Fig. 3
Placental weight mediates the effects of "vasculature-related" blood pressure SNPs on birth weight. a Venn diagram of 41 genome-wide significant single nucleotide polymorphisms (SNPs) for BP phenotypes. There is a considerable overlap of association across four BP phenotypes, with a consistent direction of the BP increasing effect, except for seven loci that present opposite (BP decreasing) or unclear effect directions for DBP (indicated by †). SNPs are classified into “vasculature-related” (red), “unlikely related” (blue), and “unknown” (green) groups, respectively, as described in the results and Additional File 1: Table S2. b Mediation analysis of the models of either the PGS with vasculature-related BP SNPs (left) and unlikely related BP SNPs (right) as an exposure variable. Maternal PGS of “vasculature-related” SNPs shows a birth weight reducing effect (total effect = − 0.06 [− 0.11, − 0.01], p = 0.032) and placental weight has a significant mediation effect (ACME = − 0.08 [− 0.12, − 0.04], p < 2.0 × 10−16). Maternal PGS of “unlikely-related” SNPs do not significantly reduce birth weight (total effect = − 0.09 [− 0.18, 0.01], p = 0.082). c Mediation analysis of the models of either the SBP-increasing PGS with “vasculature-related” SNPs (left) and with “unlikely related” SNPs (right) as an exposure variable. Of the total effect of “vasculature-related” PGS on birth weight (total effect = − 0.26 [− 0.44, − 0.11], p = 0.002), the mediated effect by placental weight accounts for the most part (ACME = − 0.26 [− 0.40, − 0.14], p < 2.0 × 10−16). Smaller portion (around 69%) of the total effect of “unlikely related” PGS (total effect = − 0.20 [− 0.38, − 0.01], p = 0.034) is mediated by placental weight (ACME = − 0.14 [− 0.26, − 0.04], p = 0.01)