Table 3.
Associations between PFAS plasma concentration and mean change in eGFR during DPPOS follow-up estimated using generalized mixed modelsa, overall and stratified according to baseline hypertension status.
| Log-2 PFAS analyte | Full | Stratified | ||
|---|---|---|---|---|
| (N=875) | Hypertensionb (N=262) | No hypertension (N=613) | ||
| β (95% CI) | β (95% CI) | β (95% CI) | pintc | |
| Total PFOS | −1.50 (−2.81, −0.19) | −2.19 (−4.92, 0.53) | −1.44 (−2.96, 0.06) | 0.68 |
| n-PFOS | −1.34 (−2.62, −0.06) | −1.98 (−4.62, 0.66) | −1.30 (−2.79, 0.17) | 0.72 |
| Sm-PFOS | −1.56 (−2.85, −0.28) | −2.15 (−4.89, 0.57) | −1.49 (−2.96, −0.02) | 0.72 |
| Sm2-PFOSd | −1.87 (−2.88, −0.86) | −2.66 (−4.73, −0.60) | −1.58 (−2.75, −0.41) | 0.24 |
| Total PFOA | −1.35 (−2.73, 0.02) | −3.18 (−5.85, −0.51) | −0.65 (−2.30, 0.99) | 0.16 |
| n-PFOA | −1.49 (−2.97, −0.01) | −3.16 (−5.98, −0.35) | −0.83 (−2.63, 0.96) | 0.24 |
| Sb-PFOA | −0.85 (−1.58, −0.12) | −1.66 (−3.11, −0.21) | −0.55 (−1.41, 0.29) | 0.20 |
| PFHxS | 0.21 (−0.79, 1.21) | −2.35 (−4.46, −0.25) | 1.24 (0.09, 2.39) | 0.01 |
| EtFOSAA | −0.15 (−0.97, 0.65) | 0.02 (−1.63, 1.68) | −0.31 (−1.27, 0.64) | 0.53 |
| MeFOSAA | −0.62 (−1.71, 0.45) | −1.56 (−3.82, 0.69) | −0.15 (−1.40, 1.09) | 0.33 |
| PFNA | 0.18 (−0.90, 1.27) | −1.85 (−4.20, 0.49) | 0.90 (−0.33, 2.14) | 0.06 |
Note:
The generalized mixed model include random intercept and random slope for year since enrollment, the crude model used repeated measures of eGFR from the 1st to the 8th annual visit during DPPOS follow-up as the dependent variable and used baseline PFAS concentration (average between baseline and year 2 measures during DPP, log-2 transformed), treatment arm, and year since randomization as the independent variables, the model also included a second and a third order spline term for year (year*year and year*year*year); the adjusted model adjusted for baseline covariates including age, sex, race/ethnicity, education, marital status, income, smoking status, menopause status, DASH diet score, hypertension status, and use of kidney medication. N=875. The beta coefficient is interpreted as the mean difference in eGFR (mL/min/1.73m2) during DPPOS follow-up per doubling of baseline PFAS concentration.
Hypertension was defined as self-reported hypertension diagnosis, use of anti-hypertensive medications, or systolic/diastolic BP ≥140/90 mmHg.
pint: p-value of the interactive term between PFAS and baseline hypertension status (p<0.1 suggests effect modification)
β (95% CI)= −3.75 (−5.83, −1.68)** for detected vs non-detected Sm2-PFOS.