Table 2.
Associations between ETS and daily urinary PdG and E1C levels.
| Crude
|
Adjusted
|
|||||
|---|---|---|---|---|---|---|
| Levels (pg/mL) | β | SE | p-Value | β | SE | p-Value |
| Nonconception cycles | ||||||
| Daily urinary log(E1C) | –0.44 | 0.08 | < 0.001 | –0.43 | 0.08 | < 0.001 |
| Daily urinary log(PdG) | –0.08 | 0.15 | 0.600 | –0.07 | 0.15 | 0.637 |
| Conception cycles | ||||||
| Daily urinary log(E1C) | –0.14 | 0.09 | 0.09 | –0.17 | 0.10 | 0.085 |
| Daily urinary log(PdG) | –0.07 | 0.14 | 0.654 | –0.10 | 0.12 | 0.408 |
Sample size: nonconception cycles, 6,880 days (344 cycles); conception cycles, 6,580 days (329 cycles). Twenty indicator variables were created to represent each day in the 20-day window and were put in the adjusted models along with other covariates: age, age squared, BMI, BMI squared, education (high/middle), shift work (yes/no), stress (low, moderate, high), noise exposure (low, middle, high), and dust exposure (low, middle, high). All models used the GEE method to adjust for intrawoman correlation in cycles. Among the other covariates, shift work was found to be associated with a lower log(E1C) levels (β= –0.25, SE = 0.12, p = 0.025); all other covariates were nonsignificant.