Figure 3


Association of exposure to OPFRs with GDM in BKMR analysis. (a) The overall effect of mixed exposure to OPFRs on GDM, while the exposure levels were higher than the median levels. (b) Association between individual OPFRs and GDM when the other OPFRs were at the 25th, 50th, and 75th percentiles. (c) The univariate exposure–response effect of individual OPFRs on GDM, while the other OPFRs were fixed at their median exposure levels. (d) The bivariate exposure–response function for one OPFR with another OPFR set at the 25th, 50th, and 75th percentiles, and the other OPFRs were fixed at their median exposure levels. Analyses were adjusted for maternal age and GDM history. BKMR Bayesian kernel machine regression, condPIP conditional posterior inclusion probability, GDM gestational diabetes mellitus, OPFRs Organic phosphate flame retardants, PIP posterior inclusion probability, TBP tributyl phosphate, TBEP Tris (2-butoxyethyl) phosphate, TCEP tris (2-chloroethyl) phosphate, TDCPP tris (1,3-dichloro-2-propyl) phosphate, TMCP tri-m-cresyl phosphate, TOCP tri-ortho-cresyl phosphate, TPHP triphenyl phosphate.