Comparing the use of silicone wristbands, hand wipes, and dust to evaluate children's exposure to flame retardants and plasticizers

Abstract

Organophosphate esters (OPEs) are applied as additive flame retardants, and along with phthalates, are also used as plasticizers in consumer products. As such, human exposure is common and chronic. Deployed as personal passive samplers, silicone wristbands have been shown to detect over a thousand industrial and consumer product chemicals; however, few studies have evaluated chemical concentrations with corresponding established exposure biomarkers, especially children. Further, little is known about how well the wristbands predict individual exposure compared to existing validated external exposure tools such as indoor air, dust, and hand wipes. Here, we analyzed wristbands worn by children (ages 4-6) for 18 OPEs and 10 phthalates and compared them to corresponding urinary biomarkers. In wristbands, 13 of 18 OPEs and all phthalates were detected in >80% of wristbands, and 6 OPEs and 4 phthalates were significantly associated with corresponding urinary metabolites (r_s=0.2-0.6, p<0.05). When compared to paired hand wipes and house dust, wristbands were found to have similar or greater correlation coefficients with respective urinary biomarkers. These results suggest that wristbands can serve as effective and quantitative assessment tools for evaluating personal exposure to some OPEs and phthalates, and for some chemicals, may provide a better exposure estimate than indoor dust.

Introduction

Organophosphate esters (OPEs) and phthalates are ubiquitously found in indoor environments because of their extensive use in consumer products as flame retardants, plasticizers, preservatives, fragrances, etc. Their partitioning characteristics allow for their presence in both the gas and condensed phases at room temperature conditions.¹ As such, human exposure to these compounds is widespread and likely chronic.^2-5 Many of these compounds are suspected to impact endocrine function, reproduction, neurodevelopment, and obesity, among other adverse health effects.^6-11 This is of particular concern for children who typically have higher exposures to these chemicals and are at greater risk for adverse health impacts that could persist into adulthood.^12,13

Over the last several years, determining how people, and specifically children, are exposed to OPEs and phthalates has been prioritized as a key first step in order to understand how one can reduce individual exposures. Inadvertent ingestion of dust, hand-to-mouth contact, inhalation, and dermal absorption have all been posited to be important pathways of exposure for these compounds.^14-18 In an effort to estimate the relative contributions of these pathways and evaluate exposure levels that can be attributed to the indoor environment, dust and indoor air have been extensively sampled.^2,3,5,19,20 For OPEs, indoor dust and air have been shown to be weakly but significantly associated with corresponding urinary metabolites, and inhalation has been posited to be a prominent pathway for exposure when compared to dust for chlorinated OPEs.^17,21-23 Similarly, phthalates in indoor dust and air have been shown to be correlated with urinary biomarkers and contribute to total phthalate exposure, in addition to contributions from dietary, inhalation and dermal absorption pathways.^24-26 Hand wipes, which are hypothesized to represent exposure via hand-to-mouth contact, inadvertent dust ingestion and dermal absorption, have also been used to reliably measure exposure to OPEs and phthalates when compared to their respective biomarkers in urine.^27,28

In the last several years, silicone wristbands have been introduced as personal passive samplers to measure a wide range of consumer product and industrial chemicals among adults and children.^29-38 In adults, concentrations of several OPEs, polycyclic aromatic hydrocarbons (PAHs) and polybrominated diphenyl ethers (PBDEs) on the wristbands have been shown to be significantly and positively associated with their exposure biomarker concentrations; in children, a similarly positive association was observed between wristband nicotine and urinary cotinine.^31,32,37,38 These studies demonstrate that wristband measurements may be representative of internal dose across three classes of compounds as well as environmental nicotine; however, children and adults may have disparate pathways and sources of exposure based on differences in product use, behaviors, and time spent in different microenvironments. A recent study suggested that the silicone wristbands integrate inhalation and dermal exposures, as shown by improved regression results when comparing wristbands to combined masses of compounds on hand wipes and air samplers.³⁹ While this work provides important insights on the exposure pathways that wristbands capture, few studies have compared wristbands and other existing exposure measurements to biomarkers, particularly in children.

Here, we utilized silicone wristbands (n=77) to evaluate children’s exposure to 28 chemicals across 2 classes of semi-volatile organic compounds (OPEs and phthalates/non-phthalate plasticizers) through associations with available urinary biomarkers in a children’s cohort recruited in central North Carolina. Our previous work in the same but expanded study population (n=203) demonstrated that hand wipes were an improved measure of exposure compared to indoor dust for OPEs and that hand wipes and dust were both useful in evaluating children's exposure to phthalates^40,41 In this study, we analyzed OPEs and phthalates on wristbands collected during this study and compared wristbands associations with paired hand wipes and house dust samples to assess their effectiveness in predicting urinary metabolite concentrations. Our primary objectives were two-fold: 1) to ascertain if wristband concentrations of OPEs and phthalates were associated with exposure biomarkers for children, and 2) to determine if associations with urinary biomarkers were similar or improved using silicone wristbands compared to more traditional exposure matrices, such as house dust and hand wipes. Identifying the best sampling tool or matrix for evaluating children’s exposure is crucial for epidemiological studies to accurately examine potential adverse health outcomes and may help to design better interventions to reduce exposure to these compounds.

Materials and methods

Study Population.

Mothers in the Newborn Epigenetics STudy (NEST), a prospective pregnancy cohort study based in central North Carolina (2005-2011), were re-contacted and asked to participate with their children in the Toddler’s Exposure to SVOCs in the Indoor Environment (TESIE) study.^42,43 Within this subset of the TESIE study population, 77 children ages 3-6 years from 74 different families wore pre-cleaned wristbands for 7 days. Silicone wristbands were added to sample collection protocols later and thus deployed among participants halfway through the study, starting in August 2015. Hoffman et al. 2018 provides a more detailed description of recruitment and enrollment of participants within the full TESIE study, which included 203 children from 190 families who participated between September 2014 and April 2016. Study team members conducted home visits with each of the enrolled families to collect environmental samples (dust and hand wipes) within the home and biospecimens (urine) from the children. The children put on the wristbands during the home visit and wore the wristbands for the subsequent 7 days. Furthermore, additional information about the home environment and children’s health and behavior were recorded using a questionnaire administered by the study team. All study protocols and related materials were reviewed and approved by the Duke Medicine Institutional Review Board. Legal guardians provided informed consent prior to any collection of samples or questionnaire data for the TESIE study, and mothers had provided informed consent prior to their participation in NEST.

Wristband Collection and Analysis.

Commercially available, adjustable wristbands were purchased in a variety of colors (diasstro adjustable silicone wristband bracelets, Amazon.com) and cleaned as described in Hammel et al. 2016, with sequential 12-hour Soxhlet extractions with 1:1 ethyl acetate/hexane (v/v) and 1:1 ethyl acetate/methanol (v/v) and passive drying in a fume hood.³⁸ Wristbands were individually wrapped in pre-cleaned aluminum foil and placed in an air-tight, labeled 40 mL amber jar. The children were asked to wear the wristbands continuously for 7 days during all daily activities including bathing and sleeping starting on the day of the home visit. For the purpose of comparison, hand wipes and dust samples were collected just prior to the child putting on the wristband during the home visit. Urine samples were collected during the 7-day sampling period. Following the sampling period, legal guardians were asked to retrieve the wristbands, rewrap them in foil, and replace them in the amber jar. Wristbands were stored at −20°C from the time of collection until analysis, which ranged from 2 to 3 years.

Wristband samples (n = 77) were extracted and analyzed for a suite of organophosphate esters and phthalates and non-phthalate plasticizers using a method that was similar to our methods for hand wipe and dust extraction (Table 2).⁴⁰ Using solvent-rinsed, stainless-steel scissors and forceps, approximately one-third of each wristband or field blank (n=8) was cut from the total wristband, with the remainder of each band being returned to its respective amber jar and stored at −20°C for future analyses. This segment was carefully cut into another 3 equal parts to facilitate extraction, accurately weighed and placed into a labeled 50 mL glass centrifuge tube. The wristband segments were spiked with the following internal standards: d₁₂-TCEP (100 ng), d₁₅-TDCIPP (100 ng), ¹³C-TPHP (100 ng), d₄-DMP (200 ng), d₄-DEP (200 ng), d₄-BBP (200 ng), and d₄-DEHP (200 ng). D₁₂-TCEP, d₁₅-TDCIPP, and ¹³C-TPHP were purchased from Wellington Laboratories (Guelph, ON). D₄-DMP, d₄-DEP, d₄-BBP, and d₄-DEHP were purchased from Cambridge Isotope Laboratories (Tewksbury, MA). Using a method similar to Hammel et al. 2018, the samples were extracted 3 times with 10 mL 1:1 hexane/dichloromethane (v/v) then concentrated to 1.0 mL using a Thermo Scientific SpeedVac Concentrator.³⁷ The 1.0 mL extracts were fractionated using Florisil^® solid-phase extraction cartridges (Supelclean ENVI-Florisil, 6 mL, 500 mg bed weight; Supelco), eluting the second fraction (F2) with 10 mL ethyl acetate for organophosphate esters and phthalates. The first (F1) and third (F3) fractions were eluted with 8 mL hexane for the brominated flame retardants and with 8 mL methanol for PFASs, respectively (data not included here). Each fraction was once again concentrated to 1.0 mL using the SpeedVac Concentrator, with the F1 and F3 fractions transferred to autosampler vials (ASVs) and stored at −20°C for future analysis. The F2 fractions were transferred to ASVs for analysis by gas chromatography/mass spectrometry in electron ionization (GC-MS-EI) for the chlorinated OPEs, phthalates, and non-phthalate plasticizers. The phthalates and non-phthalate plasticizers were quantified using a 10:1 dilution in hexane to account for the high concentrations on the wristbands. The isopropylated triaryl phosphate (ITP) and tert-butylated triaryl phosphate (TBPP) isomers were quantified using GC Orbitrap GC-MS/MS for high-resolution separation of the isomers and reduction of interferences.

Table 2.

Descriptive statistics on concentrations measured in wristbands (n=77; ng/g wristband).

Class and Compound	Detection Frequency (%)	MDL	Median	10th Percentile	95th Percentile
Organophosphate Esters (OPEs)
TCEP	81.8	8.85	35.80	ND	319.7
TCIPP	98.7	0.89	55.16	17.03	848.5
TDCIPP	98.7	0.92	179.7	38.04	1,443
EHDPHP	100.0	0.06	73.61	21.74	306.9
TPHP	98.7	1.18	872.9	224.4	3,681
ITPs
2IPPDPP	98.7	1.37	373.4	66.24	1,499
3IPPDPP	94.8	0.66	40.83	4.46	237.0
4IPPDPP	98.7	0.66	112.5	21.53	537.9
24DIPPDPP	98.7	0.88	236.0	45.01	1,644
B2IPPPP	98.7	0.66	141.1	25.62	707.4
B3IPPPP	98.7	0.66	309.8	39.83	2,053
B4IPPPP	49.4	0.66	-	ND	39.55
B24IPPPP	50.6	32.88	36.53	ND	355.0
TBPPs
2tBPDPP	5.2	0.66	-	ND	0.47
4tBPDPP	98.7	0.66	119.5	29.83	477.6
B2tBPPP	0.0	0.66	-	ND	ND
B4tBPPP	88.3	0.66	20.68	ND	137.8
T4tBPP	10.4	0.66	0.33	ND	3.91
Phthalates
DMP	98.7	0.04	22.59	8.37	127.3
DEP	98.7	1.40	739.1	254.0	4,489
DiBP	98.7	0.55	976.2	311.1	6,081
DBP	98.7	4.27	274.3	114.4	668.2
BBP	100.0	0.45	1,059	211.8	9,895
DEHP	100.0	2.24	9,010	2,198	42,296
DiNP	98.7	4.73	7,073	1,848	26,058
DEHTP	100.0	1.19	8,493	2,293	33,857
Non-phthalate plasticizers
DEHA	100.0	3.05	999.9	188.6	3,998
TOTM	98.7	0.57	124.7	26.31	419.9

Recovery of internal standards was evaluated using d₁₈-TCIPP (100 ng) for d₁₂-TCEP and d₁₅-TDCIPP; d₁₅-TPHP (100 ng) for ¹³C-TPHP; and ¹³C-DCPH (100 ng) for the deuterated phthalate standards. D₁₅-TPHP was purchased from Wellington Laboratories (Guelph, ON). D₁₈-TCIPP and ¹³C-DCPH were purchased from Cambridge Isotope Laboratories (Tewksbury, MA). For the OPE internal standards, recoveries of d₁₂-TCEP, d₁₅-TDCIPP, and ¹³C-TPHP were on average 121 ± 1%, 92.5 ± 2%, 86.0 ± 2%, respectively. For the phthalate internal standards d₄-DMP, d₄-DEP, d₄-BBP, and d₄-DEHP, recoveries were on average 70.6 ± 7%, 102 ± 13%, 118 ± 8%, and 93.2 ± 5%, respectively. In addition, lab blanks (n=5) and field blanks (n=8) were analyzed with the wristband samples for quality assurance and quality control.

Hand wipe and dust collection and extraction.

Phillips and Hammel et al. 2018 provided a more detailed description of collection and analysis for hand wipes and dust.⁴⁰ In brief, hand wipe samples were collected from each child during the home visit using pre-cleaned twill wipes which were soaked with 3 mL isopropyl alcohol. The entirety of the children’s hand surface area was wiped, from wrists to fingertips and in between the fingers. The wipes were wrapped in aluminum foil and stored at −20°C until analysis. House dust samples were collected by using a Eureka Mighty Mite vacuum fitted with a cellulose thimble in the hose attachment to vacuum the entire exposed floor area of the main living area. The thimbles were wrapped in aluminum foil and stored at −20°C until analysis. Both hand wipes and dust were spiked with internal standards and solvent extracted by sonication then cleaned using Florisil^® SPE cartridges. Field blanks for hand wipes and lab blanks and house dust standard reference material (SRM 2585, NIST, Gaithersburg, MD) for dust were analyzed alongside the samples for quality assurance and quality control. Samples were analyzed for OPEs and phthalates using GC-MS-EI. DEP and several ITP and TBPP isomers were not quantified in the hand wipe extracts due to interferences in the chromatography and lack of available standards at the time of analysis, respectively.

Urine collection and extraction.

During home visits, families were provided with collection kits for urine samples. Three spot urine samples were collected from each child during a 48-hour sampling period, after which they were stored frozen in the families’ homes until transportation back to our research laboratory where they were stored at −20°C. Prior to analysis, the samples were thawed, and equal volumes of the spot urine samples were combined to form a pooled sample. Prior to analysis, specific gravity of the pooled sample was measured using a handheld digital refractometer (Atago), and sample concentrations were corrected based on this measurement. ⁴⁴ The composite sample was analyzed for OPE metabolites in our laboratory, with a detailed description of methods previously described in Phillips and Hammel et al. 2018 ⁴⁰ This composite urine sample was also analyzed for phthalate and non-phthalate plasticizer metabolites by the CDC laboratory, as described previously in Hammel and Levasseur et al. 2019.⁴¹

Statistical analyses.

All analyses were performed using SAS statistical software (version 9.4; SAS Institute Inc., Cary, NC) for analytes detected in >60% of samples, and results were assessed at a level of α=0.05 for significance. Method detection limits (MDLs) were determined using three times the standard deviation of the average field blank levels for wristbands and hand wipes, and lab blanks for dust and urine samples. MDLs were normalized to the average field blank sample mass of wristbands (1.52 g), average mass of dust extracted (0.06 mg), or volume of urine extracted (5 mL), accordingly. Values below the MDL were replaced with MDL/2.⁴⁵ Preliminary analyses examining the chemical concentrations in the external exposure measures (wristbands, hand wipes, and dust) and biomarker concentrations in urine indicated that the levels were non-normally distributed. As such, Spearman correlations were used to assess the relationships between matrices. Certain parent phthalate compounds are metabolized to multiple metabolites (e.g., DiBP, DBP, DEHP, DiNP, and DEHTP). As such, associations were evaluated using molar sums of the metabolites, which allowed for better characterization of the relationships between the parents and total metabolites. Similarly, TCIPP has two potential metabolites, BCIPP and BCIPHIPP, which we analyzed here. We evaluated the associations for its two metabolites separately and as a molar sum. Because of the lack of associations observed with the molar sum, we present the analyses for the individual metabolites here to allow for more relevant comparisons with previously published studies.⁴⁰ Stratified analyses were considered for some of the variables (e.g., time spent in the home); however, due to the smaller study population, the sample sizes were not sufficient to allow for appropriate interpretation. As such, these analyses are not presented here.

Given that the wristbands were introduced halfway through the study, there is not a complete overlap in the number of participants for each sample type, and relationships were evaluated for the maximum number of paired samples available. Relationships utilizing hand wipes and dust were restricted to participants who had also provided wristband samples, where n=75 for OPEs and n=74 for phthalates.

Results and Discussion

Study population.

The demographics of the study population subset in which wristbands were deployed and collected are described in Table 1. Details about the full TESIE study population were more extensively described in previous publications.^40-42 This subset of the TESIE study included 77 children from 74 homes, with about two-thirds of the children being male. The median age in this study population was 57 months (4.75 years) and ranged from 50 to 67 months. Notably different from the total TESIE study population, the majority of the children’s mothers identified as Hispanic (43%), while 31% and 25% identified as non-Hispanic black and white, respectively. This was largely due to the fact that during this portion of the study recruitment period, we were conducting our home visits and questionnaires in Spanish. Also, the mothers in this subset generally had less educational attainment than the original TESIE population, with 65% in our study having attained a high school degree or less compared to the 55% in the original. Wristband collection started a year into the TESIE study, and thus, all samples described here were collected between August 2015 and April 2016.

Table 1.

Demographics of the TESIE study population that wore wristbands.

Characteristic	N	%
Child Sex
Male	46	59.7%
Female	31	40.3%
Age
50-55 months	30	39.0%
56-59 months	24	31.2%
60-67 months	23	29.9%
Ethnicity
Non-Hispanic white	19	24.7%
Non-Hispanic black	24	31.2%
Hispanic white	33	42.9%
Other	1	1.3%
Maternal education
Less than high school graduate	31	40.3%
High school graduate or GED	19	24.7%
College graduate or more	27	35.1%
	Mean	range
Child age (months)	57.1	50-67
Avg outdoor temperature at visit (°C)	15.6	−1.1-27.2
Child Body Weight (kg)	19.7	15.3-35.5
Children’s Behaviors	N	%
Time spent at home
3-5 hours	2	2.6%
6-8 hours	11	14.3%
9-12 hours	20	26.0%
13-18 hours	27	35.1%
19-22 hours	5	6.5%
23-24 hours	12	15.6%

Levels of compounds on wristbands.

Organophosphate esters.

Wristbands were analyzed for 22 organophosphate esters, and 13 of them were detected in >80% of the wristband samples (Table 2). The chlorinated alkyl phosphates (TCEP, TCIPP, TDCIPP) were detected in >80% of the wristbands, with TDCIPP being the most abundant within this subclass (median = 179.7 ng/g) (Figure S1). EHDPHP was the only OPE that was detected in every wristband collected. Measured in all but one sample, TPHP was the most abundant OPE overall on the wristbands (median = 872.9 ng/g) (Table 2, Figure S2), which is likely due to its prevalence in consumer products as both a plasticizer and flame retardant. When compared to deployed wristbands among children of a similar age range in Oregon, we observed a slightly different trend, as Kile et al. 2016 where they detected TDCIPP at the highest levels of the 4 OPEs measured (TDCIPP, TPHP, TCIPP, and TCEP).³⁶ Median values were not reported in Kile et al. 2016, and thus we were unable to compare our wristband concentrations to this previous report in children's wristbands. Also, the Oregon children were recruited from preschools, indicating that they spent part of their day in school, whereas not all of the children in our study attended school during the daytime. This suggests that the difference in where children from these two study populations spent their time could lead to the differing trends observed in their OPE exposure. Compared to OPE levels measured in an adult cohort from a similar region of central North Carolina and during the same sampling period, we again observed a dissimilar trend among the children’s wristbands.³⁸ Among adults, the most abundant OPE measured was TCIPP, which was followed by TDCIPP then TPHP. This indicates that children may have different exposure sources and pathways from adults, particularly for TCIPP, which could be due to differences in consumer product use and the shift away from using TDCIPP in products (e.g., child car seats, home furnishings) to TCIPP.⁴⁶ Additionally, the demographics of the children from this cohort (race/ethnicity and socioeconomic status) may have been different from those of the adult cohort, which could explain why the trends were different within a similar geographic region.

Of the isopropylated triaryl phosphates (ITPs), 6 of 8 were detected in >90% of the wristband samples, the other 2 being detected in ~50%. 2IPPDPP was the most abundant ITP on the wristbands (median = 373.4 ng/g), which is the dominant ITP in Firemaster^® 550 (FM550) and the ITP commercial isomer mixture (Table 2, Figure S2).⁴⁷ However, the next most abundant ITP with a similar concentration to 2IPPDPP was B3IPPPP (median = 309.8 ng/g); this compound is not a large component of any of the flame retardant mixtures previously characterized which suggests that it likely has a separate source of exposure. Only 2 of the 5 tert-butylated triaryl phosphates (TBPPs) were quantified in >80% of the wristband samples, with 4tBPDPP being the most abundant (median = 119.5 ng/g; Table 2). While 4tBPDPP is not a major component of Firemaster^® 600 (FM600), it is the dominant compound in a commercial mixture we refer to as TBPP, suggesting that FR applications could be contributing the high levels of this compound on the wristbands.⁴⁷ To the authors’ knowledge, the isopropylated and tert-butylated triaryl phosphate esters have not been previously measured on wristbands worn by children.

Phthalates and non-phthalate plasticizers.

Of the compounds evaluated here, the phthalates and non-phthalate plasticizers were measured at much higher levels on the wristbands than the OPEs. Every one of the 10 compounds quantified were measured in >98% of the wristbands (Table 2). Previous work measuring phthalates on the wristbands have reported between 50-95% detection for 6 of the phthalates presented here.^29,35 Unfortunately, we cannot make any comparisons with the previous studies as only the detection frequencies were reported, and no quantitative measurements were performed. DEHP and one of its common chemical replacements, DEHTP, were both detected in every wristband sample, were the most abundant compounds, and were measured at similar levels on the wristbands (median = 9,010 and 8,493 ng/g, respectively) (Figure S3). This suggests that DEHTP may be a prominent chemical that children are exposed to as DEHP is replaced in floorings and children’s toys; nonetheless, children’s exposure to DEHP remains high compared to the other compounds within this group. Generally, the higher molecular weight compounds were more abundant on the wristbands compared to the lower molecular weight phthalate compounds, suggesting that the wristbands may have captured a substantial concentration of particle-bound compounds, with many of these compounds being extremely abundant in the indoor environment. Based on particle-gas partitioning coefficients (Kp) calculated by Weschler et al. 2008 using saturation vapor pressures of phthalates, about 16% of BBP and 86% of DEHP would be particle-bound at room temperature.^48,49 Since DiNP and DEHTP are both less volatile than DEHP, it is likely that both compounds are being sampled by the wristbands via particles/aerosols.

Urinary metabolites.

OPE and phthalate metabolites were detected in a majority of the children’s pooled urine samples, with every metabolite except MNP detected in >65% of samples (Table S1). When comparing the urine values within this subset of the TESIE study to the entire study population, detection frequencies and geometric mean levels are very similar. Of all of the urinary metabolites measured here, MECPTP, one of the metabolites of DEHTP, was the most abundant (median = 89.69 ng/mL). Detailed discussion of the urinary metabolites and their relationships with demographic variables is presented in Hoffman et al. 2018, with additional information on indoor dust and hand wipes reported in Phillips and Hammel et al. 2018 and Hammel and Levasseur et al. 2019.^40-42

Levels of compounds on hand wipes and dust.

OPE and phthalate/non-phthalate plasticizer levels on the hand wipes and in dust within this TESIE subset were very similar to the total TESIE dataset (Table S2, S3). Several of the ITP and TBPP compounds as well as DEP were not reported on hand wipes due to the unavailability of certain standards at the time of analysis (ITPs and TBPPs) or chromatographic interferences (DEP). More extensive discussions of hand wipe and dust concentrations of OPEs are presented in Phillips and Hammel et al. 2018 and of phthalates in Hammel and Levasseur et al. 2019.^40,41

Comparing wristbands to biomarkers.

OPEs.

In general, OPE concentrations on the wristbands were positively correlated with their corresponding urinary biomarkers (Table 3). In particular, TDCIPP, 3 ITP isomers, and both TBPP isomers were significantly associated with their corresponding urinary metabolites, BDCIPP, ip-PPP, and tb-PPP, respectively (r_s=0.25-0.52, p<0.05; Figure 1). These results, particularly for TDCIPP and mono-isopropylated triaryl phosphates (2IPPDPP, 3IPPDPP, and 4IPPDPP), were similar to those evaluated in the adult pilot cohort, although in that study, the mono-isopropylated triaryl phosphates were measured on the wristbands in a semi-quantitative manner and the correlation was suggestive but not statistically significant given the sample size (r_s=0.29, p=0.07).³⁸ These results suggest that the pathways of exposure captured by the wristbands could be similar for both children and adults. The primary difference between the results for children and adults is the association between TCIPP on the wristbands and its urinary metabolites. Here, TCIPP on the wristbands was not significantly correlated with either of its primary metabolites, BCIPP and BCIPHIPP, measured in children’s urine. In the adult pilot study, BCIPP was only detected in 18% of samples and was not evaluated in statistical analyses; BCIPHIPP was significantly and positively correlated to TCIPP.³⁸ This suggests that the metabolism of TCIPP could be different in younger children, which was previously proposed by Van den Eede et al. 2015 and demonstrated by the difference in detections between adults and children from the same region. Additionally, it is possible that adults and children are not exposed to TCIPP in the same manner and could have dissimilar pathways of exposure for this compound, based on how they interact with their environments and products within their environment.

Table 3.

OPE correlation coefficients for wristband levels (ng/g) and specific gravity-corrected biomarkers (n=75), DF>60%; shading indicates potential parent-metabolite pairs.

		Urine
		BCIPP	BCIPHIPP	BDCIPP	DPHP	ip-PPP	tb-PPP
Wristband	TCEP	0.13	−0.29#	0.09	−0.06	−0.06	0.04
	TCIPP	0.18	−0.06	0.03	−0.18	0.06	−0.09
	TDCIPP	0.07	0.10	0.52†	0.04	0.14	0.16
	EHDPHP	0.13	−0.06	0.11	−0.08	0.05	0.07
	TPHP	0.05	−0.02	0.09	0.12	0.21	0.26*
	2IPPDPP	0.05	0.00	0.03	0.05	0.22	0.17
	3IPPDPP	0.05	0.04	0.11	0.07	0.27*	0.20
	4IPPDPP	0.04	0.06	0.14	0.10	0.31#	0.22
	24DIPPDPP	0.02	0.02	−0.01	0.04	0.20	0.15
	B2IPPPP	0.06	0.02	0.03	0.06	0.22	0.17
	B3IPPPP	0.05	0.02	0.03	0.07	0.25*	0.17
	4tBPDPP	−0.09	−0.06	0.03	0.03	0.18	0.35#
	B4TBPP	−0.04	0.00	0.12	0.05	0.18	0.35#
Hand Wipe	TCEP	0.13	0.07	0.14	0.10	−0.02	0.12
	TCIPP	0.24*	0.13	0.22	0.07	0.03	0.03
	TDCIPP	0.19	0.19	0.48†	0.22*	0.21	0.05
	EHDPHP	0.19	0.21	0.09	0.13	−0.03	−0.06
	TPHP	0.15	0.05	−0.01	0.18	0.11	0.14
	2IPPDPP	0.15	0.15	0.09	0.18	0.20	0.07
	4IPPDPP	0.06	0.19	0.12	0.23*	0.19	0.21
	4tBPDPP	−0.06	0.07	−0.06	−0.02	0.02	0.16
	B4tBPPP	−0.10	−0.03	−0.03	0.07	−0.03	0.28*
Dust	TCEP	−0.02	−0.15	0.16	0.14	−0.11	−0.05
	TCIPP	−0.05	−0.32#	−0.10	0.08	−0.03	−0.06
	TDCIPP	−0.06	−0.32#	0.13	−0.04	−0.05	−0.07
	EHDPHP	−0.07	−0.07	0.04	0.05	−0.10	−0.07
	TPHP	0.05	−0.10	0.13	0.00	0.03	−0.14
	2IPPDPP	−0.06	0.00	0.07	0.14	0.13	−0.08
	24DIPPDPP	−0.01	0.22	0.05	0.05	0.12	−0.06
	4tBPDPP	−0.08	−0.09	0.07	−0.04	0.19	0.05
	B4tBPPP	−0.05	−0.04	0.04	0.03	0.12	0.11

^*p<0.05

^#p<0.01

^†p<0.001

Figure 1.

Spearman correlation plot of (A) TDCIPP and (B) 4tBPDPP on wristbands with their respective urinary metabolites.

Some significant correlations were observed for non-paired parent OPEs and their metabolites. For instance, TCEP on the wristbands was negatively correlated with BCIPHIPP (r_s= −0.29, p<0.01). In this study, we did not measure a urinary metabolite for TCEP. It is unclear why this negative correlation was observed; however, this could be a result of TCEP being phased out of products due to carcinogenicity concerns, and TCIPP being used as a replacement flame retardant in similar consumer products. Wristband TPHP was also positively and significantly associated with urinary tb-PPP (r_s=0.26, p<0.05). Given that the TBPP isomers metabolize to tb-PPP, this suggests that the relationship observed here is a result of co-application of TPHP and the TBPP isomers, as is the case in FM600 commercial mixture.

The ITP isomers measured on the wristbands were all positively correlated with their corresponding urinary metabolite, ip-PPP (r_s=0.2-0.3, p<0.1; Table 3), with 3 of the 6 isomers significantly associated at α=0.05. Little is known about personal exposure to ITPs, particularly in children, and few studies have characterized exposure to ITPs in the general population, since their metabolites, namely ip-PPP, are not currently measured in U.S. National Health and Nutrition Examination Study (NHANES). Also, the sources of these compounds, other than FR applications in furniture, are largely unknown. Thus, the positive associations between ITPs on wristbands and urinary biomarkers observed here suggest that the wristbands could provide some initial insights for how humans are exposed to them in their daily environments. TBPP isomers were measured for the first time on the wristbands and were positively and significantly associated with their corresponding metabolite, tb-PPP (r_s=0.35, p=0.002; Figure 1). Like the ITP isomers, little data is available regarding personal exposure to TBPP isomers, and the positive and moderate correlation coefficient suggests that the ambient environment is a source of exposure. None of the known parent compounds for DPHP (EHDPHP, TPHP, ITPs, TBPPs) were significantly correlated with this urinary metabolite, suggesting that there could be other sources of exposure to DPHP which were not measured in this study (e.g., other OPEs or diet). DPHP is also used as a plasticizer in consumer products and recently has been reported in indoor dust.⁵⁰ Because we did not analyze for DPHP in the wristbands, we cannot rule out the possibility that DPHP in the environment could have contributed to the measured concentrations in urine. This is dissimilar to what was observed with the adult cohort where TPHP on wristbands was positively and significantly associated with DPHP in pooled urine samples (r_s=0.43, p<0.01).³⁸ TPHP was the most abundant OPE measured on the children’s wristbands, and the lack of a correlation with urinary DPHP suggests that children may be exposed to TPHP in a different manner than adults, or that children may have different metabolic processes compared to adults.

Phthalates.

Similar to OPEs, several phthalate and non-phthalate plasticizer levels on wristbands were positively correlated with their corresponding urinary metabolites (Table 4). Of the 7 plasticizer compounds that had data available for urinary biomarkers, 4 were significantly associated (r_s=0.24-0.56, p<0.05), with BBP on the wristbands being the most highly correlated (r_s=0.56, p<0.0001; Figure 2). While a few studies have evaluated phthalates on wristbands, even fewer have compared wristbands levels to biomarkers of exposure, with a recent study evaluating occupational exposure to phthalate compounds.⁵¹ These results suggest that the wristbands may capture meaningful personal exposure to some phthalates. This is particularly of interest for DEP, which is a fairly volatile compound typically used as a solvent or vehicle for fragrances and cosmetics.^27,52 Here, we observed a significant and positive association between wristband DEP and its urinary metabolite, MEP (r_s=0.53, p<0.0001; Figure 2). Because the most common use for DEP is in personal care products, children are likely exposed to DEP via such product use, potentially via inhalation or dermal absorption. For the higher molecular weight phthalates and DEHTP, the relationships between the wristbands and urinary metabolites were positive, with DiNP being significantly associated with the molar sum of its metabolites (r_s=0.24, p<0.05). DEHP on the wristbands displayed a similar positive association with the molar sum of its metabolites (r_s=0.22, p=0.06), although it did not reach statistical significance. Previous studies have estimated that the primary exposure pathway for DEHP is attributed to diet.⁵³ Our observations suggest that the wristbands capture some portion of a child’s overall DEHP exposure which could be due to products in the home environment, but may also arise from contact with food packaging materials containing DEHP. In this study, DEHTP wristband concentrations were not associated with its urinary biomarker, suggesting that the exposure pathway is not effectively captured by the wristbands (r_s=0.15, p=0.19; Table 4). No directional relationship was observed for wristband DBP levels compared to the molar sum of its specific metabolites (MHBP and MBP) or its nonspecific metabolite (MCPP).

Table 4.

Phthalates correlation coefficients for wristband levels (ng/g) and specific-gravity corrected urinary metabolites (n=74), DF>60%, shading indicates potential parent-metabolite pairings. MCNP is a non-specific metabolite of several parent phthalate compounds.

		Urine
		MEP	ΣDiBPa	ΣDBPb	MCPP	MBzP	ΣDEHPc	ΣDiNPd	MCNP	MHINCH	MCOCH	ΣDEHTPe
Wristband	DMP	0.23*	0.21	0.17	−0.09	0.15	−0.03	−0.16	−0.06	0.06	0.03	−0.01
	DEP	0.53†	0.09	0.02	0.08	−0.01	0.09	0.06	0.09	0.06	0.08	0.04
	DiBP	−0.06	0.32#	−0.07	−0.04	−0.03	0.05	0.07	−0.12	0.03	−0.02	0.05
	DBP	0.05	−0.05	0.04	−0.02	0.03	0.09	0.06	−0.03	−0.10	−0.16	0.05
	BBP	0.30#	0.13	0.22	0.09	0.56†	0.15	0.08	0.09	0.14	0.05	0.20
	DEHP	0.05	0.05	−0.12	0.00	−0.01	0.22	−0.01	−0.11	−0.02	−0.02	0.18
	DiNP	0.05	0.01	−0.16	0.10	−0.11	0.09	0.24*	0.08	0.10	0.06	0.09
	DEHTP	0.06	−0.06	−0.21	−0.21	−0.14	−0.10	−0.11	−0.17	0.05	0.05	0.15
	DEHA	0.03	0.06	−0.11	0.07	−0.02	0.08	0.12	0.00	0.16	0.14	0.07
	TOTM	−0.01	0.05	−0.17	−0.03	−0.12	0.09	−0.02	−0.08	0.18	0.17	0.14
Hand Wipe	DMP	0.07	0.04	0.00	0.14	0.02	0.02	−0.01	0.05	0.03	0.07	−0.08
	DiBP	−0.02	0.27*	0.02	−0.06	0.11	−0.06	−0.16	−0.23*	−0.01	−0.09	−0.03
	BBP	0.23*	0.36#	0.33#	0.06	0.56†	0.06	0.01	0.04	0.00	−0.11	0.11
	DEHP	−0.06	0.07	0.04	0.03	0.13	0.14	0.03	−0.10	−0.08	−0.13	−0.01
	DiNP	0.03	0.19	0.01	0.19	0.09	0.11	0.22	0.03	0.01	−0.01	−0.02
	DEHTP	0.14	0.16	0.08	−0.10	0.20	−0.03	−0.04	−0.02	0.14	0.15	0.06
	TOTM	0.00	0.23*	0.05	−0.05	0.12	−0.01	0.03	−0.03	0.20	0.14	0.14
Dust	DEP	0.17	−0.03	0.07	0.14	0.10	0.05	0.17	0.08	0.12	0.03	0.03
	DiBP	0.09	0.27*	0.29*	0.27*	0.17	0.18	0.22	0.08	0.01	−0.08	0.15
	DBP	−0.06	−0.02	0.28*	0.27*	0.18	0.15	0.21	0.12	−0.10	−0.12	0.05
	BBP	0.14	0.08	0.16	0.15	0.23*	−0.05	0.10	0.08	0.00	0.00	0.34#
	DEHP	0.00	0.06	0.14	0.17	0.07	0.16	0.03	0.08	0.08	0.11	0.11
	DiNP	−0.07	0.01	0.11	0.30#	0.04	0.19	0.20	0.10	0.10	0.09	0.11
	DEHTP	0.00	0.04	0.01	0.17	0.05	0.05	0.14	0.15	0.16	0.21	0.35#
	TOTM	0.05	0.12	0.04	0.08	0.13	0.01	−0.13	0.10	0.09	0.07	0.06

^*p<0.05

^#p<0.01

^†p<0.001

^aΣDiBP= Molar sum of MHiBP and MiBP

^bΣDBP = Molar sum of MHBP and MBP, MCPP is a non-specific metabolite

^cΣDEHP = Molar sum of MEHP, MEOHP, MEHHP, and MECPP

^dΣDiNP = Molar sum of MCOP and MONP

^eΣDEHTP = Molar sum of MECPTP and MEHHTP

Figure 2.

Spearman correlation plots of (A) DEP and (B) BBP on wristbands with their respective urinary metabolites.

Comparing wristbands to hand wipes and dust.

For OPEs measured in each matrix (wristbands, hand wipes, and dust), wristbands were generally significantly correlated with hand wipes (r_s=0.23-0.52, p<0.05), and were not correlated with house dust (Table S4). This result is similar to Wang et al. 2019, which compared wristbands to hand wipes and active and passive air samplers and observed a significant and positive association between wristbands and hand wipes for OPEs.³⁹ While larger correlation coefficients were observed for hand wipes and dust compared to hand wipes and wristbands, these correlations were not statistically significant overall. Similarly, for phthalates, wristband levels were significantly correlated with hand wipes (r_s=0.24-0.42, p<0.05), with the exception of DiBP (Table S5). Like OPEs, phthalate concentrations in dust was generally not significantly associated with wristbands or hand wipes. The lack of an association observed between house dust and urinary metabolites could, in part, be due to the child not spending as much time in the main living area of the house than other places (e.g., school, daycare, bedroom). This further suggests that house dust may not always be the best marker of exposure for SVOCs among children. Currently, wristbands have been hypothesized to effectively measure or capture inhalation and dermal absorption exposure pathways, and recent work comparing wristbands to hand wipes and air samplers supports this hypothesis across three classes of compounds (PAHs, brominated flame retardants, and OPEs).³⁹ Hand wipes are posited to capture exposures that occur via hand-to-mouth contact (i.e., inadvertent dust ingestion) and dermal absorption. As such, wristbands and hand wipes could both be capturing portions of the dermal exposure pathway, which could explain the positive and significant associations observed in this both study and Wang et al. 2019 between wristbands and hand wipes.³⁹ More research is needed for other chemical classes to confirm this statement.

Evaluating external exposure matrices in the context of urinary biomarkers.

OPEs.

In general, OPEs on wristbands and hand wipes shared similar positive correlations with their corresponding urinary metabolites, although in nearly every case, the correlation coefficient was slightly greater for wristbands compared to hand wipes (Tables 3, S6). However, the slight difference in magnitude does not indicate a statistical difference between the two matrices since the confidence intervals around the correlation coefficients are likely to overlap. The sole exceptions were TCIPP on hand wipes being significantly correlated with the BCIPP metabolite and 4IPPDPP being significantly correlated with urinary DPHP (one of its potential urinary metabolites), with no significant associations observed with the wristbands. Both of these measures were clearly more effective than house dust at evaluating children’s exposure to OPEs, with no significant and positive associations observed between dust OPEs and urinary metabolites (Table 3, S6). Table S6 includes the suggested “best media” for evaluating each of the 12 OPEs with corresponding urinary metabolites based on the Spearman correlation coefficients. For the 2 chlorinated OPEs, hand wipes appeared to be a reliable measure of exposure, with TDCIPP being roughly equivalent for both wristbands and hand wipes and not correlated with dust. In fact, dust TCIPP was significantly and negatively associated with urinary BCIPHIPP, although it is unclear why this is the case. In the case of EHDPHP and TPHP, none of the 3 exposure assessment tools utilized appeared to be effective for measuring children’s exposures to these compounds. This could be due to the fact that the metabolite used to evaluate exposure to these compounds, DPHP, has multiple parents and is also used in consumer products as a plasticizer.⁵⁴ For EHDPHP, DPHP is a possible metabolite but not the primary or most abundant one, based on previous metabolism studies.⁵⁵ For ITP and TBPP isomers, with the exception of 2IPPDPP, wristbands were observed to be a slightly improved measure compared to hand wipes. It should be noted that these isomers were quantified using different instrumentation for the wristbands compared to the hand wipes and dust, because using a higher mass resolution instrument for the wristbands allowed for a decrease in interferences and thus increased selectivity and sensitivity. Still, we interpret these results with caution, as fewer isomers were quantified on the hand wipes and dust due to the poorer sensitivity from the lower mass resolution instrument. Among the 12 OPE compounds evaluated here for children’s exposure, hand wipes and wristbands were roughly equal measures of exposure, although wristbands may be a slightly improved measure, in general. When compared to adults, the results here suggest that hand wipes may have increased utility for evaluating children’s exposure, since children have more hand-to-mouth contacts than adults and this pathway would be directly captured by hand wipes.

Phthalates.

For the plasticizer compounds, wristbands were positively correlated with their urinary metabolites for 5 of the 7 compounds in which paired data (i.e., known parents and metabolites) were available (Table 4, S6). As such, these data suggest that wristbands are the best tool available for estimating exposure to a number of these compounds, particularly the lower molecular weight phthalates. In the case of a few phthalates and DEHTP, house dust concentrations were positively correlated with the corresponding urinary metabolites (Table S6). This is particularly relevant for DBP and DEHTP, compounds for which neither hand wipes nor wristbands were significantly correlated with their urinary metabolites, suggesting that house dust may be the improved measure of children’s non-dietary exposure to these compounds given significant associations between house dust and corresponding metabolites. DEHTP was among the most abundant phthalates in house dust within this study population, and both DEHTP and DBP are used in building materials and floorings which could explain the positive correlations with house dust. The three assessment tools appeared to be roughly equivalent for DiBP, and hand wipes and wristbands were equivalent in the ability to measure exposure to BBP.

Strengths and limitations.

Our study had several important strengths, including the collection of multiple paired samples (wristband, hand wipe, house dust, urine) from each child and the use of a pooled urine sample for the child versus a single spot urine sample. The wristbands were evaluated for 28 consumer product chemicals spanning a wide range of physicochemical properties, providing additional insights on how the wristbands can be utilized in exposure assessment. Our study should also be interpreted in the context of several potential limitations. First, we only collected one wristband from each participant during the week following the home visit, and the other samples (hand wipes and dust) were collected at a single time point, with dust collected solely from the main living area; as such, we could not account for individual variability with the wristbands and hand wipes and dust from different microenvironments nor determine how exposure may fluctuate from week to week. Second, we were unable to collect detailed data regarding compliance of the wristband being worn during the sampling period and how the children’s behaviors may have influenced the measurements (e.g., swimming, bathing). Third, statistical significance was set at α = 0.05, and adjustments were not made for multiple comparisons; however, not adjusting for multiple comparisons has been recommended and generally accepted in the epidemiologic literature. ⁵⁶ While our study population was diverse and fairly representative of central North Carolina, participants were selected as a convenience sample from an established pregnancy cohort from Duke University Medical Center. This may limit the generalizability of our results to the broader population, but the sampling methodology should not impact the internal validity of this study.

Implications.

This is the first study to compare wristband concentrations of OPEs and phthalates to paired external exposure measures in their associations with biomarkers as well as evaluate wristbands for assessing exposure to some OPEs and phthalates among children. Taken together, the results of this study suggest that wristbands are an effective and quantitative exposure assessment tool for evaluating children’s exposure to OPEs and both phthalate and non-phthalate plasticizers. When evaluated in the context of contemporarily accepted exposure measures (i.e., hand wipes and house dust), wristbands were observed to be roughly equivalent to hand wipes for OPEs and an improved measure for several of the phthalate compounds. They may be advantageous over hand wipes and dust due to their ease of sample collection, improved analytical sensitivity (higher concentrations measured), and potential for supporting large-scale biomonitoring studies. While both hand wipes and wristbands capture exposure from multiple microenvironments and pathways of exposure, the wristbands are more likely to measure the integrated, average exposure over time (a week in this case), whereas hand wipes serve as more of a cross-sectional sample representing a snap-shot of exposure within the last few hours. This study supports the use of wristbands as a quantitative measure of exposure, as significant and positive associations were observed for over half of the compounds (with biomarkers) assessed in this study.