Aims/hypothesis. The incidence of Type 1 diabetes has increased 2.5 times during the time period from 1966 to 2000 in Finland -a general trend seen in almost all developed countries that can only be explained by environmental factors. The aim of this study was to test the possible effect of a changing environment on distribution of genotypes associated with disease susceptibility. Methods. HLA DRB1-DQA1-DQB1 genes and two diabetes-associated polymorphisms at IDDM2 and IDDM12 were analyzed. The frequencies of genotypes were compared between cases diagnosed with childhood-onset Type 1 diabetes during the period of 1939-1965 (n=367) and those diagnosed between 1990 and 2001 (n=736). Chi-square statistics or the Fisher's Exact test were used for the comparison of frequencies of analyzed haplotypes and genotypes in the two groups.

The diabetes predisposing effect of HLA genes is defined by a complex interaction of various haplotypes. We analyzed the disease association of HLA DRB1-DQA1-DQB1 genotypes in a large nuclear family cohort (n = 622) collected in Finland. Using the affected family based artificial control approach we aimed at characterizing all detectable disease-specific HLA haplotype and genotype effects. The DRB1*0401-DQB1*0302 haplotype was the most prevalent disease susceptibility haplotype in the Finnish population followed by (DR3)-DQA1*05-DQB1*02 and DRB1*0404-DQB1*0302. DRB1*0405-DQB1*0302 conferred the highest disease risk, although this haplotype was very rare. The DRB1*04-DQB1*0304 was also associated with increased disease risk, an effect detected for the first time in the Finnish population. The following haplotypes showed significant protection from the disease and are listed in decreasing order of the strength of their effect: (DR7)-DQA1*0201-DQB1*0303, (DR14)-DQB1*0503, (DR15)-DQB1*0602, DRB1*0403-DQB1*0302, (DR13)-DQB1*0603, (DR11/12/13)-DQA1*05-DQB1*0301, (DR1)-DQB1*0501. In addition to the DRB1*0401/0404-DQB1*0302/(DR3)-DQA1*05-DQB1*02 genotype and DRB1*04-DQB1*0302 homozygous genotypes, heterozygous combinations DRB1*0401-DQB1*0302/(DR13)-DQB1*0604, approximately /(DR8)-DQB1*04, approximately /(DR9)-DQA1*03-DQB1*0303, approximately /(DR1)-DQB1*0501 and approximately /(DR7)-DQA1*0201-DQB1*02 were also disease-associated. As a new finding in this population, the (DR3)-DQA1*05-DQB1*02 homozygous and (DR3)-DQA1*05-DQB1*02/(DR9)-DQA1*03-DQB1*0303 heterozygous genotypes conferred disease susceptibility. Similarly, the DRB1*0401-DQB1*0302/(DR13)-DQB1*0603 genotype was disease predisposing, implying that DQB*0603-mediated protection from diabetes is not always dominant. Comparison of our findings with published data from other populations indicates a significant disease-specific heterogeneity of the (DR8)-DQB1*04, (DR7)-DQA1*0201-DQB1*02 and (DR3)-DQA1*05-DQB1*02 haplotypes.

Aims/hypothesis: The aim of this study was to explore the contribution of genetic factors to the emergence of beta-cell-specific humoral autoimmunity. Subjects and methods: We analysed the effect of HLA class II, insulin (INS; −23 HphI variant) and cytotoxic T-lymphocyteassociated protein 4 (CTLA4 [+49 and CT60]) genes on the appearance of beta-cell-specific autoantibodies in a large population-based birth cohort recruited in Finland. Infants carrying increased risk HLA DQB1 genotypes were monitored for the appearance of autoantibodies (islet cell autoantibodies [ICA], insulin autoantibodies [IAA], glutamic acid decarboxylase autoantibodies [GADA] and islet antigen 2 antibodies ). Those who developed betacell-specific autoantibodies were studied (n=574, mean follow-up time: 4.9 years; range 0.5-9.3). Results: IAA emerged at a higher rate in children with the −23 HphI AA INS genotype than in those carrying AT or TT variants (hazard ratio 2.1, 95% CI 1.4-2.9, p<0.001). This effect of the INS locus was present in both HLA DQB1 risk groups. The appearance of IAA showed a strong association also with the HLA DRB1*0401 allele (hazard ratio 13.1, 95% CI 1.8-93.4, p<0.001). The development of IA-2A was also somewhat accelerated by the DRB1*0401 variant (p=0.03). Isolated ICA positivity was independent of the HLA and INS genotypes. None of the humoral immune markers showed association with the CTLA4 gene. Conclusions/interpretation: The INS and the DRB1 loci appear to contribute to the pathogenesis of type 1 diabetes by initiating/modifying insulin-specific autoimmunity. The emergence of IAA represents a crucial step in the development of beta cell autoimmunity in young children, in whom the appearance of GADA and IA-2A is linked to IAA.

Both INS and IKZF4 polymorphisms modified the probability of IAA positivity at time of T1D onset but the inverse association of IKZF4 risk allele with IAA suggests that the IKZF4 polymorphism is involved in a pathway of β-cell autoimmunity alternate to the route characterized by IAA and development of T1D in early childhood. The IKZF4 gene encodes Eos, which is implicated to play an important role in Treg programming where this gene might exert its influence on T1D risk.

In addition to the known human leukocyte antigen (HLA)-associated risk, polymorphisms of insulin gene region show association with type 1 diabetes. We analyzed possible interactions between the HLA class II genotypes and -2221 MspI (insulin) INS gene polymorphism in Finnish population, using a series of 1331 diabetic children and 2222 healthy newborns. C/C genotype was increased among diabetic children compared to the controls (83.2 vs 70.1%). This genotype was slightly more common in diabetic children with low or moderate HLA-associated risk than in those with high risk, but INS gene effect was clear in all major HLA-risk genotypes and, thus, can be used as an additional risk prediction marker, irrespective of HLA genotypes.