MR analysis
We performed MR20 analyses by using genetic instrumental variables (IVs) to proxy for candidate obesity-related factors. The use of MR enables unbiased causal effects of risk factors on respiratory outcomes to be estimated if the following assumptions are met: (1) the genetic IVs are robustly associated with the risk factor of interest; (2) the genetic IVs are not associated with any confounding factors; (3) the pleiotropic pathways through which the IVs influence the outcome are all due to the risk factors. GRS composed of risk factors associated-single-nucleotide polymorphism (SNP) was used as IVs. The dosage of the effect allele was multiplied by the regression coefficients of each SNP on associated traits divided by the mean value of all regression coefficients.
By referencing candidate SNPs from large genome-wide association studies, we collected well-known genetic variants associated with the five obesity-related factors. To avoid the pleiotropic effects of genetic variants, SNPs previously reported to be associated with asthma and confounders were excluded. After we selected SNPs for the obesity-related factors, some criteria were applied to filter the selected SNPs for quality control: (1) minor allele frequency of ≥5%, (2) genotyping call rate of >98% for all children, and (3) no linkage disequilibrium with candidate SNPs. The primers were designed using the National Genotyping Center of the Academia Sinica platform (http://lims.ngc.sinica.edu.tw/service/)21. Genotyping was performed using Sequenom iPLEX matrix-assisted laser desorption/ionization–time of flight mass spectrometry at the National Center for Genome Medicine, Taiwan. The online supplementary file provides details on how candidate SNPs were selected.