Science of the Total Environment 512–513 (2015) 364–370

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Association of polyfluoroalkyl chemical exposure with serum lipids in children Xiao-Wen Zeng a, Zhengmin Qian b, Brett Emo c, Michael Vaughn d, Jia Bao e, Xiao-Di Qin a, Yu Zhu a, Jie Li a, Yungling Leo Lee f,⁎, Guang-Hui Dong a,⁎⁎ a

Department of Preventive Medicine, School of Public Health, Sun Yat-sen University, Guangzhou 510080, China Department of Epidemiology, College of Public Health and Social Justice, Saint Louis University, Saint Louis, MO 63104, USA Department of Environmental and Occupational Health, College for Public Health and Social Justice, Saint Louis University, Saint Louis, MO 63104, USA d School of Social Work, College for Public Health and Social Justice, Saint Louis University, Saint Louis, MO 63104, USA e School of Environmental Science, Shenyang University of Technology, Shenyang 110870, China f Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, Taipei 100, Taiwan b c

H I G H L I G H T S • The association between serum PFASs and lipids was evaluated in Taiwanese children. • Eight out of ten particular PFAS chemicals were detected in most participants (N 94%). • PFOS, PFOA and PFNA were positively associated with total cholesterol, LDL and TG.

a r t i c l e

i n f o

Article history: Received 16 September 2014 Received in revised form 16 January 2015 Accepted 17 January 2015 Available online 30 January 2015 Editor: Adrian Covaci Keywords: Polyfluoroalkyl substances Polyfluoroalkyl compounds Serum lipids Children

a b s t r a c t Perfluoroalkyl and polyfluoroalkyl substances (PFASs), as well as polymers of PFASs, have been widely used in commercial applications and have been detected in humans and the environment. Previous epidemiological studies have shown associations between particular PFAS chemicals and serum lipid concentrations in adults, particularly perfluorooctane sulfonic acid (PFOS) and perfluorooctanoic acid (PFOA). There exists, however, limited information concerning the effect of PFASs have on serum lipids among children. In the present crosssectional study, 225 Taiwanese children (12–15 years of age) were recruited to determine the relationship between serum level PFASs and lipid concentration. Results showed that eight out of ten particular PFASs were detected in the serum of N 94% of the participants. Serum PFOS and perfluorotetradecanoic acid (PFTA) levels were at an order of magnitude higher than the other PFASs, with arithmetical means of 32.4 and 30.7 ng/ml in boys and 34.2 and 27.4 ng/ml in girls, respectively. However, the variation in serum PFTA concentration was quite large. Following covariate adjustment, linear regression models revealed that PFOS, PFOA, and perfluorononanoic acid (PFNA) were positively associated with total cholesterol (TC), low-density lipoprotein (LDL) and triglycerides (TG), particularly for PFOS and PFTA. Quartile analysis, with the lowest exposure quartile as a reference, yielded associations between serum PFTA and elevations in TC (p = 0.002) and LDL (p = 0.004). Though not statistically significant, high-density lipoprotein (HDL) appeared to decrease linearly across quartiles for PFOS and PFOA exposure. In conclusion, a significant association was observed between serum PFASs and lipid level in Taiwanese children. These findings for PFTA are novel, and emphasize the need to investigate the exposure route and toxicological evidence of PFASs beyond PFOS and PFOA. © 2015 Elsevier B.V. All rights reserved.

1. Introduction ⁎ Correspondence to: Y.L. Lee, Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, No. 17 Xuzhou Road, Taipei 100, Taiwan. ⁎⁎ Correspondence to: G.-H. Dong, Department of Preventive Medicine, School of Public Health, Sun Yat-sen University, 74 Zhongshan 2nd Road, Yuexiu District, Guangzhou 510080, PR China. E-mail addresses: [email protected] (Y.L. Lee), [email protected] (G.-H. Dong).

http://dx.doi.org/10.1016/j.scitotenv.2015.01.042 0048-9697/© 2015 Elsevier B.V. All rights reserved.

Perfluoroalkyl and polyfluoroalkyl substances (PFASs) are a diverse class of compounds that have an aliphatic carbon backbone in which the hydrogen atoms have substituted with fluorine. These perfluorinated chemicals include perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA), and perfluorinated sulfonates such as perfluorooctane sulfonic acid (PFOS) and perfluorohexane sulfonic acid (PFHxS). PFASs and

X.-W. Zeng et al. / Science of the Total Environment 512–513 (2015) 364–370

polymers made with the aid of PFASs have been widely used in manufactured products and are ubiquitously present in both the environment and the human body (Buck et al., 2011; Kissa, 2001). Uptake of PFASs occurs via direct exposure through diet, drinking water, and inhalation of household dust, or via indirect exposure such as PFAS contamination resulting from contact with food packaging (D'eon and Mabury, 2011). Elimination of PFASs occurs slowly without biotransformation. The estimated geometric mean half-lives of PFOA, PFOS, and PFHxS in human serum have been reported to be 3.5 years (95% CI, 3.0–4.1), 4.8 years (95% CI, 4.0–5.8), and 7.3 years (95% CI, 5.8–9.2), respectively (Olsen et al., 2007a). Two typical PFASs, PFOA and PFOS, belong to the 8-carbon backbone subgroup and have been studied more extensively than other PFASs. By virtue of its ability to bioaccumulate, its potential for long-range environmental transport, and its chemical stability and toxicity, PFOS and its salts have been added to Annex B of the Stockholm Convention of persistent organic pollutants in 2009. Additionally, PFOA, has been classified as a Group 2B carcinogen by the International Agency for Research on Cancer (IARC, 2012). Animal studies have demonstrated that high level exposure to PFOS and PFOA could induce adverse outcomes, including hepatotoxicity (Qazi et al., 2010), developmental neurotoxicity (Mariussen, 2012; Zeng et al., 2011), and carcinogenicity (Biegel et al., 2001; Butenhoff et al., 2012a,b). Although the specific mechanism of action for PFASs remains to be elucidated, peroxisome proliferator-activated receptor-α (PPARα) has been considered to play an important factor in mediating PFAS effects (Abbott et al., 2007; Wolf et al., 2012). PPARα falls within a class of ligand-activated transcription factors of the steroid/thyroid nuclear hormone receptor superfamily, and plays a role in lipid homeostasis, energy metabolism, and cell differentiation (Peraza et al., 2006). PFASs could activate PPARα with varying potencies (Buhrke et al., 2013; Wolf et al., 2008). Several studies have shown that PFOA could induce hypolipidemic effects in rodents (Loveless et al., 2006; Wang et al., 2013). However, the majority of human epidemiological studies have not reported complimentary hypolipidemic results. Epidemiological studies have suggested that exposure to PFOA and PFOS may be associated with serum lipid levels in PFAS-expose populations, though the findings are inconsistent. Positive associations have been found in adult populations exposed to relatively high levels of PFASs. An association between serum PFOA and total cholesterol was observed in two occupational populations (Costa et al., 2009; Sakr et al., 2007) and a community in a PFOA-contaminated water district (Steenland et al., 2009). However, no associations were reported in a third occupational population (Olsen and Zobel, 2007) and a separate PFOA-contaminated community cohort (Emmett et al., 2006). The association of lipids and PFOS is similar in magnitude to those with PFOA. Cross-sectional studies of the general population have found positive serum lipid associations with PFASs. Analysis of the National Health and Nutrition Examination Survey (NHANES) of the American general population noted a positive association between PFOS and PFOA and cholesterol (Nelson et al., 2010). A similar analysis of the Canadian Health Measures Survey (CHMS), Cycle 1, found only weak associations between PFOS and PFOA and serum lipids, but did observe significant associations with the PFHxS and cholesterol outcomes (Fisher et al., 2013). These observed inconsistencies in associations between PFASs and lipids may be due to differences in age and gender distributions of the participants across studies, differences in size of the study populations, or variation in PFAS exposure across study populations (Starling et al., 2014). Despite several investigations of PFASs' impact on adults, there have been relatively few studies focused on the association between PFAS exposure and lipid concentrations in children and adolescents. Studies by Frisbee et al. (2010) and Geiger et al. (2014) have found that PFOA and PFOS were significantly associated with increased total cholesterol, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) in children. However, the association between PFASs and lipids has been inconsistent. While Frisbee et al. (2010) reported a positive association

365

between PFOS and HDL in US adolescents, the inverse relationship was observed in Taiwanese adolescents (Lin et al., 2011). In addition, attention to the potential effects of exposure to other PFASs (such as PFHxS or PFNA) should not be neglected given that these have been used as alternatives to PFOS in industrial applications following the phase-out of PFOS in 2000–2002 (Beesoon et al., 2012). The renal elimination rate of various PFASs may be different, depending upon the chain length of the organic anion transport proteins involved in renal reabsorption (Han et al., 2011). Thus, the PFASs with longer carbon chain (such as PFOA) are considered to be more likely to bioaccumulate in mammals than their shorter-chained analogs (Haug et al., 2009). Previous studies in Taiwan have shown that the serum PFOS concentration in children residing in the Taipei area increased three-fold between the time periods of 2006–2008 (Lin et al., 2011) to 2009–2010 (Bao et al., 2014), despite declining serum PFAS concentration occurred globally (Fitz-Simon et al., 2013; Kato et al., 2011; Toms et al., 2014). Contrary to the concentration of PFOS in serum, PFOA concentration is much lower in Taiwanese children compared to other countries or areas (Bao et al., 2014). It is also generally accepted that individuals are more sensitive to the effects of xenobiotic exposure during periods of development, such as childhood. Furthermore, studying the potential health consequences of an environmental exposure on children rather than adults might provide unique insights because the number of underlying factors confounding the associations is likely to be smaller (Lin et al., 2011). Given that studies focusing on the association between serum PFASs and lipids in children are limited, understanding their relationship in different exposure conditions could provide further insight. Accordingly, the objective of this study was to evaluate the cross-sectional association between measures of PFAS exposure and total cholesterol, HDL, LDL, and triglycerides in Taiwanese children. The present study sample was obtained as part of a community-based child population survey in Taiwan (Tsai et al., 2010). 2. Material and methods 2.1. Study participants The study subjects were from the control group of the Genetic and Biomarkers study for Childhood Asthma (GBCA) in Taiwan. The control cohort was selected from seven public schools in the Taipei area from 2009 to 2010 (Tsai et al., 2010). These schools had diverse geographical and socioeconomic characteristics, being located in city, rural, and highaltitude communities. In each targeted school, children of the same age range and without a personal or family history of asthma were invited to participate. A total of 225 healthy children were enrolled, including 102 boys and 124 girls aged from 12 to 15 years (the response rate was 72% among those contacted by phone). A survey was used to acquire information regarding demographic variables and environmental exposure. Information was collected about the current and past smoking status of each participant's adult household members and regular household visitors. All children and their parents provided written informed consent. The study protocol was approved by the Institutional Review Board (National Taiwan University Hospital Research Ethics Committee). 2.2. Serum lipid determination The main outcome of interest was serum lipid levels. Serum was separated from red blood cells, placed in transport tubes, and refrigerated before being shipped to the analytical laboratory. Four lipid levels were measured enzymatically: total cholesterol (TC), high-density lipoprotein (HDL), low-density lipoprotein (LDL), and triglyceride (TG), and units were recorded in mg/dl. LDL was calculated using the Friedewald formula for participants when TGs were lower than

366

X.-W. Zeng et al. / Science of the Total Environment 512–513 (2015) 364–370

400 mg/dl (Friedewald et al., 1972). None of the study participants had a TG higher than 400 mg/dl in our study. 2.3. Serum PFAS measurement Details of the analytical procedure for measuring the ten PFAS chemicals, consisting of PFOS, PFOA, PFHxS, perfluorobutane sulfonate (PFBS), perfluorohexane acid (PFHxA), perfluoroheptanoic acid (PFHpA), perfluorodecanoic acid (PFDA), perfluorononanoic acid (PFNA), perfluorododecanoic acid (PFDoA) and perfluorotetradecanoic acid (PFTA) in serum samples have been described previously (Bao et al., 2010). Briefly, 0.25 M Na2CO3 and 0.5 M tetrabutylammonium hydroxide (TBAHs) were added to 0.5 ml serum and then extracted by 5 ml of methyl tert-butyl ether (MTBE) twice. The extracts were dried by N2 and resolved in methanol and 10 mM ammonium acetate (2:3, v:v), followed by filtration using a 0.22 μm nylon filter. A 25 μl aliquot of extract was analyzed by high-performance liquid chromatography coupled with triple quadrupole mass spectrometry (HPLC–MS/MS, Agilent 6410, USA). The limit of quantification (LOQ) was 0.03 ng/ml for PFOS, PFOA, and PFNA, 0.07 ng/ml for PFBS and PFHxS; 0.1 ng/ml for PFDA and PFDoA, 0.05 ng/ml for PFHxA, and 0.03 ng/ml for PFTA. Since nearly one half of the detected serum PFHpA level was below the LOQ, PFHpA was not included in the statistical analyses. Detailed information about standards and reagents, instrumental analysis, quality assurance and quality control, and recovery experiments in the present study is provided in the Supplemental material. 2.4. Statistical analysis Statistical analyses were performed using SAS software (version 9.2, SAS Institute Inc., Cary, NC, USA). All PFAS levels were natural log transformed to correct skewed distributions. PFAS concentrations were treated in two ways for the purposes of analyses: (1) as natural-log transformed continuous variables to determine the association between cholesterol and the PFAS concentration by linear regression analysis, and (2) as quartiles (with the lowest PFAS quartile as a reference group), to examine differences in serum cholesterol levels with increasing quartiles of PFAS exposure. Covariates including age, gender, body mass index (BMI), regular exercise (yes/no), parental education (less than high school, more than high school), and environmental tobacco smoke (ETS) exposure were chosen because of their established relationship with one or more lipids (independent of whether they were associated with PFASs). Each lipid parameter (TC, HDL, LDL, and TG) was treated as a continuous outcome variable in a separate model with a single PFAS exposure variable. An ordinal variable was modeled and assigned the median value for each corresponding quartile to estimate p-values for any identified trends. A p-value of b0.05 was considered statistically significant. 3. Results and discussion 3.1. Measurement of PFASs and lipids in serum The general characteristics of children who participated in the study are reported in Table 1 (n = 225). The mean age was 13.6 years. Participation was similar across gender. There was no significant difference in covariates across gender except for height, weight and regularity of exercise (p b 0.05, Table 1). Eight of the ten PFASs were detected in more than 94% of the serum specimens, the two exceptions being PFDoA (84.4%) and PFHpA (53.3%) (Dong et al., 2013). Serum PFOS and PFTA levels were an order of magnitude higher than the other PFASs, with a mean of 32.4 ng/ml and 30.7 ng/ml in boys, and a mean of 34.2 ng/ml and 27.4 ng/ml in girls (Table 1). The serum concentration of PFOS in this study is much higher than those reported in U.S. general children and adolescent' samples, while PFOA concentrations were lower than the reported serum level from the general population

Table 1 Participant characteristics. Characteristic

Boys (n = 102)

Girls (n = 123)

p-Value

Age (years) Height (cm) Weight (kg) BMI (kg/m2) Regular exercisea Yes No Parental educationa bHigh school ≥High school ETS exposurea No Ever Current PFOS (ng/ml)

13.6 ± 0.7 163.1 ± 7.2 55.9 ± 15.4 20.8 ± 4.7

13.6 ± 0.8 157.1 ± 5.5 49.6 ± 10.3 20.0 ± 3.6

0.863 b0.001 b0.001 0.139

17 (16.7) 85 (83.3)

36 (29.3) 87 (70.7)

0.027

42 (41.2) 60 (58.8)

44 (35.8) 79 (64.2)

0.406

42 (41.2) 13 (12.8) 47 (46.1) 32.4 ± 25.5 (28.8, LOQ–148.1) 1.1 ± 1.4 (0.5, LOQ–3.9)

51 (41.5) 9 (7.3) 63 (51.2) 34.2 ± 27.1 (29.9, LOQ–125.9) 0.92 ± 0.79 (0.5, LOQ–11.3) 0.4 ± 0.1 (0.5, LOQ–2.7)

0.371

1.0 ± 0.5 (1.0, LOQ–4.2) 4.4 ± 5.0 (2.4, LOQ–43.1) 0.2 ± 0.2 (0.2, LOQ–2.4) 2.1 ± 2.1 (1.4, LOQ–11.8) 0.9 ± 0.4 (0.8, 0.3–2.0) 27.4 ± 85.3 (6.0, LOQ–615.9) 156.6 ± 26.3 84 (68.3) 39 (31.7) 57.9 ± 12.3 8 (6.5) 115 (93.5) 80.7 ± 19.9 110 (89.4) 13 (10.6) 78.4 ± 30.6 119 (96.8) 4 (3.3) 2.8 ± 0.6

0.394 0.894

PFOA (ng/ml) PFBS (ng/ml) PFDA (ng/ml) PFDOA (ng/ml)

0.5 ± 0.3 (0.5, LOQ–0.81) 1.0 ± 0.5 (1.0, LOQ–5.0) 4.5 ± 7.0 (3.1, LOQ–2.2)

PFHxA (ng/ml) PFHxS (ng/ml)

0.2 ± 0.3 (0.2, LOQ–2.2) 2.1 ± 2.2 (1.2, 0.2–10.3)

PFNA (ng/ml) PFTA (ng/ml)

0.8 ± 0.3 (0.9, 0.3–2.5) 30.7 ± 77.2 (4.5, LOQ–793.6) 153.7 ± 29.1 77 (75.5) 25 (24.5) 53.2 ± 13.3 14 (13.7) 88 (86.3) 81.0 ± 23.3 92 (90.2) 10 (9.8) 86.8 ± 48.0 93 (91.2) 9 (8.8) 3.0 ± 0.9

TC (mg/dl) b170a ≥170a HDL (mg/dl) b40a ≥40a LDL (mg/dl) b110a ≥110a TG (mg/dl) b150a ≥150a TC/HDL

0.619 0.247 0.469

0.911 0.792 0.232 0.763 0.435 0.234 0.007 0.069 0.924 0.850 0.114 0.075 0.022

Values are mean ± SD. PFASs (ng/ml): mean ± SD (median, min–max). a Values are presented as numbers (%). Percentage may not total 100 because of rounding.

(Frisbee et al., 2010; Geiger et al., 2014; Schecter et al., 2012). The serum PFOS concentration was also higher than previous sampling in Taipei children (2006–2008), but the PFOA level was similar (Lin et al., 2011). Interestingly, the serum PFTA levels were much higher in this study of the Taipei area compared to what has been found regionally, suggesting a local environmental source of exposure. However, the observed variations in PFTA levels were large, ranging from LOQ to 793.6 ng/ml in boys, and from LOQ to 613.9 ng/ml in girls, respectively (Table 1). Previous studies (Lin et al., 2009a, 2014) have neglected to measure environmental PFAS exposure in the Taipei area, and long-chain PFASs in general, primarily due to the relatively high lower limit of detection and relatively few known sources compared to other homologues (Wang et al, 2014). Two likely candidates contributing to the observed high level of PFTA in serum are dietary intake and environmental exposure. Dietary intake is the known exposure pathway for the long-chain PFASs. A recent Swedish study indicated the level of PFASs with carbon chain length from C11 to C14 in fish increased substantially from 1999 to 2010 (Vestergren et al., 2012). Because of the increasing trend of fish consumption in Taiwan over the last decade (Pan et al., 2011), dietary intake may contribute to elevated serum PFTA levels. Environmental exposure may also contribute to increased serum PFTA levels. In the recent review, Wang et al. (2014) speculated that the environmental

X.-W. Zeng et al. / Science of the Total Environment 512–513 (2015) 364–370

C14-PFAS exposure might be due to the degradation of fluorotelomerbased products (indirect source) or the impurities in perfluorooctane sulfonyl fluoride (POSF)-based products (direct source). As Western countries announced an elimination of long-chain PFASs and their precursors, industrial production was on the rise in China, resulting in the production of fluorotelomer-based substances (Zhang et al., 2012; Ruan et al, 2010). However, because only a few PFAS chemicals were monitored in Taiwan, little relevant information is available. Therefore, the source of PTFA exposure should elicit concern and needs to be investigated further. The values classified as acceptable or ideal in TC (b170 mg/dl), HDL (≥ 40 mg/dl), LDL (b110 mg/dl), and fasting TG (≤ 150 mg/dl) are 75.5%, 86.3%, 90.2% and 91.2% for boys, and 68.3%, 93.5%, 89.4% and 96.8% for girls, respectively. The acceptable portion of lipid levels was higher than the 12–19-year-old adolescents from the C8 Health Project (Frisbee et al., 2010) and is similar to the 12–18 year old population from the NHAES (Geiger et al., 2014). 3.2. The association between lipids and PFASs Table 2 displays the linear regression coefficient between serum cholesterol and all PFASs. Fig. 1 exhibits the quartile analysis of selected PFASs that might have association with lipids. Results from linear regression analysis demonstrate that after adjustment for covariables, TC, LDL and TG were linearly and positively associated with PFOS and PFOA (p b 0.01) for all models, with the exception of TG and PFOS which had a coefficient of 0.05 (Table 2). A one ln-unit increase in PFOS and PFOA was associated with changes of 0.31 (95%, 0.18–0.45) and 6.57 mg/dl (95%, 2.72–10.42) increases in TC, respectively. TC was also linearly and positively associated with PFBS (p = 0.04). Each one unit increase in ln-PFNA increased TC by 12.92 (95%, 0.733–25.10), LDL by 9.63 (95%, 0.20–19.06) and TG by 23.01 (95%, 6.49–39.52) mg/dl. Serum TC level appears to be related to the extent of exposure to PFOS, PFOA, PFNA and PFTAs, with the strongest effects with PFOS and PFTA (Fig. 1A). Children in the highest PFOS quartile had significantly higher TC levels (23.1 mg/dl) than those in the lower quartile (5.5 mg/dl in the second and 12.8 mg/dl in the third quartile). In previous studies, both PFOS and PFOA exposures have been associated with TC in adolescents (Frisbee et al., 2010), and also in the North American general population (Fisher et al., 2013; Nelson et al., 2010). Adults with high PFOA exposure have also been associated with TC (Steenland et al., 2009). The reported odd ratios for

367

high TC among adolescents across quintiles of PFOS and PFOA exposure were 1.6 (95% CI, 1.4–1.9) and 1.2 (95% CI, 1.1–1.4), respectively (Frisbee et al., 2010). However, the association between TC and PFASs has not been consistently observed across previous studies. While Nelson et al. (2010) has reported PFHxS to be negatively associated with TC in the U.S. general population, a positive association between PFHxS and TC was observed in adult Canadians with no significant associations found between PFOA and PFOS and TC in weighted analyses (Fisher et al., 2013). No significant association between PFASs and serum HDL was observed (Table 2). HDL mediates the reverse cholesterol transport, a pathway of cholesterol from peripheral tissue to liver. HDL levels appeared to decrease linearly across the quartiles of PFOS and PFOA exposure, though it was not statistically significant (p = 0.75 and 0.36, respectively) (Fig. 1B). The existing literature on this topic is inconclusive, with some studies reporting a positive association between PFOA and PFOS and higher HDL levels in adolescent girls (Nelson et al., 2010), children (Frisbee et al., 2010), and pregnant women (Eriksen et al., 2013; Starling et al., 2014). A recent study from eight years of data from NHANES has suggested that PFOA and HDL exhibited a statistically significant inverse association when analyzed as continuous variables in adolescents (Geiger et al., 2014). In addition, Lin et al. (2009b) has found that increased serum PFNA had a favorable correlation with HDL (0.67 [0.45–0.99], p b 0.05) in the U.S. adolescent population. Another study examining workers exposed to high amounts of PFOA in China has also reported a significant negative association between PFOA and HDL (−0.33, p = 0.01) (Wang et al., 2012). LDL was significantly associated with serum PFOS, PFOA, and PFNA in the regression model after being adjusted for independent variables (Table 2). Comparing the lowest and highest quartile in the adjusted analysis, an increase in LDL of 20.1 mg/dl for PFOS, 12.9 mg/dl for PFOA, and 10.6 mg/dl for PFTA was observed (Fig. 1C). The associations between PFASs and LDL are consistent with the findings of previous studies in adolescents (Frisbee et al., 2010; Geiger et al., 2014). A longitudinal study of a population exposed to high levels of PFAS chemicals also found positive associations between PFOA and PFOS and LDL in several cross-sectional studies (Fitz-Simon et al., 2013). Results for TG were similar to those for LDL. Meaningful associations were observed between PFOS, PFOA, and PFNA and TG (Table 2), which is consistent with previous reports in adults (Nelson et al., 2010; Steenland et al., 2009). Increasing PFOA and PFOS quartiles was also positively associated with increased TG (Fig. 1D). Conversely, a cross-

Table 2 Change in cholesterol measure (milligrams per deciliter) per microgram per liter increase in PFAS by using multivariate linear regression analysis.

PFOS PFOA PFBS PFDA PFDoA PFHXA PFHxS PFNA PFTA

TC coefficient (95% CI, p-value)

HDL coefficient (95% CI, p-value)

LDL coefficient (95% CI, p-value)

TG coefficient (95% CI, p-value)

0.31 (0.18 to 0.45, b0.001) 6.57 (2.72 to 10.42, 0.001) 19.30 (0.60 to 38.00, 0.04) −1.29 (−9.01 to 6.42, 0.74) 0.28 (−0.34 to 0.91, 0.37) −1.28 (−16.92 to 14.37, 0.87) 1.10 (−0.71 to 2.92, 0.23) 12.92 (0.733 to 25.10, 0.04) 0.04 (−0.01 to 0.08, 0.14)

−0.01 (−0.07 to 0.05, 0.72) −1.56 (−3.20 to 0.08, 0.06) 5.78 (−2.09 to 13.65, 0.15) −1.12 (−4.40 to 2.05, 0.47) 0.05 (−0.21 to 0.32, 0.68) −0.45 (−7.00 to 6.10, 0.89) −0.23 (−0.99 to 0.53, 0.54) −2.35 (−7.49 to 2.79, 0.37) 0.01 (−0.01 to 0.03, 0.40)

0.28 (0.18 to 0.38, b0.001) 4.66 (1.67 to 7.65, 0.002) 11.02 (−3.51 to 25.55, 0.14) −0.56 (−6.53 to 5.41, 0.85) 0.19 (−0.30 to 0.68, 0.44) −1.13 (−13.24 to 10.97, 0.85) 0.99 (−0.41 to 2.39, 0.17) 9.63 (0.20 to 19.06, 0.05) 0.02 (−0.02 to 0.05, 0.32)

0.19 (0 to 0.38, 0.05) 19.63 (14.82 to 24.34, b0.001) −3.10 (−28.87 to 22.37, 0.81) 0.57 (−9.97 to 11.11, 0.92) 0.37 (−0.49 to 1.23, 0.40) 0.03 (−21.33 to 21.39, 1.00) 1.80 (−0.67 to 4.27, 0.15) 23.01 (6.49 to 39.52, 0.007) 0.02 (−0.05 to 0.08, 0.61)

All models are adjusted for age, gender, BMI, parental education level, exercise and ETS exposure. Coefficient represents the change in lipid outcome for each 1 ln-(μg/l) increase in PFAS concentration. Values with p ≤ 0.05 are indicated in bold, underlined text.

368

X.-W. Zeng et al. / Science of the Total Environment 512–513 (2015) 364–370

Fig. 1. Differences in serum cholesterol levels, with increasing quartile of PFAS exposure. Changes in TC (A), in HDL (B), in LDL (C) and in TG (D). Linear model analysis was performed to estimate the association with lipid concentration in PFAS quartiles, with the lowest PFAS quartile used as a reference group. All models are adjusted by age, gender, BMI, parental education level, regular exercise, and ETS exposure. Median PFAS levels (ng/ml) for each quartile are shown below/above the bar. Error bars stand for SEs of the difference between the quartile and the reference group, and p-values for trend are presented.

sectional study in Nunavut, Canada found a negative association with PFOS and TG, while HDL levels were positively related to PFOS (Château-Degat et al., 2010). 3.3. Interpretation of findings The present study is one of the few investigations on the relationship between PFAS exposure and lipid levels in children. As such, the present investigation adds to the sparse worldwide literature on this topic. Most striking were the findings for TC, LDL and TG. These outcomes were positively associated with PFOS, PFOA, and PFTA. Compared with findings in adults, our observations are consistent with the general trend of positive associations between PFOA, PFOS concentration and TC (Nelson et al, 2010; Steenland et al., 2009). However, we observed a greater increase in TC from the lowest to highest quartile compared to previous studies. Steenland et al. (2009) reported a 10–12 mg/dl increase in TC and a 11–12 mg/dl increase in LDL from the lowest to highest decile of serum PFOS, while our findings indicate a 22.1 mg/dl increase in TC and a 19.1 mg/dl increase from the lowest to highest quartile. Results reported in this study are also consistent with those of previous studies which have generally exhibited a trend toward a positive association between TC and LDL and PFOS and PFOA concentrations in adolescents (Frisbee et al., 2010; Geiger et al., 2014) and adults (Steenland et al., 2009). Other PFASs, particular PFBS, PFNA and PFTA, were also associated with changes in certain lipids in our study. The biological mechanism that may lead to an association between PFAS levels and lipid levels in humans is largely unknown. A putative mechanism is that the binding of PFASs to peroxisome proliferation-activated receptor alpha (PPARα) may play a role in the regulation of body lipid and glucose metabolism (Kennedy et al., 2004). Other studies have shown that the varying strengths of PPAα activation is associated with PFASs of different chain lengths (Wolf et al., 2012), suggesting the possibility of alternate mechanisms. However, most epidemiological studies in humans, unlike studies in animals report hypolipidemic results,

possibly because of a shorter half-life in rodents and the potential dependence of some animal toxicity on a peroxisome proliferation mechanism that is likely to be less influential in humans (Steenland et al., 2010). Fletcher et al. (2013) observed serum PFOS and PFOA levels to be associated with alterations in the level of expression of several genes involved in cholesterol transport or metabolism. The hypotheses are partly supported the relationship between PFASs and lipids. Given the cross-sectional nature of the study design, establishing a causal relationship between serum PFAS levels and lipid concentrations is not possible. Non-causal explanations for the findings may include unmeasured confounding or pharmacokinetics (Starling et al., 2014). This study did not consider dietary sources which may influence PFAS exposure and, subsequently, the lipid concentrations. We cannot rule out the possibility that the toxicokinetics of PFASs may behave differently in people with higher cholesterol levels (Fisher et al., 2013). Furthermore, it is possible that both cholesterol and PFASs are correlated with a third unknown substance that increases with lipids, exhibiting ‘reverse causality’ (Steenland et al., 2009). It has also been suggested that the distribution of PFASs into lipoprotein fractions or albumin might be the reason behind a non-casual positive correlation between PFASs and lipids (Olsen et al., 2007). This partitioning, however, is not supported by the study which shows maximally that 9% of PFOS distributes to lipoprotein-containing fractions, and only 1% or less distribution for PFOA (Butenhoff et al., 2012c). Nelson et al. (2010) have reported that the results are substantively the same with and without adjustment for serum albumin. 4. Conclusion In general, eight out of ten PFASs were detected in most of the children (N94%) in our study. PFOS, PFOA, PFBS, PFNA, and PFTA showed an association with lipids in the Taiwanese children of the study. The findings for PFTA are novel. This study highlights the need for additional efforts to evaluate the association of PFASs with lipid levels in prospective studies of children.

X.-W. Zeng et al. / Science of the Total Environment 512–513 (2015) 364–370

Disclosure The authors report no conflicts of interest to declare. The Authors are alone responsible for the content and writing of the paper. Acknowledgment This study was supported by grant no. 81172630 and no. 81472936 from the National Natural Science Foundation of China, grant no. 98-2314-B-002-138-MY3 from the National Science Council in Taiwan, and grant no. 201202264 from the Liaoning Province Science and Technology Foundation. The views expressed in this article are those of the authors and do not necessarily represent those of the funding source. The funding source had no role in the design or analysis of the study publication. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2015.01.042. References Abbott, B.D., Wolf, C.J., Schmid, J.E., Das, K.P., Zehr, R.D., Helfant, L., et al., 2007. Perfluorooctanoic acid induced developmental toxicity in the mouse is dependent on expression of peroxisome proliferator activated receptor-alpha. Toxicol. Sci. 98, 571–581. Bao, J., Liu, W., Liu, L., Jin, Y., Dai, J., Ran, X., et al., 2010. Perfluorinated compounds in the environment and the blood of residents living near fluorochemical plants in Fuxin, China. Environmental Science & Technology 45, 8075–8080. Bao, J., Lee, Y., Chen, P.-C., Jin, Y.-H., Dong, G.-H., 2014. Perfluoroalkyl acids in blood serum samples from children in Taiwan. Environ. Sci. Pollut. Res. 21, 7650–7655. Beesoon, S., Genuis, S.J., Benskin, J.P., Martin, J.W., 2012. Exceptionally high serum concentrations of perfluorohexanesulfonate in a Canadian family are linked to home carpet treatment applications. Environmental Science & Technology 46, 12960–12967. Biegel, L.B., Hurtt, M.E., Frame, S.R., O'Connor, J.C., Cook, J.C., 2001. Mechanisms of extrahepatic tumor induction by peroxisome proliferators in male CD rats. Toxicol. Sci. 60, 44–55. Buck, R.C., Franklin, J., Berger, U., Conder, J.M., Cousins, I.T., de Voogt, P., et al., 2011. Perfluoroalkyl and polyfluoroalkyl substances in the environment: terminology, classification, and origins. Integr. Environ. Assess. Manag. 7, 513–541. Buhrke, T., Kibellus, A., Lampen, A., 2013. In vitro toxicological characterization of perfluorinated carboxylic acids with different carbon chain lengths. Toxicol. Lett. 218, 97–104. Butenhoff, J.L., Chang, S.-C., Olsen, G.W., Thomford, P.J., 2012a. Chronic dietary toxicity and carcinogenicity study with potassium perfluorooctanesulfonate in Sprague Dawley rats. Toxicology 293, 1–15. Butenhoff, J.L., Kennedy Jr., G.L., Chang, S.-C., Olsen, G.W., 2012b. Chronic dietary toxicity and carcinogenicity study with ammonium perfluorooctanoate in Sprague–Dawley rats. Toxicology 298, 1–13. Butenhoff, J.L., Pieterman, E., Ehresman, D.J., Gorman, G.S., Olsen, G.W., Chang, S.-C., et al., 2012c. Distribution of perfluorooctanesulfonate and perfluorooctanoate into human plasma lipoprotein fractions. Toxicol. Lett. 210, 360–365. Château-Degat, M.-L., Pereg, D., Dallaire, R., Ayotte, P., Dery, S., Dewailly, É., 2010. Effects of perfluorooctanesulfonate exposure on plasma lipid levels in the Inuit population of Nunavik (Northern Quebec). Environ. Res. 110, 710–717. Costa, G., Sartori, S., Consonni, D., 2009. Thirty years of medical surveillance in perfluooctanoic acid production workers. J. Occup. Environ. Med. 51, 364–372. http://dx.doi.org/10.1097/JOM.0b013e3181965d80. D'eon, J.C., Mabury, S.A., 2011. Is indirect exposure a significant contributor to the burden of perfluorinated acids observed in humans? Environmental Science & Technology 45, 7974–7984. Dong, G., Tung, K., Tsai, C., Liu, M., Wang, D., Liu, W., et al., 2013. Serum polyfluoroalkyl concentrations, asthma outcomes, and immunological markers in a case–control study of Taiwanese children. Environ. Health Perspect. 121, 507–513. Emmett, E.A., Zhang, H., Shofer, F.S., Freeman, D., Rodway, N.V., Desai, C., et al., 2006. Community exposure to perfluorooctanoate: relationships between serum levels and certain health parameters. J. Occup. Environ. Med. 48, 771–779. http://dx.doi. org/10.1097/01.jom.0000233380.13087.37. Eriksen, H.-L.F., Kesmodel, U.S., Underbjerg, M., Kilburn, T.R., Bertrand, J., Mortensen, E.L., 2013. Predictors of intelligence at the age of 5: family, pregnancy and birth characteristics, postnatal influences, and postnatal growth. PLoS One 8, e79200. Fisher, M., Arbuckle, T.E., Wade, M., Haines, D.A., 2013. Do perfluoroalkyl substances affect metabolic function and plasma lipids?—Analysis of the 2007–2009, Canadian Health Measures Survey (CHMS) Cycle 1. Environ. Res. 121, 95–103. Fitz-Simon, N., Fletcher, T., Luster, M.I., Steenland, K., Calafat, A.M., Kato, K., et al., 2013. Reductions in serum lipids with a 4-year decline in serum perfluorooctanoic acid and perfluorooctanesulfonic acid. Epidemiology 24, 569–576.

369

Fletcher, T., Galloway, T.S., Melzer, D., Holcroft, P., Cipelli, R., Pilling, L.C., et al., 2013. Associations between PFOA, PFOS and changes in the expression of genes involved in cholesterol metabolism in humans. Environ. Int. 57–58, 2–10. Friedewald, W.T., Levy, R.I., Fredrickson, D.S., 1972. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin. Chem. 18, 499–502. Frisbee, S.J., Shankar, A., Knox, S.S., et al., 2010. Perfluorooctanoic acid, perfluorooctanesulfonate, and serum lipids in children and adolescents: results from the C8 Health Project. Arch. Pediatr. Adolesc. Med. 164, 860–869. Geiger, S.D., Xiao, J., Ducatman, A., Frisbee, S., Innes, K., Shankar, A., 2014. The association between PFOA, PFOS and serum lipid levels in adolescents. Chemosphere 98, 78–83. Han, X., Nabb, D.L., Russell, M.H., Kennedy, G.L., Rickard, R.W., 2011. Renal elimination of perfluorocarboxylates (PFCAs). Chem. Res. Toxicol. 25, 35–46. Haug, L.S., Thomsen, C., Becher, G., 2009. Time trends and the influence of age and gender on serum concentrations of perfluorinated compounds in archived human samples. Environ. Sci. Technol. 43, 2131–2136. IARC, 2012. Agents Classified by the IARC Monographs vols. 1–111. Kato, K., Wong, L.-Y., Jia, L.T., Kuklenyik, Z., Calafat, A.M., 2011. Trends in exposure to polyfluoroalkyl chemicals in the U.S. population: 1999–2008. Environ. Sci. Technol. 45, 8037–8045. Kennedy, G.J., Butenhoff, J., Olsen, G., O'Connor, J., Seacat, A., Perkins, R., et al., 2004. The toxicology of perfluorooctanoate. Crit. Rev. Toxicol. 34, 351–384. Kissa, E., 2001. Fluorinated surfactants and repellents. 2nd ed. revised and expanded. Surfactant Science Series 97Marcel Dekker, New York (NY) (640 pp.). Lin, A.Y.-C., Panchangam, S.C., Lo, C.-C., 2009a. The impact of semiconductor, electronics and optoelectronic industries on downstream perfluorinated chemical contamination in Taiwanese rivers. Environ. Pollut. 157, 1365–1372. Lin, C.-Y., Chen, P.-C., Lin, Y.-C., Lin, L.-Y., 2009b. Association among serum perfluoroalkyl chemicals, glucose homeostasis, and metabolic syndrome in adolescents and adults. Diabetes Care 32, 702–707. Lin, C.-Y., Wen, L.L., Lin, L.-Y., Wen, T.-W., Lien, G.-W., Chen, C.-Y., et al., 2011. Associations between levels of serum perfluorinated chemicals and adiponectin in a young hypertension cohort in Taiwan. Environ. Sci. Technol. 45, 10691–10698. Lin, A.-C., Panchangam, S., Tsai, Y.-T., Yu, T.-H., 2014. Occurrence of perfluorinated compounds in the aquatic environment as found in science park effluent, river water, rainwater, sediments, and biotissues. Environ. Monit. Assess. 186, 3265–3275. Loveless, S.E., Finlay, C., Everds, N.E., Frame, S.R., Gillies, P.J., O'Connor, J.C., et al., 2006. Comparative responses of rats and mice exposed to linear/branched, linear, or branched ammonium perfluorooctanoate (APFO). Toxicology 220, 203–217. Mariussen, E., 2012. Neurotoxic effects of perfluoroalkylated compounds: mechanisms of action and environmental relevance. Arch. Toxicol. 86, 1349–1367. Nelson, J.W., Hatch, E.E., Webster, T.F., 2010. Exposure to polyfluoroalkyl chemicals and cholesterol, body weight, and insulin resistance in the general U.S. population. Environ. Health Perspect. 118, 197–202. Olsen, G., Zobel, L., 2007. Assessment of lipid, hepatic, and thyroid parameters with serum perfluorooctanoate (PFOA) concentrations in fluorochemical production workers. Int. Arch. Occup. Environ. Health 81, 231–246. Olsen, G., Burris, J., Ehresman, D., Froehlich, J., Seacat, A., Butenhoff, J., et al., 2007. Half-life of serum elimination of perfluorooctanesulfonate, perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers. Environ. Health Perspect. 115, 1298–1305. Pan, W.-H., Wu, H.-J., Yeh, C.-J., Chuang, S.-Y., Chang, H.-Y., Yeh, N.-H., et al., 2011. Diet and health trends in Taiwan: comparison of two nutrition and health surveys from 1993–1996 and 2005–2008. Asia Pac. J. Clin. Nutr. 20, 238–250. Peraza, M.A., Burdick, A.D., Marin, H.E., Gonzalez, F.J., Peters, J.M., 2006. The toxicology of ligands for peroxisome proliferator-activated receptors (PPAR). Toxicol. Sci. 90, 269–295. Qazi, M.R., Abedi, M.R., Nelson, B.D., DePierre, J.W., Abedi-Valugerdi, M., 2010. Dietary exposure to perfluorooctanoate or perfluorooctane sulfonate induces hypertrophy in centrilobular hepatocytes and alters the hepatic immune status in mice. Int. Immunopharmacol. 10, 1420–1427. Ruan, T., Wang, Y., Wang, T., Zhang, Q., Ding, L., Liu, J., Wang, C., Qu, G., Jiang, G., 2010. Presence and partitioning behavior of polyfluorinated iodine alkanes in environmental matrices around a fluorochemical manufacturing plant: another possible source for perfluorinated carboxylic acids? Environ. Sci. Technol. 44, 5755–5761. Sakr, C.J., Kreckmann, K.H., Green, J.W., Gillies, P.J., Reynolds, J.L., Leonard, R.C., 2007. Cross-sectional study of lipids and liver enzymes related to a serum biomarker of exposure (ammonium perfluorooctanoate or APFO) as part of a general health survey in a cohort of occupationally exposed workers. J. Occup. Environ. Med. 49, 1086–1096. http://dx.doi.org/10.1097/JOM.0b013e318156eca3. Schecter, A., Malik-Bass, N., Calafat, A.M., Kato, K., Colacino, J.A., Gent, T.L., et al., 2012. Polyfluoroalkyl compounds in Texas children from birth through 12 years of age. Environ. Health Perspect. 120, 590–594. Starling, A.P., Engel, S.M., Whitworth, K.W., Richardson, D.B., Stuebe, A.M., Daniels, J.L., et al., 2014. Perfluoroalkyl substances and lipid concentrations in plasma during pregnancy among women in the Norwegian Mother and Child Cohort Study. Environ. Int. 62, 104–112. Steenland, K., Tinker, S., Frisbee, S., Ducatman, A., Vaccarino, V., 2009. Association of perfluorooctanoic acid and perfluorooctane sulfonate with serum lipids among adults living near a chemical plant. Am. J. Epidemiol. 170, 1268–1278. Steenland, K., Fletcher, T., Savitz, D.A., 2010. Epidemiologic evidence on the health effects of perfluorooctanoic acid (PFOA). Environ. Health Perspect. 118, 1100–1108. Toms, L.M.L., Thompson, J., Rotander, A., Hobson, P., Calafat, A.M., Kato, K., et al., 2014. Decline in perfluorooctane sulfonate and perfluorooctanoate serum concentrations in an Australian population from 2002 to 2011. Environ. Int. 71, 74–80.

370

X.-W. Zeng et al. / Science of the Total Environment 512–513 (2015) 364–370

Tsai, C., Huang, J., Hwang, B., Lee, Y., 2010. Household environmental tobacco smoke and risks of asthma, wheeze and bronchitic symptoms among children in Taiwan. Respir. Res. 11, 11. http://dx.doi.org/10.1186/1465-9921-11-11. Vestergren, R., Berger, U., Glynn, A., Cousins, I.T., 2012. Dietary exposure to perfluoroalkyl acids for the Swedish population in 1999, 2005 and 2010. Environ. Int. 49, 120–127. Wang, J., Zhang, Y., Zhang, W., Jin, Y., Dai, J., 2012. Association of perfluorooctanoic acid with HDL cholesterol and circulating miR-26b and miR-199-3p in workers of a fluorochemical plant and nearby residents. Environ. Sci. Technol. 46, 9274–9281. Wang, L., Wang, Y., Liang, Y., Li, J., Liu, Y., Zhang, J., et al., 2013. Specific accumulation of lipid droplets in hepatocyte nuclei of PFOA-exposed BALB/c mice. Sci. Rep. 3. Wang, Z., CousinsIan, T., Scheringer, M., Buck, R., Hungerbühler, K., 2014. Global emission inventories for C4–C14 perfluoroalkyl carboxylic acid (PFCA) homologues from 1951 to 2030, part I: production and emissions from quantifiable sources. Environ. Int. 62, 62–75.

Wolf, C.J., Takacs, M.L., Schmid, J.E., Lau, C., Abbott, B.D., 2008. Activation of mouse and human peroxisome proliferator-activated receptor alpha by perfluoroalkyl acids of different functional groups and chain lengths. Toxicol. Sci. 106, 162–171. Wolf, C.J., Schmid, J.E., Lau, C., Abbott, B.D., 2012. Activation of mouse and human peroxisome proliferator-activated receptor-alpha (PPARα) by perfluoroalkyl acids (PFAAs): further investigation of C4–C12 compounds. Reprod. Toxicol. 33, 546–551. Zeng, H.-C., Zhang, L., Li, Y.-Y., Wang, Y.-J., Xia, W., Lin, Y., et al., 2011. Inflammation-like glial response in rat brain induced by prenatal PFOS exposure. NeuroToxicology 32, 130–139. Zhang, L., Liu, J., Hu, J., Liu, C., Guo, W., Wang, Q., et al., 2012. The inventory of sources, environmental releases and risk assessment for perfluorooctane sulfonate in China. Environ. Pollut. 165, 193–198.