TABLE 5.

Multivariate Linear Regression Models Testing Main and Interactive Effects of Prenatal Chlorpyrifos Exposure on 12-, 24-, and 36-Month PDI Scores, Adjusted for Race, Gender, Maternal Education, Maternal IQ, Gestational Age, and Prenatal ETS Exposure

Variable	Model 1: 12 mo (n = 228)		Model 2: 24 mo (n = 227)		Model 3: 36 mo (n = 228)
	B, Mean ± SE	P	B, Mean ± SE	P	B, Mean ± SE	P
Constant	112.28 ± 34.64	.00	81.05 ± 32.45	.01	64.56 ± 36.64	.079
Prenatal chemical exposures
ETSa	0.31 ± 1.76	.86	2.83 ± 1.63	.08	−0.14 ± 1.79	.940
Chlorpyrifosb	−3.30 ± 2.11	.12	1.17 ± 1.98	.56	−6.46 ± 2.18	.003
Covariates
Race/ethnicityc	−2.00 ± 1.81	.27	2.15 ± 1.70	.21	3.88 ± 1.82	.034
Genderd	0.11 ± 1.64	.95	0.08 ± 1.54	.96	−2.95 ± 1.66	.077
Gestational age	−0.16 ± 0.56	.77	0.20 ± 0.53	.70	1.38 ± 0.64	.033
Maternal IQe	−0.71 ± 6.26	.91	0.09 ± 5.56	.99	−5.78 ± 6.08	.343
Low maternal educationf	−0.81 ± 1.77	.65	−1.26 ± 1.66	.45	1.69 ± 1.81	.350
HOME score	−0.08 ± 0.15	.61	0.09 ± 0.14	.53	0.30 ± 0.16	.057
R2	0.024		0.035		0.107

The sample included all children who had reached 3 years of age with data from the prenatal maternal interview, biomarkers of exposure, and complete data on all covariates. Sample sizes for age-specific models varied from 227 to 228 because of missed and/or unscorable assessments (<50). B indicates unstandardized regression coefficient.

^aPrenatally exposed = 1; not exposed = 0.

^bHigh exposure (>6.17 pg/g) = 1; low exposure (≤6.17 pg/g) = 0.

^cBlack = 2; Dominican = 1.

^dMale = 2; female = 1.

^eMeasured with the Test of Nonverbal Intelligence, Second Edition,⁴² natural logarithmically transformed.

^fNo high school degree = 1; high school degree = 0.