TABLE 2.
Associations of 36-Month Psychomotor Development Index Scores With Neighborhood Characteristics, Building Disrepair, and Chlorpyrifos Exposure of Children Born in New York City, NY: 1998–2002
| Model 1, B (95% CI) | Model 2, B (95% CI) | Model 3, B (95% CI) | Model 4, B (95% CI) | Model 5, B (95% CI) | Model 6, B (95% CI) | |
| % poverty | –2.6 (–3.7, −1.5) | |||||
| % high school graduates | −1.2 (–2.4, 0.1) | |||||
| % African American | −1.3 (–2.5, 0.0) | |||||
| % linguistic isolation | 0.7 (–0.1, 1.6) | |||||
| % crowded household | 0.0 (–0.9, 1.0) | |||||
| % inadequate plumbing | −0.8 (–2.0, 0.3) | |||||
| % vacant housing | 0.1 (–1.1, 1.3) | |||||
| Index of building disrepair | 0.6 (–0.4, 1.7) | 0.5 (–0.5, 1.4) | 0.6 (–0.4, 1.6) | 0.6 (–0.4, 1.7) | 0.6 (–0.4, 1.7) | 0.6 (–0.4, 1.7) |
| High chlorpyrifos exposure | –6.9 (–11.1, −2.7) | –7.0 (–11.0, −2.9) | –7.3 (–11.5, −3.0) | –7.2 (–11.3, −3.0) | –6.9 (–11.1, −2.8) | –7.1 (–11.4, −2.7) |
| Model fit (R2) | 0.126 | 0.148 | 0.129 | 0.127 | 0.126 | 0.128 |
Note. CI = confidence interval. The total sample size was N = 266. Model 1 was individual and household characteristics, model 2 was model 1 characteristics plus socioeconomic context, model 3 was model 1 characteristics plus neighborhood composition, model 4 was model 1 characteristics plus neighborhood linguistic isolation, model 5 was model 1 characteristics plus neighborhood crowding, and model 6 was model 1 characteristics plus neighborhood physical infrastructure. Regression coefficients are from generalized estimating equation models that adjust for gender, gestational age at birth, Dominican ethnicity, maternal education, maternal intelligence quotient, the presence of secondhand smoke in the home, and an index of caretaking environment quality. All neighborhood characteristics have been standardized, and the corresponding regression coefficients can be interpreted as the mean point increase in Psychomotor Development Index scores for an increase by 1 standard deviation in the neighborhood value of the given characteristic. Multiple imputation was used to fill in missing covariate values and to account for the uncertainty caused by missing data.