Silicone Wristband Passive Samplers Yield Highly Individualized

Nov 29, 2017 - Environmental Science & Technology. Xiong, Miller, Roman-White, Tasker, Farina, Piechowicz, Burgos, Joshi, Zhu, Gorski, Zydney, and Kum...
2 downloads 8 Views 1022KB Size
Article Environmental Science & Silicone Wristband Technology is published by the American Chemical Passive Samplers Yield Society. 1155 Sixteenth Street N.W., Washington, Highly 20036 by READING Subscriber Individualized accessDC provided Published by American UNIV Chemical Society. Copyright © American Chemical Society.

Pesticide Residue ExposureEnvironmental Profiles Science & Technology is published

Raf Aerts, Laure Philippe by Joly, the American Chemical Society. 1155 Sixteenth Szternfeld, Khariklia Tsilikas, Street N.W., Washington, DC 20036 Koen Deaccess Cremer, Philippe Subscriber provided by READING Published by American UNIV Chemical Society. Copyright © American Chemical Society.

Castelain, Jean-Marie Aerts, Jos Van Orshoven, Ben Environmental Science & Somers, Marijke Hendrickx, Technology is published by the American Chemical Mirjana Andjelkovic, and Society. 1155 Sixteenth An Van Nieuwenhuyse Street N.W., Washington, 20036 by READING Subscriber accessDC provided Published by American UNIV Chemical Society. Copyright © American Chemical Society.

Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b05039 • Environmental Science & Technology is published Publication Date (Web): 29 Nov 2017

by the American Chemical

DownloadedSociety. from http:// 1155 Sixteenth Street N.W., Washington, pubs.acs.org on December 3, 2017

20036 by READING Subscriber accessDC provided Published by American UNIV Chemical Society. Copyright © American Chemical Society.

Just Accepted

Environmental Science & “Just Accepted” manuscripts have been pee Technology is published online prior to technical editing, formatting fo by the American Chemical Sixteenth as a fre Society provides Society. “Just 1155 Accepted” Street N.W., Washington, dissemination of scientific material as soon a DC 20036 by Subscriber access provided READING Published by American UNIV Chemical Society. Copyright © American Chemical Society.

appear in full in PDF format accompanied by a fully peer reviewed, but should not be conside Environmental ScienceObject & readers and citable by the Digital Iden Technology is published to authors. Therefore, the “Just Accepted” W by the American Chemical Society. 1155 Sixteenthis technica in the journal. After a manuscript Street N.W., Washington, Accepted” Web site and as an ASA 20036published Subscriber accessDC provided by READING Published by American UNIV Chemical Society. Copyright © American Chemical Society.

changes to the manuscript text and/or graph and ethical guidelines that apply to the jou Science or consequences Environmental arising from the &use of info Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, 20036 by READING Subscriber accessDC provided Published by American UNIV Chemical Society. Copyright © American Chemical Society.

Page 1 ofEnvironmental 43 Science & Technology

ACS Paragon Plus Environment

Environmental Science & Technology

Page 2 of 43

1

Silicone Wristband Passive Samplers Yield Highly

2

Individualized Pesticide Residue Exposure Profiles

3

Raf Aerts,*,†,‡,ǁ Laure Joly,§ Philippe Szternfeld,§ Khariklia Tsilikas,§ Koen De Cremer,⊥ Philippe

4

Castelain,£ Jean-Marie Aerts,ǁ Jos Van Orshoven,‡ Ben Somers,‡ Marijke Hendrickx,∇ Mirjana

5

Andjelkovic,§ and An Van Nieuwenhuyse,†,∆

6



7

Contaminants,

8

Service Toxicology, and ∇Scientific Service Mycology and Aerobiology, Scientific Institute of

9

Public Health (WIV-ISP), Brussels, Belgium

Scientific Service Health and Environment,

§

Scientific Service Chemical Residues and



Operational Directorate Food, Medicines and Consumer Safety, £Scientific

10



11

ǁ

12

Engineering, Department of Biosystems (BIOSYST), and ∆Environment and Health, Department

13

of Public Health and Primary Care, University of Leuven (KU Leuven), Leuven, Belgium

Division Forest, Nature and Landscape, Department Earth and Environmental Sciences,

Measure, Model & Manage Bioresponses (M3-BIORES), Division Animal and Human Health

14 15

ABSTRACT

16

Monitoring human exposure to pesticides and pesticide residues (PRs) remains crucial for

17

informing public health policies, despite strict regulation of plant protection product and biocide

18

use.

We used 72 low-cost silicone wristbands as non-invasive passive samplers to assess

ACS Paragon Plus Environment

1

Page 3 of 43

Environmental Science & Technology

19

cumulative 5-day exposure of 30 individuals to polar PRs. Ethyl acetate extraction and LC-

20

MS/MS analysis were used for the identification of PRs. Thirty-one PRs were detected of which

21

15 PRs (48%) were detected only in worn wristbands, not in environmental controls. The PRs

22

included 16 fungicides (52%), 8 insecticides (26%), 2 herbicides (6%), 3 pesticide derivatives

23

(10%), 1 insect repellent (3%) and 1 pesticide synergist (3%). Five detected pesticides were not

24

approved for plant protection use in the EU. Smoking and dietary habits that favor vegetable

25

consumption were associated to higher numbers and higher cumulative concentrations of PRs in

26

wristbands. Wristbands featured unique PR combinations. Our results suggest both environment

27

and diet contributed to PR exposure in our study group. Silicone wristbands could serve as

28

sensitive passive samplers to screen population-wide cumulative dietary and environmental

29

exposure to authorized, unauthorized and banned pesticides.

30 31

INTRODUCTION

32

Pesticides are widely used in agriculture, industry and in the home environment as plant

33

protection products and as biocidal active substances to control weeds, fungi and pests. Strict

34

international policies and regulations with respect to market approval and maximum residue

35

limits have been implemented and enforced by authorities to minimize risk of pesticide use.

36

However, whether or not exposure to pesticide residues (PRs) in the environment, food and

37

water forms a potential threat for human health remains debated.1 The chronic and acute

38

exposure, including periconceptional, pre- and perinatal exposure, to pesticidal and biocidal

39

active substances and their residues, are increasingly associated to a number of adverse health

40

outcomes, such as Parkinson’s disease,2 prostate and liver cancer,3,4 and reduced sperm quality in

ACS Paragon Plus Environment

2

Environmental Science & Technology

Page 4 of 43

41

adults;5 cough,6 respiratory problems,7 leukemia,8,9 brain tumors,10 and neurobehavioral

42

disorders in children;11 and neurodevelopmental disorders,12,13 low birth weight,14 and congenital

43

disorders in infants.15,16 Many of these epidemiological exposure studies, however, have

44

limitations.17 Actual exposures are hard to measure as the surrogate metrics of exposure, such as

45

reported days or frequency of pesticide use, provide insufficiently precise information about

46

effective exposure. Whereas there is no “gold standard” for exposure,18 there is a need for

47

improving exposure estimation to discriminate surrogate from actual exposures.

48

Possible routes of entry of pesticides and their derivatives are transdermal absorption, inhalation

49

and oral ingestion.19 Occupational exposure and the handling of items that have been treated with

50

pesticides (e.g. cut flowers,20,21 plant cultivation infrastructure and materials22) increase the

51

probability of uptake and subsequent acute or chronic adverse health effects15,23. Such exposure

52

can be minimized by using personal protective equipment,22,24,25 by implementing engineering

53

controls during handling,25 by using proper ventilation,26 and by reducing or avoiding pesticide

54

use.1,27 Environmental exposure to contaminated water, soil or air from runoff, dust and spray

55

drift generated by domestic or agricultural pesticide applications is also an important pathway of

56

exposure to PRs.1,16,28 At the level of the general population, residential pesticide use,8,10,29

57

contact with contaminated household surfaces,30 and notably diet1,29,31-35 have been identified as

58

important exposure pathways to PRs. Short-term and cumulative dietary exposure to a number of

59

PRs have been shown to be a public health concern,36,37 in particular for children.34 PR screening

60

in food and feed and monitoring of human exposure to PRs therefore play vital roles in

61

safeguarding public health.38

62

Human biomarker monitoring, for instance urinary biomarker measurements using liquid

63

chromatography (LC) and mass spectrometry (MS),29,38,39 is the principal method to monitor

ACS Paragon Plus Environment

3

Page 5 of 43

Environmental Science & Technology

64

human exposure to PRs. Urine, sweat, saliva, breast milk, nails and hair are suitable non-

65

invasive human matrices for pesticide exposure monitoring.19,40-42

66

include waste water sampling for population-wide exposure assessment43 and (personal) passive

67

sampling devices (PSDs). Recently polyethylene, polyurethane foam and silicone matrices have

68

been proposed as passive samplers for environmental44,45 and individual assessment of exposure

69

to persistent organic pollutants, pesticides, brominated diphenyl ethers, organophosphate flame-

70

retardants and other potentially harmful substances.46-53

71

In this cross-sectional, non-interventional study, we tested the suitability of low-cost

72

commercially available silicone wristbands as personal passive sampling devices to assess

73

personal exposure to LC-amenable pesticide residues. We hypothesized that PR spectra detected

74

in the wristbands would reflect the participants’ exposure to pesticide residues, which is

75

determined by their pesticide use, outdoor and indoor environmental exposure and personal food

76

choices.

Less intrusive methods

77 78 79

MATERIALS AND METHODS Study Design.

Study participants were recruited from the Department of Earth and

80

Environmental Sciences of the University of Leuven, Leuven, Belgium (n = 30; Table S1) in

81

September 2016. The protocol for a cross-sectional, non-interventional study was reviewed and

82

approved by the Social and Societal Ethics Committee of the University of Leuven (SMEC

83

protocol number G-2016 09 636).

84

information, anonymity and the possibility of withdrawing from the study at any time without

85

the need for justification. All participants were adults and gave written prior informed consent.

Participants were assured confidentiality of personal

ACS Paragon Plus Environment

4

Environmental Science & Technology

Page 6 of 43

86

Participants agreed to wear a silicone wristband for 5 days, to place a second wristband near their

87

residence, and to complete questionnaires.

88

Wristband Cleaning. Adult-size (200 mm L × 12 mm W × 2 mm T; weight 5.33 g, SD =

89

0.10 g) silicone wristbands with embossed serial numbers were purchased (pdc Healthcare,

90

Nivelles, Belgium; wristbands.com).

91

extractions in an overhead shaker (Reax 2®, Heidolph Instruments GmbH & Co. KG,

92

Schwabach, Germany) to remove potentially present analytes of interest and compounds that

93

could cause chromatographic interference. The first extraction of 30 minutes with 1:1 ethyl

94

acetate/hexane (v/v) was followed by a second extraction of 30 minutes with 1:1 ethyl

95

acetate/methanol (v/v). After the extractions, the wristbands were dried under a nitrogen stream,

96

and stored in a dark place in labeled, neutral glass vials of 60 ml with screw closure. This

97

cleaning method was adapted from the procedure described by O’Connell et al.46 to minimize the

98

consumption of organic solvents.

The wristbands were cleaned using two successive

99

Wristband Exposure. Participants wore the wristbands continuously for a 5-day period

100

(starting on the Monday morning of a work week) during all daily activities. Stationary silicone

101

wristbands were placed by the participants on a location outside their residence (protected from

102

rain) and served as controls of the residential outdoor environment. Participants recorded start

103

and stop time of the wristband exposure in a log book. Additional stationary control bands were

104

placed in the main building of the department where the participants were recruited (n = 4,

105

occupational environment controls) and in Heverleebos-Meerdaalwoud, a high-value mixed

106

deciduous forest 3 km south of Leuven (n = 8, clean air environment controls). The purpose of

107

the stationary control bands was to identify airborne PRs. Six cleaned wristbands were kept in

108

their vials in the laboratory at room temperature to serve as blanks. At the end of the 5-day

ACS Paragon Plus Environment

5

Page 7 of 43

Environmental Science & Technology

109

period, participants placed the control and worn wristbands in their original vials and kept the

110

vials refrigerated at 4−6°C.

111

84.3−88.2 hours. In the laboratory, returned vials were stored at −20°C awaiting chemical

112

analysis.50

The wristband exposure time (95% confidence interval) was

113

Questionnaires. Participants provided three types of background information needed to

114

interpret the results: daily outdoor exposure information, demographic data and personal data.

115

To characterize the intensity of outdoor exposure, participants recorded on a daily basis

116

information on their whereabouts and travels. For each travel the participants recorded start and

117

end point addresses, start and end time, and type of transportation used. This information was

118

later used to calculate total time spent moving outdoors, and by using Google Maps routes, total

119

distance traveled (km). Participants completed a short questionnaire to provide demographic

120

data and information on their living environment, smoking habits and occupational or domestic

121

exposure to pesticides. After the experiment, participants provided information on pet ownership

122

and treatment, the number of showers taken during the 5-day period, and general dietary choices

123

(organic vs. conventional and vegetarian vs. conventional).

124

Wristband Extraction. For detection of pesticide residues, a tandem mass spectrometry

125

(MS/MS) protocol was developed, simplifying the protocol from O’Connell et al.46 To extract

126

pesticides, 40 ml of ethyl acetate was added to the vial containing the wristband. Vial, wristband

127

and extraction solvent were shaken mechanically during 30 min by an overhead shaker (Reax

128

2®, Heidolph Instruments GmbH & Co. KG, Schwabach, Germany). This procedure was

129

repeated two times with the same volume and time. The ethyl acetate phase (120 ml) was

130

evaporated with a sample concentrator at 40°C (Genevac Ltd, Ipswich, UK) and under a nitrogen

131

stream until the volume was reduced to 200 µl. To prepare the final extract, the raw extract was

ACS Paragon Plus Environment

6

Environmental Science & Technology

Page 8 of 43

132

spiked with 200 µl internal standard solution (Oxfendazole, a broad spectrum veterinary

133

anthelmintic compound, at 100 ng/ml), then diluted with 400 µl mobile phase A (water/methanol

134

(90/10 v/v) containing 5 mM ammonium acetate) and adjusted to 1 ml with methanol. An aliquot

135

(500 µl) was transferred in a filtration vial prior to injection into the UHPLC-MS/MS system.

136

Wristband Chemical Analysis. Based on production, maximum residue levels in food

137

and frequency of detection, LC-MS/MS systems have a wider scope and better sensitivity to

138

detect pesticide residues than GC-MS systems.54,55 Therefore, the screening was focused on LC-

139

amenable residues of polar plant protection products and biocides. UHPLC-MS/MS analyses

140

were performed on an UPLCTM system (Waters, Milford, MA) coupled to a Quattro PremierTM

141

mass spectrometer (Waters, Milford, MA) following the protocol of Hanot et al.56 Five microliter

142

of sample was injected onto a reverse-phase ACQUITY™ UPLC BEH C18 column (1.7 µm,

143

2.1×100 mm) (Waters, Milford, MA). The mobile phases were composed of water/methanol

144

(90/10 v/v) (phase A) and methanol/water (90/10 v/v) (phase B) both containing 5 mM

145

ammonium acetate. The chromatographic separation was performed at a flow rate of 0.45 ml

146

min-1 with an elution gradient starting with 99.9% of mobile phase A and linearly evolving until

147

0.1% in 10 min and maintained during two minutes. The column was re-equilibrated for the next

148

injection during 3 min with 99.9% of mobile phase A. Three UHPLC runs were necessary to

149

acquire two transitions for each of 200 pesticides. For the two first runs the sample was ionized

150

with an electrospray probe in the positive mode; for the third run the negative mode was used.

151

Selected quantities of 200 LC-amenable pesticide residues were added just before the vial

152

filtration. The linearity range of the method was 0.4−100 ng g-1 wristband. In all cases, good

153

linearity was achieved with correlation coefficients > 0.99. The limit of quantification (LOQ)

154

was 0.2 ng g-1 wristband.

ACS Paragon Plus Environment

7

Page 9 of 43

Environmental Science & Technology

155

Quality Control. To ensure reliability of results, a blank and a spiked sample were

156

analysed in each of 5 batches. For the spiked sample (analytical quality control sample), a multi-

157

residue solution was spiked on wristband, which was allowed to dry for 1 h, and then analysed.

158

Multi-residue solutions containing 200 LC-amenable pesticide residues were prepared in

159

methanol according to the European quality criteria for pesticide residue analysis in food and

160

feed.57 Wristbands were spiked by spraying small droplets directly on the wristbands following

161

the procedure of Toumi et al.20 The spike was calculated to reach an average concentration of

162

about 2 ng g-1 wristband for each LC-amenable pesticide residue (equivalent to a concentration

163

of 10 ng of pesticide residue on a wristband of 5 g). No pesticides were detected in the blank

164

samples with a concentration above LOD, proving an absence of pesticides in the cleaned

165

wristbands themselves, except for the insect repellent N,N-diethyl-meta-toluamide (DEET). The

166

concentration of DEET was similar in all the blank samples (0.5 ng g-1 wristband). This

167

background value was subtracted from the detected DEET values for all samples. Detections

168

below LOQ were assigned the value of ½ LOQ (0.1 ng g-1) to enable statistical analyses.

169

Statistical Analyses.

The frequencies of pesticide residue detection were compared

170

between worn and stationary wristbands using chi-square tests for independent samples. The

171

cumulative concentration of PRs (log∑PR) was calculated as the log10-transformed sum of the

172

concentrations of all PRs (ng g-1) detected in each wristband (log∑PR = log10[∑PRi+1]; the +1

173

component of the transformation was included to transform total concentrations < 1 ng g-1 to

174

positive values), with and without DEET. Differences in the average number of PRs detected

175

and differences in log∑PR between worn and stationary wristbands were assessed using Kruskal-

176

Wallis one-way analysis of variance. Generalized linear models were used to estimate the

177

effects of environmental and demographic variables on the number of PR detected and log∑PR

ACS Paragon Plus Environment

8

Environmental Science & Technology

Page 10 of 43

178

(without DEET). For the number of PRs, a Poisson log-linear model was used, with sex,

179

smoking habit and food choices as factors and age, BMI, number of showers and total distance

180

traveled as covariates. Estimated marginal means and 95% confidence intervals were obtained

181

and pairwise comparisons were made for the significant (p < 0.05) factors. A linear model with

182

the same factors and covariates was used for log∑PR. To identify groups with similar PR

183

profiles, individual PR profiles (based on log10[PR+1]-transformed values) were clustered into k

184

= 2–6 groups using the Sørensen distance measurement. The most informative number of groups

185

was determined using Indicator Species Analysis58 Nonmetric multidimensional scaling (NMS)

186

based on the Sørensen distance measure was then used to place individual participants and PR

187

profile groups in a 2-dimensional ordination space. Clustering and ordination were conducted

188

using PC-ORD 5.0 for Windows (MjM Software, Gleneden Beach, OR). All other statistical

189

analyses were performed in IBM SPSS Statistics 20 (IBM Corporation, Armonk, NY).

190 191

RESULTS AND DISCUSSION

192

All participants (n = 30, 100%) completed the exposure experiment, returned the worn

193

wristbands and controls, and completed the questionnaires. The study population comprised 16

194

males (53%) and 14 females (47%). At the time of the survey, participants were on average 32.6

195

years old (range 24−72 years) and had a normal weight (average BMI 22.8, range 18.5−28.3).

196

Participants spent on average 8.8 hours in traffic during the 5-day period (range 2.8−17.7 hours)

197

and covered on average a distance of 251 km (range 31.5−711.3 km) (Table S1).

198

occupational or accidental exposure to pesticides was reported. Four participants reported to

199

have used pet care products (spot on flea treatment and flea collars) that contained a number of

No

ACS Paragon Plus Environment

9

Page 11 of 43

Environmental Science & Technology

200

antiparasitic agents that included fipronil, methoprene, selamectine, imidacloprid and

201

flumethrine (methoprene, selamectine and flumethrine were not included in the LC-MS/MS

202

screening).

203

PRs in Wristbands. In total, 31 LC-amenable pesticide residues were detected in the 72

204

deployed wristbands. These PRs included 16 fungicides (52%), 8 insecticides (26%), 2

205

herbicides (6%), 1 insect repellent (3%), 3 pesticide derivatives (10%) and 1 pesticide synergist

206

(3%) (Table 1). Five substances which are not approved for plant protection use in the EU were

207

detected: fipronil (and its desulfinyl photoproduct and sulfone metabolite),59,60 carbendazim,

208

bendiocarb, tolylfluanid (via the metabolite DMST) and ethirimol (Table 1). Carbendazim and

209

ethirimol are also degradation products of thiophanate-methyl and bupirimate, respectively, and

210

those compounds are approved for plant protection use in the EU. Five fungicides and 4

211

insecticides had the dual use of plant protection product and biocidal active substance (BAS)

212

(Table 1). The insect repellent N,N-diethyl-meta-toluamide (DEET) was detected in 93% of all

213

worn bands and in 81% of the stationary controls. The high prevalence of DEET in the controls,

214

even in controls that were placed in forest with high ecological value, indicates that DEET is a

215

pervasive contaminant in the environment, not only in water,61,62 but also in aerosols.49,63 Apart

216

from DEET, the PRs with the highest prevalence were the pesticide synergist piperonyl butoxide

217

(PBO; 63% of the worn bands, 9.5% of the controls), the broad spectrum fungicide azoxystrobin

218

(47% of the worn bands, 2% of the controls) and the fungicide and parasiticide thiabendazole

219

(TBZ) (40% of the worn bands, 2% of the controls). PBO is added to pesticide formulations to

220

increase their effectiveness, in particular pyrethrin and pyrethroid insecticides (not included in

221

our screening), and has been detected in personal and indoor residential air6 and in PSDs.49

222

Prenatal exposure to PBO has been associated to negative health outcomes in children.6,12 The

ACS Paragon Plus Environment

10

Environmental Science & Technology

Page 12 of 43

223

fungicides azoxystrobin and TBZ are widely used as plant protection products and as biocidal

224

active substances to enhance preservation of film, fiber, leather, rubber, polymerised materials,

225

wood and construction material. Environmental exposure, via aerosols, dust or direct contact

226

with treated materials and/or food items (e.g. TBZ on orange, lemon and other citrus peels), may

227

therefore be partially responsible for the detection of these compounds.

228

Nine PRs had a significantly higher prevalence in worn wristbands than in controls (all p