TABLE 1

HLA associations in a maximum of 1,177 adult-onset autoimmune diabetic case subjects and 2,210 control subjects

SNP (allele or genotype)N (frequency)OR [95% CI]P
Case subjectsControl subjects
rs2187668 (DR3)*585 (0.25)395 (0.09)3.37 [2.92–3.90]2.31 × 10−65
rs660895 (DR4)*868 (0.37)662 (0.16)3.34 [2.94–3.81]8.65 × 10−84
DR3567 (0.25)380 (0.09)5.36 [4.53–6.34]
DR4852 (0.37)656 (0.16)4.98 [4.23–5.77]
DRX857 (0.38)3,138 (0.75)1.00 (ref.)2.02 × 10−179
3/375 (0.07)15 (0.01)37.97 [21.26–67.81]7.22 [4.07–12.79]
3/4222 (0.20)64 (0.03)26.22 [18.94–36.30]4.98 [3.65–6.80]
4/4142 (0.12)47 (0.02)22.42 [15.48–32.48]4.26 [2.98–6.10]
4/X346 (0.30)498 (0.24)5.26 [4.23–6.54]1.00 (ref.)
3/X195 (0.17)286 (0.14)5.15 [4.02–6.60]0.98 [0.78–1.23]
X/X158 (0.14)1,177 (0.56)1.00 (ref.)0.19 [0.15–0.24]
  • The SNPs rs2187668 and rs660895 were used as proxies for HLA-DRB1*03 (DR3) and HLA-DRB1*04 (DR4) alleles, respectively, and to code for the DR3/4/X genotypes (see research design and methods). The P values reported are for a model that assumes multiplicative allelic effects either for the individual SNPs or for the joint effects of both DR3 and DR4.

  • *Individual SNP associations of rs2187668 and rs660895 use all samples genotyped at the test SNP.

  • †A joint effects model including both DR3 (rs2187668) and DR4 (rs660895). Only samples genotyped at both SNPs were included (see research design and methods).

  • ‡The genotype effects model. Only samples genotyped at both SNPs were included (see research design and methods). No P value is reported for this model because a multiplicative allelic effects model was an appropriate approximation (P = 0.55). N, number of chromosomes or genotypes.