Skip to main content
. 2017 May 9;14(5):506. doi: 10.3390/ijerph14050506

Table 3.

Summary of posterior distribution of FSIQ for each exposure profile cluster and the estimated difference in FSIQ compared to lowest exposure profile cluster at age 7-year resulting from the BPR (n = 255).

Clusters Adjusted Posterior Mean FSIQ a,b (95% Credible Intervals) Mean FSIQ_c from FSIQ_CP3 a,c (95% Credible Intervals) Probability FSIQ_c < FSIQ_CP3 d
CP1 95.5 (88.6, 102.0) −6.9 (−11.7, −2.1) 0.998
CP2 100.1 (93.4, 106.3) −2.4 (−7.2, 2.5) 0.83
CP3 102.5 (95.8, 108.7) Ref Ref
CP4 100.1 (91.8, 107.9) −2.4 (−9.0, 4.3) 0.76
CP5 100.6 (93.1, 107.6) −1.9 (−7.1, 3.3) 0.77
CP6 101.3 (93.2, 108.4) −1.1 (−7.9, 3.4) 0.47
CP7 96.1 (87.7, 104.2) −6.4 (−13.1, 0.5) 0.97
CP8 102.8 (94.1, 111.3) 0.3 (−7.0, 7.8) 0.47

a Adjusted for child’s age at WISC assessment (mean centered), sex, language of assessment, maternal education, maternal intelligence (mean centered), maternal country of birth, maternal depression at 7-year visit, HOME score at 7-year visit (mean centered), household poverty level at 7-year visit, and prenatal urinary dialkyl phosphate (DAPs) metabolites (log10, mean centered); b The posterior distribution of the expected FSIQ scores for the cth cluster when holding fixed effect control variables at zero; c Difference between posterior distribution of expected FSIQ for the cth cluster compared to posterior distribution of expected FSIQ for cluster 3 (reference group) when holding control variables at zero; d Probability that the posterior distribution of expected FSIQ for the cth cluster is below the expected FSIQ for cluster 3 when holding control variables at zero.