Distribution of Novel and Well-Known Poly- and Perfluoroalkyl

Oct 22, 2017 - Statistically significant differences in PFAS concentrations with respect to age-tertiles were evaluated using a Kruskal–Wallis test...
0 downloads 12 Views 988KB Size
Subscriber access provided by READING UNIV

Article

Distribution of novel and well-known poly- and perfluoroalkyl substances (PFASs) in human serum, plasma, and whole blood Somrutai Poothong, Cathrine Thomsen, Juan Antonio Padilla Sanchez, Eleni Papadopoulou, and Line Småstuen Haug Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b03299 • Publication Date (Web): 22 Oct 2017 Downloaded from http://pubs.acs.org on October 26, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33

Environmental Science & Technology

Figure-1 500x139mm (96 x 96 DPI)

ACS Paragon Plus Environment

Environmental Science & Technology

Figure-2 200x500mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 2 of 33

Page 3 of 33

Environmental Science & Technology

Figure-3 500x179mm (96 x 96 DPI)

ACS Paragon Plus Environment

Environmental Science & Technology

Graphical Abstract 846x475mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 4 of 33

Page 5 of 33

Environmental Science & Technology

1

Distribution of novel and well-known poly- and perfluoroalkyl substances (PFASs) in

2

human serum, plasma, and whole blood

3 4

Somrutai Poothong,1* Cathrine Thomsen,1 Juan Antonio Padilla-Sanchez,1 Eleni

5

Papadopoulou,1 Line Småstuen Haug1

6 7

1

8

Health, P.O. Box 4404, Nydalen, NO-0403 Oslo, Norway.

Department of Environmental Exposure and Epidemiology, Norwegian Institute of Public

9 10

*Corresponding author:

11

Somrutai Poothong

12

Department of Environmental Exposure and Epidemiology, Norwegian Institute of Public

13

Health, P.O. Box 4404, Nydalen, NO-0403 Oslo, Norway

14

Phone number: +47 21076347

15

E-mail: [email protected]

16 17

To be submitted to – Environmental Science and Technology

ACS Paragon Plus Environment

1

Environmental Science & Technology

18 19

Page 6 of 33

Abstract Currently, there is limited knowledge on the distribution of poly- and perfluoroalkyl

20

substances (PFASs) in different blood matrices, particularly for novel PFASs such as

21

polyfluoroalkyl phosphate esters (PAPs) and perfluoroalkyl phosphonates (PFPAs). To

22

explore this, serum, plasma, and whole blood from 61 adults in Oslo, Norway were collected.

23

The largest number of PFASs were detected in whole blood. For PAPs and PFPAs, the highest

24

frequencies of detection and concentrations were observed in plasma. PAPs contributed to 8%

25

of total PFASs in plasma (median, 0.81 ng mL-1). Perfluorohexylphosphonate (PFHxPA) was

26

the dominant PFPA, regardless of blood matrix. The relative composition profiles of PFASs in

27

blood matrices differed. For some specific PFASs such as perfluorooctanesulfonamide

28

(PFOSA) and perfluorohexanoate (PFHxA), the highest concentrations were observed in

29

whole blood. The PFAS concentration ratios varied between blood matrices, depending on the

30

compounds. However, similar ratios were observed for 6:2 polyfluoroalkyl phosphate diester

31

(6:2diPAP) as well as well-known PFASs such as perfluorooctanesulfonate (PFOS) and

32

perfluorooctanoate (PFOA). Besides the determination of twenty-five PFASs in human blood,

33

this study also lead to better understanding of biomonitoring data from different blood

34

matrices, which is key knowledge for performing both exposure assessments and

35

epidemiological studies.

ACS Paragon Plus Environment

2

Page 7 of 33

36 37

Environmental Science & Technology

Introduction Poly- and perfluoroalkyl substances (PFASs, CnF2n+1 −R) are a broad range of

38

synthetic organofluorine compounds based on two structural components, a hydrophobic

39

poly- or perfluoroalkyl chain, and a hydrophilic functional group (e.g. –COOH and –SO3H).

40

These unique molecules have been very useful in surfactant and polymer industries.1-2

41

However, their widespread use and unique physicochemical properties have also resulted in

42

their ubiquitous contamination of the environment, bioaccumulation in animals,3 and presence

43

in human blood all over the world.4-7 Most studies on PFASs so far have been limited to

44

perfluoroalkyl sulfonates (PFSAs, CnF2n+1SO3H) and perfluoroalkyl carboxylates (PFCAs,

45

CnF2n+1COOH), particularly in human blood.8

46

Associations between concentrations of some specific PFSAs and PFCAs in human

47

serum or plasma and a range of health outcomes have been observed in epidemiological

48

studies.9-11 Furthermore, a range of toxicological effects have been observed in animal studies

49

including tumor development, hormonal effects, and immunotoxicity.12-13 Based on its

50

persistence in the environment as well as the bioaccumulation and toxicological potential, use

51

of perfluorooctanesulfonate (PFOS) has been banned or restricted worldwide.14-16

52

Perfluorooctanoate (PFOA) has been identified as a substance of very high concern (SVHC)

53

in the REACH regulation and has been banned in consumer products in Norway.17 Following

54

such measures for some of the well-known PFASs, a shift towards production of short-chain

55

PFSAs/PFCAs and other functional groups such as polyfluoroalkyl phosphate esters (PAPs)

56

and perfluoroalkyl phosphonates (PFPAs) has occured.18-19 Previous studies have found that

57

the well-known PFSAs and PFCAs accounted for only 33–85% of total extractable organic

58

fluorine in whole blood of general populations.20-21 Thus, there are other PFASs in human

59

blood but so far, few biomonitoring studies have assessed the exposure to for instance PFAS

60

precursors and novel PFASs due to methodological limitations of chemical analyses.

ACS Paragon Plus Environment

3

Environmental Science & Technology

61

PFPAs (CnF2n+1P(O)(OH)2) are emerging PFASs and belong to the class of

62

perfluoroalkyl acids (PFAAs) along with PFSAs and PFCAs. PFPAs have been detected in

63

the environment, such as in surface water22 and wastewater,23 but also in indoor dust.24 Wang

64

et al. recently compared PFPAs to PFSAs and PFCAs and found similarities in persistence

65

and elimination potential in rainbow trout and rats.25 A limited number of studies have

66

included PFPAs, but there is a particular need for data on PFPAs in the same samples as of

67

PFSAs and PFCAs. Previous studies have reported that human body burden of PFASs also

68

occurs through indirect exposure to PFAS precursors, such as PAPs and perfluoroalkyl

69

sulfonamides (FOSAs). PAPs are surfactants possessing at least one polyfluoroalkyl tail,

70

F(CF2)nCH2CH2, and are mainly used in paper food packaging.26 Biotransformation of PAPs

71

into PFCAs has been observed in rats.27-28 In vivo animal experiments have also reported

72

biotransformation of FOSAs to PFOS.29

Page 8 of 33

73

Blood is an important and favorable matrix for determining the internal dose of

74

PFASs, and serum and plasma are most frequently used due to practical considerations, rather

75

than favorable partitioning of PFASs. However, to back calculate PFAS concentrations in

76

serum/plasma to whole blood, knowledge on the distribution of PFASs in the different blood

77

matrices is crucial. Whole blood consists of cellular elements (~45%) suspended in plasma

78

(liquid component, ~55%). If whole blood is allowed to clog and then centrifuged, about 30–

79

50% of the volume is serum/plasma, depending on anticoagulants. If it is correct that the

80

PFAS concentrations obtained in whole blood are approximately half those in serum/plasma it

81

can be assumed that PFASs are distributed mainly to the serum/plasma and to a limited extent

82

to the blood cells. However, as human blood is a complex mixture, also plasma and serum are

83

distinct fractions of whole blood. Plasma contains fibrinogen and has a higher total protein

84

content than serum, whereas serum has a higher concentration of some important protein and

85

platelet.30 In addition, different PFASs are known to have different physicochemical

ACS Paragon Plus Environment

4

Page 9 of 33

Environmental Science & Technology

86

properties and different binding affinities to blood.31 Limiting the analyses of PFASs to

87

serum/plasma might exclude the possible amount present in blood cells, and thus

88

underestimate the body burden. It should be emphasized that determination of PFASs in both

89

human serum, plasma, and whole blood from the same individual has seldom been carried out.

90

This is particularly true for novel PFASs, while some few studies have emphasized this aspect

91

for some well-known PFASs (e.g. PFOS, PFOA, perfluorohexanesulfonate (PFHxS),

92

perfluorononanoate (PFNA), and perfluorooctanesulfonamide (PFOSA)).32-34 Some studies

93

have only considered the partitioning between plasma and whole blood,32, 34-35 and/or included

94

only small and specific groups of the population (e.g. occupational, maternal).32-34 However,

95

knowledge on distributions between different blood matrices is important when performing

96

risk assessments, and can be of high importance when evaluating the possibilities of reverse

97

causations in epidemiological studies.

98 99

The aims of the present study were to assess the presence, concentrations, composition profiles, and correlations of various PFASs in whole blood, plasma and serum in a study

100

group of the general adult population. Further, correlations and distribution ratios between

101

different blood matrices were determined for each of the PFASs. Apart from measuring the

102

well-known PFASs such as PFSAs, PFCAs, and FOSAs, this study also included novel

103

PFASs such as PAPs and PFPAs for which this information is lacking entirely.

ACS Paragon Plus Environment

5

Environmental Science & Technology

104

Materials and methods

105

Study population

106

Page 10 of 33

During winter 2013–2014, 61 women and men living in the Oslo area, Norway were

107

enrolled in the Advanced Tools for Exposure Assessment and Biomonitoring (A-TEAM)

108

project,36 where serum, plasma, and whole blood were obtained. The participants comprised

109

45 women and 16 men (~75% women) between the age of 20 and 66 years old, with a median

110

age of 41 years. The Regional Committees for Medical and Health Research Ethics in Norway

111

(2013/1269) approved the sampling campaign of the A-TEAM project. In addition, all

112

participants gave a written consent before participating.

113 114 115

Sample collection Blood samples were drawn from a single venipuncture site. Whole blood samples were

116

collected in K2-ethylenediaminetetraacetic acid (EDTA) anticoagulant vacutainer tubes

117

(Becton, Dickinson and Company, Plymouth, UK). After a fraction of the whole blood was

118

transferred to a 2 mL polypropylene tube, the remaining blood in the EDTA vacutainer tube

119

was left to clot for one hour, centrifuged for 15 min at 2200-2500 rpm and then plasma was

120

transferred to a PP bottle. Vacutainer tubes without anticoagulant were used to collect blood

121

to obtain serum from the participants. The blood in the vacutainer tube without anticoagulant

122

was left to clot for one hour, centrifuged for 15 min at 2200-2500 rpm and then serum was

123

transferred to a PP bottle. All samples were stored at -20°C until further analysis.

124 125

Chemicals and standards

126

The twenty-five PFASs included in this study were 6:2 polyfluoroalkyl phosphate monoester

127

(6:2PAP), 8:2 polyfluoroalkyl phosphate monoester (8:2PAP), 6:2 polyfluoroalkyl phosphate

128

diester (6:2diPAP), 8:2 polyfluoroalkyl phosphate diester (8:2diPAP),

ACS Paragon Plus Environment

6

Page 11 of 33

Environmental Science & Technology

129

perfluorohexylphosphonate (PFHxPA), perfluorooctylphosphonate (PFOPA),

130

perfluorodecylphosphonate (PFDPA), perfluorobutanesulfonate (PFBS), PFHxS,

131

perfluoroheptanesulfonate (PFHpS), PFOS, perfluorodecanesulfonate (PFDS),

132

perfluoropentanoate (PFPeA), perfluorohexanoate (PFHxA), perfluoroheptanoate (PFHpA),

133

PFOA, PFNA, perfluorodecanoate (PFDA), perfluoroundecanoate (PFUnDA),

134

perfluorododecanoate (PFDoDA), perfluorotridecanoate (PFTrDA), perfluorotetradecanoate

135

(PFTeDA), PFOSA, N-methyl perfluorooctanesulfonamide (MeFOSA), and N-ethyl

136

perfluorooctanesulfonamide (EtFOSA). Details on the twenty-five native PFASs and eleven

137

isotope-labeled internal standards are given in Table S1 of supporting information. Since

138

PFSAs and PFCAs have been detected in human blood worldwide, but analyses of their

139

precursors have seldom been performed. Thus, this study included PFSA and PFCA

140

precursors, which were FOSAs and PAPs, respectively. In contrast, no studies have so far

141

detected PFPAs in human blood, and therefore only PFPAs were selected in this study, not

142

their precursors compounds (i.e. PFPiAs).

143 144 145

Blood analysis PFAS concentrations in serum, plasma, and whole blood samples were determined

146

using an established method by Poothong et al.37 Briefly, 50 µL of blood (serum, plasma, or

147

whole blood) was added to a 2 mL centrifuge tube, and then 90 µL of a 5 ng mL-1 internal

148

standard solution and 90 µL of methanol were added. To precipitate the proteins, the sample

149

tubes were mixed on a whirl mixer and centrifuged for 40 min at 14000 rpm at 20°C. The

150

supernatant was transferred to a 250 µL polypropylene vial, and then 80 µL of the sample was

151

injected in an online-SPE-UHPLC-MS/MS system.

152 153

Quality assurance/quality control

ACS Paragon Plus Environment

7

Environmental Science & Technology

154

Page 12 of 33

Duplicate matrix-matched calibration standards were prepared in twelve different

155

concentrations in the range 0.006–45 ng mL-1 blood. Newborn calf serum (Invitrogen, Oslo,

156

Norway), calf plasma, and calf whole blood (Lampire Biological labs, Pipersville, USA) were

157

used for the preparation of the matrix-matched calibration standards for serum, plasma, and

158

whole blood analysis, respectively. The matrix-matched calibration standards were injected

159

before and after the samples to ensure linearity in instrument response during the analysis and

160

for monitoring possible drifts in the sensitivity. Quantification was performed using

161

concentration-weighted (1/concentration) linear regression. For quantification of PFOS, the

162

total area of linear and branched isomers was integrated. Three solvent procedure blanks and

163

three non-spiked calf serum, calf plasma or calf whole blood were included in the analysis to

164

monitor the PFASs background levels. To assure high quality of the determinations, in-house

165

quality control samples of human plasma and whole blood as well as human serum samples

166

from an interlaboratory comparison study (Arctic Monitoring and Assessment Programme,

167

AMAP) (n = 3) were analyzed along with the human serum, plasma, and whole blood

168

samples. The results of the serum, plasma, and whole blood quality control samples were

169

similar to the previous results of the application in method developments (lower than 15%

170

relative standard deviations, RSD).37 More details on the quality assurance and quality control

171

including assessment of method accuracy, repeatability, precision, and method detection

172

limits can be found in Poothong et al.37

173 174 175

Data analysis To calculate mean and median concentrations, concentrations below the method

176

detection limits (MDLs) were replaced with their MDLs divided by the square root of two

177

(MDL/√2), while values between the MDLs and method quantification limits (MQLs) were

178

used unaltered. This approach was also employed when evaluating the composition profiles

ACS Paragon Plus Environment

8

Page 13 of 33

Environmental Science & Technology

179

and correlations in the matrices. A Mann-Whitney U-test was used to assess differences in

180

PFAS concentrations between genders and to compare current levels with concentrations

181

reported in another similar study from Norway. Statistically significant differences in PFAS

182

concentrations with respect to age-tertiles were evaluated using a Kruskal-Wallis test.

183

Correlations between compounds were evaluated with Spearman's rank correlation coefficient

184

(rho) and tested for significance.

185

To evaluate the distribution ratio of PFASs between different blood matrices, only

186

paired samples from individuals with quantifiable concentrations (>MDL) in serum, plasma,

187

and whole blood were included. Since the PFAS concentration ratios between serum, plasma,

188

and whole blood are expected to be similar in all individuals, correlations between PFAS

189

concentrations in pairs of blood matrices were evaluated using Pearson correlation (pairwise

190

comparison) and tested for significance. A Wilcoxon signed ranks test was used for assessing

191

the statistically significant differences of PFASs in different blood matrices.

192

P-values < 0.05 were considered significant in all statistical tests. A Shapiro-Wilk W

193

test was used to test the data for normality. Statistical analyses were performed by using the

194

SPSS 23 software, except heat-map correlation plots which were made using the corrgram

195

package in R version 3.2.2.38, 39

ACS Paragon Plus Environment

9

Environmental Science & Technology

196

Results and discussion

197

Detection frequencies and concentrations of PFASs in serum, plasma, and whole blood

198

Page 14 of 33

Twenty-one of the twenty-five PFASs were detected in blood samples, although the

199

detection frequencies varied depending on the compounds and the blood matrices. PFHxS,

200

PFHpS, PFOS, PFOA, and PFNA were detected in all serum (n=61), plasma (n=59), and

201

whole blood (n=58) samples. Approximately 50–100% of all blood matrices had detectable

202

concentrations of 6:2diPAP, PFHxPA, PFBS, PFDS, PFDA, PFUnDA, PFDoDA, PFTrDA,

203

and PFOSA, while the detection frequency was more than 50% in some of the blood matrices

204

for 6:2PAP, PFHxA, and PFTeDA as shown in Table 1. No samples had PFDPA, PFPeA,

205

MeFOSA, and EtFOSA levels above their respective MDLs (0.0036–0.09 ng mL-1) (Table

206

S2), while 8:2diPAP, PFHpA were detected in all matrices but in less than 50% of the

207

samples.

208 209

PAPs

210

Among the PAPs, 6:2diPAP had the highest detection frequency in all blood matrices

211

(49–98%, plasma>whole blood>serum). 6:2PAP and 8:2diPAP were observed in 3–73% and

212

21–30% of the samples in all blood matrices, respectively, whereas 8:2PAP was detected only

213

in 5% of the whole blood samples. Interestingly, 6:2PAP was present in 73% of the plasma

214

samples while it has never been reported in human blood before. Very few studies have

215

previously determined the concentrations of PAPs in human blood, and most of them assessed

216

only serum. Detection frequencies of serum diPAPs in the present study were comparable to

217

other studies of PAPs in serum.6, 40 The serum concentrations of 6:2PAP, 6:2diPAP, and

218

8:2diPAP in this present study were in range of PFPAs of equal perfluorocarbon chain length.47 Assuming similar bioaccumulation

242

potentials in humans, the low levels of PFPAs in blood samples might be due to a fast

243

elimination rate in human body. However, to obtain more knowledge on this, assessments of

244

human exposure pathways for PAPs and PFPAs are needed. In addition, the relatively low

ACS Paragon Plus Environment

11

Environmental Science & Technology

Page 16 of 33

245

levels of PAPs and PFPAs observed in this study indicates that there is still a large amount of

246

organofluorine compounds in blood which are unidentified.21

247 248 249

PFSAs, PFCAs, and FOSAs Interestingly, PFHxA was detected in 100% of the whole blood samples while it was

250

not detected in any of the other blood matrices. This suggests that whole blood is the only

251

suitable blood matrix for determination of PFHxA, and that the exposure to PFHxA is

252

overlooked when assessing serum or plasma. These findings are in line with several recent

253

studies where PFHxA was not detected, or found in very few samples when assessing

254

serum/plasma.6, 42, 48-49

255

PFOSA was the only FOSA detected in any of the blood matrices (71–100%, whole

256

blood>serum>plasma), where the concentrations in whole blood were approximately one

257

order of magnitude higher than the ones in serum and plasma. The reason for detecting

258

PFOSA and PFHxA in higher frequencies and concentrations in whole blood compared to

259

serum and plasma is likely due to binding of the compounds to blood cells.

260

To assess time trends of PFSA s and PFCAs, levels from this study were compared to

261

a previous Norwegian study where samples were collected in 2007–2008 (41 women, age 28–

262

46, median 36).45 A significant decrease in the median PFOS concentration of around 35%

263

was observed (Figure S1). Surprisingly, no statistical significant change was observed for

264

PFOA (Mann-Whitney U test, p>0.05), even though PFOS and PFOA have comparable

265

elimination half-lives (PFOS: 4.8 years, PFOA: 3.5 years).50 Thus, either the exposure to

266

PFOA has not changed to the same extent as PFOS, or the direct PFOA exposure has

267

decreased but the indirect exposure through the biotransformation of precursor compounds to

268

PFOA has substantially increased. In addition, for PFUnDA no statistical change in the serum

269

concentrations was observed between the 2007/2008 study and the recent study. On the other

ACS Paragon Plus Environment

12

Page 17 of 33

Environmental Science & Technology

270

hand, increasing PFAS concentrations were observed between 2007/2008 and 2013/2014 for

271

PFHpS (74%), PFHxS (53%), PFDA (45%), and PFNA (31%). Increasing concentrations of

272

some long-chain PFCAs such as PFNA, PFDA, and PFUnDA have also been reported in

273

recent time trend studies in Danmark,51 Sweden,41, 52 and Japan.53 PAPs are metabolized into

274

long-chain PFCAs.27, 43 Thus despite direct exposure, the increasing levels of these long-chain

275

PFCAs might also relate to the indirect exposure from PAPs, which are still used in consumer

276

products.

277

Several studies have determined PFSAs, PFCAs, and FOSAs in human blood. PFOS

278

serum concentrations in the present study (median, 5.2 ng mL-1) were comparable to studies

279

from Sweden (2012)41 and Denmark (2008-2013),51 while studies in Australia (2010-2011)54

280

and the U.S. (2009)40 found PFOS concentrations in serum which were two times higher than

281

in this study. On the other hand, the PFOS median concentration reported in a study from

282

China (2009) was approximately half of that of the present study.42 The median serum

283

concentration of PFOA was 1.9 ng mL-1, which is similar to recent studies from the U.S.40 and

284

Denmark,51 while it was two times lower than in Australia.54 This study found similar median

285

serum concentrations of PFNA (0.9 ng mL-1), PFDA (0.4 ng mL-1), and PFUnDA (0.4 ng mL-

286

1

287

levels varied among the studies.

) as studies from the U.S.,40 Australia,54 Sweden,41 Denmark,51 and China,42 while PFHxS

288 289

PFAS profiles in serum, plasma, and whole blood

290

The relative composition profiles of PFASs in serum, plasma, and whole blood was

291

evaluated for PFASs with detection frequencies above 50% in each blood matrix (Table 1).

292

Total PFAS concentrations (ΣPFASs) were 10, 10, and 6.3 ng mL-1 in serum, plasma, and

293

whole blood, respectively. Significant difference of PFAS concentrations between ages and

294

genders were assessed. In whole blood, PFHxS, PFHpS, and PFOS concentrations in women

ACS Paragon Plus Environment

13

Environmental Science & Technology

295

were significantly lower than in men (Mann−Whitney U test, p ≤ 0.005, 2-tailed) (Table S3),

296

which is possibly linked to menstruation. Based on a pharmacokinetic model, menstruation

297

accounted for 30% of the difference in PFOS elimination between genders.55 No statistically

298

significant difference was observed between genders in the other compounds.

299

Page 18 of 33

The oldest age group (age-tertiles, 45) had significantly higher PFOA,

300

PFNA, PFDA, PFUnDA, PFTeDA, PFDS, and PFOSA concentrations in blood compared to

301

the other two age groups (Kruskal-Wallis test, p ≤ 0.05), while this was not the case for the

302

other compounds. The median relative PFAS compositions (ng mL-1) in serum, plasma, and whole blood

303 304

were all dominated by PFOS, followed by PFOA (Figure 1). The sum of PFOA and PFOS

305

contributed to 60–70% of the total PFAS concentrations. The contribution of PFPAs and

306

PAPs to the total PFAS concentrations in human blood were relatively low, and thus the

307

major fluorinated contaminants measured in this study in human blood are still PFSAs and

308

PFCAs. The PFAS composition profiles in serum and plasma were similar, with the

309

proportion of PFOS (48–51%) > PFOA (16-19%) > PFNA ≈ PFHxS (7–9%) > PFUnDA ≈

310

PFDA (3–4%) > PFHpS ≈ PFHxPA (2–3%) > PFTrDA ≈ PFDoDA ≈ PFDS ≈ PFBS ≈

311

PFOSA (1%). The 6:2diPAP contributed to 1% in plasma. Also the low detection frequency

312

PAP congener, 6:2PAP contributed to a greater percentage in plasma by 7%. Consequently, in

313

this study plasma proved to be a relevant matrix to quantifying PAPs and PFPAs in human

314

blood.

315

The relative composition of various PFASs in whole blood differed from serum and

316

plasma. PFHxA was only detected in whole blood, where it accounted for as much as 10% of

317

the total PFAS concentrations. Thus the composition profile of PFASs in whole blood was

318

PFOS (45%) > PFOA (15%) > PFHxA (10%) > PFNA ≈ PFHxS (6–7%) > PFUnDA ≈ PFDA

319

≈ PFTrDA ≈ PFOSA (2–3%) > PFTeDA ≈ PFDoDA ≈ PFDS ≈ PFHpS ≈ PFBS ≈ PFHxPA ≈

ACS Paragon Plus Environment

14

Page 19 of 33

Environmental Science & Technology

320

6:2diPAP (1%). Interestingly, whole blood was the matrix where the largest number of PFASs

321

was detected (16 PFASs), while serum and plasma have been the most frequently used

322

matrices in biomonitoring studies. In accordance with these differing PFAS profiles in serum,

323

plasma, and whole blood, epidemiological studies need to be taken into consideration when

324

associating PFASs levels in the blood with the health outcome.

325 326 327

Correlations between PFASs within the same blood matrix Correlations between different PFASs (detection frequency >50%) within the same

328

blood matrix were evaluated based on Spearman's rank correlation coefficients, and are

329

presented in Figure 2. Both weak positive and negative correlations were observed between

330

PAPs, as well as PFHxA with the other PFASs. However, weak but significant correlations

331

were observed between 6:2diPAP and PFHxA in whole blood (rs = 0.38, p