Article pubs.acs.org/est
Exposure to Toxic Metals and Persistent Organic Pollutants in Inuit Children Attending Childcare Centers in Nunavik, Canada Huguette Turgeon O’Brien,*,† Rosanne Blanchet,† Doris Gagné,† Julie Lauzière,† Carole Vézina,† Émilie Vaissière,† Pierre Ayotte,‡ and Serge Déry§ †
Groupe d’études en nutrition publique, Département des sciences des aliments et de nutrition, Université Laval, Québec, Québec, Canada Centre de recherche du Centre hospitalier universitaire de Québec and Université Laval, Québec, Québec, Canada § Nunavik Regional Board of Health and Social Services, Kuujjuaq, Québec, Canada ‡
S Supporting Information *
ABSTRACT: Arctic populations are exposed to substantial levels of environmental contaminants that can negatively affect children’s health and development. Moreover, emerging contaminants have never been assessed in Inuit children. In this study, we document the biological exposure to toxic metals and legacy and emerging persistent organic pollutants (POPs) of 155 Inuit children (mean age 25.2 months) attending childcare centers in Nunavik. Blood samples were analyzed to determine concentrations of mercury, lead, polychlorinated biphenyls (PCBs), pesticides, brominated flame retardants [e.g., polybrominated diphenyl ethers (PBDEs)] and perfluoroalkyl and polyfluoroalkyl substances [PFASs; e.g. perfluorooctanesulfonate (PFOS) and perfluorooctane (PFOA)]. Lead [geometric mean (GM) 0.08 μmol/L], PCB-153 (GM 22.2 ng/g of lipid), BDE-47 (GM 184 ng/g of lipid), PFOS (GM 3369 ng/L), and PFOA (GM 1617 ng/L) were detected in all samples. Mercury (GM 9.8 nmol/L) was detected in nearly all blood samples (97%). Levels of metals and legacy POPs are consistent with the decreasing trend observed in Nunavik and in the Arctic. PBDE levels were higher than those observed in many children and adolescents around the world but lower than those reported in some U.S. cities. PFOS were present in lower concentrations than in Nunavimmiut adults. There is a clear need for continued biomonitoring of blood contaminant levels in this population, particularly for PBDEs and PFASs.
■
INTRODUCTION Traditional foods are beneficial to the Inuit for cultural, social, psychological, nutritional, and economic well-being.1 However, in the Arctic, foods from mammals, birds, and fish have been documented as the main source of human exposure to contaminants.2 Despite past bans and restrictions on their uses and emissions in Arctic and other countries, concentrations of some legacy persistent organic pollutants (POPs) in top predators in the marine food web are still high enough to affect the health of wildlife and humans.3 Among their adverse health effects, toxic metals, such as mercury and lead, are recognized as significant neurotoxic agents and can lead to developmental deficiencies and learning disabilities in young children.4,5 Exposure to legacy POPs [e.g., polychlorinated biphenyls (PCBs) and chlorinated pesticides (OCs)] has also been associated with negative health effects, including impaired neurodevelopment, depressed immunity, and altered levels of thyroid hormones.6−8 Moreover, evidence suggests that mercury, lead, and POPs can affect human health at lower levels of exposure than previously thought.9 Children might be more vulnerable to negative health effects related to contaminants because of their physiological and anatomical immaturity and rapid development.1 Concentrations of metals and legacy POPs have decreased in many Arctic populations during the past years.10 In Nunavik, temporal variations of toxic metals and POPs have been © 2012 American Chemical Society
monitored in the Inuit population over the last two decades. Between the early 1990s and the mid-2000s, levels of mercury, lead, and some legacy POPs have decreased significantly in umbilical cord blood and in adults.11−13 However, to the best of our knowledge, only one study investigating contaminant blood level of 4−6-year-old children has been reported in the literature,4 making it impossible to determine temporal trends in Inuit children from Nunavik. Furthermore, exposure to environmental contaminants has not been measured in 1−3-yearold children from Nunavik. In addition to the legacy POPs, a growing number of compounds are now being detected in humans from the Arctic regions and are of concern because they share many of the chemical properties of legacy POPs.5 These so-called emerging contaminants include brominated flame retardants [BFRs; e.g., polybrominated diphenyl ethers (PBDEs)] and perfluoroalkyl and polyfluoroalkyl substances [PFASs; e.g., perfluorooctanesulfonate (PFOS) and perfluorooctane (PFOA)]. These synthetic chemicals have been used in a variety of consumer and industrial applications for over 50 years. Among BFRs, pentaBDE has been added to polyurethane foams used in mattresses, Received: Revised: Accepted: Published: 4614
October 12, 2011 March 13, 2012 March 15, 2012 March 15, 2012 dx.doi.org/10.1021/es203622v | Environ. Sci. Technol. 2012, 46, 4614−4623
Environmental Science & Technology
Article
period (unpublished data from Pereg in Donaldson et al.).1 Decreasing trends in PFAS concentrations were also observed in the early to mid-2000s in the United States, Sweden, and Norway.38,39,41−44 In Arctic children, exposure to PFASs has been examined only in the Faroe Islands.45 To our knowledge, the exposure to PFASs and other emerging contaminants has never been investigated in young Inuit children. The aim of this study is to document the exposure to metals and persistent organic pollutants (legacy and emerging) of Inuit children attending childcare centers in Nunavik.
upholstered furniture, and carpet padding, and octaBDE has been used in plastics, such as those found in the casings of computer monitors and televisions.14,15 At the end of 2004, production of pentaBDE and octaBDE was voluntarily discontinued in the United States,16 whereas Canada banned the use of these products in 2006.17 As for PFASs, they were mainly used as stain, grease, and water repellents, in paper treatment (e.g., for food packaging), and as lubricants.18 In Canada, use of PFOS has been banned since 2009,19 whereas PFOA will be phased out by 2015.20 The definite health risks of these compounds have not been determined yet, but experimental toxicology studies suggest thyroid hormone perturbations, neurobehavioral effects, liver toxicity, and possible reproductive effects.7 PBDEs, PFOS, and PFOA are now being found in wildlife species that are the traditional foods of Arctic Canadians.21 According to the 2009 AMAP report, in remote communities with fewer consumer products, exposure through the food chain is more likely to be the main contributor of PBDE exposure.10 However, PBDEs were not related to traditional food consumption in the Nunavik Inuit Health Survey (NIHS) conducted in 2004, whereas consumption of marine mammal fat and fish was a source of exposure to PFOS.12 Until recently, there were only a few data available on human concentrations of emerging contaminants such as PBDEs in Arctic regions.10 Although PBDE levels have been reported in 7-year-old children from the Faroe Islands,22 concentrations of PBDEs have never been reported in Inuit children. A representative sample of the Inuit adult population of Nunavik revealed that PBDE levels were very similar to those observed among Inuit, Dene/Métis, and Caucasians from Inuvik and Baffin regions but lower than those reported in southern Québec and in the United States.1,12,23 Temporal trends of BFRs in the Arctic environment showed increasing levels of PBDEs in both abiotic (e.g., air, freshwater sediment) and biotic samples (e.g., seabirds, marine mammals).15 Increasing levels of PBDE concentrations have been observed in blood samples from Norway (1977−1999), and in breast milk samples from Sweden (1967−1997) and the Faroe Islands (1987−1999).24−26 On the other hand, studies have shown a leveling off of curves in the late 1990s and a recent decline in PBDE levels in breast milk samples from Sweden and Norway.27−30 In Nunavik, breast milk samples collected in 1990 and 2002 and analyzed in 2003 revealed that PBDE congeners increased approximately 3-fold during this time period.31 However, since PBDEs were more extensively used in North America and most were only banned later than in Europe,32 it is still unclear when PBDE levels will stabilize and decline in Nunavik as observed in Sweden and Norway. With the exception of the study of Pereg et al.31 on breast milk in Nunavik, time trends seem to be nonexistent in Inuit populations. In addition, many studies showed that PBDE levels are significantly higher in children than in adults in various populations worldwide,24,33−35 indicating the need for ongoing monitoring of PBDEs in Inuit children. The amount of data on PFASs concentrations has increased considerably over the past decade.36 However, PFASs have only been measured in a few studies carried out in the Canadian Arctic.1 In the Nunavik Inuit Health Survey (NIHS), concentrations of PFOS were similar or slightly lower than those observed in southern Canada and in many countries including the United States and Denmark.12,37−40 Temporal comparisons between data obtained in women of childbearing age and pregnant women from Nunavik in 1992 and 2004 showed that PFOS concentrations decreased significantly over this time
■
EXPERIMENTAL SECTION Study Design and Data Collection. A prospective study was conducted between 2006 and 2010 in young children attending childcare centers in Nunavik (northern Québec, Canada) to document the benefits of a nutrition program implemented in different Inuit communities. Information regarding the nutrition intervention and details on data collection can be found in the Supporting Information. This paper presents data that were collected between 2006 and 2008. Briefly, eight communities were visited at least once from 2006 to 2008. Data collections took place in the autumn, a season where most families remained in the community and were not gone fishing or hunting. Parents of Inuit children aged 11−54 months and attending childcare centers were contacted and invited to participate by an Inuk, usually from their own community. Although the number of potential participants was not systematically collected throughout the project, it is clear that the participation rate was influenced by many factors (e.g., flu vaccination clinics, other ongoing research projects with Nunavimmiut families) and was highly variable between communities and years of data collection. Parents interested in the project were met by a research team member. Information on the study was provided individually, orally or through a DVD available in three languages (Inuktitut, English, and French). Parents agreeing for their child to participate in the study gave their written informed consent. This study was approved by the Research Ethics Board of the Centre Hospitalier de l′Université Laval (CHUQ-CHUL), Québec, Canada. Blood Sampling and Laboratory Analyses. Blood samples (6 mL) were collected by venipuncture in a vial containing ethylenediaminetetraacetic acid (EDTA) as the anticoagulant (BD 367863, Becton Dickinson and Co.). Blood samples were either kept at room temperature for a maximum of 20 min, kept in an insulated container with ice packs, or refrigerated at 4 °C prior to processing. A sample of whole blood was aliquoted for metals analysis, and then tubes containing the remaining blood sample were centrifuged and the plasma was isolated and aliquoted. Within 3 h of collection, aliquots were frozen and stored at −18/−20 °C. Frozen aliquots were kept in insulated containers with ice packs during transportation to the Laboratoire de toxicologie of the Institut national de santé publique du Québec (INSPQ). Supelco hexane-rinsed glass vials were used to store plasma samples for POPs determination, whereas whole-blood samples for metal analysis were stored in plastic tubes. Although reports on adsorption of PFOS to glass exist, there is no indication that PFASs in human blood transfer to glass surface during storage at −20 °C for at least 4 months.46 Total mercury and lead concentrations were determined in whole blood samples by inductively coupled plasma mass spectrometry (ICP-MS).47 PCB, toxaphen, OC, and BFR concentrations were measured in plasma samples by gas 4615
dx.doi.org/10.1021/es203622v | Environ. Sci. Technol. 2012, 46, 4614−4623
Environmental Science & Technology
Article
chromatography/mass spectrometry (GC/MS).48 Plasma PFAS determination was performed by high-performance liquid chromatography tandem mass spectrometry (HPLC-MS/ MS).48 Information regarding the analytical procedures for these contaminant analyses can be found in the Supporting Information. Limits of detection (LOD) are presented in Table S1 (Supporting Information). Cholesterol and triglyceride analyses were performed on a Hitachi 917 auto analyzer with reagents from Roche Diagnostics. Concentration of total plasma lipids were estimated according to the formula developed by Phillips et al.49 The accuracy and precision of the analysis are evaluated on a regular basis through the laboratory participation in external quality assessment schemes, including the interlaboratory comparison program for metals in biological matrices (PCI), the German external quality assessment scheme (G-EQUAS), and the AMAP ring test for persistent organic pollutants in human serum. Interday precision results ranged between 2.5% and 25.2% and are presented in Table S1 (Supporting Information) for each pollutant. Statistical Analyses. Statistical analyses were performed with SAS 9.2 (SAS Institute Inc., Cary, NC). Contaminant concentrations followed a log-normal distribution and were transformed accordingly. Therefore, results are presented as geometric means (GM). A value equal to half the limit of detection of the analytical method was attributed to nondetected contaminants in biological samples. The sum of PCBs includes all congeners with a detection frequency ≥50% (congeners 118, 138, 146, 153, 163, 170, 180, 183, 187, and 194) whereas the sum of PBDEs includes congeners 47, 99, 100, and 153 (detection frequency ≥50%). Total PCBs expressed as Aroclor 1260 was estimated by multiplying the sum of PCB-138 and PCB-153 concentrations by 5.2. Breastfeeding duration was reported in months for children who had been breastfed. Nonbreastfed children had a breastfeeding duration equal to 0, whereas children who were still breastfed at the time of the interview had a breastfeeding duration equal to the child’s age.
had been breastfed and one-fifth were still being breastfed at the time of the interview. Mean breastfeeding duration was 8.3 ± 10.7 months (mean ± SD, data not shown). Although smoking was forbidden inside the house in the majority of households (84.6%), children were likely to be exposed to the harmful effects of passive smoking in their environment, as smoking was prevalent in the participants’ households (87.1% had at least one smoker). Percentages of detection and concentrations of mercury, lead, and several organic contaminants are presented in Table 2. Concentrations of most contaminants are also expressed in wet weight or in empirical units (Table S1, Supporting Information). Lead, PCB-153, BDE-47, PFOS, and PFOA were detected in all samples (Figure 1), whereas mercury was detected in almost all blood samples (97%). Ten of the 24 PCB congeners were detected in ≥50% of the children; PCB-153 was the most prevalent and the most abundant (100%; GM 22.2 ng/g of lipid). More than half the pesticides were detected in ≥50% of the children, and three pesticides were not detected (aldrin, α-chlordane, and γ-chlordane). Among pesticides, p,p′-DDE had the highest concentration (GM 79.7 ng/g of lipid), which was almost 5 times higher than the second most abundant, trans-nonachlor (GM 16.1 ng/g of lipid). Both toxaphen congeners were detected in more than 60% of participants. Among PBDEs, four congeners were detected in more than 65% of the samples, and three were not detected (BDE-15, -17, and -33). BDE-47 was the most prevalent and the most abundant congener (100%; GM 34.6 ng/g of lipid), followed by BDE-99 (94%; GM 10.3 ng/g of lipid). PFOS and PFOA were detected in all participants, whereas perfluorohexanesulfonate (PFHxS) was detected in 50% of participants. The geometric mean of BDE-47 concentration was 1.6 times higher than that of PCB-153. Nearly two-thirds of participating children (64%) had a higher concentration of BDE-47 than PCB-153 (result not shown). Nonetheless, ∑PCBs was still slightly higher than ∑PBDEs (GM 72.1 versus 59.8 ng/g of lipid, respectively). Among PBDEs, the ratio of BDE-47 to BDE-153 was 6.6 (IC 5.7−7.5). On a wet weight basis, PFOS concentration (GM 3369 ng/L) was 2-, 18-, and 29-fold greater than those of PFOA, BDE-47, and PCB 153, respectively (Table S1, Supporting Information). Blood mercury, lead, and PCB levels were compared to blood guidance values (Table 3). As indicated in Table 3, 13−22% of the children had values that were equal to or exceeded the level of concern for blood mercury,50,51 and 13% of the children were above the PCB level of concern.52 There is no level of concern threshold available for lead. None of the children exceeded the levels of action for mercury, lead, and PCBs.51−53
■
RESULTS One hundred fifty-five children were recruited between 2006 and 2008. Blood samples were available for 86−129 of the children, depending on the contaminant examined. Table 1 presents selected characteristics of the participating children. Children were aged between 11 and 54 months (mean ± SD = 25.2 ± 9.6 months). About half of the participants were male and lived on the Hudson coast. About three-quarters of them
■
Table 1. Descriptive Characteristics of Participants mean (SD)a or %
characteristics
DISCUSSION In this cohort of preschool-aged children, we found lower levels of metals and legacy POPs than those observed in other surveys carried out in Nunavik residents of various age groups. The average blood mercury concentrations (8.9 nmol/L) was almost 9 times lower than levels found in older Inuit children (mean age = 5.4 years) (79.2 nmol/L)4 and 6 times lower than levels found in pregnant women from the Hudson Bay Coast (51.8 nmol/L)54 and adults who participated in the NIHS 2004 (51.2 nmol/L).13 The average blood lead concentrations (0.08 μmol/L) was 2.5 times lower than those reported in the Inuit children (0.20 μmol/L)4 and adults (0.19 μmol/L)13 mentioned previously. In agreement with other studies,4,12 PCB-153 was detected in all participants, whereas PCB-118,
n
Child Characteristics age (months) 25.2 (9.6) gender (% male) 51.0 coast of origin (% Hudson) 53.5 ever breastfed (% yes) 74.5 currently breastfed (% yes) 20.3 Household Characteristics smoking status (% with ≥1 smoker in the 87.1 household) restriction about smoking (% forbidden 84.6 inside the home) a
155 155 155 153 153 155 152
SD, standard deviation. 4616
dx.doi.org/10.1021/es203622v | Environ. Sci. Technol. 2012, 46, 4614−4623
Environmental Science & Technology
Article
Table 2. Descriptive Statistics for Metals and Persistent Organic Pollutants n mercury (nmol/L) lead (μmol/L) Aroclor 1260 28 66 74 99 101 105 118 128 138 146 153 156 163 167 170 178 180 183 187 194 201 203 206 ∑PCBsb p,p′-DDE p,p′-DDT β-HCH γ-HCH hexachlorobenzene Mirex cis-nonachlor trans-nonachlor oxychlordane Parlar 26 Parlar 50 PBB IUPAC 153 BDE IUPAC 25 BDE IUPAC 28 BDE IUPAC 47 BDE IUPAC 99 BDE IUPAC 100 BDE IUPAC 153 ∑PBDEsd perfluorooctanesulfonate (PFOS) perfluorooctane (PFOA) perfluorohexanesulfonate (PFHxS)
% detected
GM
95% CI
Metals 96.9 8.9 100.0 0.08 PCB (IUPAC Number)c (ng/g of lipid) 126 97.6 174.7 126 8.7 126 0.8 126 27.8 126 42.9 126 6.4 126 34.1 126 88.1 5.9 126 9.5 126 95.2 11.4 126 54.8 3.3 126 100.0 22.2 126 47.6 126 60.3 3.9 126 17.5 126 68.3 3.9 126 36.5 126 89.7 9.6 126 50.8 2.1 126 66.7 4.4 126 50.0 2.2 126 38.1 126 43.7 126 12.7 126 72.1 Chlorinated Pesticidesc (ng/g of lipid) 126 89.7 79.7 126 17.5 126 54.8 2.7 126 1.6 126 54.8 12.3 126 33.3 126 71.4 2.8 126 92.9 16.1 126 95.2 9.6 Toxaphen (ng/g of lipid) 126 61.9 2.2 126 77.8 3.5 Brominated Flame Retardantsc (ng/g of lipid) 109 4.6 126 1.6 126 6.4 126 100.0 34.6 126 94.4 10.3 126 73.0 6.1 109 67.9 5.5 109 59.8 Perfluoroalkyl and Polyfluoroalkyl Substances (ng/L) 86e 100.0 3369 86e 100.0 1617 86e 50.0 331 129 129
rangea
7.1−11.3 0.07−0.09
0.2−164.5 0.02−0.48
135.0−226.0
7.0−6931.8 2.6−40.8 1.4−5.9 1.4−39.8 1.6−71.9 1.4−10.2 0.5−15.7 0.5−90.2 0.5−5.6 0.7−340.9 0.5−159.1 1.4−988.6 0.5−147.7 0.5−215.9 0.5−12.5 0.5−227.3 0.5−71.6 0.5−738.6 0.5−39.8 0.5−215.9 0.5−101.1 0.5−125.0 0.5−62.5 0.5−18.2 7.7−3060.2
4.7−7.3 8.9−14.6 2.6−4.2 17.1−28.8 3.0−5.0 3.1−5.0 7.3−12.5 1.8−2.5 3.4−5.6 1.8−2.7
56.7−91.7 61.4−103.5
2.1−3.7 12.1−21.5 7.1−12.8
5.6−1864.4 2.3−62.7 0.5−39.7 0.5−2.1 2.1−241.4 0.5−23.9 0.3−81.0 0.7−517.2 0.4−344.8
1.7−2.9 2.6−4.6
0.3−81.0 0.3−121.4
30.1−39.9 8.9−11.9 5.1−7.1 4.6−6.6 51.2−69.9
0.9−6.3 1.4−8.5 1.4−16.2 8.3−358.5 1.1−104.3 1.1−207.5 1.0−301.9 14.6−950.9
3798−4056 1412−1852 272−403
910−31 000 460−11 000 150−4400
2.2−3.4 9.8−15.4
a Range values cover all values. b∑PCBs includes PCB congeners 118, 138, 146, 153, 163, 170, 180, 183, 187, and 194. cPCB-52, aldrin, α-chlordane, γ-chlordane, PBDE 15, PBDE 17, and PBDE 33 were not detected. d∑PBDEs includes PBDE congeners 47, 99, 100, and 153. eA smaller number of subjects was available for analyses due to the challenges related to drawing blood from young children.
4617
dx.doi.org/10.1021/es203622v | Environ. Sci. Technol. 2012, 46, 4614−4623
Environmental Science & Technology
Article
Figure 1. Blood concentration of metal and persistent organic pollutants (POPs) detected in all participating children: (A) Lead, (B) PCB-153, (C) BDE-47, (D) PFOS, (E) PFOA.
PCB-180, p,p′-DDE, cis-nonachlor, and oxychlordane were detected in at least 70% of the study samples. However, observed levels were lower in our study than in Inuit children and pregnant women mentioned above. Finally, aldrin, α-chlordane, and γ-chlordane were not detected in our subjects and hardly detected in the other studies conducted in Nunavik.4,54 The proportion of children exceeding the blood guidance values for mercury is considerably lower in the present study (13−22%) compared to data obtained in women of childbearing age (72%) participating in the NIHS.13 For blood lead level, none of the children exceeded the Health Canada level of action (100 μg/L),53 compared to 10% in 4−6-year-old Inuit children from Nunavik.4,55 However, this threshold is questioned since recent scientific evidence has not been able to confirm
the existence of a safe level for some of the adverse neurological effects of lead in children.56,57 For example, deficits in several fine motor tasks have been observed in Nunavik children at blood lead concentrations below 100 μg/L.4 As for PCBs (Aroclor 1260), the proportion of participants exceeding the Health Canada level of concern is similar to the one observed by Dewailly et al.12 for women of childbearing age which is in agreement with simulation models developed by Quinn et al.58 Indeed, most of these women were born before the PCBs production phase-out, which led to a high lifetime exposure as indicated by the high proportion of women exceeding the threshold. However, our participants were born after the phaseout and experienced high prenatal and postnatal exposure to PCBs, explaining the high proportion exceeding the threshold 4618
dx.doi.org/10.1021/es203622v | Environ. Sci. Technol. 2012, 46, 4614−4623
Environmental Science & Technology
Article
Table 3. Proportion of Participants Exceeding Selected Canadian and American Blood Guidelines for Contaminants guidance level
target population
cutoff value
% of samples ≥ cutoff (n)
source
5.8 μg/L (29 nmol/L) 8 μg/L (40 nmol/L)a 40 μg/L (200 nmol/L)a
22 (28) 13 (17) 0 (0)
50 51 51
not available 100 μg/L (0.48 μmol/L)
0 (0)
53 53
5 μg/L (13 nmol/L) 100 μg/L (260 nmol/L)
13 (17) 0 (0)
52 52
Mercury level of concern proposed level of concern proposed level of action
women of childbearing age females birth−49 years; males ≤18 years females birth−49 years; males ≤18 years
level of concern level of action
women of childbearing age women of childbearing age
level of concern level of action
women of childbearing age women of childbearing age
Lead
PCBsb
a
Harmonized provisional interim blood guidance value. bAs Aroclor 1260.
and in new products built with recycled PBDE-containing material.66−68 Moreover, the widely used, unregulated decaBDE product may, to some extent, break down into more lower brominated congeners (penta and octa congeners) with enhanced toxicity and ability to bioaccumulate relative to the parent.69 In the present study, the concentration of BDE-47 was 6 times higher than that of BDE-153, indicating a greater exposure to the pentaBDE compared to the octaBDE formulation. This reflects the higher proportion of the pentaBDE compared to the octaBDE in the total North American PBDE market (24.4% versus 4.1% in 1999).70 Also, in the main United States penta-formulation, BDE-47 and BDE-99 contribute >70% and BDE-153