Table 4.
| PFHxSd | PFOSd | PFOAd | PFNAd | NMeFOSAAe | PFPeAe | PFDAe | PFUnDAe | |
|---|---|---|---|---|---|---|---|---|
| % difference in PFAS concentrationf (95% CI) | ||||||||
| Age (years) | −0.3 (−1.6, 1.1) | −0.2 (−1.8, 1.3) | −0.4 (−2.4, 1.7) | −0.3 (−2.2, 1.6) | −1.0 (−6.1, 4.3) | 0.8 (−2.1, 3.8) | −1.0 (−4.5, 2.6) | 3.1 (−1.5, 7.8) |
| Education (Ref = Less than high school) | ||||||||
| High school | 4.9 (−10.2, 22.6) | −0.2 (−16.4, 19.2) | 6.5 (−15.2, 33.8) | 1.2 (−18.4, 25.4) | −26.0 (−58.0, 30.4) | 5.8 (−24.0, 47.3) | 13.0 (−23.3, 66.6) | 0.5 (−40.2, 69.0) |
| Some college | 5.0 (−11.6, 24.7) | 2.1 (−16.1, 24.4) | 5.9 (−17.8, 36.4) | 5.7 (−16.8, 34.1) | −44.4 (−69.4, 0.9) | −2.4 (−31.2, 38.4) | 19.5 (−23.7, 87.2) | −2.5 (−46.8, 78.6) |
| College and above | 9.5 (−13.2, 38.2) | 25.8 (−3.5, 64.0) | 35.7 (−3.5, 90.9) | 57.8 (14.4, 118)** | −17.7 (−63.1, 83.5) | −20.6 (−51.2, 29.1) | 47.9 (−19.5, 172) | 78.1 (−15.0, 273) |
| Poverty income ratio (%) (Ref < 100) | ||||||||
| 100–150 | 8.4 (−6.2, 25.3) | 17 (−1.1, 38.3) | 19.2 (−3.9, 47.9) | 15 (−5.8, 40.5) | 2.1 (−38.7, 70.0) | 4.8 (−24.0, 44.6) | −0.9 (−33.9, 48.6) | 42.2 (−13.9, 135) |
| ≥ 150 | 14.4 (−2.0, 33.6) | 13.8 (−4.2, 35.2) | 15.4 (−7.6, 44.2) | 13.3 (−8.4, 40.0) | −21.0 (−54.4, 37.0) | −7.7 (−32.4, 26.0) | 5.7 (−26.5, 52.1) | 12.3 (−30.4, 81.2) |
| Married or cohabitating (Ref = No) | ||||||||
| Yes | 2.5 (−7.8, 14.0) | 2.3 (−9.4, 15.6) | 6.5 (−8.9, 24.5) | 10.7 (−4.4, 28.3) | −14.0 (−40.2, 23.7) | −0.1 (−22.3, 28.5) | 5 (−20.3, 38.3) | 14.9 (−18.7, 62.2) |
| Insurance (Ref = Private) | ||||||||
| Medicaid | 4.6 (−12.8, 25.4) | 4.4 (−15.1, 28.4) | −2.6 (−25.4, 27.2) | 4.6 (−18.7, 34.5) | 14.3 (−39.8, 117) | −9.5 (−40.1, 36.8) | −5.9 (−41.6, 51.4) | −19.1 (−54.2, 42.9) |
| Hospital (Ref = Emory) | ||||||||
| Grady | 4.2 (−9.8, 20.3) | −0.7 (−15.8, 17.1) | 16.4 (−5.7, 43.8) | −0.8 (−18.7, 21.1) | 38.0 (−17.5, 131) | 3.3 (−23.9, 40.0) | 5.0 (−26.1, 49.3) | −25.2 (−53.8, 21.3) |
| Sampling year (Ref =2014) | ||||||||
| 2015 | −9.4 (−22.1, 5.4) | −12.8 (−26.7, 3.6) | 12.4 (−10.0, 40.4) | -20.6 (−35.6, −2.1)* | -62.8 (−76.9, −40.1)** | 43 (2.3, 100)* | -42.9 (−59.5, −19.5)** | 64.6 (5.4, 157)* |
| 2016 | -57.6 (−63.7, −50.4)** | -50.2 (−58.3, −40.5)** | −4.4 (−23.9, 20.1) | −15.1 (−31.5, 5.3) | -63.6 (−84.0, −17.3)* | 59 (−3.0, 161) | -87.2 (−93.2, −75.9)** | 110.2 (2.0, 333)* |
| 2017–2018 | -35.5 (−44.7, −24.9)** | -54.4 (−61.7, −45.7)** | -37.4 (−50.0, −21.7)** | -60.3 (−67.8, −51.0)** | -59.4 (−74.4, −35.6)** | 208.8 (119, 337)** | -89.7 (−92.8, −85.4)** | 43.8 (−6.9, 122) |
| Parity (#) (Ref = 0) | ||||||||
| 1–2 | −10.5 (−20.3, 0.6) | −6.7 (−18.3, 6.5) | -19.6 (−32.2, −4.6)* | −10.2 (−23.5, 5.5) | −3.8 (−35.9, 44.5) | 1.1 (−21.2, 29.8) | −14.1 (−34.9, 13.4) | −12.5 (−42.1, 32.1) |
| 3–4 | -19.1 (−33.2, −2.0)* | −6.8 (−25.1, 16.1) | -25.8 (−44.0, −1.7)* | −11.4 (−32.1, 15.5) | −9.6 (−54.0, 77.5) | 13.9 (−26.3, 76.2) | −22.8 (−53.5, 28.0) | −20 (−47.5, 82.8) |
| BMI (kg/m2) | -0.9 (−1.6, −0.3)** | -1.2 (−2.0, −0.5)** | −0.3 (−1.3, 0.6) | −0.4 (−1.3, 0.5) | 0.1 (−2.1, 2.4) | −0.9 (−2.3, 0.4) | −0.6 (−2.2, 1.1) | -2.4 (−4.5, −0.3)* |
| Tobacco use (Ref = Never) | ||||||||
| Ever | −22.9 (−34.9, −8.7)** | -25.2 (−38.4, −9.2)** | −18.9 (−36.6, 3.8) | −18.1 (−35.1, 3.3) | 20.6 (−34.1, 121) | 34.9 (−8.5, 98.9) | −18.5 (−50.6, 34.5) | −26.1 (−58.0, 29.9) |
| Alcohol use (Ref = Not during pregnancy) | ||||||||
| During pregnancy | 5.9 (−12.9, 28.6) | 20.8 (−3.3, 50.8) | 25.8 (−5.4, 67.5) | 19.4 (−8.8, 56.3) | 28.4 (−34.4, 151) | −32.8 (−57.8, 7.1) | 1.6 (−37.9, 66.1) | 79.1 (−5.8, 240) |
| Marijuana use (Ref = Not during pregnancy) | ||||||||
| During pregnancy | 9.8 (−4.4, 26.1) | 21.8 (3.9, 42.7)* | 38.8 (13.3, 70.0)** | 42.4 (17.6, 72.4)** | 41.9 (−11.4, 127) | −8.6 (−33.4, 25.3) | 20.9 (−17.8, 77.9) | 23.4 (−21.5, 93.9) |
CI = confidence interval;
*
p-value < 0.05;
**
p-value < 0.01.
a
The models were mutually adjusted for all the variables listed above.
b
The values below LODs were imputed by a lognormal probability distribution and maximum likelihood estimation.
c
The missing values were multiply imputed by chained equations with logistic and polytomous logistic regression imputation approaches.
d
Sample size = 453.
e
Sample size = 356.
f
Percent change in PFAS concentrations associated with each predictor by exponentiating regression coefficients, subtracting 1, and multiplying by 100%.