In our sample of patient probands with CD and control probands without CD, race/ethnicity was significantly related to genotype (n=480 excluding 2 Asian youth; χ2=15.31; df=6; p=0.02) and marginally related to caseness (χ2=5.42; df=3; p=0.14) raising concerns about population substructure; however, allele frequencies were generally similar between Caucasian and Hispanic subjects (examining patient probands with CD and control probands without CD n=428; χ2=0.58; df=2; p=0.75). We completed case-control analyses within our sample of Caucasian and Hispanic subjects; a significant association was seen between rs279871 and CD (p=0.02); other case-control analyses were non-significant. Power analyses for our case-control tests were conducted in PBAT; assumptions included: (1) population prevalence of disease = 0.1; (2) α = 0.05; (3) an A-allele frequency of 0.53; (4) disease locus = marker locus; and, (5) a dominant model.
