TABLE 4

Constructed CAPN10 haplotype combinations and the risk of type 2 diabetes

Constructed haplotype combination SNP-44, -43, -19, -63Frequency (%)
OR (95% CI)P
Type 2 diabetesControl subjects
1111/11111.51.01.53 (0.38–6.15)0.55
1111/11122.01.71.22 (0.39–3.76)0.73
1111/11219.07.41.23 (0.70–2.14)0.47
1111/21115.15.70.89 (0.46–1.72)0.72
1111/12215.47.80.67 (0.37–1.24)0.21
1112/11121.51.41.14 (0.32–4.08)0.84
1112/21113.62.41.54 (0.61–3.86)0.36
1121/112110.55.71.93 (1.07–3.47)0.028*
1112/11216.73.71.85 (0.90–3.81)0.095
1112/12215.73.71.55 (0.74–3.25)0.25
1121/211111.114.20.75 (0.47–1.18)0.21
2111/21116.24.71.32 (0.67–2.60)0.42
1221/12216.48.80.71 (0.40–1.26)0.24
1121/122115.418.20.81 (0.54–1.22)0.32
2111/12218.712.20.67 (0.42–1.13)0.14
Other1.01.4NANA
Total389296
  • *

    * Frequencies of the 1121/1121 haplotype combination in type 2 diabetes patients vs. control subjects in Botnia I sample were 11.2 vs. 6.3% (P = 0.085) and in Botnia II were 9.9 vs. 4.8% (P = 0.12).

  • Frequency of the 1121/1121 or 1121/1112 haplotype combinations in all type 2 diabetes patients vs. all control subjects was 17.2 vs. 9.5% (P = 0.0036); for Botnia I, 16.0 vs. 8.4% (P = 0.023); and for Botnia II, 18.3 vs. 11.4% (P = 0.12). All SNPs were genotyped in 395 patients with type 2 diabetes and in 298 control subjects. As 0.3% of the genotypes could not be provided despite repeated genotyping, haplotypes are missing for 6 patients and 2 control subjects. Haplotypes were constructed using information on TDT trios with data on parental alleles. A strong linkage disequilibrium between the four SNPs was found, with the five most common haplotypes accounting for 99.2% of all haplotypes in our TDT material. The possibility of the rare haplotypes (1211, 1212, 2121, 2112, 2221) accounting for <0.2% each was ignored while constructing the presented haplotype combinations. NA, not analysed.