The initially determined risk-allele (A) of rs2302382 was more frequent in obese children (23.8%) and adults (22.4%) than in the controls (19.1%; pFishers exact test = 0.013). Using the log-additive genetic model in the joint analyses of all cases against the control group, we e.g. estimated that individuals homozygous for the A-allele have an 1.54-fold increase in the odds for obesity compared with individuals not carrying the A-allele (95%CI 1.09;2.19, pCA-Test = 0.014). The estimated odds ratio for the heterozygous CA carriers was 1.24 (95%CI 1.04;1.48, Table 5). When the children were analysed separately, the odds ratios were descriptively larger than those observed for the joint analyses.
