Associations between repeated measure of plasma perfluoroalkyl substances and cardiometabolic risk factors

http://www.diva-portal.org

This is the published version of a paper published in Environment International.

Citation for the original published paper (version of record):

Donat-Vargas, C., Bergdahl, I A., Tornevi, A., Wennberg, M., Sommar, J. et al. (2019) Associations between repeated measure of plasma perfluoroalkyl substances and cardiometabolic risk factors

Environment International, 124: 58-65 https://doi.org/10.1016/j.envint.2019.01.007

Access to the published version may require subscription.

N.B. When citing this work, cite the original published paper.

Permanent link to this version:

http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-156228

Contents lists available at ScienceDirect

Environment International

journal homepage: www.elsevier.com/locate/envint

Associations between repeated measure of plasma per fluoroalkyl substances and cardiometabolic risk factors

Carolina Donat-Vargas ^a , Ingvar A. Bergdahl ^b , Andreas Tornevi ^b , Maria Wennberg ^c , Johan Sommar ^b , Jani Koponen ^d , Hannu Kiviranta ^d , Agneta Åkesson ^a,⁎

a

Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden

b

Department of Public Health and Clinical Medicine, Occupational and Environmental Medicine, Umeå University, Umeå, Sweden

c

Department of Public Health and Clinical Medicine, Nutritional Research, Umeå University, Umeå, Sweden

d

Department for Health Security, Environmental Health Unit, National Institute for Health and Welfare, Kuopio, Finland

A R T I C L E I N F O

Handling Editor: Lesa Aylward Keywords:

Plasma perfluoroalkyl substances Cardiometabolic risk factors Lipids

Hypertension

Environmental epidemiology Repeated measurements Prospective assessment

A B S T R A C T

Background: Per fluoroalkyl substances (PFAS) are persistent synthetic chemicals that may affect components of metabolic risk through the peroxisome proliferator-activated receptor but epidemiological data remain scarce and inconsistent.

Objective: To estimate associations between repeated measurements of the main PFAS in plasma and total cholesterol, triglycerides and hypertension among the control subjects from a population-based nested case- control study on diabetes type 2 in middle-aged women and men.

Methods: Participants (n = 187) were free of diabetes at both baseline and follow-up visits to the Västerbotten Intervention Programme, 10 years apart: during 1990 to 2003 (baseline) and 2001 to 2013 (follow-up).

Participants left blood samples, completed questionnaires on diet and lifestyle factors, and underwent medical examinations, including measurement of blood pressure. PFAS and lipids were later determined in stored plasma samples. Associations for the repeated measurements were assessed using generalized estimating equations.

Results: Six PFAS exceeded the limit of quantitation. Repeated measures of PFAS in plasma, cardiometabolic risk factors and confounders, showed an average decrease of triglycerides from −0.16 mmol/l (95% confidence interval [CI]: −0.33, 0.02 for PFOA) to −0.26 mmol/l (95% CI: −0.50, −0.08 for PFOS), when comparing the highest tertile of PFAS plasma levels with the lowest. Associations based on average PFAS measurements and follow-up triglycerides revealed similar inverse associations, although attenuated. The estimates for cholesterol and hypertension were inconsistent and with few exception non-signi ficant.

Conclusions: This study found inverse associations between PFAS and triglycerides, but did not support any clear link with either cholesterol or hypertension.

1. Introduction

Increasing prevalence and clustering of cardiometabolic risk factors across age groups has been described in many areas of the world (Atlas Writing et al., 2018). Knowledge on the potential role of speci fic en- vironmental contaminants in the development of these risk factors is required to enable prevention of the high number of deaths from car- diovascular diseases (CVD) considered to be attributed to environ- mental contaminants (Collaborators, 2016).

An issue of concern is that of the per- and poly-fluoroalkyl sub- stances (PFAS), a large group of fluorinated synthetic chemicals that

due to their unique characteristics – including water and oil repellence and thermal stability, as well as their film forming and surfactant properties – have been leaked, both intentionally and unintentionally, to the environment through both industrial and consumer use. PFAS are persistent and bioaccumulative substances and currently are ubiquitous in the environment and have been detected in humans all around the world (Kannan et al., 2004). Human exposure to PFAS occurs mainly through contaminated food, drinking water, and dust, as well as through direct contact with daily consumer products containing PFAS, including food-contact materials (Begley et al., 2005; D'Eon and Mabury, 2011).

https://doi.org/10.1016/j.envint.2019.01.007

Received 18 September 2018; Received in revised form 21 December 2018; Accepted 3 January 2019

Abbreviations: PFAS, perfluoroalkyl substances; PFOA, perfluorooctanoic acid, 8 carbons; PFOS, perfluorooctane sulfonic acid, 8 carbons; PFNA, perfluorononanoic acid, 9 carbons; PFHxS, per fluorohexane sulfonic acid, 6 carbons; PFDA, perfluorodecanoic acid, 10 carbons; PFUnDA, perfluoroundecanoic acid, 11 carbons

⁎

Corresponding author at: Karolinska Institutet, Institute of Environmental Medicine, Box 210, SE-171 77 Stockholm, Sweden.

E-mail address: Agneta.Akesson@ki.se (A. Åkesson).

Available online 10 January 2019

0160-4120/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).

T

Longer-chain PFAS (e.g., perfluorononane acid [PFNA], per- fluorodecanoic acid [PFDA], perfluoroundecanoic acid [PFUnDA]) seems to be more bioaccumulative than the shorter ones (e.g., per- fluorooctanoic acid [PFOA]). Likewise, the sulfonates (e.g., per- fluorohexane sulfonic acid [PFHxS]) seem more bioaccumulative than the carboxylates of the same fluorinated carbon chain length ( Conder et al., 2008). The mean serum elimination half-lives (in years) of PFOS, PFOA, and PFHxS in humans were estimated to be 5.4, 3.8, and 8.5 years, respectively, in a highly exposed occupational cohort (24 males and 2 females) (Olsen et al., 2007). More recently, the half-lives (in years) were 5.8 (PFOS), 1.5 (PFOA), 7.1 (PFHxS), 1.7 (PFNA), 4.0 (PFDA) and 4.0 (PFDUnDA) in ~20 young females and 18 (PFOS), 1.2 (PFOA), 25 (PFHxS), 3.2 (PFNA), 7.1 (PFDA), and 7.4 (PFDUnDA) in

~60 males and older females (Zhang et al., 2013).

PFAS have structural similarities with fatty acids and may interfere with fatty acid metabolism and lipid synthesis in the liver (Frayn et al., 1996; Pilz et al., 2006) and they activate the peroxisome proliferator- activated receptor alpha (PPARα), expressed predominantly in the liver, heart and endothelial cells and whose endogenous ligands are fatty acids. PPARα is a major regulator of a wide variety of target genes involved in lipid metabolism (Barger and Kelly, 2000; Berger et al., 2005; Ren et al., 2009; Rosen et al., 2008). Most studies of rodents exposed to high concentrations of PFOA, although not all (Rebholz et al., 2016), have resulted in manifest reduction in both total choles- terol and triglycerides through altered expression of lipid metabolism- related genes (Guruge et al., 2006; Haughom and Spydevold, 1992;

Loveless et al., 2006; Martin et al., 2007). PFAS' interference with the signaling pathways of the thyroid hormones (Butenhoff et al., 2002; Lau et al., 2007) – involved in energy metabolism and blood pressure reg- ulation – might be another mechanism of action of PFAS. Moreover, PFAS exposure has been associated with oxidative stress (Liu et al., 2007; Yao and Zhong, 2005) and endothelial dysfunction (Hu et al., 2003; Qian et al., 2010), both connected to the development of CVD.

Among the observed PFAS-associated health effect in humans, much attention has been on the increase in cholesterol (mainly for PFOA) (Fitz-Simon et al., 2013; Steenland et al., 2010; Sunderland et al., 2018;

Winquist and Steenland, 2014). Opposite effects has been observed in rodents in most studies. Epidemiological studies involving occupa- tionally exposed workers, people living near PFAS-emitting facilities, as well as general population with lower PFAS plasma concentrations, have mainly been based on cross-sectional assessments (Chain et al., 2018). As regards blood pressure, PFOA was associated with higher prevalence of hypertension in U.S. NHANES (Min et al., 2012), but this was not confirmed by two subsequent prospective studies (Geiger et al., 2014b; Winquist and Steenland, 2014) and even inverse association have been reported (Christensen et al., 2016). The hazard pro file of individual PFAS could vary depending on chain length, functional groups and on species (Wolf et al., 2008; Wolf et al., 2012). Human data on other longer-chain PFAS, introduced later than PFOA and PFOS, remain scarce (Chain et al., 2018).

The purpose of the present study was to estimate associations be- tween the main PFAS detected in human plasma and total cholesterol, triglycerides and hypertension, among the control subjects from a po- pulation-based nested case-control study on diabetes type 2 (Donat- Vargas et al., 2019) in middle-aged women and men.

2. Material and methods 2.1. Study design and population

The study used data and samples from the Västerbotten Intervention Programme (VIP), a sub-cohort in the Northern Sweden Health and Disease Study initiated in 1985 (Norberg et al., 2010). Briefly, in- habitants within Västerbotten County were invited to a health ex- amination together with a questionnaire on diet and lifestyle the year they became 40, 50 or 60 years old (until 1996 also those becoming 30 years old). The participation rate exceeded 56%, but was often around 70%, of which the vast majority (90.5%) donated blood samples for research.

The present study was originally initiated to disentangle whether persistent organic pollutants, including PFAS, were associated with risk of diabetes, applying a nested case-control study design with repeated Fig. 1. PFAS temporal trends (median per each calendar year) from 1991 to 2013 (374 measurements).

RC: the median (interquartile range, IQR) within-person 10-year relative change (%) for the four PFAS with all the measurements > LOD.

C. Donat-Vargas et al.

Environment International 124 (2019) 58–65

59

sampling (Donat-Vargas et al., 2019). The study included T2D cases identified in the DiabNorth register (Rolandsson et al., 2012) that had donated blood samples to the biobank on at least two occasions, of which at least one prior to a diabetes diagnosis. Diabetes cases were matched (1:1) according to gender, age, sample date ( ± 90 days) and type of questionnaire at baseline examination with VIP non-diabetic participants (controls), who also donated blood samples to the biobank on at least two occasions. PFAS were measured twice in plasma of 187 T2D case-control pairs.

The analyses in the present study were limited to the controls alone (i.e., we excluded the diabetes cases) to minimize the possibility that the preclinical T2D stage of cases affected plasma PFAS concentrations and CVD risk markers. The rationale for this restriction was i) the un- known e ffect of the diabetes disease or its treatment on the exposure, distribution or excretion of PFAS, ii) the high use of lipid lowering and antihypertensive treatment among the diabetes cases (> 50%), iii) the signi ficant interaction detected between PFAS levels and diabetes status (i.e., being a future T2D case or a control) for the cardiometabolic outcomes (P values for the baseline interactions between the exposure to the sum of the 6 PFAS and T2D for triglycerides, cholesterol and hypertension were < 0.05, respectively); and finally, because iv) con- trols are more representative of the overall source population.

Therefore, we considered that the interpretation of the associations based on subjects free of diabetes was the most reliable; and the study population was restricted to the 187 controls, sampled twice. Oral/

written informed consent was obtained and the regional ethics review board approved the study.

2.2. Exposure and covariate assessment

The participants' first medical examination and blood sampling were carried out during 1990 to 2003 (baseline) and later during 2001 to 2013 (follow-up). All blood samples were collected after an 8-h overnight fasting. PFAS were measured at the National Institute for Health and Welfare in Finland, using a method based on liquid chro- matography-triple quadrupole mass spectrometry (LC-MS/MS) (Koponen et al., 2013). The limit of quantitation (LOQ) ranged from 0.15 to 1.0 ng/ml. Out of the 13 di fferent PFAS measured in plasma, baseline and follow-up concentrations of PFOA, PFOS, PFNA and PFHxS were > LOQ (0.15 ng/ml) in all subjects. Conversely, baseline and follow-up concentrations of per fluorododecane acid (PFDoA), perfluorotridecane acid (PFTrA), perfluorotetradecane acid (PFTA), PFHxA, perfluoroheptane sulfonate (PFHpS), perfluoroheptane acid (PFHpA) and per fluorodecane sulfonate (PFDS) were < LOQ in all subjects (LOQ was 0.30 ng/ml except for PFHpS that was 1.0 ng/ml).

Finally, PFDA and PFUnDA were < LOQ (0.15 ng/ml) in 26% and 42%

of the subjects respectively at baseline (n = 374) and 10% and 34%

(n = 374) respectively at follow-up.

Thus, ultimately, six individual PFAS were available to be properly assessed, four whose concentrations were > LOQ in all subjects (PFOS, PFOA, PFHxS and PFNA), and two (PFDA and PFUnDA) to which the corresponding LOQ/2 value was imputed in those samples where con- centrations were < LOQ.

The covariates included were age, gender, education level, smoking habits, body mass index (BMI), Cambridge index for physical activity, history of certain diseases, medication, alcohol consumption, phytos- terols and other validated food intakes (Johansson et al., 2001;

Johansson et al., 2002) from which the healthy diet score was retrieved (Nettleton et al., 2013).

2.3. Outcome assessment

Total cholesterol and triglycerides in plasma were analysed in a sample collected (after an 8-h overnight fasting) at the same time as the sample in which PFAS later were measured. Reflotron was used up to September 2009, and the clinical chemical laboratory after September 2009. Systolic and diastolic blood pressure were measured at baseline and at follow-up with a mercury sphygmomanometer after 5 min rest with the subject in a supine or sitting position, depending on whether it was measured before September 2009 or after, respectively. Accurate equations taking into account age and gender (developed by Umeå University, Sweden, and available from the corresponding author on reasonable requests) were applied for both lipids and blood pressure measurements to make reliable comparisons before and after 2009 (Norberg et al., 2010). Hypertension was defined as any of: i) self-re- ported diagnosis, ii) use of antihypertensive drugs, or iii) measured systolic/diastolic blood pressure ≥140 or ≥90 mmHg.

2.4. Statistical analyses

To illustrate the temporal trends we estimated the median PFAS for each calendar year over the whole study period (1991–2013) as well as the median (and corresponding interquartile range, IQR) within-person 10-year change (%) for the PFAS with all the measurements > LOD, i.e., PFOS PFOA, PFHxS and PFNA. Correlations between different PFAS were assessed using Spearman's (rho).

We assessed PFAS plasma concentrations as a continuous variable (per one standard deviation [SD] increment) and categorized them into Table 1

Characteristics of the study subjects and by sampling occasion.

Characteristics Baseline

(1990–2003)

Follow-up (2001−2013)

Sample size 187 187

Female, % 46 46

Age (yrs.) 46 ( ± 6) 56 ( ± 6)

Body mass index (kg/m

²

) 25.3 ( ± 3.2) 26.3 ( ± 3.7)

Education > 12 yrs., % 72 74

Smoking status, %

Current 21 13

Former 38 45

Physical activity, %

Inactive 58 53

Healthy diet (1–22 score) 11 ( ± 3) 13 ( ± 4)

Alcohol consumption, %

0.1–5 g/day 66 63

5.1–15 g/day 26 29

> 15 g/day 3 4

BP lowering medication, % 9 25

Cholesterol lowering medication, % 0 7

Hypertensive

^a

, % 37.5 53

Laboratory analyses Mean (SD) Mean (SD)

Total cholesterol (mmol/l) 5.5 ( ± 1.1) 5.4 ( ± 1.11)

Triglycerides (mmol/l) 1.3 ( ± 0.6) 1.4 ( ± 0.7)

Systolic BP (mmHg) 126.5 ( ± 16) 131 ( ± 17)

Diastolic BP (mmHg) 80 ( ± 10.2) 81 ( ± 11)

Plasma PFAS levels (ng/ml) Median (IQR) Median (IQR)

PFOA 2.9 (2.2–4.2) 2.7 (1.9–3.6)

PFOS 20 (15–26) 15 (9.7–21)

PFNA 0.53 (0.42–0.74) 0.83 (0.64–1.1)

PFHxS 1.0 (0.74–1.4) 1.2 (0.82–1.5)

PFDA 0.23 (0.08–0.31) 0.33 (0.25–0.45)

PFUnDA 0.19 (0.08–0.28) 0.22 (0.08–0.37)

Note: Continuous variables are shown as mean ( ± standard deviation) and categorical variables as percentage (%). PFAS are depicted as median (inter- quartile range) to minimize the influence of extreme values.

Abbreviations: PFOA, per fluorooctanoic acid; PFOS, perfluorooctane sulfonate;

PFNA, per fluorononanoic acid; PFHxS, perfluorohexane sulfonic acid; PFDA, perfluorodecanoic acid; PFUnDA, perfluoroundecanoic acid.

a

Hypertension de fined as any of: i) self-reported diagnosis, ii) use of anti- hypertensive drugs, or iii) measured systolic/diastolic blood pressure ≥140 or

≥90 mmHg.

tertiles, which relax the linearity assumption. The association between individual PFAS and each cardiometabolic outcome was first assessed i) cross-sectionally at each sampling occasion (presented as Supplemental Material). In order to improve the statistical power (minimizing type 2 error) and allow for time-dependent variables potentially linked to both

PFAS exposures and outcomes, we performed as a principal examina- tion ii) repeated measures analyses based on measurements (of PFAS, outcome variables and confounders) in the two sample occasions (at baseline and at follow-up) using generalized estimating equations (GEE), which take the over-time correlations into account. A strength of Table 2

Associations between repeated measurements (baseline and at follow-up) of PFAS plasma concentrations and total cholesterol, triglycerides and hypertension, estimated using generalized estimated equation and adjusting for confounders at both sampling occasions.

Total cholesterol

(n

1

= 186 and n

2

= 172; 358 measurements)

T1 T2 T3 1-SD increase

^a

Mean differences in total cholesterol (β coefficients, 95% CI, mmol/l)

PFOA Ref. 0.03

(−0.24, 0.31) −0.11

(−0.42, 0.20) −0.12

(−0.23, −0.00)

PFOS Ref. −0.31

(−0.58, −0.04)

−0.33 (−0.67, 0.00)

−0.09 (−0.22, 0.03)

PFNA Ref. −0.01

(−0.26, 0.24)

−0.01 (−0.32, 0.30)

−0.09 (−0.23, 0.05)

PFHxS Ref. −0.02

(−0.33, 0.28)

−0.51 (−0.86, −0.15)

−0.11 (−0.24, 0.02)

PFDA Ref. −0.23

(−0.51, 0.06)

−0.12 (−0.40, 0.17)

−0.12 (−0.27, 0.03)

PFUnDA Ref. 0.15

(−0.12, 0.41)

−0.14 (−0.45, 0.16)

−0.07 (−0.22, 0.08)

Triglycerides

(n

1

= 164 and n

2

= 172; 336 measurements)

T1 T2 T3 1-SD increase

^a

Mean differences in total triglycerides (β coefficients, 95% CI, mmol/l)

PFOA Ref. 0,03

(−0.15, 0.20)

−0.16 (−0.33, 0.02)

−0.07 (−0.13, −0.01)

PFOS Ref. −0.14

(−0.32, 0.05)

−0.29 (−0.50, −0.08)

−0.10 (−0.18, −0.02)

PFNA Ref. −0.02

(−0.19, 0.16)

−0.20 (−0.39, −0.02)

−0.09 (−0.17, −0.01)

PFHxS Ref. −0.06

(−0.25, 0.12)

−0.21 (−0.39, −0.02)

−0.03 (−0.08, 0.01)

PFDA Ref. −0.09

(−0.28, 0.11)

−0.26 (−0.42, −0.09)

−0.09 (−0.16, −0.03)

PFUnDA Ref. −0.08

(−0.25, 0.08)

−0.23 (−0.40, −0.07)

−0.13 (−0.21, −0.05)

Hypertension

^b

(n

1

= 184 and n

2

= 186; 370 measurements)

T1 T2 T3 1-SD increase

^a

Risk of hypertension (OR, 95% CI)

PFOA Ref. 0.79

(0.47, 1.33)

0.87 (0.50, 1.52)

0.94 (0.76, 1.16)

PFOS Ref. 0.82

(0.48, 1.40)

0.73 (0.41, 1.30)

0.71 (0.56, 0.89)

PFNA Ref. 0.96

(0.59, 1.56)

0.90 (0.52, 1.57)

0.99 (0.81, 1.21)

PFHxS Ref. 1.16

(0.58, 2.33)

0.54 (0.25, 1.18)

1.05 (0.84, 1.32)

PFDA Ref. 1.55

(0.91, 2.65)

0.85 (0.48, 1.51)

1.03 (0.82, 1.28)

PFUnDA Ref. 0.85

(0.50, 1.43)

1.08 (0.62, 1.85)

0.86 (0.68, 1.08) Notes: Estimations are adjusted for gender, age, education, sample year, body mass index, smoking habit, alcohol consumption, physical activity and healthy diet score. Estimations of individual PFAS are not mutually adjusted due to collineraity. Subjects taking cholesterol-lowering medication were excluded from the analysis of lipids.

Abbreviations: PFOA, per fluorooctanoic acid; PFOS, perfluorooctane sulfonate; PFNA, perfluorononanoic acid; PFHxS, perfluorohexane sulfonic acid; PFDA, per- fluorodecanoic acid; PFUnDA, perfluoroundecanoic acid; T1: first tertile; T2: second tertile; T3: third tertile; β, beta coefficient; OR, odds ratio; CI, confidence interval; SD, standard deviation.

a

1-SD increase in ng/ml is: 1.38 for PFOA, 8.62 for PFOS, 0.49 for PFNA, 0.70 for PFHxS, 0.20 for PFDA and 0.19 for PFUnDA.

b

Hypertension de fined as any of: i) self-reported diagnosis, ii) use of antihypertensive drugs, or iii) measured systolic/diastolic blood pressure ≥140 or

≥90 mmHg.

C. Donat-Vargas et al.

Environment International 124 (2019) 58–65

61

this approach is that individuals are considered as a random factor, where each individual is a cluster with the two measurements; thus, GEE includes an intra-subject factor and the e ffect due to time-variant differences between individuals is controlled for. When lipids were the outcome, we excluded subjects using cholesterol-lowering medication.

Additionally, iii) we prospectively evaluated the average PFAS ex- posure (of baseline and follow-up concentrations) in relation to lipids measured at follow-up adjusting for baseline confounders. We did not prospectively assess odds of hypertension because of too few new cases in relation to baseline prevalence. Results were displayed as regression coefficients (β) and Odds Ratio (OR) with their corresponding 95%

con fidence intervals (CI). The goodness-of-fit of the models was checked by studying the behaviour of the residuals.

All analyses were adjusted for established risk factors (gender, age, education, sample year, BMI, smoking habits, alcohol consumption, physical activity and a healthy diet score (Nettleton et al., 2013) of cardiovascular disease and variables potentially associated with PFAS concentrations.

The level of statistical signi ficance was set at 0.05 and all tests were two-tailed. The statistical software STATA/SE version 14.0 (Stata Corp LP, College Station, TX, USA) was used for statistical analyses.

3. Results

Temporal trends of PFAS over the whole study period (1991–2013)

with the averaged within person 10-year relative change for the four PFAS with all measurements > LOD are presented in Fig. 1. PFOA and PFOS decreased by 15% (IQR: −33% to 11%) and 29% (IQR: −42 to

−8%), respectively; while PFNA and PFHxS increased by 53% (IQR:

23% to 94%) and 13% (IQR: −15% to 37%), respectively. Generally, the di fferent PFAS were low to moderately correlated (rho ≤0.33) with the exception of the correlations between PFOA-PFOS and PFNA-PFDA- PFUnDA (rho 0.63–0.76) (Table A1).

Table 1 summarizes the characteristics of the study population at both sampling occasions. In general, no major differences were ob- served in the main characteristics between baseline and follow-up, with the exception of age (on average, 10 years di fference between samples), and the use of anti-hypertensive and cholesterol lowering drugs, which increased considerably at follow-up. Smoking had decreased and the healthy diet score improved two scores on average.

In the repeated measures analysis using GEE (Table 2), including two measurements of exposure, outcome and confounders, we observed a clear pattern of inverse associations between all assessed PFAS and triglycerides. For PFOS, the di fference of levels of triglycerides was

−0.29 (95% CI: −0.50, −0.08) mmol/l when comparing tertile 3 with tertile 1 and − 0.10 (95% CI: −0.18, −0.02) mmol/l per one SD in- crease of PFOS. This pattern was similar for the di fferent PFAS, cate- gorized into tertiles or expressed as continuous variables, as well as for the separate cross-sectional assessments (see Supplemental Table A2).

Although associations between PFAS and cholesterol were also inverse, Table 3

Prospective associations between long-term PFAS plasma concentrations

^a

and total cholesterol and triglycerides at follow-up.

Total cholesterol (n = 172) T1 T2 T3 1-SD increase

^b

Mean differences in total cholesterol (β coefficients [95% CI], mmol/l)

PFOA Ref. 0.00

(−0.40, 0.40)

0.16 (−0.30, 0.62)

−0.04 (−0.22, 0.14)

PFOS Ref. 0.01

(−0.40, 0.43)

0.20 (−0.25, 0.65)

0.05 (−0.15, 0.21)

PFNA Ref. 0.11

(−0.31, 0.54)

0.22 (−0.22, 0.65)

−0.02 (−0.22, 0.17)

PFHxS Ref. −0.11

(−0.55, 0.32) −0.28

(−0.75, 0.20)

0.02 (−0.15, 0.19)

PFDA Ref. −0.10

(−0.53, 0.32)

0.05 (−0.39, 0.49)

0.05 (−0.14, 0.24)

PFUnDA Ref. 0.23

(−0.20, 0.66)

0.02 (−0.43, 0.46)

0.01 (−0.18, 0.21)

Triglycerides (n = 172) T1 T2 T3 1-SD increase

^b

Mean differences in total cholesterol (β coefficients [95% CI], mmol/l)

PFOA Ref. 0.23

(−0.02, 0.48)

−0.19 (−0.49, 0.10)

−0.10 (−0.22, 0.02)

PFOS Ref. −0.05

(−0.31, 0.22)

−0.27 (−0.56, 0.02)

−0.14 (−0.27, −0.02)

PFNA Ref. −0.11

(−0.38, 0.16) −0.25

(−0.53, 0.02) −0.13

(−0.26, −0.01)

PFHxS Ref. −0.16

(−0.44, 0.12)

−0.26 (−0.56, 0.04)

−0.05 (−0.16, 0.06)

PFDA Ref. −0.04

(−0.32, 0.23)

−0.29 (−0.57, −0.01)

−0.12 (−0.24, −0.00)

PFUnDA Ref. −0.42

(−0.69, −0.16)

−0.38 (−0.66, −0.10)

−0.15 (−0.27, −0.03) Notes: Estimations are adjusted for gender, age, education, sample year, body mass index, smoking habit, alcohol consumption, physical activity and healthy diet score. Estimations of individual PFAS are not mutually adjusted due to collineraity. Subjects taking cholesterol-lowering medication have been excluded for the analysis of lipids.

Abbreviations: PFOA, per fluorooctanoic acid; PFOS, perfluorooctane sulfonate; PFNA, perfluorononanoic acid; PFHxS, perfluorohexane sulfonic acid; PFDA, per- fluorodecanoic acid; PFUnDA, perfluoroundecanoic acid; T1: first tertile; T2: second tertile; T3: third tertile; β, beta coefficient; OR, odds ratio; CI, confidence interval; SD, standard deviation.

a

Long-term PFAS are the average of baseline and follow-up concentrations and the outcomes are measured at follow-up.

b

1-SD increase in ng/ml is: 1.19 for PFOA, 7.42 for PFOS, 0.42 for PFNA, 0.79 for PFHxS, 0.18 for PFDA and 0.19 for PFUnDA.

most of them were not significant. Likewise, the observed overall non- significant inverse association between PFAS and hypertension using tertiles, practically disappeared when PFAS were assessed as con- tinuous variable (1-SD increase) (Table 2).

In the prospective assessment (Table 3), associations between average long-term PFAS and triglycerides at the end of the follow-up showed a similar inverse pattern, although slightly diluted as compared to the repeated measures analysis. For cholesterol, the corresponding estimations were non-signi ficant.

4. Discussion

In this population-based study among a non-diabetic population, using repeated measurements of both PFAS and cardio-metabolic out- comes, overall we found no indication of any associations between PFAS and increased blood lipids – as has been suggested by some pre- vious studies (Fitz-Simon et al., 2013; Steenland et al., 2010; Winquist and Steenland, 2014). We observed, by contrast, that PFAS were asso- ciated with lower levels of triglycerides. For total cholesterol and hy- pertension, the associations were mainly non-significant.

Manufacturing and use of PFOS and PFOA led to increased con- centrations in human blood from the 1970s to the mid-1990s followed by a downward trend since the early 2000s, but then, with an upward trend of other longer chain-length PFAS including PFNA, PFHxS, PFDA and PFUnDA (Paul et al., 2009). These temporal fluctuations were de- tected in our data and PFAS concentrations were analogous to those observed in other populations worldwide, including U.S. (Kato et al., 2011), Sweden (Glynn et al., 2012), Germany (Schroter-Kermani et al., 2013), Norway (Haug et al., 2009), Australia (Toms et al., 2009), Japan (Harada et al., 2007) and China (Jin et al., 2007). Similarly to previous population-based studies (CDC-, 2017; Christensen et al., 2016; Eriksen et al., 2013; Fisher et al., 2013; Glynn et al., 2012; Lin et al., 2009), in this examination, PFOS and PFOA showed the highest plasma con- centrations over the study period, followed by PFHxS and PFNA and, in considerably lower concentrations, PFDA and PFUnDA.

Our findings of inverse associations between PFAS and triglycerides are in the same direction as those from animal studies, including pri- mates (Seacat et al., 2002), where reductions of cholesterol, but also triglycerides are the prevailing findings ( Guruge et al., 2006; Haughom and Spydevold, 1992; Loveless et al., 2006; Martin et al., 2007). This hypolipidemic effect of PFAS was, however, not observed in previous epidemiological studies, which have generally shown positive rather than inverse associations between PFAS and lipids (Frisbee et al., 2010;

Steenland et al., 2010; Sunderland et al., 2018; Winquist and Steenland, 2014). The latest report from EFSA Panel on Contaminants in the Food Chain has concluded that it is likely that associations between serum PFOS and PFOA levels and serum cholesterol are causal (Chain et al.

2018). However, the report advocated caution (and it has been con- sidered provisional) because i) the majority of the studies are cross- sectional and the reverse causality cannot be ruled out, ii) findings in humans are not consistent with the results in experimental animals, which show a PPARa-mediated decrease in serum lipids after admin- istration of PFOS or PFOA (at high dose); and iii) confounding by diet has not been properly addressed (Chain et al. 2018).

We observed the strongest inverse associations between PFAS and triglycerides. These findings agree with a cross-sectional study using NHANES data from adolescent participants (12–19 yrs of age), which reported inverse associations between PFNA and metabolic syndrome (Lin et al., 2009). In analogy, another cross-sectional study with Inuit people of Nunavik, showed cholesterol/HDL-C ratio and triglycerides levels to be inversely associated with PFOS plasma levels after adjust- ment for n-3 PUFAs in blood (Chateau-Degat et al., 2010). A later cross- sectional study overall found no associations between PFAS and cho- lesterol (Christensen et al., 2016).

By contrast, five other cross-sectional general population studies – three in adults (Eriksen et al., 2013; Nelson et al., 2010; Skuladottir

et al., 2015) and two in adolescents (Fisher et al., 2013; Geiger et al., 2014a) – reported statistically significant positive associations between PFAS and cholesterol concentrations in serum. Among occupationally exposed workers, three (Costa et al., 2009; Olsen et al., 2003; Sakr et al., 2007) of the four existing longitudinal studies with more than one measurement reported positive associations of PFOA and PFOS with cholesterol; and only one (Olsen et al., 2012) observed no associations of PFOA and PFOS with non-high-density lipoprotein. Only two pro- spective studies have been published. One of them observed that a re- duction of PFOA and PFOS predicted a fall in low-density lipoprotein (LDL) cholesterol but not in triglycerides (in 560 adults living in an area with contaminated public water) (Fitz-Simon et al., 2013). The other found a positive association between retrospectively estimated serum PFOA concentration (generated through fate, transport and exposure modeling) and self-resported hypercholesterolemia (Winquist and Steenland, 2014). Altogether, clear conclusions cannot be drawn on PFAS and human blood lipids from previous literature, due to the conflicting observations reported and the cross-sectional nature of most of the studies. As for our data, the observed PFAS-cholesterol associa- tions were mainly non-significant and therefore considered incon- clusive.

As regards blood pressure, data remain very limited for associations with PFAS. In vitro (Qian et al., 2010) and animal (Cui et al., 2009) studies support a plausible link between PFAS and vascular injury and hypertension. Conversely, in a prospective study involving people from the state of Wisconsin (Christensen et al., 2016), all PFAS were asso- ciated with lower odds of hypertension, although only PFNA reached statistical signi ficance. A cross-sectional NHANES study performed in children found no association between serum levels of PFOA and PFOS and hypertension (Geiger et al., 2014b), however, the corresponding study among adults reported positive associations between PFOA and systolic blood pressure and hypertension (Min et al., 2012). In our data, despite observing a trend of negative associations between PFAS and hypertension, it was not possible to draw any conclusion about these observations.

Although the biological pathways a ffected by PFAS are uncertain, the activation of the nuclear receptor PPARα seems to be a key com- ponent mediating their possible metabolic e ffects ( Bjork and Wallace, 2009). PPAR α is a major regulator of the expression of lipid metabo- lism-related genes (Guruge et al., 2006; Haughom and Spydevold, 1992; Latru ffe et al., 2000 ; Loveless et al., 2006) and PFAS with a chain length of C

4

–C

12

have been shown to activate PPAR α with varying potencies in mice (Buhrke et al., 2013). This PPARα-mediated me- chanism may, however, differ between humans and animals. It has been observed that the mouse PPAR α is more sensitive to PFAS than the human PPARα (Wolf et al., 2008; Wolf et al., 2012), which is expressed only to 1/10 of the PPARα in rodents (Klaunig et al., 2003). In addition to the di fferent doses of exposure between animal and population stu- dies, this interspecies variability could explain the discrepancies ob- served between PFAS' effect on lipids in laboratory animals and human studies.

Interestingly, fibrates, which exert their effect by activating PPARα, have been extensively used as hypotriglyceridemic drugs. Fibrates de- crease plasma triglyceride concentrations by activating enzymes re- sponsible for β-oxidation of long chain fatty acids, reducing triglyceride synthesis and activating lipoprotein lipase (Auwerx et al., 1996; Staels et al., 1998). However, a paradoxal increase in total or LDL-cholesterol in fibrate treated subjects is not uncommon ( Pasternak et al., 1996;

Turpin and Bruckert, 1996) and opposite regulation of murine and human lipid metabolism-related genes expression after treatment has been reported (Berthou et al., 1996). In mice, genetically modi fied to express specific human genes (Bighetti et al., 2009), the fibrate treat- ment reduced triglyceride levels independently of the lipemic pheno- type, while the e ffect on LDL and HDL cholesterol was dependent on the genotype and the presence of certain proteins (apolipoprotein CIII and the cholesteryl ester transfer protein, CETP), which strongly affect the

C. Donat-Vargas et al.

Environment International 124 (2019) 58–65

63

lipid metabolism (Bighetti et al., 2009). These data support our main findings of an inverse association between PFAS and decreased trigly- ceride levels and could explain the discrepancies between human and animal studies as well as the inconsistencies within the human studies.

Any failure to fully adjust for confounding leading to biased asso- ciation estimates as an alternative explanation for the con flicting results observed in human studies needs to be considered. A special emphasize should be on the diet as a common cause of both PFAS exposure and altered blood lipids. Such confounding may di ffer between populations.

In the present population, fish, regarded as healthy food, is one source of PFAS. In other populations, food packaging containing PFAS that can leach into food may have been a source (Schaider et al., 2017). Al- though, by adjusting for dietary factors such as a healthy diet score, we strived to remove such confounding (blocking the backdoor path from PFAS to the lipids via the common cause); still we cannot rule out the possibility of residual or unmeasured confounding by diet.

Some main limitations of this study need to be considered. First, the cardiometabolic risk factors were not the primary endpoint of the ori- ginal nested case-control study and participants were selected to be controls in a diabetes study. To avoid any impact of the diabetes disease in the associations, we limited the analysis to the non-diabetics. This selection resulted in a smaller sample size a ffecting the statistical power and potentially contributing to not having reached statistical sig- nificance for some associations.

An important strength is the large number of PFAS measured in plasma at two occasions, allowing for repeated measures analysis taking into account time-varying factor impact. Moreover, the data collected including diet, anthropometric measurements and clinical parameters, allowed us to control for important potential confounders that have not been accounted for in some previous studies. Finally, data on PFAS relation to hypertension and triglycerides are limited in the literature.

5. Conclusion

PFAS were consistently associated with lower triglycerides in blood.

We did not find any clear link with either cholesterol or hypertension.

Larger studies with repeated measures are needed to strengthen these results.

Acknowledgements and ethical considerations

We acknowledge the Northern Sweden Diet Database and the funds supporting it; the Swedish Research Council for Health, Working Life and Welfare [FORTE number 2012-0758] and the Västerbotten County Council, which supports the Västerbotten Intervention Programme. We also acknowledge the Swedish Research Council (VR) [grant number 2017-00822] and Fundación Ramón Areces (Spain) for funding grant to Carolina Donat-Vargas. All necessary ethical permits were obtained from the Regional Ethical Review Board at Umeå University Dnr 2013/

414-31, 2014/147-32M.

Financial interest's declaration

The authors declare they have no actual or potential competing fi- nancial interests.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://

doi.org/10.1016/j.envint.2019.01.007.

References

Atlas Writing, G., Timmis, A., Townsend, N., Gale, C., Grobbee, R., Maniadakis, N., et al., 2018. European society of cardiology: cardiovascular disease statistics 2017. Eur.

Heart J. 39, 508–579.

Auwerx, J., Schoonjans, K., Fruchart, J.C., Staels, B., 1996. Transcriptional control of triglyceride metabolism: fibrates and fatty acids change the expression of the LPL and apo c-iii genes by activating the nuclear receptor PPAR. Atherosclerosis 124 (Suppl), S29–S37.

Barger, P.M., Kelly, D.P., 2000. Ppar signaling in the control of cardiac energy metabo- lism. Trends Cardiovasc. Med. 10, 238–245.

Begley, T.H., White, K., Honigfort, P., Twaroski, M.L., Neches, R., Walker, R.A., 2005.

Perfluorochemicals: potential sources of and migration from food packaging. Food Addit. Contam. 22, 1023–1031.

Berger, J.P., Akiyama, T.E., Meinke, P.T., 2005. PPARs: therapeutic targets for metabolic disease. Trends Pharmacol. Sci. 26, 244–251.

Berthou, L., Duverger, N., Emmanuel, F., Langouet, S., Auwerx, J., Guillouzo, A., et al., 1996. Opposite regulation of human versus mouse apolipoprotein a–i by fibrates in human apolipoprotein a–i transgenic mice. J. Clin. Invest. 97, 2408–2416.

Bighetti, E.J., Patricio, P.R., Casquero, A.C., Berti, J.A., Oliveira, H.C., 2009. Ciprofibrate increases cholesteryl ester transfer protein gene expression and the indirect reverse cholesterol transport to the liver. Lipids Health Dis. 8, 50.

Bjork, J.A., Wallace, K.B., 2009. Structure-activity relationships and human relevance for perfluoroalkyl acid-induced transcriptional activation of peroxisome proliferation in liver cell cultures. Toxicol. Sci. 111, 89–99.

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., Costa, G., Elcombe, C., Farrar, D., Hansen, K., Iwai, H., et al., 2002. Toxicity of ammonium perfluorooctanoate in male cynomolgus monkeys after oral dosing for 6 months. Toxicol. Sci. 69, 244–257.

CDC-2017 Cdc (centers for disease control and prevention), 2017. Fourth National Report on Human Exposure to Environmental Chemicals.

Chain EPoCitF, Knutsen, H.K., Alexander, J., Barregård, L., Bignami, M., Brüschweiler, B., et al., 2018. Risk to human health related to the presence of perfluorooctane sulfonic acid and perfluorooctanoic acid in food. EFSA J. 16, e05194.

Chateau-Degat, M.L., Pereg, D., Dallaire, R., Ayotte, P., Dery, S., Dewailly, E., 2010.

Effects of perfluorooctanesulfonate exposure on plasma lipid levels in the Inuit po- pulation of Nunavik (northern Quebec). Environ. Res. 110, 710–717.

Christensen, K.Y., Raymond, M., Thompson, B.A., Anderson, H.A., 2016. Perfluoroalkyl substances in older male anglers in Wisconsin. Environ. Int. 91, 312–318.

Collaborators GBDRF, 2016. Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990–2015: a systematic analysis for the global burden of disease study 2015.

Lancet 388, 1659–1724.

Conder, J.M., Hoke, R.A., De Wolf, W., Russell, M.H., Buck, R.C., 2008. Are PFCAs bioaccumulative? A critical review and comparison with regulatory criteria and persistent lipophilic compounds. Environ. Sci. Technol. 42, 995–1003.

Costa, G., Sartori, S., Consonni, D., 2009. Thirty years of medical surveillance in per- fluooctanoic acid production workers. J. Occup. Environ. Med. 51, 364–372.

Cui, L., Zhou, Q.F., Liao, C.Y., Fu, J.J., Jiang, G.B., 2009. Studies on the toxicological effects of PFOA and PFOS on rats using histological observation and chemical ana- lysis. Arch. Environ. Contam. Toxicol. 56, 338–349.

D'Eon, J.C., Mabury, S.A., 2011. Exploring indirect sources of human exposure to per- fluoroalkyl carboxylates (PFCAs): evaluating uptake, elimination, and bio- transformation of polyfluoroalkyl phosphate esters (PAPs) in the rat. Environ. Health Perspect. 119, 344–350.

Donat-Vargas, C., Bergdahl, I., Tornevi, A., Wennberg, M., Sommar, J., Kiviranta, H., et al., 2019. Perfluoroalkyl substances and risk of type ii diabetes: a prospective nested case-control study. Environ. Int. 123, 390–398.

Eriksen, K.T., Raaschou-Nielsen, O., McLaughlin, J.K., Lipworth, L., Tjonneland, A., Overvad, K., et al., 2013. Association between plasma PFOA and PFOS levels and total cholesterol in a middle-aged Danish population. PLoS One 8, e56969.

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.

Frayn, K.N., Williams, C.M., Arner, P., 1996. Are increased plasma non-esterified fatty acid concentrations a risk marker for coronary heart disease and other chronic dis- eases? Clin. Sci. (Lond.) 90, 243–253.

Frisbee, S.J., Shankar, A., Knox, S.S., Steenland, K., Savitz, D.A., Fletcher, T., 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., 2014a. The as- sociation between PFOA, PFOS and serum lipid levels in adolescents. Chemosphere 98, 78–83.

Geiger, S.D., Xiao, J., Shankar, A., 2014b. No association between perfluoroalkyl che- micals and hypertension in children. Integr. Blood Press. Control 7, 1–7.

Glynn, A., Berger, U., Bignert, A., Ullah, S., Aune, M., Lignell, S., et al., 2012.

Perfluorinated alkyl acids in blood serum from primiparous women in Sweden: serial sampling during pregnancy and nursing, and temporal trends 1996–2010. Environ.

Sci. Technol. 46, 9071–9079.

Guruge, K.S., Yeung, L.W., Yamanaka, N., Miyazaki, S., Lam, P.K., Giesy, J.P., et al., 2006.

Gene expression profiles in rat liver treated with perfluorooctanoic acid (PFOA).

Toxicol. Sci. 89, 93–107.

Harada, K., Koizumi, A., Saito, N., Inoue, K., Yoshinaga, T., Date, C., et al., 2007.

Historical and geographical aspects of the increasing perfluorooctanoate and

perfluorooctane sulfonate contamination in human serum in Japan. Chemosphere 66, 293–301.

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.

Haughom, B., Spydevold, O., 1992. The mechanism underlying the hypolipemic effect of perfluorooctanoic acid (PFOA), perfluorooctane sulphonic acid (PFOSA) and clofibric acid. Biochim. Biophys. Acta 1128, 65–72.

Hu, W., Jones, P.D., DeCoen, W., King, L., Fraker, P., Newsted, J., et al., 2003. Alterations in cell membrane properties caused by perfluorinated compounds. Comp. Biochem.

Physiol. C: Toxicol. Pharmacol. 135, 77–88.

Jin, Y., Saito, N., Harada, K.H., Inoue, K., Koizumi, A., 2007. Historical trends in human serum levels of perfluorooctanoate and perfluorooctane sulfonate in Shenyang, China. Tohoku J. Exp. Med. 212, 63–70.

Johansson, G., Wikman, A., Ahren, A.M., Hallmans, G., Johansson, I., 2001.

Underreporting of energy intake in repeated 24-hour recalls related to gender, age, weight status, day of interview, educational level, reported food intake, smoking habits and area of living. Public Health Nutr. 4, 919–927.

Johansson, I., Hallmans, G., Wikman, A., Biessy, C., Riboli, E., Kaaks, R., 2002. Validation and calibration of food-frequency questionnaire measurements in the northern Sweden health and disease cohort. Public Health Nutr. 5, 487–496.

Kannan, K., Corsolini, S., Falandysz, J., Fillmann, G., Kumar, K.S., Loganathan, B.G., et al., 2004. Perfluorooctanesulfonate and related fluorochemicals in human blood from several countries. Environ. Sci. Technol. 38, 4489–4495.

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.

Klaunig, J.E., Babich, M.A., Baetcke, K.P., Cook, J.C., Corton, J.C., David, R.M., et al., 2003. PPARalpha agonist-induced rodent tumors: modes of action and human re- levance. Crit. Rev. Toxicol. 33, 655–780.

Koponen, J., Rantakokko, P., Airaksinen, R., Kiviranta, H., 2013. Determination of se- lected perfluorinated alkyl acids and persistent organic pollutants from a small vo- lume human serum sample relevant for epidemiological studies. J. Chromatogr. A 1309, 48–55.

Latruffe, N., Cherkaoui Malki, M., Nicolas-Frances, V., Clemencet, M.C., Jannin, B., Berlot, J.P., 2000. Regulation of the peroxisomal beta-oxidation-dependent pathway by peroxisome proliferator-activated receptor alpha and kinases. Biochem.

Pharmacol. 60, 1027–1032.

Lau, C., Anitole, K., Hodes, C., Lai, D., Pfahles-Hutchens, A., Seed, J., 2007. Perfluoroalkyl acids: a review of monitoring and toxicological findings. Toxicol. Sci. 99, 366–394.

Lin, C.Y., Chen, P.C., Lin, Y.C., Lin, L.Y., 2009. Association among serum perfluoroalkyl chemicals, glucose homeostasis, and metabolic syndrome in adolescents and adults.

Diabetes Care 32, 702–707.

Liu, C., Yu, K., Shi, X., Wang, J., Lam, P.K., Wu, R.S., et al., 2007. Induction of oxidative stress and apoptosis by PFOS and PFOA in primary cultured hepatocytes of fresh- water tilapia (Oreochromis niloticus). Aquat. Toxicol. 82, 135–143.

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.

Martin, M.T., Brennan, R.J., Hu, W., Ayanoglu, E., Lau, C., Ren, H., et al., 2007.

Toxicogenomic study of triazole fungicides and perfluoroalkyl acids in rat livers predicts toxicity and categorizes chemicals based on mechanisms of toxicity. Toxicol.

Sci. 97, 595–613.

Min, J.Y., Lee, K.J., Park, J.B., Min, K.B., 2012. Perfluorooctanoic acid exposure is as- sociated with elevated homocysteine and hypertension in us adults. Occup. Environ.

Med. 69, 658–662.

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.

Nettleton, J.A., Hivert, M.F., Lemaitre, R.N., McKeown, N.M., Mozaffarian, D., Tanaka, T., et al., 2013. Meta-analysis investigating associations between healthy diet and fasting glucose and insulin levels and modification by loci associated with glucose home- ostasis in data from 15 cohorts. Am. J. Epidemiol. 177, 103–115.

Norberg, M., Wall, S., Boman, K., Weinehall, L., 2010. The Vasterbotten intervention programme: background, design and implications. Glob. Health Action 3.

Olsen, G.W., Burris, J.M., Burlew, M.M., Mandel, J.H., 2003. Epidemiologic assessment of worker serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) concentrations and medical surveillance examinations. J. Occup. Environ. Med. 45, 260–270.

Olsen, G.W., Burris, J.M., Ehresman, D.J., Froehlich, J.W., Seacat, A.M., Butenhoff, J.L., et al., 2007. Half-life of serum elimination of perfluorooctanesulfonate, per- fluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers. Environ. Health Perspect. 115, 1298–1305.

Olsen, G.W., Ehresman, D.J., Buehrer, B.D., Gibson, B.A., Butenhoff, J.L., Zobel, L.R., 2012. Longitudinal assessment of lipid and hepatic clinical parameters in workers involved with the demolition of perfluoroalkyl manufacturing facilities. J. Occup.

Environ. Med. 54, 974–983.

Pasternak, R.C., Brown, L.E., Stone, P.H., Silverman, D.I., Gibson, C.M., Sacks, F.M., 1996.

Effect of combination therapy with lipid-reducing drugs in patients with coronary heart disease and “normal” cholesterol levels. A randomized, placebo-controlled trial.

Harvard atherosclerosis reversibility project (harp) study group. Ann. Intern. Med.

125, 529–540.

Paul, A.G., Jones, K.C., Sweetman, A.J., 2009. A first global production, emission, and environmental inventory for perfluorooctane sulfonate. Environ. Sci. Technol. 43, 386–392.

Pilz, S., Scharnagl, H., Tiran, B., Seelhorst, U., Wellnitz, B., Boehm, B.O., et al., 2006. Free fatty acids are independently associated with all-cause and cardiovascular mortality in subjects with coronary artery disease. J. Clin. Endocrinol. Metab. 91, 2542–2547.

Qian, Y., Ducatman, A., Ward, R., Leonard, S., Bukowski, V., Lan Guo, N., et al., 2010.

Perfluorooctane sulfonate (PFOS) induces reactive oxygen species (ROS) production in human microvascular endothelial cells: role in endothelial permeability. J. Toxicol.

Environ. Health A 73, 819–836.

Rebholz, S.L., Jones, T., Herrick, R.L., Xie, C., Calafat, A.M., Pinney, S.M., et al., 2016.

Hypercholesterolemia with consumption of PFOA-laced western diets is dependent on strain and sex of mice. Toxicol. Rep. 3, 46–54.

Ren, H., Vallanat, B., Nelson, D.M., Yeung, L.W., Guruge, K.S., Lam, P.K., et al., 2009.

Evidence for the involvement of xenobiotic-responsive nuclear receptors in tran- scriptional effects upon perfluoroalkyl acid exposure in diverse species. Reprod.

Toxicol. 27, 266–277.

Rolandsson, O., Norberg, M., Nystrom, L., Soderberg, S., Svensson, M., Lindahl, B., et al., 2012. How to diagnose and classify diabetes in primary health care: lessons learned from the diabetes register in northern Sweden (DiabNorth). Scand. J. Prim. Health Care 30, 81–87.

Rosen, M.B., Lee, J.S., Ren, H., Vallanat, B., Liu, J., Waalkes, M.P., et al., 2008.

Toxicogenomic dissection of the perfluorooctanoic acid transcript profile in mouse liver: evidence for the involvement of nuclear receptors PPAR alpha and car. Toxicol.

Sci. 103, 46–56.

Sakr, C.J., Leonard, R.C., Kreckmann, K.H., Slade, M.D., Cullen, M.R., 2007. Longitudinal study of serum lipids and liver enzymes in workers with occupational exposure to ammonium perfluorooctanoate. J. Occup. Environ. Med. 49, 872–879.

Schaider, L.A., Balan, S.A., Blum, A., Andrews, D.Q., Strynar, M.J., Dickinson, M.E., et al., 2017. Fluorinated compounds in U.S. fast food packaging. Environ. Sci. Technol. Lett.

4, 105–111.

Schroter-Kermani, C., Muller, J., Jurling, H., Conrad, A., Schulte, C., 2013. Retrospective monitoring of perfluorocarboxylates and perfluorosulfonates in human plasma ar- chived by the German environmental specimen bank. Int. J. Hyg. Environ. Health 216, 633–640.

Seacat, A.M., Thomford, P.J., Hansen, K.J., Olsen, G.W., Case, M.T., Butenhoff, J.L., 2002.

Subchronic toxicity studies on perfluorooctanesulfonate potassium salt in cyno- molgus monkeys. Toxicol. Sci. 68, 249–264.

Skuladottir, M., Ramel, A., Rytter, D., Haug, L.S., Sabaredzovic, A., Bech, B.H., et al., 2015. Examining confounding by diet in the association between perfluoroalkyl acids and serum cholesterol in pregnancy. Environ. Res. 143, 33–38.

Staels, B., Dallongeville, J., Auwerx, J., Schoonjans, K., Leitersdorf, E., Fruchart, J.C., 1998. Mechanism of action of fibrates on lipid and lipoprotein metabolism.

Circulation 98, 2088–2093.

Steenland, K., Fletcher, T., Savitz, D.A., 2010. Epidemiologic evidence on the health ef- fects of perfluorooctanoic acid (PFOA). Environ. Health Perspect. 118, 1100–1108.

Sunderland, E.M., Hu, X.C., Dassuncao, C., Tokranov, A.K., Wagner, C.C., Allen, J.G., 2018. A review of the pathways of human exposure to poly- and perfluoroalkyl substances (PFASs) and present understanding of health effects. J. Expo. Sci. Environ.

Epidemiol. https://doi.org/10.1038/s41370-018-0094-1. (Epub ahead of print).

Toms, L.M., Calafat, A.M., Kato, K., Thompson, J., Harden, F., Hobson, P., et al., 2009.

Polyfluoroalkyl chemicals in pooled blood serum from infants, children, and adults in Australia. Environ. Sci. Technol. 43, 4194–4199.

Turpin, G., Bruckert, E., 1996. Efficacy and safety of ciprofibrate in hyperlipoproteinae- mias. Atherosclerosis 124 (Suppl), S83–S87.

Winquist, A., Steenland, K., 2014. Modeled PFOA exposure and coronary artery disease, hypertension, and high cholesterol in community and worker cohorts. Environ.

Health Perspect. 122, 1299–1305.

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 (PPARalpha) by perfluoroalkyl acids (PFAAs): further investigation of c4-c12 compounds. Reprod. Toxicol. 33, 546–551.

Yao, X., Zhong, L., 2005. Genotoxic risk and oxidative DNA damage in hepg2 cells ex- posed to perfluorooctanoic acid. Mutat. Res. 587, 38–44.

Zhang, Y., Beesoon, S., Zhu, L., Martin, J.W., 2013. Biomonitoring of perfluoroalkyl acids in human urine and estimates of biological half-life. Environ. Sci. Technol. 47, 10619–10627.

C. Donat-Vargas et al.

Environment International 124 (2019) 58–65

65