Summary statistics for the association between DNA methylation and obesity in whole blood and sperm.

We regressed DNA methylation onto obesity status in our combined replication dataset, separately in whole blood and sperm, controlling for estimated blood cell type proportions in the blood analysis. We furthermore used a linear mixed effects model across the combined discovery and replication datasets, regressing DNA methylation onto obesity status, tissue type and batch while controlling for interindividual variation. Summary statistics for both analyses are reported—the LME results are restricted to sites available in both the discovery and replication datasets. IlmnID = Illumina CpG identifier, chr = chromosome, location = position on chromosome in hg19 reference, P_blood = p-value in blood analysis, effect_blood = effect size in whole blood, P_sperm = p-value in sperm analysis, effect_sperm = effect size in sperm, P_mix = p-value in the mixed effects model, effect_mix = effect size in the mixed effects model. All effect sized are reported using the lean men as reference group. (ZIP)