Vol 62, No 5 (2011)
Original paper
Published online: 2011-11-08
Increased risk of type 1 diabetes in Polish children — association with INS-IGF2 5’VNTR and lack of association with HLA haplotype
Endokrynol Pol 2011;62(5):436-442.
Abstract
Background: Human leukocyte antigens (HLA) complex and INS-IGF2 5’VNTR loci are principal determinants of the risk of type 1 diabetes
mellitus (T1DM). Carriage of class III allele is protective, while class I/I homozygosity increases the risk of T1DM.
Material and methods: HLA and 5’VNTR allele frequencies were summarised and multivariate logistic regression models with interaction evaluation were employed to determine the presence and types of allele effect interdependency. The study group was planned to number 590 children who would undergo genotyping of 5’VNTR and HLA.
Results: 590 patients (302 with T1DM and 288 controls) were recruited. Frequencies of HLA risk alleles were: 117 carriers of DR3-DQ2; 130 carriers of DR4-DQ8 including 43 DR3-DQ2/DR4-DQ8 heterozygotes. In all cases, risk alleles were vastly overrepresented in the T1DM group compared to the controls (p < 0.0001 in all cases). The most frequent protective haplotype was DQB1 × 0602 observed in 24 controls and two T1DM cases (p < 0.001). Class I 5’VNTR homozygotes constituted 58% of the control group (n = 174) and 78% (n = 224) of T1DM patients [OR = 2.63 (95% CI: 1.79–3.57)]. Interactions between 5’VNTR and DR3-DQ2 or DR4-DQ8 variants did not reach statistical significance for risk of developing T1DM (p = 0.54 and 0.24) or age at its diagnosis (p = 0.14 and 0.67 respectively).
Conclusions: Interactions between HLA and 5’VNTR genotype are not of multiplicative character. Class I homozygosity at 5’VNTR is a significant risk factor of T1DM and acts independently from HLA haplotype in determining the actual risk of diabetes in children. (Pol J Endocrinol 2011; 62 (5): 436–442)
Material and methods: HLA and 5’VNTR allele frequencies were summarised and multivariate logistic regression models with interaction evaluation were employed to determine the presence and types of allele effect interdependency. The study group was planned to number 590 children who would undergo genotyping of 5’VNTR and HLA.
Results: 590 patients (302 with T1DM and 288 controls) were recruited. Frequencies of HLA risk alleles were: 117 carriers of DR3-DQ2; 130 carriers of DR4-DQ8 including 43 DR3-DQ2/DR4-DQ8 heterozygotes. In all cases, risk alleles were vastly overrepresented in the T1DM group compared to the controls (p < 0.0001 in all cases). The most frequent protective haplotype was DQB1 × 0602 observed in 24 controls and two T1DM cases (p < 0.001). Class I 5’VNTR homozygotes constituted 58% of the control group (n = 174) and 78% (n = 224) of T1DM patients [OR = 2.63 (95% CI: 1.79–3.57)]. Interactions between 5’VNTR and DR3-DQ2 or DR4-DQ8 variants did not reach statistical significance for risk of developing T1DM (p = 0.54 and 0.24) or age at its diagnosis (p = 0.14 and 0.67 respectively).
Conclusions: Interactions between HLA and 5’VNTR genotype are not of multiplicative character. Class I homozygosity at 5’VNTR is a significant risk factor of T1DM and acts independently from HLA haplotype in determining the actual risk of diabetes in children. (Pol J Endocrinol 2011; 62 (5): 436–442)
Keywords: autoimmunityhuman leukocyte antigensproinsulin gene
