TABLE 4.
Multiple Logistic Regression Models Testing Effects of Piperonyl Butoxide in Personal Air on the Odds of Delayed Mental Development (MDI ≤85) (n = 230)
| Median Piperonyl Butoxide, ng/m3a | Subjects With MDI Score >85, n | Subjects With MDI Score ≤85, n | Odds Ratio (95% CI) | P | |
|---|---|---|---|---|---|
| Model 1a | |||||
| (ln)PBO | 0.51 | 170 | 60 | 1.32 (1.06–1.66) | 0.01 |
| Model 2b | |||||
| Reference | — | 42 | 15 | — | — |
| 1st Quartile vs reference | 0.27 | 32 | 8 | 0.61 (0.21–1.74) | 0.36 |
| 2nd Quartile vs reference | 0.45 | 41 | 8 | 0.58 (0.21–1.64) | 0.31 |
| 3rd Quartile vs reference | 0.87 | 32 | 10 | 0.97 (0.95–2.60) | 0.95 |
| 4th Quartile vs reference | 4.33 | 23 | 19 | 2.49 (0.95–6.54) | 0.06 |
| Model 3c | |||||
| 4th Quartile vs all other categories | 4.33 | 23 | 19 | 3.11 (1.38–6.98) | 0.006 |
All models adjusted for gender, ethnicity, gestational age, maternal nonverbal intelligence (TONI), education, quality of the home environment (HOME), and prenatal exposure to ETS and plasma chlorpyrifos.
a
Exposure assessed as (ln)PBO; geometric mean is presented.
b
5 categories of PBO exposure: 0 indicates reference (less than LOD); quartiles (1–4) of values in the detectable range.
c
2 categories of PBO exposure: 0 indicates reference (less than LOD) and lowest 3 quartiles; 1, highest quartile of values.