Fig. 1
Results of the logistic regression analysis of BAT-positive/negative and SNP genotypes. The forest plot shows the association between genotype groups and BAT prevalence in the FDG-PET/CT population (n = 399). Binary variables representing BAT-negative or BAT-positive statuses were coded as 0 or 1, respectively. Sex (0 = female, 1 = male), the interaction term between age and sex (female: 0 × age, male: 1 × age), a dummy variable representing the study month (0 = December and March, 1 = January and February), and the genotype were included as independent variables. To consider multiple tests of the number of SNPs and genetic models, the significance level for the genotypes was set at P = 0.0015. 1Identifiers in the Single Nucleotide Polymorphism Database at the National Center for Biotechnology Information are shown. 2Putatively selected alleles of each SNP are indicated. 3The tested genetic models were as follows: A, additive model; D, dominant model; and R, recessive model for the putatively selected allele. SNP, single nucleotide polymorphism; OR, odds ratio; 95% CI, 95% confidence interval. Black dots and horizontal bars represent ORs in each model and their 95% confidence intervals (CIs), respectively