Catching Flame Retardants and Pesticides in Silicone Wristbands: Evidence of Exposure to Current and Legacy Pollutants in Uruguayan Children

Abstract

Children are exposed to many potentially toxic compounds in their daily lives and are vulnerable to the harmful effects. To date, very few non-invasive methods are available to quantify children’s exposure to environmental chemicals. Due to their ease of implementation, silicone wristbands have emerged as passive samplers to study personal environmental exposures and have the potential to greatly increase our knowledge of chemical exposures in vulnerable population groups. Nevertheless, there is a limited number of studies monitoring children’s exposures via silicone wristbands. In this study, we implemented this sampling technique in ongoing research activities in Montevideo, Uruguay which aim to monitor chemical exposures in a cohort of elementary school children. The silicone wristbands were worn by 24 children for 7 days; they were quantitatively analyzed using gas chromatography with tandem mass spectrometry for 45 chemical pollutants, including polychlorinated biphenyls (PCBs), pesticides, polybrominated diphenyl ethers (PBDEs), organophosphorus flame retardants (OPFRs), and novel halogenated flame-retardant chemicals (NHFRs). All classes of chemicals except NHFRs were identified in the passive samplers. The average number analytes detected in each wristband was 13 ± 3. OPFRs were consistently the most abundant class of analytes detected. Median concentrations of ΣOPFRs, ΣPBDEs, ΣPCBs, and dichlorodiphenyltrichloroethane (DDT) and its metabolites (dichlorodiphenyldichloroethylene (DDE) and dichlorodiphenyldichloroethane (DDD)) were 1020, 3.00, 0.52 and 3.79 ng/g wristband, respectively. Two major findings result from this research; differences in trends of two OPFRs (TCPP and TDCPP) are observed between studies in Uruguay and the United States and the detection of DDT, a chemical banned in several countries, suggests that children’s exposure profiles in these settings may differ from other parts of the world. This was the first study to examine children’s exposome in South America using silicone wristbands and clearly points to a need for further studies.

1. Introduction

The use of silicone wristbands to measure personal exposure was first published in 2014¹, as a novel and appealing non-invasive method for measuring human exposure to environmental chemicals. The wristbands have grown in popularity due to their advantages as passive samplers, and to date, have been used in over 30 epidemiological and toxicological studies.^1-33 Specifically, silicone wristbands are popular because they are non-invasive and represent low burden to study participants.^{4, 16, 17, 19, 25, 33} Nevertheless, there are a limited number of studies using passive sampling with silicone wristbands to measure chemical exposure in children, especially outside of the United States (U.S.).

A handful of studies in the U.S. have used silicone wristbands to asses children’s exposure to polybrominated diphenyl ethers (PBDEs)^{4, 33}, organophosphorus flame retardants (OPFRs)^{4, 16, 29, 33}, phthalates²⁹, polycyclic aromatic hydrocarbons (PAHs)²⁵, pesticides¹⁷, and nicotine exposure from second hand smoke¹⁹. Household characteristics were associated with differences in total exposure to PBDEs and OPFRs (i.e. age of house, vacuuming frequency)⁴, and PAH levels (stove/vent type, commuting via car)²⁵. Sex differences were observed in exposure to some PAHs.²⁵ Additionally, a dose-dependent relationship between higher OPFR exposures and teacher ratings of less responsible behavior (p = 0.07) and more external behavior problems (p = 0.03) was observed in children aged 3-5 years, while children rated by their teachers as less assertive had higher PBDE exposure (p = 0.007).³³

Outside the U.S., data are lacking on exposure monitoring via wristbands for both children and adults. In low and middle-income countries, it is estimated that 250 million children do not reach their developmental potential due to poverty and other disparities, including higher burdens of environmental exposure.³⁴ For certain classes of organic chemicals like flame retardants, children have higher exposures than adults^{16, 35-38}, necessitating exposure monitoring specifically targeted to this age group. Furthermore, assessing personal exposure in children is critical to understand the potential effects of chemicals on their growth and development.

Given the limited research on children’s personal exposure in lower income countries using wristbands, our objective was to qualitatively and quantitatively characterize organic chemical exposures among Montevideo school children. Our study, named the Scola-Exposome, leveraged ongoing research activities within a cohort of elementary school children in Montevideo, Uruguay that aims to understand the effects of metal exposures on children’s cognitive and behavioral function.^39-50 Outcomes of this study will provide valuable information on exposures to organic contaminants in children outside of the U.S.

2. Materials and Methods

2.1. Study Design

Participants of the Scola-Exposome study included 24 children aged 6.0-7.8 years, attending first grade in 9 elementary schools in different areas of Montevideo, Uruguay. The children were participating in a larger environmental cohort study of school-aged children called Salud Ambiental Montevideo. Families recently recruited into Salud Ambiental Montevideo were screened for eligibility for Scola-Exposome. Families were first asked if they were willing to participate in the additional study and second, if they had a refrigerator in the home to store samples for a separate study purpose.

Children and their parents attended a meeting with the study coordinator at the Catholic University of Uruguay, to receive the bands and further instructions. Children took the bands out of the storage jars, placed them on their own wrists, and then were asked to wear the bands continuously during play, wash, and sleep for seven days. After the deployment period, children met again with the study coordinator during a scheduled home visit, took off the bands and placed them back in the respective jars. Detailed demographic information was collected as part of ongoing research activities in Salud Ambiental Montevideo, including maternal age, education and occupation; family structure and income. The child and family’s health-related behaviors were also queried, including the child’s personal hygiene habits (daily frequency of hand and face-washing), smokers in the house and child’s contact with smokers. Additionally, during the scheduled activities for Salud Ambiental Montevideo, a nurse took a fasting blood sample from the child’s antecubital vein for the determination of hemoglobin (an indicator of iron status & anemia) and blood lead (BLL) concentrations. The analytical procedures for both measures have been described in detail elsewhere.⁴⁷

All study protocols and materials related to the study were approved by the University at Buffalo Institutional Review Board and the Ethics Committee of the Catholic University of Uruguay. Parents/legal guardians provided written consent prior to involvement in the study; children provided verbal assent.

2.2. Chemicals

Analytical standards for ¹³C₁₂-PBDEs, ¹³C₁₂-polychlorniated biphenyls (PCBs), ¹³C₆-hexabromobenzene, and ¹³C₁₂-1,2-bis(2,4,6-tribromophenoxy)ethane were obtained from Wellington Labs Inc. (Guelph, ON, Canada) and ¹³C₁₂-p,p’-dichlorodiphenyldichloroethylene, d₁₀-diazinon, and deuterated-OPFRs were acquired from Cambridge Isotope Laboratories Inc. (Tewksbury, MA) (Table S1). Acetonitrile (ACN), ethyl acetate, hexanes, isooctane, and methanol were purchased from Fisher Scientific (Pittsburg, PA). Sep-Pak^™ C18 solid-phase extraction (SPE) cartridges (500 mg, 3 cc) were obtained from Waters Inc. (Milford, MA).

2.3. Wristband Collection

Silicone wristbands (100% silicone, https://24hourwristbands.com/, Houston, TX) were cleaned prior to deployment in the second half of 2018 using the protocol described by O’Connell et al.¹ Briefly, ≤65 g of silicone were soaked in 800 mL of mixed organic solvent (see below) for five extractions for a minimum of 2.5 hours each at 60 rotations per minute using an Innova 2000 (New Brunswick Scientific Co., Edison, NJ) platform shaker. The first three pre-cleaning extractions consisted of ethyl acetate/hexanes (1:1, v:v) and the last two extractions used ethyl acetate/methanol (1:1, v:v) as additional pre-cleanings. Cleaned wristbands were air-dried in a fume hood and then stored in 60-mL certified contaminant free glass amber jars (VWR International, Radnor, PA) until deployment. After participants wore the wristbands for a 7-day period, the bands were placed back in their respective jars and stored at −20°C at the Catholic University of Uruguay, and shipped back to the University at Buffalo, where samples were stored at −40°C until analysis.

2.4. Wristband Extraction.

Using solvent-rinsed surgical scissors, wristbands were cut into 8 equal pieces and two pieces (1 g) were transferred to 50-mL acid-washed glass centrifuge tubes for extraction. Samples were then spiked with 5 ng of surrogate standards dropwise directly onto the samples. Extraction and cleanup was performed using a method previously described by Kile et al.⁴, which was modified to decrease the amount of organic solvent used. Briefly, extraction was performed twice, each using 25 mL of ethyl acetate for two hours on an orbital shaker at 60 rotations per minute. Ethyl acetate extracts were combined and concentrated to 300 μL, then ACN (3 mL) was added to samples prior to SPE cleanup. The SPE cartridges (C18, 500 mg, 3 cc) were rinsed with ACN (6 mL). Each of the extracts was passed through an SPE cartridge, collecting the eluent in a 10-mL acid-washed glass centrifuge tube and further eluted with ACN (6 mL) into the same collection vessel. Samples were then evaporated to dryness and reconstituted in 200 μL of ¹³C₁₂-PCB-138 in isooctane and transferred to 2-mL amber vials. An aliquot (50 μL) of each sample was transferred to a 200-μL conical insert for analysis.

2.5. Instrumental Analysis

Wristband samples were analyzed for a total of 45 chemicals including 14 PCBs, 13 PBDEs, 6 OPFRs, 8 novel halogenated flame retardants (NHFRs), and 4 pesticides. Analysis of samples was performed using a Trace GC Ultra^™ coupled to a TSQ Quantum XLS^™ triple quadrupole mass spectrometer (Thermo Fisher Scientific, West Palm Beach, FL) with electron ionization mode. A programmable temperature vaporization inlet was utilized for injection and was set at an initial temperature of 89°C, then increased at a rate of 7.5°C/min for 0.2 minutes. The analytes were then transferred to the column at 330°C for 1 minute. The split flow was set at 60 mL/min with a splitless time of 1.13 minutes. The flow rate remained constant at 1.0 mL/min with helium (99.999% purity) as the carrier gas. Separation was achieved on a 15-m DB-5HT^™ capillary column with a 0.25 mm internal diameter, and a 0.10 μm film thickness (Agilent, Santa Clara, CA). The oven temperature program was as follows: initial temperature of 100°C was held for 2 minutes, ramped at 10°C/min to 210°C, followed by a second ramp at 20°C/min to a final temperature of 330°C, and was held for 10 minutes. Retention times, mass fragmentation characteristics, and optimized collision energies are presented in Table S2.

2.6. Quality Assurance/Quality Control.

Pre-baked (250°C) aluminum foil was used to weigh wristbands to prevent any contamination; and all tools were rinsed thoroughly with ethyl acetate prior to and between handling of each wristband. Isotope dilution was utilized for the quantitation of PBDEs, PCBs, OPFRs, NHFRs, and pesticides using isotopically labelled standards (Table S1). Analytical figures of merit are provided in the Table S2; method limits of detection (LODs) were determined based on 3 times the standard deviation of the lowest observed concentration spiked in wristband matrix (n=7) and ranged from 0.001 to 0.014 ng/g. Limits of quantitation (LOQs) were determined based on 10 times the standard deviation of the lowest observed concentration spiked in wristband matrix (n=7) ranging from 0.003 to 0.047 ng/g. Recovery experiments were performed by spiking (n=3) cleaned wristbands (1g) with a standard solution containing 5 ng/g of each analyte. The overall recoveries ranged from 50 to 122; more than 86% of the target analytes have recoveries greater than 70%. Laboratory blanks (n=5) were used to subtract any potential interferences. To confirm positive detection of a targeted compound, a retention time match of ± 0.25 min with the reference standard was used.

2.7. Statistical Analysis

Descriptive statistics (geometric mean, median, interquartile range) for the analytes were performed using R 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria). Further analysis was performed for chemicals with a detection frequency >50%. For compounds below the LOQ, values were assigned equal to the LOD divided by the square root of 2. Spearman correlations for the analytes and descriptive statistics for the sociodemographic variables were performed using Stata 14 (StataCorp LLC., College Station, TX).

3. Results and discussion

3.1. Study Outline

Twenty-four silicone wristbands were deployed and all participants completed the study according to previously described guidelines. One wristband could not be analyzed due to high chromatographic interferences, reducing the sample to 23 wristbands that were quantitatively analyzed for 45 chemicals. Participant characteristics are provided in Table 1. The study population was 60.9% female and participant age ranged 6.0 - 7.8 years. Furthermore, 39% of families reported having a car, 100% cleaned their floors more than once per week, and 18 out of 23 families reported cleaning dust more than one time per week. Of note, the children had overall good nutritional status, with little evidence of anemia and all had BLL < 10 μg/dL. On the other hand, 13% had BLL ≥ 5 μg/dL.

Table 1:

Characteristics of children participating in Scola-Exposome with passive sampler data

Characteristic	N	% or Mean ± Standard Deviation	Range
Child Characteristics
% Boys	23	39.1%
Age (months)	23	82.2 ± 4.7	72 – 94
Child always washes hands prior to eating	23	73.9%
Child does not share space with smokers	23	73.9%
Hemoglobin (g/dL)	23	13.2 ± 0.9	11.3 – 14.6
Blood lead concentration (μg/dL)	23	3.9 ± 1.3	3.3 – 7.6
BLL ≥ 5 μg/dL		13.0%
Parent/Household Characteristics
Mother’s age (y)	23	32.2 ± 5.2	24 – 43
Mother continued her education after 14 y of age	23	30.4%
Parents are married or cohabitating	23	73.9%
Child lives with:	22
Both parents		40.9%
Mother only		13.6%
Other arrangement		45.4%
Mother currently smokes	23	43.5%
Monthly family income < 15,0001 Uruguayan pesos	23	52.2%

¹In the second half of 2018, when the study was conducted, this monthly income was equivalent to roughly 465 U.S. dollars; according to CEIC (https://www.ceicdata.com/en), the average monthly household income in Montevideo was just shy of 74,000 Uruguayan pesos.

With the exception of NHFRs, at least one analyte from the five groups of chemicals (PCBs, PBDEs, OPFRs, NHFRs, and pesticides) in the method was detected, with each extract from the wristbands containing 8 to 19 detected chemicals (13 ± 3, mean ± standard deviation (SD)). Overall, 13 out of 45 compounds were detected in over 50% of the samples (Figure 1, Table 2). For the analytes detected at less than 50% detection frequency, quantities of individual target analytes in each wristband sample are further provided in the supplemental material (Tables S3-6). OPFRs were found at the highest concentration and were the most frequently detected group of compounds, contributing 94 ± 2% of total chemical concentrations.

Figure 1:

Thirteen analytes belonging to five different classes of chemicals were detected in >50% of wristband samples from the Scola-Exposome study in Montevideo, Uruguay. Classes of chemicals include polychlorinated biphenyls (PCBs), polybrominated diphenyl ethers (PBDEs), organophosphorus flame retardants (OPFRs) and pesticides.

Table 2:

Concentrations of 13 analytes (ng/g wristband) with >50% detection frequency in wristbands from the Scola-Exposome study in Montevideo, Uruguay (n=23). LOQ represents the limit of quantitation. For the analytes detected at less than 50% detection frequency, quantities of individual target analytes in each wristband sample are further provided in the supplemental material (Tables S3-6).

Compound	Percent detect (n)	LOQ	Median	Geometric mean	Percentiles		Maximum
					25th	75th
PCB-180 1	69.6 (16)	0.028	0.33	0.13	<LOQ	0.52	2.32
ΣPCBs	-	-	0.52	0.48	0.19	1.18	8.35
BDE-47	95.7 (22)	0.019	1.09	1.08	0.57	1.63	138
BDE-99	69.6 (16)	0.019	0.81	0.22	<LOQ	1.47	190
ΣPBDEs	-	-	3.00	2.73	1.24	5.06	433
TBP	82.6 (19)	0.006	66.4	12.8	39.4	136	510
TCEP	87.0 (20)	0.023	20.8	8.99	3.77	77.3	1420
TCPP	82.6 (19)	0.017	208	65.3	158	1010	4480
TDCPP	56.5 (13)	0.009	5.10	0.64	<LOQ	78.5	1500
TPhP	95.7 (22)	0.007	85.4	97.0	57.2	429	8920
EHDPHP	100 (23)	0.009	290	337	131	661	4820
ΣOPEs	-	-	1020	1370	806	2970	12300
Diazinon	95.7 (22)	0.020	12.7	12.9	7.68	32.4	340
p,p’-DDE	100 (23)	0.027	1.30	1.32	0.85	2.12	5.19
p,p’-DDD	78.3 (18)	0.020	0.67	0.27	0.19	1.46	4.60
p,p’-DDT	87.0 (20)	0.007	1.63	0.80	0.95	3.67	13.7

¹Abbreviations are as follows: 2,2’,3,4,4’,5,5’-heptachlorobiphenyl (PCB-180), 2,2’,4,4’-tetrabromodiphenyl ether (BDE-47), 2,2’,4,4’,5-pentabromodiphenyl ether (BDE-99), tributyl phosphate (TBP), tris(2-carboxyethyl)phosphine) (TCEP), tris (chloroisopropyl) phosphate (TCPP), tris(1,3-dichloroisopropyl) phosphate (TDCPP), triphenyl phosphate (TPhP), 2-ethylhexyl diphenyl phosphate (EHDPHP), p,p’-dichlorodiphenyldichloroethane (p,p’-DDD), p,p’-dichlorodiphenyldichloroethylene (p,p’-DDE), p,p’-dichlorodiphenyltrichloroethane (p,p’-DDT).

3.2. Polychlorinated Biphenyls

PCBs were detected in 19 out of 23 wristbands and exhibited the lowest concentrations of all compound classes with a median sum of PCBs (ΣPCBs) concentration of 0.52 (range, 0.19 – 8.35) ng/g wristband (Table 2). On average, two PCB congeners were detected in each wristband and the maximum number of congeners in a single wristband analyzed was four. PCB-180 was the most frequently detected, in 16 wristbands, and contributed over 50% of ΣPCB concentrations. PCB-170 had the second highest detection frequency in 5 out of 23 wristbands, however it only contributed 5% of ΣPCB concentrations. All other congeners were detected in only one or two wristbands, furthermore, there were three wristbands that did not contain any PCBs. Interestingly, the three samples with the highest ΣPCB concentrations also had the greatest number of PCBs observed, with three or four congeners.

To the best of our knowledge, there has been no quantitative reports of PCB concentrations in wristband samplers worn by children in any country. However, one study detected five PCBs (PCB-102, PCB-88, PCB-93, PCB-95, PCB-98) in wristbands of urban participants, mostly adults, living in the U.S. in a qualitative screening.¹⁴ The general lack of detection of PCBs in other studies is most likely due the location and different exposure profiles of various settings; however, analytical differences and limits of detection of targeted methods versus large chemical screening can also play a role.

In Uruguay, the entry and trade of PCBs had not been regulated prior to the Stockholm Convention calling for a Global Monitoring Program (GMP) of persistent organic pollutants (POPs) in 2007. Prior to that, it was estimated that 40,000 transformers, which are a major source of PCBs, were operating in Uruguay in 2006.⁵¹ The country has had limited experience regarding management of PCBs due to lack of infrastructure and equipment to contain and dispose of them in an environmentally appropriate manner. To address this concern, a project aimed at implementing systems for the safe disposal of PCBs was crafted in 2008.⁵¹

Previous studies have aimed to research PCBs in Uruguay and as a part of the GMP performed in 2007, PCBs were detected in breast milk of Uruguayan women at levels above 25 ng/g lipid, and in air samples in Montevideo above 200 pg/m³.⁵² Air concentrations were the second highest level of the seven Latin American countries tested since the GMP began. Results of efforts to remove and properly dispose of PCBs in Uruguay have shown to be effective as concentrations in air samples slightly decreased between 2010 to 2011 but have not been reported since⁵²; nonetheless concentrations were detected in the wristbands from the Scola-Exposome sample.

3.3. Polybrominated Diphenyl Ethers

The presence of PBDEs was confirmed in 22 out of 23 wristbands, with a median concentration of the sum of PBDEs (ΣPBDEs) of 3.00 (range, 0.26 – 433) ng/g wristband (Table 2). Each wristband contained 0 – 9 congeners (3 ± 2, mean ± SD). BDE-47 was most frequently detected, in 22 wristbands, followed by BDE-99 in 16 and BDE-100 in 9 out of 23 wristbands. These three congeners made up an average of 95% of ΣPBDE concentrations. BDE-47 had the greatest contribution, an average of 56%, to ΣPBDE concentrations, followed by BDE-99 and −100 at 26% and 13%, respectively. These three congeners are the most abundant BDEs detected globally, since they comprised the largest percent of one of the most commonly manufactured PBDE mixtures.^{53, 54}

Results from this study are similar to trends reported in preschoolers’ wristbands from Oregon, where BDE-47 and −99 also had the highest concentrations.⁴ However, ΣPBDE concentrations were higher compared to our findings, with an average of 4.49 ± 5.59 ng/g/day and reaching up to 171 ng/g/day.⁴ BDE-47 was also the most abundant congener in wristband samplers of adults from North Carolina, followed by BDE-99 and −100.¹¹ Furthermore, both adults in North Carolina and children in Oregon had greater mean concentrations of BDE-47, 55.9¹¹ and 30.4 ng/g⁴ respectively, than observed in our study.

One finding from household surveys may provide some explanation for lower concentrations of PBDEs in our study compared to the U.S. Caregivers in the Salud Ambiental Montevideo cohort reported that 93.5% of families never used carpet cleaner, suggesting that carpets might not be present in the home. PBDEs are used significantly in polyurethane foams that are recycled and used in the production of carpet padding.⁵⁵ Higher levels of PBDEs have also been observed in serum samples from carpet installation workers compared to the general U.S. population.⁵⁶

3.4. Organophosphorus Flame Retardants

A mean of 5 ± 1 (range, 2 – 6) OPFR compounds were detected in each wristband; 11 wristbands contained all six OPFRs analyzed. Total OPFR (ΣOPFR) concentrations ranged from 208 to 12,300 ng/g, with a median of 1020 ng/g band (Table 2). All OPFRs were detected at greater than 50% frequency in samples and 2-ethylhexyl diphenyl phosphate (EHDPHP) had the highest median concentration of 290 ng/g and was the only OPFR detected in all 23 wristbands. A previous study by Hammel et al.²⁹ has also reported 100% detection frequency of EHDPHP in wristbands of children from North Carolina. In our study, tris(1-chloro-2-propyl) phosphate (TCPP) had the second highest median concentration of 208 ng/g and was detected in 19 of 23 wristbands, followed by triphenyl phosphate (TPhP) which was found in 22 of 23 children’s wristbands. Tris(1,3-dichloro-2-propyl) phosphate (TDCPP) had the lowest median concentration of 5.10 ng/g and was least frequently detected, in 13 wristbands. Comparatively, in wristbands from children in Oregon, TDCPP had the highest mean concentrations of 154 ± 171 ng/g/day followed by TPhP at 134 ± 290 ng/g/day.⁴ Additionally, in wristbands worn by children from North Carolina, TPhP had the greatest median concentration of 872.9 ng/g band, followed by TDCPP with 179.7 ng/g.²⁹ Our study reveals differences in exposure patterns of OPFRs among children in Uruguay compared to the U.S. In sum, OPFR concentrations were lower in wristbands of Montevideo children, and specifically, had lowest levels of TDCPP compared to children in other studies, but Uruguayan children also had highest exposure to TCPP.

A recent study reported that in Southern California, longer commute times were associated with higher TDCPP exposures.²⁷ Separate studies also reported that TDCPP is the dominant OPFR found in cars in the United Kingdom⁵⁷, and is frequently detected in dashboard dust worldwide^{58, 59}. In the Scola-Exposome population, only 9 of 23 families reported having cars, indicating that a majority of the children either walk or utilize public transportation, possibly explaining the lower levels of TDCPP observed in our study. On the other hand, it has been observed that TCPP concentrations exceeded all other OPFRs in living room dust⁵⁹ and is also used widely in furniture in the U.S.⁶⁰ The lower concentrations of TDCPP and higher concentrations of TCPP found in wristbands from Uruguay children compared to the U.S. suggest that children in Uruguay travel by car less frequently and possibly spend more time in or around the house. The differences in OPFR concentrations among settings could arise from variations in manufacturing and additives used in products from different countries, as well as in differences in socioeconomic status and lifestyles of the participants. The country-specific patterns of children’s exposure highlight the need for studies in lower income countries to understand exposures and outcomes worldwide.

3.5. Pesticides

Four pesticides were analyzed in wristband extracts, including p,p’-dichlorodiphenyltrichloroethane (DDT), its metabolites (p,p’-dichlorodiphenyldichloroethylene (DDE), p,p’-dichlorodiphenyldichloroethane (DDD)), and the insecticide diazinon. All four compounds were detected in more than 75% of samples (Table 2). The primary metabolite of DDT, DDE, was identified in 100% of wristbands with a median concentration of 1.30 ng/g. The parent compound DDT was detected in 20 wristbands, with median concentration of 1.63 ng/g. In 16 out of 23 wristbands, the concentration of p,p’-DDT was higher than DDE. To estimate exposures to recent applications of DDT, the ratio of p,p’-DDT/( p,p’-DDE + p,p’-DDD) was calculated as previously performed in other wristband studies.^{2, 6} A ratio greater than one suggests exposure to a recent application of DDT^61-63, and was observed in ten of 23 wristbands in our study.

DDT is also included in the Stockholm Convention’s GMP in Uruguay.⁵² In 2009, the sum of DDT and its 5 total metabolites and isomers were identified in breast milk of Uruguayan women at levels greater than 120 ng/g lipid.⁵² In addition, Montevideo air samples in 2011 were above 1.0 pg/m³, both with DDE as the major isomer identified.⁵² High levels of pesticides observed in the study in Uruguay may have been due to lack of regulation and proper disposal facilities for outdated and banned chemicals,⁶⁴ thus leaving waste on farms or near water resources, and in some cases burned or utilized for other purposes.⁶⁵ Additionally, dicofol, an organochlorine pesticide known to contain DDT (due to its use as intermediate in dicofol production), may be a secondary source of DDT in Uruguay. Dicofol has been documented as a source of exposure to DDT in a number of countries.^{66, 67} Currently, there are no studies specifically linking dicofol to DDT exposure in Uruguay, but it appears that in 2011 there was one commercial formulation approved for use in the country.⁶⁸ The specific reasons for exposure to recent applications of DDT in urban Uruguayan children are unknown and should be investigated further.

DDE was detected in 55.7% of wristbands worn by Latina adolescent females from California.¹⁷ However outside the U.S., a study in Peru had a higher frequency of detection of DDT and its metabolites in 63 of 65 wristbands, with concentrations ranging from 8.8 to 5400 ng/g.⁶ Four of the wristbands from Scola-Exposome, were within the reported range from the wristbands in Peru, however most of the observed concentrations in wristbands of Uruguayan children were lower. We observed lower concentrations of DDT and metabolites in urban school-age children, whereas the demographic in Peru consisted of mainly adults that most commonly reported their occupation as “farm worker”; only 12 of 68 the Peruvian participants were students. Additionally, Bergmann et al., observed that 48% of Peruvian wristbands had p,p’-DDT/(p,p’-DDE + p,p’-DDD) ratios greater that one.⁶ Our study found similar results, with ten of 23 wristbands (44%) having DDT ratios greater than one, suggesting that despite ceased production, both urban and farming populations in both countries may have relatively recent exposures to DDT. The complementary results from both settings indicate the need to further investigate exposure to DDT and its metabolites across South America.

Diazinon was detected in 22 of 23 wristbands (Table 2) with a median concentration of 12.7 ng/g. Diazinon was used as a household insecticide since 1956 in the European Union, until 2007 when this use was prohibited.⁶⁹ Further review of literature focused on environmental analysis of diazinon in Uruguay, revealed that it has been detected in 12% of surface water samples from two lagoons (Laguna de Rocha and Laguna de Castillos) in Uruguay approximately 200-250 km east of Montevideo, potentially explaining the high detection frequency in humans.⁷⁰ To the best of our knowledge, there is no quantitative data available to compare diazinon concentrations in silicone wristbands in other studies.

3.6. Spearman correlation coefficients

Spearman correlation coefficients (ρ) in Table 3 were calculated for analytes that had at least 50% detection frequency. Strong correlations between chemicals can tell us how these compounds, as a class or individually, are used in the industry in the specific region and point to potentially common sources of exposure to different chemicals. Moreover, comparing correlations between different study populations is valuable in uncovering differences in sources or patterns of exposure. BDE-47 and 99 were positively correlated (ρ = 0.51, p < 0.05), as expected given that PBDEs are manufactured in mixtures and previous studies have found that BDE-47 and −99 are the two major congeners in the penta-BDE formulation.^{53, 54} Hammel et al. also observed similar correlations between these two congeners in wristband samples from adults in North Carolina.¹¹ Furthermore, DDT and its metabolites were highly positively correlated (ρ 0.88 – 0.93, p < 0.01). TPhP and EHDPHP were also positively correlated (ρ = 0.79, p < 0.01), likely due to their structural similarity. EHDPHP is used primarily in food-packaging plastics, rubber, paints, textile coatings, and adhesives⁷¹, while TPhP is used primarily in polyvinyl chloride (PVC), polyurethane foams, hydraulic fluids, and themroplastics.⁷² Even though these two chemicals have different primary uses, they are both applied in a wide variety of products such as electronics, many of which may overlap, thus leading to the observed correlation. No other compounds in the OPFR class significantly correlated with each other. In contrast, wristbands from adults in Southern California showed significant positive correlations between TDCPP and tributyl phosphate, tris (2-butoxyethyl) phosphate, TCPP and EHDPHP.²⁷ Correlations between TDCPP and TCPP observed in Southern California are attributed to co-application in products containing polyurethane foams as well as dust from vehicles.²⁷ Different correlations among compounds in studies conducted in the U.S. and Uruguay point towards the need for further exposome research worldwide.

Table 3:

Spearman correlations among chemical compounds detected in at least 50% of silicone wristbands in Scola-Exposome study.

	PCB-180	BDE-47	BDE-99	TBP	TCEP	TCPP	TDCPP	TPhP	EHDPHP	Diazinon	p,p’-DDE	p,p’-DDD
BDE-471	0.20	-	-	-	-	-	-	-	-	-	-	-
BDE-99	0.23*	0.51**	-	-	-	-	-	-	-	-	-	-
TBP	0.06	0.18	0.43*	-	-	-	-	-	-	-	-	-
TCEP	−0.19	−0.08	0.20	0.42*	-	-	-	-	-	-	-	-
TCPP	−0.40*	−0.15	−0.03	0.16	0.29	-	-	-	-	-	-	-
TDCPP	0.28	−0.11	0.21	0.23	0.32	0.24	-	-	-	-	-	-
TPhP	0.22	0.16	0.07	0.35*	0.05	0.19	−0.25	-	-	-	-	-
EHDPHP	−0.10	0.08	−0.09	0.26	0.34	0.30	−0.38	0.79***	-	-	-	-
Diazinon	−0.27	0.10	0.07	0.30	0.31	0.08	0.03	0.10	0.31	-	-	-
p,p’-DDE	0.04	−0.14	0.22	0.04	−0.01	−0.18	0.07	0.09	0.03	−0.17	-	-
p,p’-DDD	0.11	−0.18	0.18	0.10	−0.15	−0.25	−0.05	0.12	0.05	−0.07	0.89***	-
p,p’-DDT	0.09	−0.12	0.17	0.15	−0.09	−0.16	−0.01	0.20	0.15	−0.14	0.88***	0.93***

^*p<0.1

^**p<0.05

^***p<0.01

¹Abbreviations are as follows: 2,2’,4,4’-tetrabromodiphenyl ether (BDE-47), 2,2’,4,4’,5-pentabromodiphenyl ether (BDE-99), p,p’-dichlorodiphenyldichloroethane (p,p’DDD), p,p’-dichlorodiphenyldichloroethylene (p,p’DDE), p,p’-dichlorodiphenyltrichloroethane (p,p’DDT), 2-ethylhexyl diphenyl phosphate (EHDPHP), 2,2’,3,4,4’,5,5’-heptachlorobiphenyl (PCB-180), tributyl phosphate (TBP), tris(2-carboxyethyl)phosphine) (TCEP), tris (chloroisopropyl) phosphate (TCPP), tris(1,3-dichloroisopropyl)phosphate (TDCPP), and triphenyl phosphate (TPhP).

3.7. Future Directions

Given the scarcity of studies on exposure assessment in children via silicone wristbands, more research is needed to better understand to what extent chemical concentrations in the wristbands reflect body burdens of exposure. Moreover, several issues with respect to study design need to be clarified, including the duration and frequency of sampling to reflect typical levels and variability of exposures. Finally, epidemiological studies are needed to understand how the levels of exposure observed in our study and other related research, relate to both disease processes and functional health outcomes.

4. Conclusions

Most studies using silicone wristbands currently focus on chemical exposures in adults living in the U.S. Additional studies are needed to understand the extent and patterns of chemical exposures in children, specifically in settings other than the U.S. Information on low and middle-income countries, where exposures are potentially greater due to unregulated use of chemicals, is particularly lacking and should be prioritized. This study is the first to focus on children’s exposure using silicone wristbands in South America, which reveals that legacy pollutants remain a concern for Uruguayan children. High levels of TCPP and lower concentrations of TDCPP in Scola-Exposome wristbands highlight lifestyle and sociodemographic differences in study samples from Uruguay and the U.S. This study represents an important stride towards determining children’s exposome outside of the Global North. Furthermore, results provide data on chemical exposure of a continually studied population, vital to unveiling the exposome and the health effects of children’s exposures to chemical mixtures.