Table 3.
Associations between candidate mediators and PLE-total outcomes.
| Total score | Total persistence | |||||
|---|---|---|---|---|---|---|
| Variable | β (SE) | 95% CI | p | Odds ratio | 95% CI | p |
| PM2.5 | 0.072 (0.021) | 0.029, 0.113 | 0.001* | 1.228 | 1.112, 1.356 | <0.001* |
| NO2 | 0.043 (0.021) | 0.001, 0.085 | 0.046 | 1.099 | 0.996, 1.213 | 0.061 |
| Proximity to roads | −0.029 (0.012) | −0.052, −0.006 | 0.014* | 0.942 | 0.893, 0.994 | 0.028* |
| Lead paint | 0.045 (0.013) | 0.020, 0.070 | 0.001* | 1.069 | 1.009, 1.132 | 0.023* |
| Families in poverty | 0.105 (0.012) | 0.081, 0.128 | <0.001* | 1.246 | 1.180, 1.316 | <0.001* |
| Income disparity | 0.116 (0.012) | 0.092, 0.140 | <0.001* | 1.251 | 1.183, 1.322 | <0.001* |
| Violent offenses | 0.031 (0.030) | −0.029, 0.092 | 0.301 | 1.092 | 0.938, 1.270 | 0.257 |
| Marijuana sales | 0.039 (0.028) | −0.017, 0.096 | 0.171 | 1.119 | 0.973, 1.287 | 0.116 |
Associations were corrected for effect of age, sex-at-birth, parental risk for psychosis and include the random intercepts for family nested in site.
False Discovery Rate (FDR) corrected p < 0.05.