Perfluorinated organic compounds in human blood serum and

seminal plasma: a study of urban and rural tea worker populations

in Sri Lanka

Keerthi S. Guruge,*a Sachi Taniyasu,b Nobuyoshi Yamashita,b Sumedha Wijeratna,c

Keerthi M. Mohotti,d Harsha R. Seneviratne,c Kurunthachalam Kannan,e Noriko Yamanakaa

and Shigeru Miyazakia

aToxico-Biochemistry Section, National Institute of Animal Health, Kannondai 3-1-5, Tsukuba,Ibaraki 305-0856, Japan. E-mail: guruge@affrc.go.jp; Fax: þ81-29-838-7825;Tel: þ81-29-838-7822

bEnvironmental Measurement Group, National Institute of Advance Industrial Science andTechnology, Onogawa 16-1, Tsukuba, Ibaraki 305-8569, Japan

cDepartment of Obstetrics & Gynecology, Faculty of Medicine, University of Colombo,Colombo 8, Sri Lanka

dTea Research Institute of Sri Lanka, Talawakele, Sri LankaeWadsworth Center, New York State Department of Health and Department of EnvironmentalToxicology and Health, State University of New York, D547 Empire State Plaza,P.O. Box 509, Albany, NY 12201-0509, USA

Received 13th August 2004, Accepted 11th February 2005First published as an Advance Article on the web 7th March 2005

Concentrations and accumulation of 13 fluorinated organic compounds (FOCs) in human sera andseminal plasma were measured in an Asian developing country, Sri Lanka. Six of the FOCs, PFOS(perfluorooctanesulfonate), PFHS (perfluorohexanesulfonate), PFUnA (perfluoroundecanoic acid), PFDA(perfluorodecanoic acid), PFNA (perfluorononanoic acid) and PFOA (perfluorooctanoic acid), were detectedin all of the sera samples. Measurable quantities of two main perfluorosulfonates, PFOS and PFHS, werefound in all seminal plasma samples. The detection frequency of the predominant perfluoroalkylcarboxylate,PFOA, in seminal plasma was 470%. Accumulation of PFOS in sera was significantly positively correlatedwith PFOA, PFHS and PFNA. Positive linear regressions were also found between PFNA and PFUnAand PFNA and PFDA suggesting that these compounds may have a similar origin of exposure andaccumulation. Significantly positive associations were observed for partitioning of both PFOS and PFNAbetween sera and seminal plasma. The accumulation of FOCs was not significantly different in sera fromColombo (urban population) and Talawakele (rural conventional tea workers). However, the Haldummullapopulation (rural organic tea workers) had relatively lower exposure to FOCs compared to the other twogroups, urban and rural conventional tea workers. Concentrations of FOCs in Sri Lanka were similar tothose reported for industrialized countries suggesting that human exposure to such chemicals is widespreadeven in developing countries. The novel finding of FOCs in human seminal plasma implies that furtherstudies are needed to determine whether long-term exposure in humans can result in reproductiveimpairments.

Introduction

Occurrence of perfluorinated chemicals in the environment andwildlife recently has raised considerable public health concernson the effects of these chemicals. Fluorinated organic com-pounds (FOCs) are constituents in a wide range of applicationssuch as liquid repellants for paper, packaging, textile, leather,adhesives, insecticides and other industrial products.1 Themain two groups of FOCs, perfluoroalkylsulfonates and per-fluoroalkylcarboxylates, were found in many environmentalcompartments including water, sediment and biota. It wasreported that FOCs were found in several species of wildlifefrom various locations including some remote areas.2–4 Fishand aquatic animals accumulated greater concentrations ofPFOS and PFOA with no clear age- or sex-related differ-ences.5,6 FOCs bind to the serum albumin and are found inthe protein fraction of blood.7 Studies have reported theoccurrence of FOCs in non-occupationally exposed humansin developed countries such as the USA, Japan and selectedEuropean countries.6,8–10 High concentrations of perfluoro-octanesulfonate (PFOS) and perfluorooctanoic acid (PFOA),

of 13 and 114 mg ml�1, respectively, were found in the sera ofoccupationally exposed populations.11,12

Potential health effects of FOCs were summarized by Giesyand Kannan13 reporting that these compounds can cause liverdamage, hypolipidemia, peroxisome proliferation, and tumorpromotion in laboratory experiments. Long chain perfluori-nated fatty acid analogues can interfere with lipid metabolismby increasing peroxisomal fatty acid b-oxidation, and byinducing several liver enzyme activities.14 Cardiac malforma-tions and maternal and developmental toxicity have beendemonstrated in both rats and mice exposed to PFOSin utero.15 Chronic exposure of primates to PFOS resulted inaltered blood cholesterol concentrations.16

A number of animal and human studies suggest possibleassociations of exposure to organohalogen compounds withaltered reproductive parameters and functions.17,18 However,to our knowledge, no study has been carried out to characterizethe distribution of FOCs in seminal plasma and its relation toother body fluids. Moreover, apart from a few compounds,occurrence in human blood of long-chain FOCs such asPFNA, PFDA, and PFUnA is not well documented. To date,

P A P E R

ww

w.rsc.o

rg/jem

DO

I:1

0.1

03

9/b

41

25

32

k

J . E n v i r o n . M o n i t . , 2 0 0 5 , 7 , 3 7 1 – 3 7 7 3 7 1T h i s j o u r n a l i s & T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5

FOC contamination in developing countries has been paidlittle or no attention. Hence, the purpose of the present studywas to report the prevalence of 13 FOCs in human blood seraand seminal plasma and to observe serum to seminal plasmarelationships in three groups of volunteers belonging to threelocations in Sri Lanka.

Experimental procedures

Human blood sera and seminal plasma were collected fromvolunteer donors in 2003 in Sri Lanka. Donors representedthree population groups from different backgrounds. The firstgroup was from Colombo, which is the capital and mostdensely populated urban area in Sri Lanka (n ¼ 10; 29–48years). The second group (n ¼ 10; 30–45 years) representedworkers at a conventional tea cultivation area in Talawakele,and the third group (n ¼ 10; 24–61 years) represented thosewho were involved in organic tea cultivation in Haldummulla.The latter two groups represented up-country rural popula-tions in Sri Lanka. All volunteers were asked to complete aquestionnaire about age, weight, parity, and occupation. Theidentity of individual donors was not revealed. The tea workersfrom Talawakele had been occupationally exposed to variouspesticides in tea plantations for more than 10 years. Theorganic tea cultivation in Haldummulla consisted of a popula-tion which had been away from organic pesticides for nearly 18years. The blood samples were collected by venipuncture. Priorto semen collection, the subjects were instructed to abstainfrom sexual activities for 3 days. At the study visit mencollected semen samples by masturbation at clinics in ColomboUniversity or in the tea estates. Samples were collected into50 ml polypropylene tubes pre-cleaned with analytical grademethanol. Sera and seminal plasma were prepared by centri-fugation at 3000 rpm for 15 min according to the World HealthOrganization’s criteria, and they were kept at �20 1C untilanalysis.19

Samples were analyzed for 13 FOCs: perfluorobutanesulfo-nate (PFBS), perfluorohexanesulfonate (PFHS), perfluorooc-tanesulfonate (PFOS), perfluoropentanoic acid (PFPeA),perfluorohexanoic acid (PFHxA), perfluoroheptanoic acid(PFHpA), perfluorooctanoic acid (PFOA), perfluorononanoicacid (PFNA), perfluorodecanoic acid (PFDA), perfluorounde-canoic acid (PFUnA), perfluorododecanoic acid (PFDoA),perfluorooctanesulfonylamide (PFOSA) and 1H,1H,2H,2H-perfluorooctanesulfonate (THPFOS). Extraction was carriedout by an ion-pairing method, which is described elsewhere.6,8

Briefly, 1 ml of serum or seminal plasma was mixed with 1 mlof 0.5 M tetra-n-butylammonium hydrogen sulfate solutionand 2 ml of buffer (pH 10, 0.25 M) in a polypropylene (PP)tube. The sample mixture was extracted with 5 ml of methyltert-butyl ether (MTBE) by shaking for 20 min followed bycentrifugation. A 4 ml aliquot of MTBE was removed from thesolution and placed in a second PP tube. The extraction wasperformed for twice more, 5 ml of MTBE was removed eachtime and combined in the second PP tube. The final extract wasconcentrated under nitrogen after adding 0.5 ml of methanol.The sample was passed through a 0.1 mm nylon filter beforeinjection.

Analysis of FOCs was performed using a high performanceliquid chromatograph-tandem mass spectrometer (HPLC-MS/MS), comprising an Agilent HP1100 liquid chromatographinterfaced with a Micromasss (Beverly, MA, USA) QuattroUltima Pt mass spectrometer operated in the electrospraynegative ionization (ESNI) mode. A 10 ml aliquot of the sampleextract was injected into a guard column (XDB-C8, 2.1 mmi.d. � 12.5 mm, 5 mm; Agilent Technologies, Palo Alto, CA)connected sequentially to a Betasil C18 column (2.1 mm i.d. �50 mm length, 5 mm; Termo Hypersil-Keystone, Bellefonte,PA) with 2 mM ammonium acetate–methanol as mobile phase,starting at 10% methanol. At a flow rate of 300 ml min�1, the

gradient was increased to 30% methanol at 0.1 min, 75%methanol at 7 min, and 100% methanol at 10 min. The systemwas switched back to original conditions at 12 min and waskept at these conditions until 20 min. The capillary was held at1.2 kV. Cone-gas and desolvation-gas flows were kept at 60and 650 l h�1, respectively. Source and desolvation tempera-tures were kept at 120 and 420 1C, respectively. MS/MSparameters were optimized so as to transmit the [M–K]� or[M–H]� ions as shown in Table 1. Eight calibration curvepoints bracketing the concentrations in samples were preparedroutinely, to check for linearity. To reduce the detection limit,all accessible PTFE tubes in the solvent inlet filter unit andHPLC system were replaced with stainless steel or PEEK tubeswhile no degasser or solvent selection valves were used. Ad-ditionally suitable PP tubes and septa were selected afterthorough blank checking. The mass determination, proceduralblank values, and recovery of target analytes are given in Table1. The blank value was subtracted from the sample data. Forthe data calculation, concentrations less than the blank valueswere not included.

Results and discussion

Concentrations in blood sera

The concentrations of FOCs in sera and seminal plasma aregiven in Table 2. The sample concentrations of PFBS, PFPeA,PFOSA, and THPFOS were less than those for blank concen-trations (i.e., below LOQ). Hence the data for these com-pounds are not presented. Of the thirteen FOCs monitored,PFOA was the predominant compound in two groups with amean value of 9.5 and 9.1 ng ml�1, respectively, in the urbanand rural conventional tea worker populations. Although over50% of the urban subjects had PFOA 410 ng ml�1 in bloodsera, only three individuals from rural areas contained similarconcentrations of PFOA to the urban population. However,the greatest PFOA concentration (23.5 ng ml�1) was found inthe serum of a rural conventional tea worker from Talawakele.The highest PFOA concentration in the sera of adults (20–69 y)in the US general population was 52 ng ml�1, which was morethan twice the highest concentration measured in Sri Lanka.20

The greatest PFOA concentration found in sera of ruralorganic tea workers was 0.855 ng ml�1. The PFOA exposurein urban and rural conventional tea workers was nearly 18-foldgreater than that in rural organic tea workers. In the urbangroup, 7 out of 10 subjects had PFOS concentrations greaterthan PFOA concentrations. However, the overall PFOA con-centration (6.4 ng ml�1) in Sri Lankan blood sera was greater

Table 1 Mass determination of analyte, procedural solvent blank and

recovery (n ¼ 5)

Analyte

Primary and

product ions (m/z)

Solvent

blank/pg ml�1Recovery

(%)

THPFOS 426.7 4 406.7 2 � 1a 93 � 9a

PFOSA 497.7 4 77.7 10 � 3 88 � 9

PFOS 498.6 4 79.7 10 � 2 80 � 9

PFHS 398.7 479.7 2 � 0.6 85 � 13

PFBS 298.7479.7 3 � 0.6 61 � 9

PFDoA 612.7 4 568.8 4 � 1 64 � 9

PFUnA 563 4519 5 � 2 69 � 10

PFDA 512.8 4 468.8 9 � 3 85 � 11

PFNA 462.7 4 418.8 8 � 3 86 � 4

PFOA 413 4 368.7 72 � 17 110 � 10

PFHpA 362.8 4 318.8 15 � 4 95 � 7

PFHxA 312.8 4 268.8 20 � 11 60 � 26

PFPeA 262.8 4 218.7 3 � 3 75 � 3

a Mean � S.D.

3 7 2 J . E n v i r o n . M o n i t . , 2 0 0 5 , 7 , 3 7 1 – 3 7 7

Table

2Concentrationsoffluorinatedorganic

compoundsin

humanserum

andseminalplasm

a(ngml�

1)from

SriLanka

Location

Age

PFOS

PFHS

PFDoA

PFUnA

PFDA

PFNA

PFOA

PFHpA

PFHxA

(year)

Serum

SPa

Serum

SP

Serum

SP

Serum

SP

Serum

SP

Serum

SP

Serum

SP

Serum

SP

Serum

SP

Urban

38

1.50

0.043

0.169

0.016

0.014

o0.004

0.006

0.002

0.009

0.006

0.120

0.010

0.63

0.009

0.045

0.022

o0.020

o0.020

Colombo

48

13.7

0.162

1.395

0.024

0.020

o0.004

0.184

0.004

0.007

0.003

0.470

0.018

12.1

0.470

0.243

0.093

0.135

o0.020

32

6.20

0.106

0.457

0.145

0.016

o0.004

0.235

0.003

0.185

0.009

0.210

o0.008

4.71

0.140

0.029

0.012

0.120

o0.020

38

8.00

0.200

1.127

0.030

0.016

o0.004

0.244

0.004

0.198

0.006

0.449

0.005

22.8

0.211

1.359

0.028

3.300

o0.020

29

7.80

0.078

0.350

0.039

0.012

o0.004

0.178

0.001

0.193

0.007

0.252

0.007

17.6

0.320

0.126

0.016

0.155

o0.020

43

6.40

0.170

1.032

0.031

0.012

o0.004

0.188

0.002

0.169

0.003

0.250

0.009

4.95

0.143

0.167

0.019

0.825

o0.020

32

4.95

0.140

0.400

0.023

0.016

o0.004

0.205

0.002

0.201

0.007

0.345

0.007

11.9

0.156

1.447

0.016

3.000

o0.020

32

3.16

0.110

0.215

0.019

0.009

o0.004

0.119

o0.005

0.174

o0.009

0.080

0.020

1.95

0.214

0.100

0.025

0.100

o0.020

33

8.10

0.128

1.443

0.024

0.014

o0.004

0.153

0.003

0.091

0.006

0.377

0.015

7.31

0.507

0.134

0.006

0.330

o0.020

31

18.2

0.110

1.200

0.018

0.087

o0.004

0.980

0.015

0.680

o0.009

1.300

0.027

11.3

0.210

0.157

0.079

0.230

o0.020

Mean

36

7.80

0.125

0.779

0.037

0.022

o0.004

0.249

0.004

0.191

0.005

0.385

0.012

9.54

0.241

0.381

0.032

0.820

o0.020

Median

33

7.10

0.119

0.745

0.024

0.015

o0.004

0.186

0.002

0.180

0.006

0.299

0.010

9.32

0.210

0.146

0.021

0.193

o0.020

S.D

.b6.1

4.91

0.046

0.506

0.039

0.023

o0.004

0.266

0.004

0.188

0.003

0.347

0.008

7.06

0.179

0.543

0.029

1.251

o0.020

Ruralc

44

1.20

0.042

0.132

0.009

0.003

o0.004

0.047

0.003

0.023

0.006

0.063

0.002

0.40

0.114

o0.015

o0.015

0.012

0.025

Haldummulla

29

1.03

0.010

0.077

0.008

o0.004

o0.004

0.034

o0.005

0.017

0.006

0.031

0.025

0.86

0.108

o0.015

o0.015

0.010

0.013

37

0.35

o0.010

0.055

0.006

o0.004

o0.004

0.004

o0.005

0.002

o0.009

0.003

0.006

0.32

0.145

o0.015

o0.015

0.041

0.025

36

1.20

o0.010

0.137

0.009

0.006

o0.004

0.065

0.001

0.028

0.007

0.062

0.004

0.52

0.107

o0.015

o0.015

o0.020

0.016

28

1.33

0.140

0.070

0.035

0.004

o0.004

0.079

0.003

0.038

0.007

0.040

0.016

0.53

0.180

o0.015

o0.015

0.017

0.032

39

0.82

o0.010

0.055

0.002

o0.004

o0.004

0.020

o0.005

0.012

o0.009

0.017

o0.008

0.50

0.045

o0.015

o0.015

o0.020

o0.020

50

1.02

0.005

0.085

0.020

0.002

o0.004

0.039

0.003

0.024

0.004

0.055

0.002

0.65

0.081

o0.015

o0.015

0.002

0.020

49

0.86

0.040

0.061

0.048

o0.004

o0.004

0.051

0.004

0.036

0.004

0.045

0.001

0.40

0.117

o0.015

o0.015

0.058

0.028

61

0.64

0.150

0.060

0.017

o0.004

o0.004

0.024

o0.005

0.019

0.001

0.034

o0.008

0.39

0.090

o0.015

o0.015

o0.020

0.027

24

1.15

0.005

0.092

0.094

0.006

o0.004

0.052

o0.005

0.030

0.006

0.090

o0.008

0.89

0.101

o0.015

o0.015

0.123

0.005

Mean

40

0.96

0.039

0.082

0.025

0.002

o0.004

0.042

0.001

0.023

0.004

0.044

0.006

0.53

0.109

o0.015

o0.015

0.026

0.019

Median

38

1.02

0.008

0.074

0.013

0.001

o0.004

0.043

0.001

0.024

0.005

0.043

0.002

0.51

0.108

o0.015

o0.015

0.011

0.023

S.D

.11

0.30

0.058

0.030

0.028

0.003

o0.004

0.022

0.002

0.011

0.003

0.025

0.008

0.23

0.066

o0.015

o0.015

0.039

0.010

Rurald

32

1.88

0.039

0.111

o0.002

0.004

o0.004

0.018

o0.005

0.022

o0.009

0.048

o0.008

1.93

o0.072

0.008

0.005

0.032

0.026

Talawakele

45

7.51

0.529

1.032

0.055

0.007

o0.004

0.048

0.005

0.045

0.002

0.064

0.004

10.9

1.030

0.026

0.004

0.031

0.018

41

2.58

0.084

0.580

o0.002

0.008

o0.004

0.024

o0.005

0.028

o0.009

0.074

o0.008

4.21

o0.072

0.007

0.002

0.040

0.012

30

5.34

0.192

0.458

0.166

0.009

o0.004

0.077

0.026

0.068

0.019

0.141

0.026

3.83

2.130

0.016

0.006

0.074

0.012

40

7.33

0.100

0.683

o0.002

0.016

o0.004

0.082

o0.005

0.080

o0.009

0.160

o0.008

9.06

0.118

0.011

0.003

0.095

0.013

43

11.7

0.191

1.436

0.027

0.006

o0.004

0.035

o0.005

0.041

o0.009

0.086

0.002

15.6

0.567

0.014

0.002

0.043

0.013

40

17.5

0.488

2.050

0.064

0.008

o0.004

0.062

0.003

0.053

o0.009

0.089

0.005

23.5

1.710

0.035

0.030

0.045

0.054

35

1.73

0.014

0.199

0.004

0.006

o0.004

0.022

o0.005

0.110

o0.009

0.068

o0.008

3.74

o0.072

0.006

0.004

0.103

o0.020

40

3.36

0.096

0.813

0.007

0.007

o0.004

0.037

0.006

0.031

0.003

0.057

0.004

7.89

0.294

0.008

0.035

0.032

0.046

37

4.29

0.170

0.975

0.012

0.007

o0.004

0.036

0.003

0.039

o0.009

0.077

0.007

9.86

0.346

0.009

0.002

0.049

0.012

Mean

38

6.32

0.190

0.834

0.034

0.008

o0.004

0.044

0.004

0.051

0.002

0.087

0.005

9.06

0.621

0.014

0.009

0.054

0.021

Median

40

4.82

0.135

0.748

0.010

0.007

o0.004

0.037

0.002

0.043

o0.009

0.076

0.003

8.47

0.320

0.010

0.004

0.044

0.013

S.D

.4.8

4.99

0.178

0.583

0.052

0.003

o0.004

0.023

0.008

0.028

0.006

0.036

0.008

6.60

0.785

0.010

0.012

0.027

0.017

Overall(n¼

30)

Mean

38

5.03

0.118

0.565

0.032

0.011

o0.004

0.112

0.003

0.088

0.004

0.172

0.007

6.38

0.323

0.132

0.014

0.300

0.013

Median

38

3.26

0.103

0.376

0.020

0.007

o0.004

0.052

0.002

0.039

0.003

0.077

0.005

4.02

0.134

0.010

0.004

0.047

0.012

S.D

.7.9

4.92

0.125

0.553

0.040

0.016

o0.004

0.179

0.005

0.129

0.004

0.248

0.008

6.84

0.513

0.351

0.022

0.791

0.015

aSP¼

seminalplasm

a.bS.D

.¼

standard

deviation

cOrganic

teaworkers.

dConventionalteaworkers.

J . E n v i r o n . M o n i t . , 2 0 0 5 , 7 , 3 7 1 – 3 7 7 3 7 3

than the PFOS concentration (5.0 ng ml�1). One possiblereason for this observation was the greater recovery of PFOAcompared to that of PFOS. Generally, PFOS concentrationswere greater than those of PFOA in human sera in theliterature, although PFOA concentrations were greater thanPFOS occasionally.8 Occupationally exposed individuals hadhigher sera PFOA concentrations than PFOS concentra-tions.11,12 The overall PFOA concentration found in this studywas well within the range of PFOA concentrations in generalpopulations in developed countries such as the USA, butgreater than concentrations reported for human sera fromItaly and India (Table 3).

Although there is individual variation in the concentrationof FOCs, PFOS was the dominant compound in most samples.The mean PFOS concentration in the urban population (7.8 ngml�1) was 1.2 and 8 times higher than that of rural conven-tional tea workers (6.3 ng ml�1) and rural organic tea workers(0.96 ng ml�1). The PFOS concentration was nearly 2-foldgreater than that of PFOA in rural organic tea workers, whichis similar to values reported for human sera from the USA.Overall, only four volunteers, two each from the urban groupand the rural conventional tea workers, had over 10 ng ml�1

PFOS in sera. The highest PFOS concentration in Sri Lankanmales was 18.2 ng ml�1. This value was several times lowerthan the mean PFOS concentration found for the USA.9,20,21

The overall mean PFOS concentration in Sri Lanka (5.0 ngml�1, n ¼ 30) was greater than that found in Italy (4.4 ng ml�1)and India (1.7 ng ml�1).9

PFHS was found in all sera samples. The concentration ofPFHS in sera was similar in urban and conventional tea workerpopulations with 0.780 and 0.834 ng ml�1, respectively (Table2). The PFHS concentration in sera of organic tea workers(0.082 ng ml�1) was nearly 10-fold lower than that of the othertwo groups. The overall mean PFHS concentration in thepresent study was lower than those reported for the USA,Italy, and India (Table 3).

Reports of concentrations of FOCs such as PFDoA, PFU-nA, PFDA, PFNA, PFHpA and PFHxA are scarce in humanbody fluids. A recent study showed that such long-chain FOCsare found in remote wild animals.22 Concentrations of all ofthese long-chain FOCs in the urban population were severaltimes higher than those in rural groups. The concentration ofPFHxA (0.82 ng ml�1) was greater than PFNA (0.385 ngml�1), followed by PFHpA (0.381 ng ml�1), PFUnA (0.249ng ml�1), PFDA (0.191 ng ml�1) and PFDoA (0.022 ng ml�1)in urban sera (Table 2). A 100% detection frequency wasobserved for PFNA, PFUnA and PFDA in sera samples(n ¼ 30) while PFHpA (o0.015 ng ml�1) was not detected inany of the sera samples from rural organic tea workers. ThePFNA concentration in Sri Lankan sera was lower than, whilePFUnA was similar to, those from the USA.23 However, the

accumulation pattern was different in remote wild animals likepolar bears, in which PFOA concentration was a few timeslower than that of PFNA, PFDA, and PFUnA.22 This suggeststhat physico-chemical and accumulation properties andsources of exposures are important factors for human contam-ination. Overall, rural organic tea workers from Haldummullahad less exposure to FOCs compared to rural conventional teaworkers from Talawakele. Urban inhabitants from Colombohad the greatest exposure. None of the FOC concentrations insera were positively correlated with age.

Concentrations in seminal plasma

The data on the accumulation of FOCs in semen was novel tothis study. PFOA and PFOS concentrations in the seminalplasma of conventional tea workers were found to be thehighest among all the three groups studied. Mean concentra-tion of PFOA (0.241 ng ml�1) was nearly 2-fold greater thanthat of PFOS (0.125 ng ml�1) in urban seminal plasma. ThePFOA concentration in the seminal plasma of conventional teaworkers was (0.62 ng ml�1), which was 3-fold higher than thePFOS concentration (0.19 ng ml�1). The overall detectionfrequency for PFOS, PFHS and PFOA in seminal plasmawas 90%. The mean PFHS concentration measured in seminalplasma was 0.032 ng ml�1, while the greatest concentration was0.166 ng ml�1 found in the rural conventional tea workerpopulation in Talawakele. PFHpA was detected in all seminalplasma from the urban and conventional tea worker popula-tions while this compound was not found in the semen ofworkers from the organic tea plantation. PFUnA, PFDA, andPFNA were found in most urban seminal plasma samples with

Table 3 Concentrations (ng ml�1) of FOCs in human sera in various

countries

Country PFOA PFOS PFHS Ref.

USA 6.1–57 10

1.9–52.3 o4.3–1656 o1.4–66.3 20

1.4–16.7 o3.4 –175 o1.4–40.3 21

o5.0–35.2 6.7–81.5 o5.0–21.4 8

o3.0–14.7 o1.3–124 o1.3–13.6 9

Italy o3.0 o1.0–10.3 o1.0–1.4 9

India o3.0–3.5 o1.0–3.1 o1.0–1.8 9

Japana 19–41 o1.0–3.8a 6

a Whole blood.Fig. 1 Relationship between PFOS and PFHS concentrations inhuman sera (ng ml�1) from Sri Lanka.

Fig. 2 Relationship between PFOS and PFOA concentrations inhuman sera (ng ml�1) from Sri Lanka.

3 7 4 J . E n v i r o n . M o n i t . , 2 0 0 5 , 7 , 3 7 1 – 3 7 7

greater than 80% detection frequency, while they were rarelyfound in rural samples. PFDoA was not found in any seminalplasma samples (Table 2). These data suggest that similar toother human body fluids, seminal plasma also accumulatesmeasurable concentrations of various FOCs.

Accumulation properties of FOCs

The accumulation of PFOS and other main FOC residues suchas PFHS (Fig. 1, n ¼ 30, R2 ¼ 0.769, p o 0.01), PFOA (Fig. 2,n ¼ 30, R2 ¼ 0.636, p o 0.01) and PFNA (Fig. 3, n ¼ 30, R2 ¼0.466, p o 0.01) in sera was significantly positively correlated.A similar observation was reported for PFOS and PFOA inhuman sera in the USA.20,21 Positive linear regressions werefound between PFNA and PFUnA (Fig. 4, n ¼ 30, R2 ¼ 0.927,p o 0.01), and PFNA and PFDA (Fig. 5, n ¼ 30 R2 ¼ 0.791,p o 0.01) suggesting that these compounds may have similarorigins of exposure. However, excluding the highest concen-tration point in the regression, PFNA was not positivelycorrelated with PFUnA and PFDA (n ¼ 29). Nevertheless,correlation was positive for PFNA and PFUnA in eachpopulation such as R2 ¼ 0.356, 0.423 and 0.771 for Colombo(n ¼ 9, p o 0.05), Talawakele (n ¼ 10, p o 0.02) andHaldummulla (n ¼ 10, p o 0.01), respectively. Likewise,PFNA and PFDA regressions were positively correlated inboth rural populations as Talawakele (n ¼ 10, R2 ¼ 0.449, p o0.02) and Haldummulla (n ¼ 10, R2 ¼ 0.221, p4 0.05). Martinet al.22 reported that long-chain perfluorocarboxylates such asPFNA, PFDA and PFUnA were commonly found in Arctic

mammals and their accumulations were positively correlated.PFOS is a major metabolic product of POSF (perfluoroocta-nesulfonyl fluoride)-based compounds, which are widely usedas surfactants. Fluorotelomer alcohols are degraded to fluoro-alkyl carboxylic acids such as PFOA, PFNA, PFDA andPFUnA. The aqueous film-forming foams (AFFF), which areused to extinguish fires, contain mixtures of FOCs includingfluorotelomer sulfonates.24 PFOA, PFOS and PFHS are alsoimpurities in fire fighting foams. Occurrence of FOCs inaquatic products including water and fish may be a source ofexposure to humans.6,25 We also found that beef cattle canaccumulate all of the above FOCs up to a few hundred pg ml�1

concentration in their blood suggesting that ingestion ofanimal products may be another source of human exposure.26

However, n-ethyl perfluorooctanesulfonamide, which is aninsecticide (Sulfluramide) used to control termites and ants,was not used in Sri Lanka.Significant positive correlations were observed for PFOS

(Fig. 6, n ¼ 27, R2 ¼ 0.3199, p o 0.001) and PFNA (Fig. 7,n ¼ 22, R2 ¼ 0.2161, p o 0.01) concentrations in blood seraand seminal plasma. This could be explained by the long serumhalf-life of PFOS and higher interaction with fatty-acid bind-ing proteins.27 However, a weak partitioning (n ¼ 27, R2 ¼0.152, p 4 0.01) was observed for PFOA between serum andseminal plasma. This may be indicative of a lack of equilibriumin PFOA partitioning between serum and seminal plasma. Theblood serum to seminal plasma concentration ratios for majorFOCs such as PFOS, PFOA, PFHS and PFNA were 42 : 1,16 : 1, 21 : 1 and 20 : 1, respectively. These ratios wereonly calculated for donors whose serum and seminal plasma

Fig. 3 Relationship between PFOS and PFNA concentrations inhuman sera (ng ml�1) from Sri Lanka.

Fig. 4 Relationship between PFUnA and PFNA concentrations inhuman sera (ng ml�1) from Sri Lanka.

Fig. 5 Relationship between PFDA and PFNA concentrations inhuman sera (ng ml�1) from Sri Lanka.

Fig. 6 Relationship of PFOS concentrations between blood sera andseminal plasma (ng ml�1) from Sri Lanka.

J . E n v i r o n . M o n i t . , 2 0 0 5 , 7 , 3 7 1 – 3 7 7 3 7 5

concentrations were above the detection limits. Liver to serumconcentration ratios of PFOS have been found to be 1.4 : 1 forhumans suggesting that FOC partitioning to seminal plasmawas not preferential.10

It is expected that humans and animals that live in morepopulated and industrialized locations have greater exposurelevels of FOCs.2 However, mammals feeding at higher trophiclevels in remote areas had greater concentrations of FOCs.23 Inthe present study, urban males from Colombo and ruralconventional tea workers in Talawakele had significantly high-er PFOS, PFHS, PFOA and PFNA concentrations in their seracompared to rural organic tea workers (p o 0.01, 2-tailedt-test). We did not observe any statistical significance in FOCconcentrations in seminal plasma among these three groups ofsubjects (p 4 0.01, 2-tailed t-test). The greater exposure levelsfound in Colombo are indicative of sources from variousindustries and frequent use of products that contain FOCssuch as paper, packing products, carpet spray, stain-resistanttextiles, cosmetics, electronics and fire-fighting foams, whichare widespread in an urban community. On the contrary, someindividuals of the rural population in Sri Lanka, especiallyamong rural conventional tea workers may also have highexposures; the sources in such rural areas are yet to beelucidated.

So far, no clear link has been established between FOCexposures and adverse health impairments in humans or wild-life. Occupationally exposed workers had no association be-tween PFOS and PFOA exposure and hepatic enzymes,lipoproteins, and cholesterol irregularity and clinical hepatictoxicity.11,12 A population of workers with high exposure jobshad an increased number of deaths from bladder cancer, butthis was not clearly attributed to fluorochemical exposures.28

Studies with mammals suggest that lowered serum cholesterolmay be an early finding, with cumulative toxicity resulting inmortality. Studies with primates showed that serum cholesterollevels began to decline when PFOS concentration exceeded100 mg ml�1 in serum.16 The highest PFOS concentrationfound in this study was 18.2 ng ml�1 which was approximately3 orders of magnitude lower than the level that resulted inlowered cholesterol levels in primates.

Since Taves29 reported the occurrence of organic forms offluoride in human blood, numerous papers have been pub-lished on background concentrations of FOCs in humans andwildlife. However, FOC data in developing countries are stillmeager. Our data indicate that human contamination byvarious FOCs is widespread even in developing countries,similar to that in industrialized nations. Exposure to organo-halogen compounds may be associated with reproductiveeffects, as these compounds can be transferred to reproductiveorgans including semen.30 Renal elimination of long-chain

FOCs is lower than that of short-chain FOCs, which isregulated by testosterone.31 Kennedy et al.32 showed an in-crease of tumors in the liver, pancreas, and testicular Leydig-cells, in rats fed with 430 parts per million (ppm) of PFOA.However, the mechanism of Leydig-cell tumorigenesis causedby FOCs and evidence for enzymes that are critical to sexualhormone biosynthesis and metabolism are still not clear. SinceFOCs exhibit a characteristic toxicokinetic behavior in highertrophic positions, exposures in humans should be consideredwith special consideration in future risk assessments, specifi-cally focusing on semen quality and reproductive impairments.

Acknowledgements

This study was partly funded by a Grant-in-Aid by theEnvironmental Ministry of Japan (Year 2004–2009). We alsogratefully acknowledge the Director, Tea Research Institute ofSri Lanka, tea estate managers of S. Coombs, Talawakele andNeedwood and Stassens Bio Tea Project, Haldummulla for thesupport in conducting the health assessment of estate workersand the donors who provided the samples and the medical stafffor support during sampling. Dr S. P. Galbada Arachchige ofITI Sri Lanka is thanked for arrangement of this researchwork.

References

1 E. Kissa, Fluorinated Surfactants and Repellents, Marcel Dekker,New York, 2nd edn., 2001.

2 J. P. Giesy and K. Kannan, Environ. Sci. Technol., 2001, 35, 1339.3 K. Kannan, J. C. Franson, W. W. Bowerman, K. J. Hansen, P. D.

Jones and J. P. Giesy, Environ. Sci. Technol., 2001, 35, 3065.4 K. Kannan, J. Koistinen, K. Beckmen, T. Evans, J. F. Gorzelany,

K. J. Hansen, P. D. Jones, E. Helle, M. Nyman and J. P. Giesy,Environ. Sci. Technol., 2001, 35, 1593.

5 K. Kannan, J. Newsted, R. S. Halbrook and J. P. Giesy, Environ.Sci. Technol., 2002, 36, 2566.

6 S. Taniyasu, K. Kannan, Y. Horii, N. Hanari and N. Yamashita,Environ. Sci. Technol., 2003, 37, 2634.

7 X. Han, T. A. Snow, R. A. Kemper and G. W. Jepson, Chem. Res.Toxicol., 2003, 16, 775.

8 K. J. Hansen, L. A. Clemen, M. E. Ellefson and H. O. Johnson,Environ. Sci. Technol., 2001, 35, 766.

9 K. Kannan, K. Senthil Kumar, S. Corsolini and K. M. Aldous,Organohalogen Compd., 2003, 64, 29.

10 G. W. Olsen, K. J. Hansen, L. A. Stevenson, J. M. Burris andJ. H. Mandel, Environ. Sci. Technol., 2003, 37, 888.

11 G. W. Olsen, J. M. Burris, J. H. Mandel and L. R. Zobel,J. Occup. Environ. Med., 1999, 41, 799.

12 G. W. Olsen, J. M. Burris, M. M. Burlew and J. H. Mandel, DrugChem. Toxicol., 2000, 23, 603.

13 J. P. Giesy and K. Kannan, Environ. Sci. Technol., 2002, 36, 146A.14 A. K. Sohlenius, K. Andersson, A. Bergstrand, O. Spydevold and

J. W. De Pierre, Biochem. Pharmacol., 1996, 51, 821.15 J. R. Thibodeaux, R. G. Hanson, J. M. Rogers, B. E. Grey, B. D.

Barbee, J. H. Richards, J. L. Butenhoff, L. A. Stevenson andC. Lau, Toxicol. Sci., 2003, 74, 369.

16 A. M. Seacat, P. J. Thomford, K. J. Hansen, G. W. Olsen, M. T.Case and J. L. Butenhoff, Toxicol. Sci., 2002, 68, 249.

17 L. E. Gray, C. Wolf, P. Mann and J. S. Ostby, Toxicol. Appl.Pharmacol., 1997, 146, 237.

18 E. V. Younglai, W. G. Foster, E. G. Hughes, K. Trim and J. F.Jarrell, Arch. Environ. Contam. Toxicol., 2002, 43, 121.

19 World Health Organization, Laboratory manual for the examina-tion of human semen and sperm–cervical interaction, CambridgeUniversity Press, Cambridge, UK, 4th edn., 1999.

20 G. W. Olsen, T. R. Church, J. Miller, J. M. Burris, K. J. Hansen,J. K. Lundberg, J. B. Armitage, R. M. Herron, Z. Medhdizadeh-kashi, J. B. Nobiletti, E. M. O’Neill, J. H. Mandel and L. R.Zobel, Environ. Health Perspect., 2003, 111, 1892.

21 G. W. Olsen, T. R. Church, E. B. Larson, G. van Belle, J. K.Lundberg, K. J. Hansen, J. M. Burris, J. H. Mandel and L. R.Zobel, Chemosphere, 2004, 54, 1599.

22 J. W. Martin, M. M. Smithwick, B. M. Braune, P. F. Hoekstra, D.C. Muir and S. A. Mabury, Environ. Sci. Technol., 2004, 38, 373.

Fig. 7 Relationship of PFNA concentrations between blood sera andseminal plasma (ng ml�1) from Sri Lanka.

3 7 6 J . E n v i r o n . M o n i t . , 2 0 0 5 , 7 , 3 7 1 – 3 7 7

23 Z. Kuklenyk, J. A. Reich, J. S. Tully, L. L. Needham and A. M.Calafat, Environ. Sci. Technol., 2004, 38, 3698.

24 M. M. Schultz, D. F. Barofsky and J. A. Field, Environ. Sci.Technol., 2004, 38, 1828.

25 K. Harada, N. Saito, K. Sasaki, K. Inoue and A. Koizumi, Bull.Environ. Contam. Toxicol., 2003, 71, 31.

26 K. S. Guruge, S. Taniyasu, S. Miyazaki, N. Yamanaka andN. Yamashita, Organohalogen Compd., 2004, 66, 4029.

27 D. J. Luebker, K. J. Hansen, N. M. Bass, J. L. Butenhoff andA. M. Seacat, Toxicology, 2002, 176, 175.

28 B. H. Alexander, G. W. Olsen, J. M. Burris, J. H. Mandel and J. S.Mandel, Occup. Environ. Med., 2003, 60, 722.

29 D. Taves, Nature, 1968, 217, 1050.30 T. E. Arbuckle and L. E. Sever, Crit. Rev. Toxicol., 1998, 28,

229.31 N. Kudo, E. Suzuki, M. Katakura, K. Ohmori, R. Noshiro and Y.

Kawashima, Chem.–Biol. Interact., 2001, 134, 203.32 G. L. Kennedy, Jr., J. L. Butenhoff, G. W. Olsen, J. C. O’Connor,

A. M. Seacat, R. G. Perkins, L. B. Biegel, S. R. Murphy andD. G. Farrar, Crit. Rev. Toxicol., 2004, 34, 351.

J . E n v i r o n . M o n i t . , 2 0 0 5 , 7 , 3 7 1 – 3 7 7 3 7 7