TABLE 4

Analysis of heterogeneity between Pima Indians and Caucasians in the association of sentinel SNPs in CDKAL1, SLC30A8, HHEX, IGF2BP2, CDKN2B, and FTO with type 2 diabetes

SNPGenePima Indian risk allele
Caucasian risk allele
Heterogeneity
PowerFrequencyOR (95% CI)FrequencyOR (95% CI)PowerP value
rs7756992CDKAL10.590.321.10 (0.97–1.25)0.251.14 (1.11–1.18)0.490.59
rs13266634SLC30A80.290.911.04 (0.84–1.29)0.751.14 (1.11–1.18)0.220.41
rs1111875HHEX0.580.391.04 (0.92–1.18)0.561.13 (1.09–1.16)0.490.20
rs4402960IGF2BP20.430.171.08 (0.93–1.26)0.291.13 (1.10–1.17)0.330.59
rs10811661CDKN2B0.380.941.13 (0.86–1.49)0.791.21 (1.16–1.26)0.270.63
rs8050136FTO0.550.151.20 (1.02–1.41)0.451.17 (1.12–1.22)0.450.80
• ORs and 95% CIs are given per copy of the risk allele, as determined in studies of Caucasians. Results for Pima Indians are derived from the present study, whereas those from Caucasians represent a combined estimate from other published studies (26,18) (see research design and methods). The heterogeneity power is defined as the power needed to detect significant heterogeneity at P < 0.05 given the Caucasian OR, the SE of its logarithm, and the SE of the logarithm of the Pima OR, under the assumption that the “true” OR in the Pima Indians is 1. The heterogeneity P value is given for the null hypothesis that the ORs in Pima Indians and Caucasians are the same.