorly understood. In defining these mechanisms, the various phases of the disease process resulting in the clinical presentation of type 1 diabetes have to be considered. The diagnosis of type 1 diabetes is preceded by a preclinical phase of β-cell autoimmunity characterized by islet autoantibodies appearing into the peripheral circulation . After the initiation and spreading of β-cell autoimmunity to cover multiple autoantigens, β-cell destruction results in progression to clinical disease in the great majority of such subjects, but the progression rate is highly individual varying from a few months up to more than 20 years . Natural history studies on disease-associated autoimmunity with their prospective follow-up cohorts of infants and young children at increased genetic risk for type 1 diabetes provide a valuable tool for defining whether individual genes exert their effect in the generation of autoimmunity or in the later phase of the disease process. The Finnish Diabetes Prediction and Prevention (DIPP) study is a birth cohort of subjects carrying HLA class II genotypes associated with type 1 diabetes risk who are closely monitored for the initiation of β-cell autoimmunity and development of clinical diabetes. In the current analysis, we set out to characterize the role of 39 single nucleotide polymorphisms (SNPs) in 37 confirmed non-HLA risk regions previously associated with type 1 diabetes risk  in the initiation of β-cell autoimmunity and in progression to clinical diseases after seroconversion to autoantibody positivity.
Materials and methods
Discussion Similar analyses of non-HLA single nucleotide polymorphisms effect on the appearance of beta-cell autoimmunity have been published recently from the prospective TEDDY cohort collected from Finland, Sweden, Germany and USA , and in addition, the BABYDIAB study group in Germany and the DAISY study group in Colorado, USA have explored the effects of diabetes-associated polymorphisms on both the appearance of autoantibodies and the progression rate of β-cell autoimmunity , , , . Findings from our current analysis and from the previously reported cohorts are partly consistent. However, the importance of some single SNPs seems to vary between the cohorts and in some cases, discrepancies can be seen in the affected phase between the reports. A part of these differences can be explained by the relatively large number of SNPs tested in small sample series causing by chance associations as well as lacking detection due to the low power of the studies. Moreover, the differences in the sampling intervals between the protocols affect the sensitivity of more detailed analyses of the disease pathogenesis; long sampling intervals hamper the attempt to define the exact seroconversion moment and consequently the estimation of the duration of the time period from seroconversion to clinical disease. In our analysis, polymorphisms in PTPN22, INS and IFIH1 showed an effect on the appearance of autoantibodies. For the INS SNP rs689 this finding is in line with earlier observations in the DIPP follow-up cohort , , in the observational cohort from Germany  and the international TEDDY cohort . In DAISY cohort borderline significance was observed for autoantibody production  in the initial report and for the progression rate in the subsequent expanded series published . The present analyses also confirmed the results of our recent study, which demonstrated that the INS gene polymorphism was associated with autoantibody appearance only in those subjects with IAA as the first autoantibody initiating the disease process leading to advanced autoimmunity . This is in good agreement with the mechanistic explanation of the association with INS and β-cell autoimmunity based on the phenomenon that the INS risk genotype relates to reduced fetal expression of insulin in the thymus leading to non-effective deletion of insulin-specific autoreactive T cells , , .