Renal function and isomers of perfluorooctanoate (PFOA) and perfluorooctanesulfonate (PFOS): Isomers of C8 Health Project in China

Chemosphere 218 (2019) 1042e1049

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Renal function and isomers of perfluorooctanoate (PFOA) and perfluorooctanesulfonate (PFOS): Isomers of C8 Health Project in China Jia Wang a, 1, Xiao-Wen Zeng a, 1, Michael S. Bloom a, b, c, 1, Zhengmin Qian d, Leslie J. Hinyard e, Rhonda Belue f, Shao Lin b, c, Si-Quan Wang g, Yan-Peng Tian a, Mo Yang a, Chu Chu a, Namratha Gurram a, c, Li-Wen Hu a, Kang-Kang Liu a, Bo-Yi Yang a, Dan Feng a, Ru-Qing Liu a, Guang-Hui Dong a, * a

Guangdong Provincial Engineering Technology Research Center of Environmental Pollution and Health Risk Assessment, Guangzhou Key Laboratory of Environmental Pollution and Health Risk Assessment, Department of Preventive Medicine, School of Public Health, Sun Yat-sen University, Guangzhou, 510080, China b Department of Environmental Health Sciences, University at Albany, State University of New York, Rensselaer, NY, 12144, USA c Department of Environmental Health Sciences & Epidemiology and Biostatistics, University at Albany, State University of New York, Rensselaer, NY, 12144, USA d Department of Epidemiology and Biostatistics, College for Public Health & Social Justice, Saint Louis University, Saint Louis, 63104, USA e Center for Health Outcomes Research, Saint Louis University, Saint Louis, 63104, USA f Department of Health Management and Policy, College for Public Health & Social Justice, Saint Louis University, Saint Louis 63104, USA g Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, MA, 02115, USA

h i g h l i g h t s
A large population-based study was conducted to assess the association between PFASs exposure and renal function and CKD.
Branched PFOS isomers were positively associated with lower eGFR and higher CKD.
Branched PFOS isomers show stronger effects than linear PFOS.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 8 September 2018 Received in revised form 23 November 2018 Accepted 27 November 2018 Available online 28 November 2018

Perfluoroalkyl substances (PFASs) are widely-utilized synthetic chemicals commonly found in industrial and consumer products. Previous studies have examined associations between PFASs and renal function, yet the results are mixed. Moreover, evidence on the associations of isomers of PFASs with renal function in population from high polluted areas is scant. To help to address this data gap, we used high performance liquid chromatography-mass spectrometry to measure serum isomers of perfluorooctanoate (PFOA), perfluorooctanesulfonate (PFOS), and other PFASs from 1612 adults residing in Shenyang, China, and characterized their associations with estimated glomerular filtration rate (eGFR) and chronic kidney disease (CKD). Results showed that after adjusted for multiple confounding factors, most of the higher fluorinated PFASs, except for PFOA and PFDA, were negatively associated with eGFR and positively associated with CKD. Compared with linear PFOS (n-PFOS), branched PFOS isomers (Br-PFOS) were more strongly associated with eGFR (Br-PFOS; b ¼ 1.22, 95%CI: 2.02, 0.42; p ¼ 0.003 vs. n-PFOS; b ¼ 0.16, 95%CI: 0.98, 0.65; p ¼ 0.691) and CKD (Br-PFOS; OR ¼ 1.27; 95% CI: 1.02, 1.58; p ¼ 0.037 vs. n-PFOS; OR ¼ 0.98; 95% CI: 0.80, 1.20; p ¼ 0.834). In conclusion, branched PFOS isomers were negatively associated with renal function whereas their linear counterparts were not. Given widespread exposure to PFASs, potential nephrotoxic effects are of great public health concern, Furthermore, longitudinal

Handling Editor: A. Gies Keywords: Polyfluoroalkyl substances (PFASs) PFASs isomers Estimated glomerular filtration rate Renal function Isomers of C8 health project

* Corresponding author. Guangdong Provincial Engineering Technology Research Center of Environmental Pollution and Health Risk Assessment, Guangzhou Key Laboratory of Environmental Pollution and Health Risk Assessment, Department of Preventive Medicine, School of Public Health, Sun Yat-sen University, 74 Zhongshan 2nd Road, Yuexiu District, Guangzhou, 510080, China. E-mail addresses: [email protected], [email protected] (G.-H. Dong). 1 These authors contributed equally to this work and should both be list as the first author. https://doi.org/10.1016/j.chemosphere.2018.11.191 0045-6535/© 2018 Published by Elsevier Ltd.

J. Wang et al. / Chemosphere 218 (2019) 1042e1049

1043

research on the potential nephrotoxic effects of PFASs isomers will be necessary to more definitively assess the risk. © 2018 Published by Elsevier Ltd.

1. Introduction Perfluoroalkyl substances (PFASs) are a class of highly persistent, synthetic chemicals that have been widely used in commerce and industry (Wang et al., 2017; KEMI, 2015). PFASs are found in air and water worldwide (Guelfo and Adamson, 2018; Lu et al., 2018; Lau, 2012), including the Arctic (AMAP, 2009), and have been widely detected in human and wildlife biospecimens, given in vivo halflives measured in years (Bangma et al., 2019; Choi et al., 2017; Buck et al., 2011; Olsen et al., 2007). Biomonitoring studies have suggested that some PFASs exposure has decreased over the past 10e15 years, following manufacturer phase outs, including perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) (Olsen et al., 2017; Stubleski et al., 2016). China, however, still continue to produce PFOS, PFOA and other PFASs, which led to a sharp rise in PFOS manufacturing from the 2003 to 2006 followed by a plateau (Wang et al., 2014). As recently as 2011, approximately 230 tons of PFOS-related chemicals were still manufactured in China (Xie et al., 2013). The impact of the continued presence of PFASs in the environment in China remains a significant concern (Lam et al., 2016). Chronic kidney disease (CKD) is a stage early in the renal disease continuum in which could lead to renal disease. It is known to be a growing public health problem (Webster et al., 2017; Gansevoort et al., 2013). The prevalence of CKD has not only been increasing in economically developed countries, but also in developing countries, including China (Hill et al., 2016; Nugent et al., 2011). In China, CKD affects about 119.5 million people with a total prevalence of 10.8% (Zhang et al., 2012). A large quantity of experimental animal studies implicates PFASs as renal toxicants (DeWitt, 2015; Lau et al., 2007). The kidney appears to be a target organ for PFAS-associated health effects, given its primary role in excreting PFASs (Zhong et al., 2019; Stanifer et al., 2018). Thus, PFAS exposure might contribute to CKD. While several epidemiological studies have explored associations between PFASs and renal function and CKD, the results have been inconclusive (Stanifer et al., 2018; ATSDR, 2015). Furthermore, to the best of our knowledge, there have been no human studies to assess the relative renal effects of the different linear and branched PFAS isomers created by various production processes, despite reported differences in their bioaccumulation and toxicity (Liu et al., 2019; Fang et al., 2016; O'Brien et al., 2011). To help to address the pending data gap, we investigated associations between PFOS and PFOA isomers and renal function in The Isomers of C8 Health Project (C8). The C8 study was designed to explore health effects of PFAS isomers among Chinese government employees residing in Shenyang, the largest city in northeast China. Average PFOA and PFOS concentrations in serum, the two most prevalent PFASs in Chinese biospecimens, were higher in Shenyang residents than in other parts of China (Zeng et al., 2015; Yeung et al., 2006, 2008). We hypothesized that branched PFOA and PFOS isomers would be more strongly associated with a lower eGFR and higher prevalence of CKD than linear PFOA and PFOS.

described in detail (Bao et al., 2017). We recruited Chinese government employees and retirees from Shenyang from July 2015 to October 2016 as part of the C8 study. We also enrolled community dwelling residents from Shenyang to evaluate different exposure source of PFASs throughout city sectors. One community was randomly selected from each of five geographic zones (north, south, east, west, and central Shenyang), and 100 residents at least 35 years of age, and with at least five years residency at their present addresses, were picked randomly from each community. An overall n ¼ 1753 (1253 government staffs and 500 general residents) were recruited, and n ¼ 1612 completed the study (1228 government staffs (98.0%) and 384 general residents (76.8%); 1204 men and 408 women). The ratio of women: men for community-dwelling participants was similar to that for government staffs and with few exceptions, they had similar serum PFAS levels (Table S9; Bao et al., 2017). To maximize statistical power, we combined the government and community-dwelling subgroups for subsequent data analysis. Study participants completed an interviewer-administered questionnaire, and we collected anthropometric data and blood samples after each participant signed informed consent. The questionnaire queried information on demographic characteristics, likely environmental sources of PFAS exposure, and current and past tobacco smoking behaviors. The blood samples were taken after 8 h of fasting and were immediately processed for serum and refrigerated before being shipped to the analytical laboratory. This study was approved by the IRB of Sun Yat-sen University Research Ethics committee and complied with the principles of the Helsinki Declaration for the protection of human subjects. 2.2. Serum PFASs and isomeric PFOA and PFOS measurements The analytic method was previously described in detail (Li et al., 2017; Benskin et al., 2007; Kuklenyik et al., 2004). Briefly, after mixing 0.3 mL serum with 2 mL 0.1 M formic acid to prepare for extraction, 0.5 ng of each mass-labeled internal standards were added and thoroughly vortexed. All standards used above, containing the mixtures of characterized isomeric PFOS and PFOA, were purchased from Wellington Laboratories on Guelph, Canada. Mixture of seventeen linear standards for PFASs is PFAC-MXB and another nine mass-labeled internal standards of linear PFASs is MPFAC-MXA. Then, after a series of extraction processes, 10 mL of the extracts was used to detect PFASs and isomers of PFOS and PFOA levels by a high-performance liquid chromatographer (HPLC) coupled to a Triple Quadrupole mass spectrometer (MS) (Agilent, Palo Alto and Santa Clara, CA). More detailed methods could be looked over in Supplemental Material. The detection rates of 18 individual PFASs, including seven structural PFOA and PFOS isomers, were >30% and then were selected for analysis in present study. The limit of detection (LOD) is defined as the peak analyte level requested to reach a signal-to-noise ratio of 3:1. Table S1 presents the nomenclature and abbreviations for the measured PFASs. The mass transitions and compound-specific tandem MS parameters are presented in Table S2.

2. Materials and methods 2.3. Study outcomes 2.1. Study participants Participants enrollment and data collection were previously

Serum creatinine levels were detected using the modified kinetic Jaffe reaction. The GFR is the most widely used and most

1044

J. Wang et al. / Chemosphere 218 (2019) 1042e1049

perfect comprehensive indicator for the level of renal function (NKF, 2002) and is estimated rather than measured in clinical practice. We measured the estimated glomerular filtration rate (eGFR) with the Modification of Diet in Renal Disease study equation (Matsuo et al., 2009): eGFR (mL/min/1.73 m2) ¼ 194  serum creatinine (mg/dL)1.094  age0.287 (  0.739 for women). CKD was defined as estimated glomerular filtration rate (eGFR) is less than 60 mL/min/1.73 m2, according to previous studies and guidelines (NKF, 2002; Levey et al., 2005).

sensitivity analyses. We also conducted repeated the analysis when excluding smokers, alcohol drinkers and those using medication to account for strong associations with kidney disease (Xia et al., 2017; Markowitz et al., 2015). We defined significance as P < 0.05 for a two-tailed test, for major effects and interactions. Analyses were performed in SAS version 9.4 (Carey, NC).

2.4. Potential confounders

The demographic and renal characteristics of 1612 study participants are shown in Table 1. Participants were 22e96 years of age (mean ¼ 55.1 years) at enrollment. The mean level of eGFR was 78.0 mL/min/1.73 m2 and the prevalence of CKD was 11.7%. Table 2 shows the distribution of PFASs concentrations in serum. The descending order of median concentrations was total PFOS (24.22 ng/mL), total branched PFOS isomers (12.16 ng/mL), linear PFOS (11.37 ng/mL), S3þ4þ5 m ePFOS (8.23 ng/mL), and total PFOA (6.19 ng/mL). We detected multiple, mostly positive moderate to strong and statistically significant intercorrelations among the measure PFASs (Supplementary material Figure S1).

Potential confounders were selected a priori according to our literature review. These included: age, sex, annual income, education attainment, smoking status (Xia et al., 2017), drinking status (Thawornchaisit et al., 2015), regular exercise and family history of CKD which were all from the questionnaire. Smokers were defined as people who smoked one or more cigarettes per day over the past one year (Liu et al., 2016). Drinking status were defined as alcohol consumption of weekly drinking of beer, wine or hard liquor. A person was considered regular exercise if they reported averagely one or more hours of exercise per day for the past year. A person was considered to have a family history of CKD if any member, including biological parent, grandparents, brothers or sisters, had a CKD diagnosis. Since increased body mass index (BMI) was recognized as a risk factor of kidney disease (Saydah et al., 2007), BMI was also included in the analyses. Weight and height were measured using a standardized protocol (WHO, 1995) and BMI was calculated as weight in kg/height in m2. Moreover, total cholesterol was recognized to be a possible covariate given that lowering cholesterol concentration might decrease the rate of renal function loss (Anavekar and Pfeffer, 2004). Serum total cholesterol levels were measured enzymatically and units were recorded in mg/dl. 2.5. Statistical analysis For serum PFAS levels were lower than the limit of detection (LOD), we substituted these values with the LOD divided by the square root of 2 (Hornung and Reed, 1990). We characterized the distributions of PFASs as well as covariates, and we evaluated associations among the measured PFASs using Spearman correlation coefficients. Serum PFASs concentrations were log-transformed and analyzed as continuous variables to maximize statistical power for detecting modest associations. We also categorized exposure as quartiles to allow for non-linear dose-response associations. We used linear regression models to examine the cross-sectional associations between individual PFASs and continuous eGFR, and logistic regression models to estimate the odds ratios for dichotomous CKD. In an initial set of regression models, we assessed the impact of PFAS concentration without adjustment for covariates. In a second set of regression models adjusted for age (<44, 44e60, or >60 years), sex (man or woman), annual income (<30,000, 30,000e100000, or >100,000 RMB/year), education (
3. Results

3.1. Associations between serum PFAS concentrations and renal function Table 3 shows the results of linear regression model evaluating associations between individual continuous natural log transformed PFAS concentrations and eGFR. In the adjusted models, eGFR was negatively associated with higher concentrations of Total-PFOS, Br-PFOS, iso-PFOS, S3þ4þ5 m ePFOS, Sm2-PFOS, PFBA, PFPeA and PFHxA. In contrast, we found eGFR was positively associated with higher n-PFOA, Total-PFOA, PFDA as well as PFUdA. Compared to linear PFOS, branched PFOS isomers has stronger associations with eGFR. For example, a ln-unit higher of n-PFOS was associated with a mean lower eGFR of 0.16 mL/min/1.73 m2 (95% CI: 0.98, 0.65; p ¼ 0.691) and a ln-unit higher of Br-PFOS was associated with a mean lower eGFR of 1.22 mL/min/1.73 m2 (95% CI: 2.02, 0.42; p ¼ 0.003). Fig. 1 presents that the positive trends Table 1 Demographic and renal characteristics of the Isomers of C8 Health Project (C8) participants. Characteristic

Overall (n ¼ 1612)

Age [years (mean ± SD)] Sex - Male [n (%)] Race - Han [n (%)] Education level < High school [n (%)] Occupation - Government worker [n (%)] Average annual income [n (%)] 30,000 RMBa 30,000 RMB - 100,000 RMBa 100,000 RMBa Regular exercise [n (%)] Smoke [n (%)] Drink alcohol [n (%)] BMI [kg/m2 (mean ± SD)] Family history of CKDb [n (%)] Medication [n (%)] Total cholesterol [mg/dL (mean ± SD)] Serum creatinine [mg/dL (mean ± SD)] eGFR [mL/min/1.73m2 (mean ± SD)] CKDb [n (%)]

55.1 ± 16.4 1204 (74.7) 1481 (91.9) 336 (20.8) 1228 (76.2) 267 (16.6) 1070 (66.4) 275 (17.1) 606 (37.6) 288 (17.9) 613 (38.0) 24.8 ± 3.2 22 (1.4) 432 (26.8) 193.3 ± 36.6 0.8 ± 0.2 78.0 ± 16.0 188 (11.7)

Abbreviations: BMI, body mass index; eGFR, estimated glomerular filtration rate; CKD, chronic kidney disease. a RMB, Chinese Yuan. b CKD was defined as eGFR <60 mL/min/1.73 m2.

J. Wang et al. / Chemosphere 218 (2019) 1042e1049 Table 2 Serum concentrations (ng/ml) of perfluoroalkyl substances quantified in The Isomers of C8 Health Project (C8) participants. PFASs

Total-PFOS n-PFOS Br-PFOS 1m-PFOS iso-PFOS S3þ4þ5m-PFOS Sm2-PFOS Total-PFOA n-PFOA iso-PFOA PFBA PFPeA PFHxA PFHxS PFNA PFDA PFDS PFUdA PFDoA PFTrDA PFTeDA

LOD (ng/mL)

0.02 0.02 0.02 0.03 0.03 0.09 0.10 0.02 0.02 0.09 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.24 0.24

Overall (n ¼ 1612) Median (Q1, Q3)

mean ± SD

24.22 (14.62, 37.19) 11.37 (7.21, 18.09) 12.16 (6.68, 18.43) 1.31 (0.76, 2.00) 2.29 (1.14, 4.26) 8.23 (4.22, 11.90) 0.07 (0.07, 0.16) 6.19 (4.08, 9.31) 6.08 (3.98, 9.12) 0.06 (0.06, 0.13) 0.15 (0.01, 0.51) 0.01 (0.01, 0.06) 0.03 (0.01, 1.55) 0.71 (0.01, 2.68) 1.96 (1.11, 3.07) 0.86 (0.51, 1.45) 0.01 (0.01, 0.11) 0.50 (0.01, 0.95) 0.12 (0.05, 0.19) 0.41 (0.17, 0.88) 0.17 (0.17, 0.35)

30.55 ± 28.12 15.04 ± 15.10 15.51 ± 17.11 1.64 ± 1.63 3.44 ± 3.87 10.23 ± 13.07 0.20 ± 0.66 8.02 ± 7.98 7.83 ± 7.44 0.19 ± 1.88 0.97 ± 4.69 0.06 ± 0.29 0.97 ± 1.93 1.65 ± 2.32 2.34 ± 1.83 1.20 ± 1.15 0.07 ± 0.34 0.71 ± 1.08 0.18 ± 0.32 1.00 ± 6.41 0.38 ± 0.91

Abbreviation: PFASs and branched PFOS/PFOA isomers, see in supplemental table S1.

for the associations between isomers of PFOS and eGFR were highly significant, however, we did not detect evidence for non-linear associations when using categorical PFASs as predictors of eGFR (Table S3) and there was no evidence for an interaction by sex (Table S4). Table 4 shows the results of logistic regression model evaluating associations between individual natural log-transformed PFAS levels and CKD, adjusted for confounding factors. We found that branched PFOS isomers (e.g., Br-PFOS OR ¼ 1.27; 95% CI: 1.02, 1.58; p ¼ 0.037) were more strongly associated with CKD than linear PFOS (e.g., n-PFOS OR ¼ 0.98; 95% CI: 0.80, 1.20; p ¼ 0.834). In contrast, higher linear PFOA and PFDA were associated with lower odds for CKD (e.g., n-PFOA OR ¼ 0.75; 95% CI: 0.59, 0.97; p ¼ 0.025),

1045

whereas there was no association for total branched PFOA (OR ¼ 1.01; 95% CI: 0.77, 1.31); p ¼ 0.967). We did not detect evidence for non-linear associations when using categorical PFASs as predictors of CKD (Table S5). 3.2. Sensitivity analyses We also conducted several sensitivity analysis, as shown in Table 5, we found that excluding 432 participants using medicines did not alter the associations of PFASs with eGFR, except for an attenuated and non-significant effect estimate for PFOA association (b ¼ 0.54; 95% CI: 0.52, 1.61; p ¼ 0.316), or excluding smokers (n ¼ 288), alcohol drinkers (613), or excluding cholesterol variable from covariates (Supplementary material Table S6, S7 and S8). 4. Discussion In the present investigation, we observed many associations between PFASs and eGFR levels, and CKD prevalence, yet the pattern was mixed. While positive associations were found for branched PFOS, negative associations were found for linear PFOA. In general, branched PFOS isomers showed stronger effects for renal function and CKD in comparison to linear PFOS, underscoring the importance of differentiating PFAS structure during epidemiologic investigations. However, few previous studies have assessed PFAS isomers in human serum, and to the best of our knowledge, none examined their association with renal function and CKD. Previous evidence of association between PFASs and renal function and CKD is limited to few published human studies (Shankar et al., 2011; Watkins et al., 2013; Kataria et al., 2015; Dhingra et al., 2016; Steenland and Woskie, 2012; Steenland et al., 2015). Shankar et al. (2011) examined the cross-sectional relation of serum PFASs and CKD in 4587 adults (51.1% women) using the data from the National Health and Nutrition Examination Survey (NHANES) in 1999e2000 and 2003e2008. They found higher serum PFOA and PFOS levels were negatively associated with eGFR levels (Q4 v. Q1: PFOA (b ¼ 5.7, 95%CI: 7.9, 3.5; p for trend <0.0001), PFOS (b ¼ 6.7, 95%CI: 8.9, 4.4; p for trend <0.0001)), and positively associated with CKD (Q4 v. Q1: PFOA (OR ¼ 1.73, 95%

Table 3 Mean changes in eGFR (mL/min/1.73 m2) per 1 ln-serum PFASs (ng/mL) using linear regression. PFASs

Unadjusted b (95%CI)<

P value

Adjusted ba (95%CI)

P value

Total-PFOS n-PFOS Br-PFOS 1m-PFOS iso-PFOS S3þ4þ5m-PFOS Sm2-PFOS Total-PFOA n-PFOA iso-PFOA PFBA PFPeA PFHxA PFHxS PFNA PFDA PFDS PFUdA PFDoA PFTrDA PFTeDA

4.18 (5.16, 3.19) 2.05 (2.99, 1.12) 4.33 (5.14, 3.52) 2.16 (2.99, 1.34) 3.92 (4.64, 3.21) 3.95 (4.71, 3.20) 4.03 (5.05, 3.02) 1.10 (0.00, 2.20) 1.09 (0.02, 2.16) 1.86 (3.11, 0.61) 1.50 (1.91, 1.08) 2.57 (3.43, 1.72) 0.86 (1.19, 0.54) 0.40 (0.09, 0.71) 2.16 (3.09, 1.22) 0.57 (0.36, 1.49) 0.77 (1.45, 0.08) 0.77 (0.35, 1.19) 0.97 (1.63, 0.31) 1.79 (2.61, 0.98) 1.41 (2.56, 0.26)

<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.051 0.047 0.004 <0.001 <0.001 <0.001 0.012 <0.001 0.231 0.028 0.001 0.004 <0.001 0.017

¡0.91 (-1.83, 0.00) 0.16 (0.98, 0.65) ¡1.22 (-2.02, -0.42) 0.24 (0.52, 1.00) ¡1.28 (-1.95, -0.61) ¡1.12 (-1.86, -0.38) ¡1.92 (-2.78, -1.06) 1.23 (0.30, 2.17) 1.18 (0.27, 2.09) 0.37 (1.41, 0.67) ¡0.49 (-0.84, -0.14) ¡1.17 (-1.88, -0.45) ¡0.29 (-0.56, -0.01) 0.24 (0.02, 0.50) 0.36 (0.46, 1.18) 1.04 (0.27, 1.81) 0.07 (0.63, 0.50) 0.40 (0.05, 0.74) 0.01 (0.56, 0.54) 0.40 (1.08, 0.28) 0.69 (1.64, 0.26)

0.050 0.691 0.003 0.538 <0.001 0.003 <0.001 0.010 0.011 0.485 0.006 0.001 0.041 0.071 0.390 0.008 0.815 0.026 0.975 0.253 0.153

Abbreviation: BMI, body mass index; eGFR, estimated glomerular filtration rate; CKD, chronic kidney disease. PFASs and branched PFOS/PFOA isomers, see in Table S1. NOTE: Number in bold indicates p < 0.05. a Adjusted for age, sex, BMI, education, annual income, regular exercise, cigarette smoking, drinking alcohol, family history of CKD and total cholesterol.

1046

J. Wang et al. / Chemosphere 218 (2019) 1042e1049 Table 5 Mean changes in eGFR (mL/min/1.73 m2) per 1 ln-serum PFASs (ng/mL) using linear regression excluding medicine takers. PFASs

Adjusted ba (95%CI)

P value

Total-PFOS n-PFOS Br-PFOS 1m-PFOS iso-PFOS S3þ4þ5m-PFOS Sm2-PFOS Total-PFOA n-PFOA iso-PFOA

¡1.38 (-2.39, -0.37) 0.77 (1.70, 0.15) ¡1.51 (-2.39, -0.63) 0.40 (1.26, 0.45) ¡1.40 (-2.14, -0.66) ¡1.40 (-2.23, -0.58) ¡1.21 (-2.18, -0.23) 0.52 (0.54, 1.59) 0.54 (0.52, 1.61) 0.79 (1.92, 0.35)

0.008 0.103 0.001 0.355 <0.001 0.001 0.015 0.337 0.316 0.176

Abbreviation: BMI, body mass index; eGFR, estimated glomerular filtration rate; CKD, chronic kidney disease. PFASs and branched PFOS/PFOA isomers, see in supplemental table S1. NOTE: Number in bold indicates p < 0.05. a Adjusted for age, sex, BMI, education, annual income, regular exercise, cigarette smoking, drinking alcohol, family history of CKD and total cholesterol.

Fig. 1. Differences in serum eGFR (mL/min/1.73 m2) with increasing quartile of PFOS exposure (ng/mL). Changes in n-PFOS (A) and in Br-PFOS (B). Linear model analysis was performed to estimate the association with eGFR levels in PFOS isomers quartiles, with the lowest PFOS isomers quartile used as a reference group. All models are adjusted by age, sex, BMI, education, annual income, regular exercise, cigarette smoking, drinking alcohol, family history of CKD and total cholesterol. p-Values for trend were calculated using categories representing the median values of corresponding quartiles (Q1: quartile 1 - reference category; Q2: quartile 2; Q3: quartile 3; Q4: quartile 4 with boxes representing the median of each quartile and whiskers representing the 95% confidence interval). *p < 0.05 compared with quartile 1.

CI: 1.04, 2.88; p for trend ¼ 0.015), PFOS (1.82, 95%CI: 1.01, 3.27; p for trend ¼ 0.019). However, the magnitude of effects of PFOS exposure was stronger than ours, and the effects of PFOA was contradictory to ours, despite lower median serum PFOA (4.1 ng/ mL) and PFOS levels (18.7 ng/mL) for that study were lower than ours. Other investigators also reported negative associations between serum PFOS/PFOA and eGFR levels using nationally representative U.S. data for adolescents from 2003 to 2010 (Kataria et al., 2015), and with no associations for PFNA and PFHxS (Q4 v. Q1: PFOA (b ¼ 6.84 mL/min/1.73 m2; 95%CI: 11.48, 2.19), PFOS (b ¼ 9.69 mL/min/1.73 m2; 95% CI: 14.78, 4.59). A large crosssectional investigation of children and adolescents exposed to high PFOA levels (median: 28.3 ng/mL, range: 0.7e2071 ng/mL) via contaminated drinking water in the U.S. Mid-Ohio Valley also reported decrements in eGFR in association with higher serum concentrations of PFOA, PFOS, PFNA and PFHxS (Watkins et al., 2013). In an investigation of workers (n ¼ 5791) exposed to high level of PFOA (median concentration: 580 ng/mL) at the DuPont chemical

Table 4 Odds ratios (95% CI) for CKDb per 1 ln-PFASs (ng/mL) using logistic regression. PFASs

Unadjusted OR (95%CI)

P value

Adjusted ORa(95%CI)

P value

Total-PFOS n-PFOS Br-PFOS 1m-PFOS iso-PFOS S3þ4þ5m-PFOS Sm2-PFOS Total-PFOA n-PFOA iso-PFOA PFBA PFPeA PFHxA PFHxS PFNA PFDA PFDS PFUdA PFDoA PFTrDA PFTeDA

1.56 (1.27, 1.91) 1.11 (0.91, 1.34) 1.76 (1.46, 2.12) 1.22 (1.03, 1.46) 1.75 (1.48, 2.06) 1.71 (1.42, 2.05) 1.86 (1.58, 2.19) 0.73 (0.59, 0.90) 0.75 (0.61, 0.91) 1.17 (0.94, 1.45) 1.21 (1.11, 1.31) 1.17 (0.99, 1.37) 1.05 (0.99, 1.12) 0.99 (0.93, 1.05) 1.12 (0.93, 1.36) 0.71 (0.60, 0.85) 0.98 (0.85, 1.12) 0.96 (0.89, 1.04) 1.04 (0.91, 1.18) 1.00 (0.85, 1.17) 0.87 (0.69, 1.11)

<0.001 0.304 <0.001 0.024 <0.001 <0.001 <0.001 0.004 0.004 0.170 <0.001 0.059 0.106 0.621 0.240 <0.001 0.721 0.319 0.585 0.949 0.262

1.17 (0.92, 1.49) 0.98 (0.80, 1.20) 1.27 (1.02, 1.58) 0.93 (0.76, 1.13) 1.34 (1.11, 1.61) 1.26 (1.01, 1.55) 1.76 (1.44, 2.14) 0.73 (0.57, 0.95) 0.75 (0.59, 0.97) 1.01 (0.77, 1.31) 1.07 (0.98, 1.18) 1.02 (0.86, 1.22) 0.99 (0.92, 1.07) 1.01 (0.94, 1.07) 0.86 (0.7, 1.07) 0.73 (0.60, 0.88) 0.92 (0.79, 1.06) 1.01 (0.93, 1.11) 0.95 (0.82, 1.11) 0.88 (0.74, 1.06) 0.83 (0.63, 1.09)

0.205 0.834 0.037 0.446 0.002 0.037 <0.001 0.019 0.025 0.967 0.137 0.810 0.837 0.849 0.186 0.001 0.248 0.798 0.530 0.189 0.187

Abbreviation: BMI, body mass index; eGFR, estimated glomerular filtration rate; CKD, chronic kidney disease. PFASs and branched PFOS/PFOA isomers, see Table S1. NOTE: Number in bold indicates p < 0.05. a Adjusted for age, sex, BMI, education, annual income, regular exercise, cigarette smoking, drinking alcohol, family history of CKD and total cholesterol. b CKD was defined as eGFR <60 mL/min/1.73 m2.

J. Wang et al. / Chemosphere 218 (2019) 1042e1049

facility in West Virginia, Steenland and Woskie (2012) observed an ascending in deaths caused by CKD when compared with other regional DuPont workers (standardized mortality ratio (SMR) ¼ 3.11, 95% CI: 1.66, 5.32). However, another cohort incidence study of workers (n ¼ 3713) also exposed to PFOA at the DuPont chemical facility in West Virginia (Steenland et al., 2015) reported no association between the cumulative PFOA levels and the risk of CKD. Similarly, Dhingra et al. (2016) also found no relationship between PFOA exposure and CKD using a longitudinal analysis among adults (n ¼ 32,254) in a Mid-Ohio Valley community cohort exposed to high PFOA levels. These contradicted observations might result from the differences between study design, population and PFASs exposure level. Watkins et al. (2013) noted that the association was likely to be attributed to reverse causation, in which slower excretion led to higher serum PFAS levels among participants with compromised kidney function. Dhingra et al. (2017) later made a similar suggestion by comparing associations estimated in cross-sectional and prospective approaches, using measured and modeled serum PFOA in association with eGFR, respectively. While our study is also vulnerable to reverse causation given its cross-sectional nature, we detected positive associations for PFOA and PFDA with higher eGFR and lower CKD, and contradictory negative associations for PFOS, which suggests otherwise. The reasons for the opposite results between PFOA/PFDA and PFOS are unknown but are probably caused by different toxicokinetic effects. A 2-compartment clearance-volume model showed that PFOA had approximately 75 times higher efficient elimination in renal that those for PFOS in fish (Consoer et al., 2016). Zhao et al. (2017) suggested that some membrane transporters, such as organic anion transporting polypeptides (OATPs), apical sodiumdependent bile salt transporter (ASBT) were capable of contributing to the enterohepatic circulation of perfluoroalkyl sulfonates. In addition, Popovic et al. (2014) found that PFOA was a strong inhibitor of OATP 1D1 whereas PFOS was a high affinity substrate in zebrafish. These evidences may be an explanation for the opposite association observed in PFOA and PFOS. However, we need to interpret this result cautiously because of the cross-sectional study design. Still, we cannot rule out the possibility for reverse causality and a future longitudinal study will be necessary for a more definitive interpretation of our results. In current study, we also observed less association between longer chain PFASs (C > 9, i.e. PFUdA, PFDoA, PFTrDA and PFTeDA) and eGFR and CKD, whereas the association was more obvious in shorter chain PFASs. A recent study has investigated the association between peroxisome proliferator-activated receptor alpha (PPARa) activity and cellular uptake of 14 PFASs in HepG2 cells (Rosenmai et al., 2018). Results showed that higher induction of PPARa activity was observed after treatment of cells with short-chain PFASs compared with PFASs with longer chain (C > 9), even though the cellular uptake of PFASs was lower in the latter one. Additionally, it is been reported that short-chain PFASs could be more easily transported to the nucleus by human liver fatty acid binding protein compared to long-chain PFASs (Zhang et al., 2013). These evidences could provide a possible explanation for the different observation between the short-chain PFASs and long-chain PFASs in our study. Our results suggested the health risk of short-chain PFASs is needed to take attention. However, this speculation remains to be verified. Electro-chemical fluorination is one of the most commonly used PFASs manufacture processes which can produce up to 30% branched PFOS isomers and 20% branched PFOA isomers (Benskin et al., 2010). Even the global production of PFOS has declined, some developing countries, such as China has continued the production of PFOS (Xie et al., 2013). It is worth noting that some of the

1047

branched PFAS isomers may have different behavior according to its isomeric and enantiomeric nature (Miralles-Marco and Harrad, 2015). Although there is scant research on the associations between PFAS isomers and renal function and CKD which can be compared with our results, we found strong associations between other health outcomes and exposure to branched PFOS/PFOA isomers in some studies. For example, in a prior C8 study, Bao et al. (2017) identified stronger association for branched PFOS isomers and hypertension (e.g. for Br-PFOS, OR ¼ 1.25; 95% CI: 1.11, 1.42; for n-PFOS, OR ¼ 1.11; 95% CI: 0.97, 1.27). In another recent birth cohort study in Guangzhou, China (n ¼ 321), Li et al. (2017) reported stronger negative effects for branched PFOS isomers (b ¼ 126.3 g; 95%CI: 195.9, 56.8) than linear PFOS (b ¼ 57.2 g; 95%CI: 103.1, 11.3) on birth weight. A recent study investigated the association between the isomers of PFOA/PFOS and some serum biochemistry biomarkers in adults (n ¼ 1871) by analyzing data from NHANES in cycle of 2013e2014 (Liu et al., 2018). In this report, Liu et al. (2018) revealed that serum isomers of PFOA and PFOS were associated with glucose homeostasis, serum protein as well as lipid profiles in adults. However, Yu et al. (2015) found that the linear PFOS contributed the most of risk (83.0e90.2% of the risk quotient) of total PFOS in connection with thyroid hormone perturbation in drinking water in China. Therefore, prospective investigations to explore the relationship between renal function and CKD and isomeric PFASs are required to further clarify the importance of PFAS structure on associations with renal function and to more definitively assess the risks to human health. Biologic mechanisms underlying associations between PFASs and renal function is not well understood. In a recent review, Stanifer et al. (2018) has thoroughly summarized the understanding between PFAS exposure and kidney health and has shown that PFAS exposure altered several pathways linked to renal function, including oxidative stress response, PPAR pathways, and partial epithelial mesenchymal transition (EMT). Wen et al. (2016) demonstrated that PFOS exposure induced renal tubular epithelial cell toxicity via PPARg inactivation. Chou et al. (2017) further demonstrated that PFOS increased the expression of renal injury biomarkers and found that the increased PPARg deacetylation and inactivation played a crucial role in the cell transformation. However, the toxicokinetic properties and activities of PFAS may be isomer-specific. In animals, pharmacokinetic studies have shown the preferential bioaccumulation of linear PFOA and PFOS (Benskin et al., 2009a, 2009b; De Silva et al., 2009; Fang et al., 2016; Sharpe et al., 2010). Although branched PFOS isomers were eliminated more than linear PFOS, they had different elimination mechanisms. A recent study has shown that the proportion of branched PFOS isomers were significantly higher in kidney than those in ECF industrial product and other organs in fish (Zhong et al., 2019). The result showed that branched PFOS isomers was preferentially eliminated via the urine whereas linear PFOS was mainly eliminated via the feces (Zhong et al., 2019), suggesting potential different elimination pathway exposure to branch PFOS isomers. In addition, using embryonic chicken hepatocytes, O'Brien et al. (2011) indicated that branched PFOS isomers mixture produced a higher transcriptional response in genes than linear PFOS, including the downregulation of several antioxidant genes involved in oxidative stress pathway (e.g. glutathione-S-transferases genes). More experimental research to explore the possible different biological mechanisms of branched PFOS isomers are required in the further study. Our study has several important limitations. As noted above, the possibility of reverse causality is a major limitation of our study, and so will require confirmation in a future, longitudinal study. Additionally, we were unable to adjust for some important variables such as dietary factors and serum albumin (Han et al., 2003;

1048

J. Wang et al. / Chemosphere 218 (2019) 1042e1049

White et al., 2011), which may have biased our results. Finally, the moderate to strong intercorrelations among the measured PFASs precluded a simultaneous analysis of the impact of the PFAS ‘mixture’ in a single model. More elegant statistical approaches that accommodate multiple highly correlated predictors will be necessary to explore these effects in a future investigation. Despite the limitations, our large sample size allowed for ample statistical power to detect modest associations with a high degree of accuracy. We also collected comprehensive covariate data which allow us to consider multiple sources of confounding in the analysis. In addition, we also considered a more extensive panel of PFASs and structural isomers of PFOS and PFOA than previous studies related to renal function. Finally, our results were robust to several sensitivity analyses, in which we evaluated the impact of potential biasing factors. 5. Conclusion In conclusion, the current study provides novel findings to suggest that exposure to branched PFOS isomers may be in association with a decline renal function and higher risk of CKD. To the best of our knowledge, this is the first study to report a link among isomers of PFOS and PFOA with renal function and CKD. However, given the limitations of our study design, prospective studies are further required to confirm these association and to clarify the relevant biological mechanism. Competing financial interests The authors declare they have no actual or potential competing interests. Acknowledgments: This study was funded by Grants from the National Natural Science Foundation of China (No.81872582, 81472936, 81673127, and 81001255), the National Key Research and Development Program of China (2016YFC0207000), the Fundamental Research Funds for the Central Universities (No. 17ykpy16, and 17ykzd14), and Science and Technology Planning Project of Guangdong Province (No. 2016A030313342, 2017A050501062, 2018B05052007). The views expressed in this manuscript belong the authors only. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.chemosphere.2018.11.191. References AMAP, 2009. Arctic Pollution 2009. Arctic Monitoring and Assessment Programme, Oslo (xi þ 83 pp). Anavekar, N.S., Pfeffer, M.A., 2004. Cardiovascular risk in chronic kidney disease. Kidney Int. Suppl. S11eS15. ATSDR, 2015. Draft Toxicological Profile for Perfluoroalkyls. Available at: https:// www.atsdr.cdc.gov/toxprofiles/tp200.pdf(2015). Bangma, J.T., Ragland, J.M., et al., 2019. Perfluoroalkyl substances in diamondback terrapins (Malaclemys terrapin) in coastal South Carolina. Chemosphere 215, 305e312. Bao, W.W., Qian, Z.M., et al., 2017. Gender-specific associations between serum isomers of perfluoroalkyl substances and blood pressure among Chinese: isomers of C8 Health Project in China. Sci. Total Environ. 607e608, 1304e1312. Benskin, J.P., Bataineh, M., et al., 2007. Simultaneous characterization of perfluoroalkyl carboxylate, sulfonate, and sulfonamide isomers by liquid chromatography-tandem mass spectrometry. Anal. Chem. 79, 6455e6464. Benskin, J.P., De Silva, A.O., et al., 2009a. Disposition of perfluorinated acid isomers in Sprague-Dawley rats; part 1: single dose. Environ. Toxicol. Chem. 28, 542e554. Benskin, J.P., Holt, A., et al., 2009b. Isomer-specific biotransformation rates of a

perfluorooctane sulfonate (PFOS)-precursor by cytochrome P450 isozymes and human liver microsomes. Environ. Sci. Technol. 43, 8566e8572. Benskin, J.P., De Silva, A.O., et al., 2010. Isomer profiling of perfluorinated substances as a tool for source tracking: a review of early findings and future applications. Rev. Environ. Contam. Toxicol. 208, 111e160. Buck, R.C., Franklin, J., et al., 2011. Perfluoroalkyl and polyfluoroalkyl substances in the environment: terminology, classification, and origins. Integrated Environ. Assess. Manag. 7, 513e541. Choi, J., Knudsen, L.E., et al., 2017. Identification of exposure to environmental chemicals in children and older adults using human biomonitoring data sorted by age: results from a literature review. Int. J. Hyg Environ. Health 220, 282e298. Chou, H.C., Wen, L.L., et al., 2017. From the cover: l-carnitine via PPARgamma- and sirt1-dependent mechanisms attenuates epithelial-mesenchymal transition and renal fibrosis caused by perfluorooctanesulfonate. Toxicol. Sci. 160, 217e229. Consoer, D.M., Hoffman, A.D., et al., 2016. Toxicokinetics of perfluorooctane sulfonate in rainbow trout (Oncorhynchus mykiss). Environ. Toxicol. Chem. 35, 717e727. De Silva, A.O., Benskin, J.P., et al., 2009. Disposition of perfluorinated acid isomers in Sprague-Dawley rats; part 2: subchronic dose. Environ. Toxicol. Chem. 28, 555e567. DeWitt, J.C. (Ed.), 2015. Toxicological Effects of Perfluoroalkyl and Polyfluoroalkyl Substances. Humana Press Heidelberg, New York, Dordrecht, London. Dhingra, R., Lally, C., et al., 2016. Perfluorooctanoic acid and chronic kidney disease: longitudinal analysis of a Mid-Ohio Valley community. Environ. Res. 145, 85e92. Dhingra, R., Winquist, A., et al., 2017. A study of reverse causation: examining the associations of perfluorooctanoic acid serum levels with two outcomes. Environ. Health Perspect. 125, 416e421. Fang, S., Zhang, Y., et al., 2016. Bioaccumulation of perfluoroalkyl acids including the isomers of perfluorooctane sulfonate in carp (Cyprinus carpio) in a sediment/ water microcosm. Environ. Toxicol. Chem. 35, 3005e3013. Gansevoort, R.T., Correa-Rotter, R., et al., 2013. Chronic kidney disease and cardiovascular risk: epidemiology, mechanisms, and prevention. Lancet (N. Am. Ed.) 382, 339e352. Guelfo, J.L., Adamson, D.T., 2018. Evaluation of a national data set for insights into sources, composition, and concentrations of per- and polyfluoroalkyl substances (PFASs) in U.S. drinking water. Environ. Pollut. 236, 505e513. Han, X., Snow, T.A., et al., 2003. Binding of perfluorooctanoic acid to rat and human plasma proteins. Chem. Res. Toxicol. 16, 775e781. Hornung, R.W., Reed, L.D., 1990. Estimation of average concentration in the presence of nondetectable values. Appl. Occup. Environ. Hyg 5 (1), 46e51. Hill, N.R., Fatoba, S.T., et al., 2016. Global prevalence of chronic kidney disease - a systematic review and meta-analysis. PloS One 11, e158765. Kataria, A., Trachtman, H., et al., 2015. Association between perfluoroalkyl acids and kidney function in a cross-sectional study of adolescents. Environ. Health 14, 89. KEMI (Swedish Chemicals Agency), 2015. Occurrence and Use of Highly Fluorinated Substances and Alternatives. Kuklenyik, Z., Reich, J.A., et al., 2004. Automated solid-phase extraction and measurement of perfluorinated organic acids and amides in human serum and milk. Environ. Sci. Technol. 38, 3698e3704. Lam, J.C., Lyu, J., et al., 2016. Perfluoroalkyl substances (PFASs) in marine mammals from the south China sea and their temporal changes 2002-2014: concern for alternatives of PFOS? Environ. Sci. Technol. 50, 6728e6736. Lau, C., 2012. Perfluorinated compounds. EXS 101, 47e86. Lau, C., Anitole, K., et al., 2007. Perfluoroalkyl acids: a review of monitoring and toxicological findings. Toxicol. Sci. 99, 366e394. Levey, A.S., Eckardt, K.U., et al., 2005. Definition and classification of chronic kidney disease: a position statement from Kidney Disease: improving Global Outcomes (KDIGO). Kidney Int. 67, 2089e2100. Li, M., Zeng, X.W., et al., 2017. Isomers of perfluorooctanesulfonate (PFOS) in cord serum and birth outcomes in China: Guangzhou Birth Cohort Study. Environ. Int. 102, 1e8. Liu, J., Zhong, G., et al., 2019. Isomer-specific biotransformation of perfluoroalkyl sulfonamide compounds in aerobic soil. Sci. Total Environ. 651, 766e774. Liu, R.Q., Qian, Z., et al., 2016. Poor sleep quality associated with high risk of hypertension and elevated blood pressure in China: results from a large population-based study. Hypertens. Res. 39, 54e59. Liu, H.S., Wen, L.L., et al., 2018. Association among total serum isomers of perfluorinated chemicals, glucose homeostasis, lipid profiles, serum protein and metabolic syndrome in adults: NHANES, 2013-2014. Environ. Pollut. 232, 73e79. Lu, Z., Lu, R., et al., 2018. Correction to: risk exposure assessment of per- and polyfluoroalkyl substances (PFASs) in drinking water and atmosphere in central eastern China. Environ. Sci. Pollut. Res. Int. 25, 9321. Markowitz, G.S., Bomback, A.S., et al., 2015. Drug-induced glomerular disease: direct cellular injury. Clin. J. Am. Soc. Nephrol. 10, 1291e1299. Matsuo, S., Imai, E., et al., 2009. Revised equations for estimated GFR from serum creatinine in Japan. Am. J. Kidney Dis. 53, 982e992. Miralles-Marco, A., Harrad, S., 2015. Perfluorooctane sulfonate: a review of human exposure, biomonitoring and the environmental forensics utility of its chirality and isomer distribution. Environ. Int. 77, 148e159. National Kidney Foundation, 2002. K/DOQI clinical practice guidelines for chronic

J. Wang et al. / Chemosphere 218 (2019) 1042e1049 kidney disease: evaluation, classification, and stratification. Am. J. Kidney Dis. 39 (2 Suppl. 1), S1eS266. Nugent, R.A., Fathima, S.F., et al., 2011. The burden of chronic kidney disease on developing nations: a 21st century challenge in global health. Nephron Clin. Pract. 118, c269ec277. O'Brien, J.M., Austin, A.J., et al., 2011. Technical-grade perfluorooctane sulfonate alters the expression of more transcripts in cultured chicken embryonic hepatocytes than linear perfluorooctane sulfonate. Environ. Toxicol. Chem. 30, 2846e2859. Olsen, G.W., Burris, J.M., et al., 2007. Half-life of serum elimination of perfluorooctanesulfonate,perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers. Environ. Health Perspect. 115, 1298e1305. Olsen, G.W., Mair, D.C., et al., 2017. Per- and polyfluoroalkyl substances (PFAS) in American Red Cross adult blood donors, 2000-2015. Environ. Res. 157, 87e95. Popovic, M., Zaja, R., et al., 2014. Interaction of environmental contaminants with zebrafish organic anion transporting polypeptide, Oatp1d1 (Slco1d1). Toxicol. Appl. Pharmacol. 280, 149e158. Rosenmai, A.K., Ahrens, L., et al., 2018. Relationship between peroxisome proliferator-activated receptor alpha activity and cellular concentration of 14 perfluoroalkyl substances in HepG2 cells. J. Appl. Toxicol. 38, 219e226. Saydah, S., et al., 2007. Prevalence of chronic kidney disease and associated risk factors d United States, 1999e2004. MMWR (Morb. Mortal. Wkly. Rep.) 56, 161e165. Shankar, A., Xiao, J., et al., 2011. Perfluoroalkyl chemicals and chronic kidney disease in US adults. Am. J. Epidemiol. 174, 893e900. Sharpe, R.L., Benskin, J.P., et al., 2010. Perfluorooctane sulfonate toxicity, isomerspecific accumulation, and maternal transfer in zebrafish (Danio rerio) and rainbow trout (Oncorhynchus mykiss). Environ. Toxicol. Chem. 29, 1957e1966. Stanifer, J.W., Stapleton, H.M., et al., 2018. Perfluorinated chemicals as emerging environmental threats to kidney health: a scoping review. Clin. J. Am. Soc. Nephrol. 13, 1479e1492. Steenland, K., Woskie, S., 2012. Cohort mortality study of workers exposed to perfluorooctanoic acid. Am. J. Epidemiol. 176, 909e917. Steenland, K., Zhao, L., et al., 2015. A cohort incidence study of workers exposed to perfluorooctanoic acid (PFOA). Occup. Environ. Med. 72, 373e380. Stubleski, J., Salihovic, S., et al., 2016. Changes in serum levels of perfluoroalkyl substances during a 10-year follow-up period in a large population-based cohort. Environ. Int. 95, 86e92. Thawornchaisit, P., de Looze, F., et al., 2015. Health-risk factors and the prevalence of chronic kidney disease: cross-sectional findings from a national cohort of 87,143 Thai open university students. Global J. Health Sci. 7, 59e72. Wang, Z., Cousins, I.T., et al., 2014. Global emission inventories for C4-C14 perfluoroalkyl carboxylic acid (PFCA) homologues from 1951 to 2030, Part I:

1049

production and emissions from quantifiable sources. Environ. Int. 70, 62e75. Wang, Z., DeWitt, J.C., et al., 2017. A never-ending story of per- and polyfluoroalkyl substances (PFASs)? Environ. Sci. Technol. 51, 2508e2518. Watkins, D.J., Josson, J., et al., 2013. Exposure to perfluoroalkyl acids and markers of kidney function among children and adolescents living near a chemical plant. Environ. Health Perspect. 121, 625e630. Webster, A.C., Nagler, E.V., et al., 2017. Chronic kidney disease. Lancet (N. Am. Ed.) 389, 1238e1252. Wen, L.L., Lin, C.Y., et al., 2016. Perfluorooctanesulfonate mediates renal tubular cell apoptosis through PPARgamma inactivation. PloS One 11, e155190. White, S.S., Fenton, S.E., et al., 2011. Endocrine disrupting properties of perfluorooctanoic acid. J. Steroid Biochem. Mol. Biol. 127, 16e26. World Health Organization, 1995. Physical Status: the Use and Interpretation of Anthropometry. WHO, Geneva, Switzerland. Xia, J., Wang, L., et al., 2017. Cigarette smoking and chronic kidney disease in the general population: a systematic review and meta-analysis of prospective cohort studies. Nephrol. Dial. Transplant. 32, 475e487. Xie, S., Lu, Y., et al., 2013. Estimation of PFOS emission from domestic sources in the eastern coastal region of China. Environ. Int. 59, 336e343. Yeung, L.W., Miyake, Y., et al., 2008. Perfluorinated compounds and total and extractable organic fluorine in human blood samples from China. Environ. Sci. Technol. 42, 8140e8145. Yeung, L.W., So, M.K., et al., 2006. Perfluorooctanesulfonate and related fluorochemicals in human blood samples from China. Environ. Sci. Technol. 40, 715e720. Yu, N., Wang, X., et al., 2015. Distribution of perfluorooctane sulfonate isomers and predicted risk of thyroid hormonal perturbation in drinking water. Water Res. 76, 171e180. Zeng, X.W., Qian, Z., et al., 2015. Human serum levels of perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA) in Uyghurs from Sinkiang-Uighur Autonomous Region, China: background levels study. Environ. Sci. Pollut. Res. Int. 22, 4736e4746. Zhang, L., Wang, F., et al., 2012. Prevalence of chronic kidney disease in China: a cross-sectional survey. Lancet (N. Am. Ed.) 379, 815e822. Zhang, L., Ren, X.M., et al., 2013. Structure-based investigation on the interaction of perfluorinated compounds with human liver fatty acid binding protein. Environ. Sci. Technol. 47, 11293e11301. Zhao, W., Zitzow, J.D., et al., 2017. Organic anion transporting polypeptides contribute to the disposition of perfluoroalkyl acids in humans and rats. Toxicol. Sci. 156, 84e95. Zhong, W., Zhang, L., et al., 2019. Probing mechanisms for bioaccumulation of perfluoroalkyl acids in carp (Cyprinus carpio): impacts of protein binding affinities and elimination pathways. Sci. Total Environ. 647, 992e999.