Birth weight (BW) was transformed to a z score ([BW value – mean BW]/s.d. BW) to facilitate comparison of the data across studies. Multiple births and, where information was available (see Supplementary Table 1), preterm births (gestational age <37 weeks) were excluded from all analyses. The association between each SNP and birth weight was assessed in each study sample using linear regression of birth weight z score against genotype using an additive genetic model, with sex and, where available, gestational age as covariables. Since gestational age was not available in all studies, we later performed a sensitivity analysis, excluding the studies that did not have this covariable (see below). The GWA analysis was performed using SNPTEST23, mach2qtl24, PLINK25 (http://pngu.mgh.harvard.edu/purcell/plink/), GenABEL26 or ProbABEL27. Details of any additional corrections for study-specific population structure are given in Supplementary Table 1. The data annotation, exchange and storage were facilitated by the SIMBioMS platform (http://www.simbioms.org)28.
