- Email: [email protected]
Impact of WHO 2005 revised toxic equivalency factors for dioxins on the TEQs in serum, household dust and soil
Chemosphere 76 (2009) 727–733
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Impact of WHO 2005 revised toxic equivalency factors for dioxins on the TEQs in serum, household dust and soil Biling Hong a,*, David Garabrant a, Elizabeth Hedgeman a, Avery Demond b, Brenda Gillespie c, Qixuan Chen c, Chiung-Wen Chang a, Timothy Towey b, Kristine Knutson a, Alfred Franzblau a, James Lepkowski d, Peter Adriaens b a
Department of Environmental Health Sciences and the Risk Science Center, University of Michigan School of Public Health, Ann Arbor, MI, USA Department of Civil and Environmental Engineering, University of Michigan College of Engineering, Ann Arbor, MI, USA Department of Biostatistics, University of Michigan School of Public Health, Ann Arbor, MI, USA d Institute for Social Research, University of Michigan, Ann Arbor, MI, USA b c
a r t i c l e
i n f o
Article history: Received 7 January 2009 Received in revised form 13 April 2009 Accepted 25 May 2009 Available online 1 July 2009 Keywords: Dioxins Toxic equivalency factors Furans PCBs UMDES Serum
a b s t r a c t Background: In 2005, the World Health Organization (WHO) – International Programme on Chemical Safety reevaluated the toxic equivalency factors (TEFs) for dioxin-like compounds and made changes that affect the calculation of the total toxic equivalent (TEQ). The impact of these changes on the TEQs for human blood and abiotic matrices such as soil and household dust has not been widely assessed or reported. Methods and results: Using a major exposure study which examined blood, household dust, and soil levels of dioxin-like compounds in several regions of Michigan, we found the mean total TEQ was significantly reduced by 26%, 12% and 14% for serum, household dust, and soil, respectively, when the TEQ was based on the 2005 TEFs compared to the 1998 TEFs. The decrease in the serum total TEQ was largely due to the down-weighting of the TEFs for the majority of mono–ortho PCBs. In contrast, the decrease in the soil total TEQ was mostly due to the down-weighting of the TEF for 2,3,4,7,8-pentachlorodibenzofuran (PeCDF) (1998 TEF = 0.5, 2005 TEF = 0.3). For household dust, the decrease in total TEQ was not due to any single TEF but was due to small changes in a number of compounds. There was a dramatic decrease (88%) in the mean and 95th percentile for mono–ortho PCB TEQ due to the 2005 TEFs. Discussion: These findings suggest that comparisons between studies based on the TEQ–WHO98 and TEQ–WHO05 may need to consider an appropriate conversion factor to assure comparability. Furthermore, the changes in TEFs may have impact in locations where regulations of soil contamination are triggered by specific TEQ levels. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction The toxic equivalent (TEQ) concept was developed to measure the combined toxicity of complex mixtures of dioxin and dioxinlike compounds through the toxic equivalency factors (TEFs) methodology, which assesses the toxicity of each congener relative to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Mathematically, the total TEQ is defined as the weighted sum, over all WHO 29 dioxin congeners, of the TEF multiplied by the congener concentration. In 2005, the World Health Organization (WHO) – International Programme on Chemical Safety reevaluated the TEFs for dioxin-
* Corresponding author. Address. Department of Environmental Health Sciences and the Risk Science Center, University of Michigan School of Public Health, 1420 Washington Heights, Room 6529, Ann Arbor, MI 48109-2029, USA. Tel.: +1 734 936 0726; fax: +1 734 763 7170. E-mail address: [email protected] (B. Hong). 0045-6535/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2009.05.034
like compounds and made changes that affect the calculation of the total TEQ (Van den Berg et al., 2006). Objectives of these revisions included harmonization of the TEFs internationally and insuring that studies published since the previous revision in 1997 (Van den Berg et al., 1998) were evaluated. Although no compounds were added or removed from the list of 29 dioxin-like compounds for which TEFs were agreed in 1997, two important changes were made in the WHO 2005 TEFs: (1) The WHO 1998 TEFs were assigned in increments of 0.01, 0.05, 0.1, etc., whereas the new 2005 TEFs used the half order of magnitude increments on a logarithmic scale of 0.03, 0.1, 0.3, etc. (2) TEFs for 14 out of 29 congeners were changed in 2005 (Table 1). Octachlorodibenzo-p-dioxin (OCDD) and octachlorodibenzofuran (OCDF) were both slightly up-weighted from 0.0001 to 0.0003; 1,2,3,7,8-PeCDF was down-weighted
728
B. Hong et al. / Chemosphere 76 (2009) 727–733
Table 1 Updated TEFs for 14 dioxin-like congeners during the 2005 reevaluation. Congener
WHO 1998 TEF
WHO 2005 TEF
Dibenzo-p-dioxins OCDD
0.0001
0.0003
Dibenzofurans 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF OCDF
0.05 0.5 0.0001
0.03 0.3 0.0003
Non-ortho PCBs PCB 81 PCB 169
0.0001 0.01
0.0003 0.03
Mono-ortho PCBs PCB 105 PCB 114 PCB 118 PCB 123 PCB 156 PCB 157 PCB 167 PCB 189
0.0001 0.0005 0.0001 0.0001 0.0005 0.0005 0.00001 0.0001
0.00003 0.00003 0.00003 0.00003 0.00003 0.00003 0.00003 0.00003
The objective of this paper is to present the impact of the 2005 TEF changes on TEQ levels, as well as on the percent contributions of PCDDs, PCDFs, non-ortho PCBs and mono–ortho PCBs to the total TEQ, in serum, soil, and household dust samples from UMDES. 2. Materials and methods
Note: The italicised congeners were up-weighted in 2005; the rest of the congeners were down-weighted in 2005.
from 0.05 to 0.03; 2,3,4,7,8-PeCDF was down-weighted from 0.5 to 0.3; two non-ortho PCBs were slightly up-weighted (from 0.0001 to 0.0003 for PCB 81, and from 0.01 to 0.03 for PCB 169); a single value (0.00003) was assigned for all mono–ortho PCBs, which resulted in downweighting for the majority of mono–ortho PCBs (except for PCB 167). The WHO expert panel recommended caution in applying the TEQ approach to abiotic matrices because the TEFs and TEQ approach is primarily intended for estimating exposures and risks from dietary ingestion. Nonetheless, this approach is widely applied to environmental matrices such as soils and sediments because it gives a single value to represent contamination from complex and highly variable mixtures of dioxin-like compounds. Although these revisions were estimated to result in a 10–25% reduction in the human serum total TEQ, we are not aware of studies that have evaluated their actual impact in human populations or in abiotic matrices such as soil and household dust. This paper examines the effects of the changes in TEFs on the TEQ using data from a large population-based study of dioxin exposure in Michigan. The University of Michigan Dioxin Exposure Study(UMDES) (www.umdioxin.org) was undertaken in response to concerns that the discharge of dioxin-like compounds from the Dow Chemical Company facilities in Midland, Michigan (USA) has resulted in contamination of soils in the Tittabawassee River flood plain and areas of the City of Midland, leading to an increase in residents’ body burdens of polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs) and dioxin-like polychlorinated biphenyls (PCBs). Serum samples were obtained from 946 participants in Michigan by using a complex sample design in 2004–2005 (Garabrant et al., 2008) and analyzed for World Health Organization (WHO) 29 dioxin-like compounds (Van den Berg et al., 1998), including 2,3,7,8-TCDD. Soil samples from eligible participants’ homes and dust samples from inside homes were also collected and analyzed for the same set of congeners. The study is uniquely valuable because of its large size and its evaluation of dioxin-like compounds in serum, soil, and household dust from a population-based sample. As a result, the inferences from these samples can be applied to the general population that was sampled.
The study included Michigan populations who live in Midland County, Saginaw County, and part of Bay County, both in and out of the Tittabawassee River flood plain, and who live in Jackson and Calhoun Counties. Adults age 18 or over who had lived in their current residence for at least five years were eligible to participate. A total of 946 participants who met the Red Cross criteria for blood donation gave an 80 mL whole blood sample for analysis. In order to be eligible for household dust and soil sampling, the participant had to be an owner of the property. A total of 764 participants provided household dust samples, and 766 participants provided soil samples from around the house perimeter. The complete UMDES protocol, including the blood collection, household dust and soil sampling protocols, is available at www.umdioxin.org. Serum, household dust, and soil samples were analyzed by Vista Analytical Laboratory (El Dorado Hills, CA) for the 29 dioxin-like congeners, using modified US EPA protocols 8290 (US Environmental Protection Agency 1994) and 1668 (US Environmental Protection Agency 1999). For blood samples, congeners were extracted from serum, cleaned following a multi-column protocol and quantified using high resolution mass spectrometry (HRMS). Final analyte concentrations were provided on both a whole weight and lipid-adjusted weight basis, expressed in parts per trillion (ppt) (equivalent to picograms gram1). Household dust samples were collected using high volume small surface samplers (HVS3s) manufactured by CS-3, Inc. (Sandpoint, ID, USA), and equipped with a cyclone and fine particle filter capable of capturing 99.95% of particles above 0.3 lm aerodynamic mean diameter. One 10 g composite sample was taken in each household from frequently used areas, from both hard and soft surfaces with carpets and area rugs being preferred, according to the American Society for Testing and Materials (ASTM) method ‘‘Standard Practice for Collection of Floor Dust for Chemical Analysis” (ASTM 2000). The soil samples were archived in 4 oz dioxin-grade amber glass containers to avoid photolytic degradation reactions, and stored in dedicated 4 °C cold rooms prior to analysis. For all residences, the top 1 in. composited sample from the perimeter was submitted for analysis using HRGC/HRMS. The results for dioxin concentrations that fell below the limit of p detection (LOD) were estimated by LOD/ 2 (Hornung and Reed, 1990). Statistical analyses, which accounted for the complex survey design, were performed using SAS version 9.1.3 (SAS Institute, Inc., Cary, NC, 2002–2003) and R package ‘‘survey” (R Release 2.6.1, 2007).
3. Results Among those 946 participants who donated blood, 93% were white and 57% were female, with a mean age of 51 years and a mean BMI of 29 kg m2. Table 2 presents the comparisons between TEQ–WHO98 and TEQ–WHO05 for serum, household dust, and soil samples, respectively, with descriptive statistics giving the mean (and 95% confidence interval), 95th percentile (and its 95% confidence interval), and the percent changes in the mean and 95th percentile TEQs. The 95th percentile serum TEQs are also compared with the values reported by Patterson et al. (2008) based on the NHANES 2001–2002 serum dioxin concentrations. The mean total TEQ–WHO05 levels in serum, household dust, and soil in the UMDES data were reduced by 26%, 12% and 14%, respectively, com-
729
B. Hong et al. / Chemosphere 76 (2009) 727–733 Table 2 Comparison between TEQ–WHO98 and TEQ–WHO05 for samples in the UMDES study and in NHANES 2001–2002 (unit: ppt). N
UMDES serum data Total PCDDs PCDFs Non-ortho PCBs Mono-ortho PCBs
946 946 946 946 946
TEQ based on WHO 1998 TEFs
UMDES household dust data Total 764 PCDDs 764 PCDFs 764 Non-ortho PCBs 764 Mono-ortho PCBs 764
b
95% CI for P95
Mean
95% CI for mean
75.3 35.6 10.6 11.9 24.6
68.3–104.0 32.4–42.7 9.6–12.3 8.9–16.9 21.1–29.9
23.9 15.3 3.6 4.0 1.1
22.1–25.7 14.2–16.4 3.4–3.8 3.4–4.5 1.0–1.2
68.9 34.8 12.3 11.0 18.0
62.9–80.8 28.7–43.3 11.0–14.4 9.7–12.0 15.1–20.4
Mean
95% CI for mean
P95
32.4 15.2 4.9 3.4 8.9
30.0–34.8 14.1–16.3 4.7–5.2 2.9–3.9 8.0–9.7
NHANES 2001–2002 serum datac Total 1194 PCDDs 1194 PCDFs 1194 Non-ortho PCBs 1194 Mono-ortho PCBs 1194
TEQ based on WHO 2005 TEFs a
% Change 95% CI for P95
In mean
In P95a
58.6 35.8 7.5 13.1 3.0
50.8–73.3 32.5–43.1 6.9–8.3 10.2–18.1 2.6–4.0
26.0 0.7 27.0 18.0 88.0
22.0 0.6 29.0 10.0 88.0
56.1 34.8 8.9 11.9 2.6
47.6–65.4 28.7–43.4 7.7–10.2 10.8–12.9 2.3–3.0
P95
a
b
18.6 0.0 27.6 8.2 85.6
41.9 22.3 4.9 7.0 7.7
32.0–51.9 16.4–28.2 4.1–5.8 3.5–10.5 4.0–11.5
142.5 70.2 13.5 10.4 19.1
95.4–275.7 46.2–99.9 12.6–18.7 7.3–167.6 10.1–81.1
36.7 24.0 4.2 7.1 1.5
28.4–45.1 17.6–30.3 3.5–5.0 3.6–10.6 0.85–2.1
126.4 74.3 12.7 10.5 3.8
90.6–220.8 49.1–110.7 10.3–17.5 7.4–168.6 2.3–30.6
12.4 7.6 14.3 1.4 80.5
11.3 5.8 5.9 1.0 80.1
UMDES house perimeter soil data Total 766 16.6 PCDDs 766 6.2 PCDFs 766 8.2 Non-ortho PCBs 766 1.4 Mono-ortho PCBs 766 0.8
12.7–20.5 5.0–7.4 5.2–11.2 0.8–2.0 0.2–1.5
73.3 25.4 23.9 3.7 1.2
51.1–85.5 15.5–48.7 13.5–63.9 2.3–4.9 0.7–2.6
14.3 6.4 6.3 1.4 0.2
11.2–17.4 5.1–7.6 4.1–8.6 0.8–2.0 0.03–0.3
59.4 25.8 18.5 3.7 0.3
50.1–74.7 16.2–49.3 11.3–48.9 2.3–5.0 0.1–0.5
13.9 3.2 23.2 0.0 75.0
19.0 1.6 22.6 0.0 75.0
Note: All results accounted for complex survey design. a P95 is the 95th percentile of the distribution. b The 95% CI (Confidence Interval) for P95 is calculated by R using method of Woodruff (1952). c From Patterson et al. (2008) Table 1.
pared to the mean total TEQ–WHO98 levels. The 95th percentile for serum total TEQ was reduced by 22% in UMDES data, compared to 19% in NHANES 2001–2002 data (Patterson et al., 2008). We also examined the TEQ for the PCDD congeners alone, for the PCDF congeners, the non-ortho PCBs, and the mono–ortho PCBs. The PCDD TEQ showed only modest changes when comparing the 1998 and 2005 TEFs at both the mean and at the 95th percentile (the mean PCDD TEQ increased 0.7% in serum, 7.6% in household dust, and 3.2% in soil). In contrast, the mean PCDF TEQ decreased substantially at both the mean and at the 95th percentile in serum and soil, with less pronounced changes in the household dust (the mean PCDF TEQ decreased 27% in serum, 14% in household dust, and 23% in soil). For the non-ortho PCBs, the mean serum TEQ increased by 18%, but showed almost no change in household dust or soil. The most dramatic effect of the 2005 TEFs was seen for the mono–ortho PCB TEQ, for which the mean serum level decreased by 88%, household dust decreased by 81%, and soil decreased by 75%. Similar changes were seen at the 95th percentile. Comparison of the UMDES and NHANES serum results showed virtually identical changes for the 2005 TEFs at the 95th percentile TEQ. We also evaluated the serum total TEQs by age groups (Table 3). The serum TEQ levels increased with age using either the 1998 TEFs or the 2005 TEFs. The mean serum TEQ for the oldest age group (1998 TEFs: 54.4 ppt for age 60+; 2005 TEFs: 39.6 ppt for age 60+) is about five times larger than the mean serum TEQ in the youngest age group (1998 TEFs: 9.5 ppt for age 18–29; 2005 TEFs: 8 ppt for age 18–29). The trends by age are similar for the 95th percentile for serum total TEQ. Patterson’s results (Patterson et al., 2008) on serum TEQs by age group show a similar pattern. Table 3 also shows the serum mean TEQs decreased in all age groups when based on the 2005 TEFs compared to the 1998 TEFs (16% in age 18–29, 22% in age 30–44, 28% in age 45–59, and 27% in age 60+). The pattern of decreases was similar for the serum 95th percentile (P95) TEQ across age groups, except for in the
age 60+ group which showed only 18% decrease after the implementation of the 2005 TEFs. Table 4 presents the mean percent contribution of PCDDs, PCDFs, non-ortho PCBs, and mono–ortho PCBs to total TEQ– WHO98 and TEQ–WHO05 for serum, household dust, and soil samples in the study. The results for serum samples are also compared with the values reported by Patterson et al. (2008) based on the NHANES 2001–2002 serum dioxin concentrations. Table 5 provides the mean percent contribution to the total TEQ by each specific congener for serum, household dust and soil samples. Fig. 1 provides the percent contribution of the seven important congeners (i.e., 2,3,7,8-TCDD, 1,2,3,7,8-PeCDD, 1,2,3,6,7,8-HxCDD, 2,3,4,7,8PeCDF, PCB 126, PCB 118, and PCB 156) to the serum total TEQ– WHO98 and TEQ–WHO05 in UMDES data, shown using box plots. For serum TEQ, the mean percent contribution of PCDDs increased 16% in the TEQ–WHO05 compared to the TEQ–WHO98 (Table 4), while the contribution of the mono–ortho PCBs decreased 21%. There were small changes in the contribution of the PCDFs (0.4%) and non-ortho PCBs (5.3%). The Patterson et al. (2008) results showed similar results (for PCDDs: there was a 14% increase; for PCDFs: there was a 1% decrease; for non-ortho PCBs: there was a 6% increase, and for mono–ortho PCBs there was a 19% decrease). The reduction of the contribution in mono–ortho PCBs was due to the down-weighting of the TEFs for the majority of mono–ortho PCBs. For example, PCB 156 contributed 1.2% to the serum TEQ– WHO05 compared to 14% to the serum TEQ–WHO98 (Table 5). These changes in the TEFs for the mono–ortho PCBs had a dramatic impact on which congeners were the major contributors to the TEQ. Seven congeners accounted for more than 80% of the serum TEQ–WHO98 (i.e., 2,3,7,8-TCDD, 1,2,3,7,8-PeCDD, 1,2,3,6,7,8HxCDD, 2,3,4,7,8-PeCDF, PCB 126, PCB 118, and PCB 156) in our study. These results are in agreement with the NHANES 2001– 2002 data (Patterson et al., 2008). With the new 2005 TEFs, both PCB 118 and PCB 156 dropped from the top seven contributors
730
B. Hong et al. / Chemosphere 76 (2009) 727–733
Table 3 Comparison between serum total TEQ–WHO98 and TEQ–WHO05 by age groups in the UMDES study and in NHANES 2001–2002 (unit: ppt). N
UMDES serum data Age 18–29 55 Age 30–44 200 Age 45–59 378 Age 60+ 313
Total TEQ based on WHO 1998 TEFs a
Total TEQ based on WHO 2005 TEFs 95% CI for P95
Mean
95% CI for mean
P95
9.5 18.6 32.2 54.4
8.5–10.4 17.0–20.1 29.1–35.2 49.6–59.3
13.6 28.2 66.5 106.7
12.1–21.1 26.0–47.1 48.2–96.1 97.2–128.1
34.7 56.7 102
29.9–40.7 46.4–82.6 86.9–114
NHANES 2001–2002 serum datac Age 20–39 443 Age 40–59 362 Age 60+ 389
b
% Change 95% CI for P95
In mean
In P95a
11.2 21.3 48.8 87.2
10.4–17.8 20.6–40.5 36.1–70.6 70.6–103.8
15.8 22.0 28.0 27.2
17.6 24.5 26.6 18.3
26.2 46.9 79.7
23.7–32.5 36.4–66.1 68.2–96.3
Mean
95% CI for mean
P95
8.0 14.5 23.2 39.6
7.2–8.7 13.3–15.8 21.1–25.4 35.8–43.4
a
b
24 17 22
Note: All results accounted for complex survey design. a P95 is the 95th percentile of the distribution. b The 95% CI (confidence interval) for P95 is calculated by R using method of Woodruff (1952). c From Patterson et al. (2008) Table 8.
Table 4 Mean percent contribution to the total TEQ for PCDDs, PCDFs, non-ortho PCBs and mono–ortho PCBs (unit: %). Percent contribution to the TEQ (%)
Serum PCDDs PCDFs Non-ortho PCBs Mono-ortho PCBs Household dust PCDDs PCDFs Non-ortho PCBs Mono-ortho PCBs House perimeter soil PCDDs PCDFs Non-ortho PCBs Mono-ortho PCBs
NHANES 2001–2002 dataa
UMDES data WHO 1998
WHO 2005
Changes
WHO 1998
WHO 2005
Changes
48.7 16.8 9.4 25.1
64.7 16.4 14.7 4.1
16.1 0.4 5.3 21.0
48.0 15.0 14.0 23.0
62.0 14.0 20.0 4.0
14.0 1.0 6.0 19.0
59.5 17.0 12.4 11.2
66.8 15.6 14.6 3.0
7.3 1.3 2.2 8.2
49.4 36.2 11.1 3.2
54.4 32.1 12.8 0.8
4.9 4.1 1.7 2.4
Note: All results accounted for complex survey design. a These results are from Patterson et al. (2008) Table 7.
and only contributed about 3% to the serum TEQ–WHO05 (Table 5 and Fig. 1). However, the TEQ–WHO05 was still dominated by 2,3,7,8-TCDD, 1,2,3,7,8-PeCDD, 1,2,3,6,7,8-HxCDD, 2,3,4,7,8-PeCDF, and PCB 126, which contributed 9.6%, 27%, 20%, 8.7% and 11% to the serum TEQ, respectively. The new two congeners among the top seven contributors to the TEQ–WHO05 were PCB 169 and 1,2,3,7,8,9-HxCDD, which contributed 3.3% and 3.2% to the serum TEQ, respectively. For household dust TEQ, the mean percent contribution of PCDDs increased about 7.3% in the TEQ–WHO05 compared to the TEQ–WHO98, while the contributions of mono–ortho PCBs decreased about 8.2% (Table 4). There were small changes in the contribution of the PCDFs (1.3%) and non-ortho PCBs (2.2%). For each specific congener, the percent contribution changed slightly within a range of 2.9 to 4.5% (Table 5). The largest impact was on OCDD, which contributed 4.5% more for household dust in the TEQ– WHO05 compared to the TEQ–WHO98 (Table 5), due to the 3-fold increase in the TEF from 0.0001 to 0.0003. The major contributors for the household dust TEQ–WHO05 were 1,2,3,7,8-PeCDD (9.8%), 1,2,3,6,7,8-HxCDD (9.8%), 1,2,3,4,6,7,8-HpCDD (31%), and PCB 126 (14%). For the house perimeter soil samples, the mean percent contribution of PCDDs increased 4.9%, while the contributions of PCDFs decreased 4.1% (Table 4). There were small changes in the contribution of the non-ortho PCBs (1.7%) and mono–ortho PCBs
(2.4%). The reduction of the contribution in PCDFs was mainly due to the down-weighting of 2,3,4,7,8-PeCDF (1998 TEF = 0.5, 2005 TEF = 0.3), which accounted for 18% of the soil TEQ–WHO98 but 12% of TEQ–WHO05 (Table 5). The major contributors for soil TEQ–WHO05 were 2,3,7,8-TCDD (15%), 1,2,3,7,8-PeCDD (15%), 1,2,3,4,6,7,8-HpCDD (10%), 2,3,4,7,8-PeCDF (12%), and PCB 126 (12%).
4. Discussion The mean total TEQ–WHO05 levels in serum, household dust, and soil samples in our study were reduced, by 26%, 12% and 14%, respectively, after the implementation of the 2005 TEFs compared to the 1998 TEFs (Table 2). The change in the serum TEQ was at the upper end of the range of 10–25% for the general population predicted by Van den Berg et al. (2006). Although these changes are appreciable, it should be kept in mind that laboratory variation in congener measurements may vary by roughly the same amount. We performed duplicate GC/MS analyses of our serum samples in cooperation with the National Center for Environmental Health laboratory and found that variation was within about two standard deviations (mean variation was 11% and maximum variation was 27%). Our results for the change in serum TEQ 95th percentile were quite similar to the
731
B. Hong et al. / Chemosphere 76 (2009) 727–733 Table 5 Mean percent contribution to the total TEQ for each WHO dioxin congener in UMDES study (unit: %). Compound
Serum
Household dust
WHO 1998 2378_TCDD 12378_PeCDD 123478_HxCDD 123678_HxCDD 123789_HxCDD 1234678_HpCDD OCDD 2378_TCDF 12378_PeCDF 23478_PeCDF 123478_HxCDF 123678_HxCDF 123789_HxCDF 234678_HxCDF 1234678_HpCDF 1234789_HpCDF OCDF PCB_77 PCB_81 PCB_126 PCB_169 PCB_105 PCB_114 PCB_118 PCB_123 PCB_156 PCB_157 PCB_167 PCB_189
WHO 2005
c
c
7.3 20.6c 2.0 14.6c 2.5 1.6 0.1 0.2 0.1 10.9c 2.2 2.1 0.4 0.5 0.4 0b 0b 0b 0b 8.6c 0.8 0.9 2.0 4.2c 0.1 14.4c 3.3 0.1 0.2
9.6 27.3c 2.6 19.5c 3.2c 2.1 0.4 0.3 0.1 8.7c 2.9 2.8 0.5 0.6 0.5 0.1 0b 0b 0b 11.4c 3.3c 0.4 0.2 1.7 0b 1.2 0.3 0.2 0.1
Changes
a
WHO 1998
WHO 2005
c
2.3 6.7 0.6 4.9 0.8 0.5 0.3 0.1 0b 2.2 0.7 0.7 0.1 0.1 0.1 0b 0b 0b 0b 2.8 2.5 0.5 1.9 2.5 0b 13.2 3.0 0.2 0.1
House perimeter soil
c
4.2 9.2c 1.5 9.3c 3.6 29.7c 2.1 1.3 0.5 6.4c 1.9 1.9 0.4 1.8 2.6 0.1 0.1 0.2 0b 12.1c 0.1 2.0 0.6 4.7c 0.1 3.1 0.7 0b 0b
4.5 9.8c 1.6 9.8c 3.8 30.7c 6.6c 1.4 0.3 4.2c 2.1 2.1 0.5 2.0 2.7 0.1 0.2 0.2 0b 14.1c 0.3 0.8 0.1 1.8 0b 0.2 0.1 0.1 0b
Changes
a
WHO 1998 c
0.3 0.6 0.1 0.5 0.2 1.1 4.5 0.1 0.2 2.2 0.2 0.2 0b 0.1 0.2 0b 0.2 0b 0b 2.0 0.2 1.3 0.6 2.9 0b 2.9 0.6 0b 0b
14.6 14.4c 1.8 4.5c 3.8c 9.6c 0.8 3.4 1.1 18.0c 3.1 3.5 0.7 2.7 3.3 0.2 0.1 0b 0b 10.9c 0.2 0.6 0.1 1.1 0b 1.1 0.3 0b 0b
WHO 2005 c
15.4 15.4c 1.9 4.8c 4.0c 10.2c 2.6 4.0 0.8 12.3c 3.5 3.8 0.8 2.9 3.6 0.2 0.2 0.1 0b 12.2c 0.5 0.2 0b 0.4 0b 0.1 0b 0b 0b
Changesa 0.8 1.0 0.1 0.3 0.3 0.6 1.8 0.6 0.4 5.7 0.4 0.3 0.1 0.3 0.3 0b 0.1 0b 0b 1.3 0.4 0.4 0b 0.7 0b 1.0 0b 0b 0b
Note: All results accounted for complex survey design. a Percent contribution to TEQ–WHO05 – percent contribution to TEQ–WHO98 (the italicised values indicate the largest change due to the 2005 TEFs). b Values less than 0.05. c Indicate the top seven contributed congeners to the TEQ.
UMDES serum data
Percent contribution to Total TEQ
60
40
20
Percent contribution to TEQ-WHO98
PCB_156
PCB_118
PCB_126
23478_PeCDF
123678_HxCDD
12378_PeCDD
2378_TCDD
PCB_156
PCB_118
PCB_126
23478_PeCDF
123678_HxCDD
12378_PeCDD
2378_TCDD
0
Percent contribution to TEQ-WHO05
Fig. 1. The percent contribution of seven important congeners to the serum total TEQ–WHO98 versus TEQ–WHO05 in UMDES data, shown using box plots. The plus sign indicates the arithmetic mean, the horizontal line across the box indicates the 50th percentile; the lower and upper margins of the box indicate the 25th and 75th percentiles, respectively; the lower and upper ticked lines extend to the 1st and 99th percentile, respectively; the stars show values below the 1st or above 99th percentile.
results reported by Patterson et al. (2008) based on the NHANES 2001–2002 data. The changes in TEFs will make comparisons be-
tween older publications (Flesch-Janys et al., 1995; DeVito et al., 1995; Wittsiepe et al., 2000; Millette, 2000) and newer publica-
732
B. Hong et al. / Chemosphere 76 (2009) 727–733
tions (based on TEQ–WHO05) (Wittsiepe et al. 2007) more challenging. In our study, the serum TEQs were calculated from both sets of TEFs (WHO 1998 TEFs and 2005 TEFs), and the full results for each of our study populations by age group are provided on the study website (www.umdioxin.org) in an effort to facilitate comparisons to past and future studies. The contribution of mono–ortho PCBs to the serum TEQ was reduced when the TEQ–WHO05 was compared to the TEQ–WHO98 (21%) (Table 4), due to the down-weighting of the majority of mono–ortho PCBs. The largest decrease in the serum TEQ was due to PCB 156, which alone accounted for 14% of the TEQ– WHO98 versus 1.2% of the TEQ–WHO05 (Table 5). The serum TEQ–WHO05 was still dominated by the five major congeners (2,3,7,8-TCDD, 1,2,3,7,8-PeCDD, 1,2,3,6,7,8-HxCDD, 2,3,4,7,8-PeCDF, and PCB 126), with the highest contribution coming from 1,2,3,7,8-PeCDD (27%). The impact of the changes in TEFs on total TEQ was not as substantial for soil as it was for serum in the present study. The 4.1% reduction in the contribution of PCDFs in soil (Table 4) was mainly due to the down-weighting of 2,3,4,7,8-PeCDF (the TEF decreased from 0.5 to 0.3), which was the largest contributor to TEQ–WHO98 (accounting for 18%). The contribution to the TEQ–WHO98 or TEQ–WHO05 of the other two PCDFs for which the TEFs changed (1,2,3,7,8-PeCDF and OCDF) was small (less than 1.5%). The impact of the changes in TEFs on total TEQ was small for household dust. The TEQ approach is intended to estimate exposure to dioxin mixtures via dietary intake. Interpreting TEQs from environmental matrices such as soil and household dust in settings where there is no ingestion may have no relevance to risk assessment unless the bioavailability and environmental fate and transport of specific congeners are taken into account. Because of this concern, we have provided the congener-specific data that underlies the TEQs and recommend that for purposes of risk assessment the congener specific concentrations be used. In the instance of household dust, where toddlers may ingest material, the TEQ may have direct relevance to risk assessment. It is also worthwhile to examine the percent contributions of the different groups of congeners (PCDDs, PCDFs, non-ortho PCBs or mono–ortho PCBs) to the total TEQs (Table 4). The mean percent contribution of PCDDs increased for all three types of samples for the TEQ–WHO05, compared to the TEQ–WHO98 (16% for serum, 7.3% for household dust, and 4.9% for soil). This was due to the proportional decrease in the contribution by the PCDFs and mono–ortho PCBs. For the PCDDs, the only change in TEFs in 2005 was for OCDD, which increased from 0.0001 in 1998 to 0.0003 in 2005. The contribution of the PCDFs was slightly reduced for serum (0.4%) and household dust (1.3%), but somewhat greater (4.1%) for soil. The changes in TEQ–WHO05 compared to the TEQ–WHO98 may have an impact in locations where regulations of soil contamination are triggered by specific TEQ levels. For example, the Michigan Department of Environmental Quality (MDEQ) residential soil direct contact criterion is 90 ppt TEQ (, Lansing, MI, 1999). In the area of Midland, Michigan downwind of the Dow facilities, 38.1% of properties had a soil TEQ–WHO98 greater than 90 ppt in one or more samples (mean = 121.2 ppt, median = 59.2 ppt, 75th percentile = 114.1 ppt, 95th percentile = 270.7 ppt, maximum = 925.8 ppt), compared to 35.8% using the TEQ–WHO05 (mean = 109.2 ppt, median = 58.2 ppt, 75th percentile = 111.9 ppt, 95th percentile = 257.2 ppt, maximum = 745.5 ppt) (Demond et al., 2008). In the Jackson/Calhoun referent area, 1.8% of properties had a soil TEQ–WHO98 greater than 90 ppt (mean = 8.3 ppt, median = 3.6
ppt, 75th percentile = 8.5 ppt, 95th percentile = 23.8 ppt, maximum = 329.7 ppt), compared to 0.3% using the TEQ–WHO05 (mean = 6.9 ppt, median = 3.6 ppt, 75th percentile = 7.6 ppt, 95th percentile = 22.6 ppt, maximum = 186.2 ppt) (Demond et al., 2008). The change in TEFs had a substantial impact (1.8% versus 0.3%) on the proportion of properties above the regulatory limit in the latter population. In contrast, it had a small effect on the proportion of properties above this regulatory limit in the contaminated area (38.1% versus 35.8%). The changes in mean TEQ–WHO05 compared to the TEQ– WHO98 seen in serum were quite similar between the populations in Midland/Saginaw and the Jackson/Calhoun referent population (total TEQ: 26% to 23% for Midland/Saginaw populations and 28% for Jackson/Calhoun population; for PCDD TEQ: 0%–0.7% for Midland/Saginaw populations and 0.8% for Jackson/Calhoun population; for PCDF TEQ: 28% to 29% for Midland/Saginaw populations and 26% for Jackson/Calhoun population; for non-ortho PCB TEQ: 14% to 22% for Midland/Saginaw populations and 13% for Jackson/Calhoun population; for mono–ortho PCB TEQ: 88% to 87% for Midland/Saginaw populations and 87% for Jackson/Calhoun population), suggesting that the effect of the new TEFs on the TEQ is fairly stable across populations. This pattern may not be the same in settings where there is excessive exposure to a specific pattern of congeners that is different than what is commonly seen in the general population. The change in TEFs in 2005 was principally a reflection of updated information on toxicity of specific congeners. Acknowledgments Financial support came from The Dow Chemical Company through an unrestricted grant to the University of Michigan. The authors acknowledge Ms. Sharyn Vantine for her assistance, and Drs. Linda Birnbaum, Ronald Hites, Paolo Boffetta and Marie Haring Sweeney for their guidance as members of our Scientific Advisory Board. Drs. Garabrant and Franzblau declare they have competing financial interests. Drs. Garabrant and Franzblau have at times been retained as consultants and served as expert witnesses for the Dow Chemical Company. Appendix A WHO TEFs TEQ PeCDF TEQ–WHO98 TEQ–WHO05 TCDD OCDD OCDF UMDES PCDDs PCDFs PCBs HRMS ppt HVS3s ASTM LOD MDEQ NHANES
World Health Organization toxic equivalency factors toxic equivalent pentachlorodibenzofuran TEQ based on the WHO 1998 TEFs TEQ based on the WHO 2005 TEFs tetrachlorodibenzo-p-dioxin octachlorodibenzo-p-dioxin octachlorodibenzofuran University of Michigan Dioxin Exposure Study polychlorinated dibenzo-p-dioxins polychlorinated dibenzofurans polychlorinated biphenyls high resolution mass spectrometry parts per trillion high volume small surface samplers American Society for Testing and Materials limit of detection Michigan Department of Environmental Quality the National Health and Nutrition Examination Survey
B. Hong et al. / Chemosphere 76 (2009) 727–733
References ASTM, 2000. Standard Practice for Collection of Floor Dust for Chemical Analysis. D 5438-00, 1–7. American Society for Testing and Materials, West Conshohocken, PA. (ref. type: report). Demond, A., Adriaens, P., Towey, T., Chang, S.-C., Hong, B., Chen, Q., Chang, C.-W., Franzblau, A., Garabrant, D., Gillespie, B., Hedgeman, E., Knutson, K., Lee, C.-Y., Lepkowski, J., Olson, K., Ward, B., Zwica, L., Luksemburg, W., Maier, M., 2008. Statistical comparison of residential soil concentrations of PCDDs, PCDFs, and PCBs from two communities in Michigan. Environmental Science and Technology 42 (15), 5441–5448. doi:10.1021/es702554. DeVito, M., Birnbaum, L., Farland, W., Gasiewicz, T., 1995. Comparisons of estimated human body burdens of dioxinlike chemicals and TCDD body burdens in experimentally exposed animals. Environmental Health Perspectives 103, 820– 831. EPA, 1994. Method 8290, Polychlorinated Dibenzodioxins (PCDDs) and Polychlorinated Dibenzofurans (PCDFs) by High-Resolution Gas Chromatography/High-Resolution Mass Spectrometry (HRGC/HRMS). US Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington, DC. EPA, 1999. Method 1668, Revision A: Chlorinated Biphenyl Congeners in Water, Soil, Sediment, and Tissue by HRGC/HRMS. US Environmental Protection Agency, Office of Water, Washington, DC. Flesch-Janys, D., Berger, J., Gurn, P., Manz, A., Nagel, S., Waltsgott, H., Dwyer, J., 1995. Exposure to polychlorinated dioxins and furans (PCDD/F) and mortality in a cohort of workers from a herbicide-producing plant in Hamburg, Federal Republic of Germany. The American Journal of Epidemiology 142, 1165–1175. Garabrant, D.H., Franzblau, A., lepkowski, J., Gillespie, B.W., Adriaens, P., Demond, A., Ward, B., LaDronka, K., Hedgeman, E., Knutson, K., Zwica, L., Olson, K., Towey, T.,
733
Chen, Hong, B., 2008. The University of Michigan dioxin exposure study: methods for an environmental exposure study of polychlorinated dioxins, furans and biphenyls. Environmental Health Perspectives. doi:10.1289/ ehp.11777.
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Impact of WHO 2005 revised toxic equivalency factors for dioxins on the TEQs in serum, household dust and soil Biling Hong a,*, David Garabrant a, Elizabeth Hedgeman a, Avery Demond b, Brenda Gillespie c, Qixuan Chen c, Chiung-Wen Chang a, Timothy Towey b, Kristine Knutson a, Alfred Franzblau a, James Lepkowski d, Peter Adriaens b a
Department of Environmental Health Sciences and the Risk Science Center, University of Michigan School of Public Health, Ann Arbor, MI, USA Department of Civil and Environmental Engineering, University of Michigan College of Engineering, Ann Arbor, MI, USA Department of Biostatistics, University of Michigan School of Public Health, Ann Arbor, MI, USA d Institute for Social Research, University of Michigan, Ann Arbor, MI, USA b c
a r t i c l e
i n f o
Article history: Received 7 January 2009 Received in revised form 13 April 2009 Accepted 25 May 2009 Available online 1 July 2009 Keywords: Dioxins Toxic equivalency factors Furans PCBs UMDES Serum
a b s t r a c t Background: In 2005, the World Health Organization (WHO) – International Programme on Chemical Safety reevaluated the toxic equivalency factors (TEFs) for dioxin-like compounds and made changes that affect the calculation of the total toxic equivalent (TEQ). The impact of these changes on the TEQs for human blood and abiotic matrices such as soil and household dust has not been widely assessed or reported. Methods and results: Using a major exposure study which examined blood, household dust, and soil levels of dioxin-like compounds in several regions of Michigan, we found the mean total TEQ was significantly reduced by 26%, 12% and 14% for serum, household dust, and soil, respectively, when the TEQ was based on the 2005 TEFs compared to the 1998 TEFs. The decrease in the serum total TEQ was largely due to the down-weighting of the TEFs for the majority of mono–ortho PCBs. In contrast, the decrease in the soil total TEQ was mostly due to the down-weighting of the TEF for 2,3,4,7,8-pentachlorodibenzofuran (PeCDF) (1998 TEF = 0.5, 2005 TEF = 0.3). For household dust, the decrease in total TEQ was not due to any single TEF but was due to small changes in a number of compounds. There was a dramatic decrease (88%) in the mean and 95th percentile for mono–ortho PCB TEQ due to the 2005 TEFs. Discussion: These findings suggest that comparisons between studies based on the TEQ–WHO98 and TEQ–WHO05 may need to consider an appropriate conversion factor to assure comparability. Furthermore, the changes in TEFs may have impact in locations where regulations of soil contamination are triggered by specific TEQ levels. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction The toxic equivalent (TEQ) concept was developed to measure the combined toxicity of complex mixtures of dioxin and dioxinlike compounds through the toxic equivalency factors (TEFs) methodology, which assesses the toxicity of each congener relative to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Mathematically, the total TEQ is defined as the weighted sum, over all WHO 29 dioxin congeners, of the TEF multiplied by the congener concentration. In 2005, the World Health Organization (WHO) – International Programme on Chemical Safety reevaluated the TEFs for dioxin-
* Corresponding author. Address. Department of Environmental Health Sciences and the Risk Science Center, University of Michigan School of Public Health, 1420 Washington Heights, Room 6529, Ann Arbor, MI 48109-2029, USA. Tel.: +1 734 936 0726; fax: +1 734 763 7170. E-mail address: [email protected] (B. Hong). 0045-6535/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2009.05.034
like compounds and made changes that affect the calculation of the total TEQ (Van den Berg et al., 2006). Objectives of these revisions included harmonization of the TEFs internationally and insuring that studies published since the previous revision in 1997 (Van den Berg et al., 1998) were evaluated. Although no compounds were added or removed from the list of 29 dioxin-like compounds for which TEFs were agreed in 1997, two important changes were made in the WHO 2005 TEFs: (1) The WHO 1998 TEFs were assigned in increments of 0.01, 0.05, 0.1, etc., whereas the new 2005 TEFs used the half order of magnitude increments on a logarithmic scale of 0.03, 0.1, 0.3, etc. (2) TEFs for 14 out of 29 congeners were changed in 2005 (Table 1). Octachlorodibenzo-p-dioxin (OCDD) and octachlorodibenzofuran (OCDF) were both slightly up-weighted from 0.0001 to 0.0003; 1,2,3,7,8-PeCDF was down-weighted
728
B. Hong et al. / Chemosphere 76 (2009) 727–733
Table 1 Updated TEFs for 14 dioxin-like congeners during the 2005 reevaluation. Congener
WHO 1998 TEF
WHO 2005 TEF
Dibenzo-p-dioxins OCDD
0.0001
0.0003
Dibenzofurans 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF OCDF
0.05 0.5 0.0001
0.03 0.3 0.0003
Non-ortho PCBs PCB 81 PCB 169
0.0001 0.01
0.0003 0.03
Mono-ortho PCBs PCB 105 PCB 114 PCB 118 PCB 123 PCB 156 PCB 157 PCB 167 PCB 189
0.0001 0.0005 0.0001 0.0001 0.0005 0.0005 0.00001 0.0001
0.00003 0.00003 0.00003 0.00003 0.00003 0.00003 0.00003 0.00003
The objective of this paper is to present the impact of the 2005 TEF changes on TEQ levels, as well as on the percent contributions of PCDDs, PCDFs, non-ortho PCBs and mono–ortho PCBs to the total TEQ, in serum, soil, and household dust samples from UMDES. 2. Materials and methods
Note: The italicised congeners were up-weighted in 2005; the rest of the congeners were down-weighted in 2005.
from 0.05 to 0.03; 2,3,4,7,8-PeCDF was down-weighted from 0.5 to 0.3; two non-ortho PCBs were slightly up-weighted (from 0.0001 to 0.0003 for PCB 81, and from 0.01 to 0.03 for PCB 169); a single value (0.00003) was assigned for all mono–ortho PCBs, which resulted in downweighting for the majority of mono–ortho PCBs (except for PCB 167). The WHO expert panel recommended caution in applying the TEQ approach to abiotic matrices because the TEFs and TEQ approach is primarily intended for estimating exposures and risks from dietary ingestion. Nonetheless, this approach is widely applied to environmental matrices such as soils and sediments because it gives a single value to represent contamination from complex and highly variable mixtures of dioxin-like compounds. Although these revisions were estimated to result in a 10–25% reduction in the human serum total TEQ, we are not aware of studies that have evaluated their actual impact in human populations or in abiotic matrices such as soil and household dust. This paper examines the effects of the changes in TEFs on the TEQ using data from a large population-based study of dioxin exposure in Michigan. The University of Michigan Dioxin Exposure Study(UMDES) (www.umdioxin.org) was undertaken in response to concerns that the discharge of dioxin-like compounds from the Dow Chemical Company facilities in Midland, Michigan (USA) has resulted in contamination of soils in the Tittabawassee River flood plain and areas of the City of Midland, leading to an increase in residents’ body burdens of polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs) and dioxin-like polychlorinated biphenyls (PCBs). Serum samples were obtained from 946 participants in Michigan by using a complex sample design in 2004–2005 (Garabrant et al., 2008) and analyzed for World Health Organization (WHO) 29 dioxin-like compounds (Van den Berg et al., 1998), including 2,3,7,8-TCDD. Soil samples from eligible participants’ homes and dust samples from inside homes were also collected and analyzed for the same set of congeners. The study is uniquely valuable because of its large size and its evaluation of dioxin-like compounds in serum, soil, and household dust from a population-based sample. As a result, the inferences from these samples can be applied to the general population that was sampled.
The study included Michigan populations who live in Midland County, Saginaw County, and part of Bay County, both in and out of the Tittabawassee River flood plain, and who live in Jackson and Calhoun Counties. Adults age 18 or over who had lived in their current residence for at least five years were eligible to participate. A total of 946 participants who met the Red Cross criteria for blood donation gave an 80 mL whole blood sample for analysis. In order to be eligible for household dust and soil sampling, the participant had to be an owner of the property. A total of 764 participants provided household dust samples, and 766 participants provided soil samples from around the house perimeter. The complete UMDES protocol, including the blood collection, household dust and soil sampling protocols, is available at www.umdioxin.org. Serum, household dust, and soil samples were analyzed by Vista Analytical Laboratory (El Dorado Hills, CA) for the 29 dioxin-like congeners, using modified US EPA protocols 8290 (US Environmental Protection Agency 1994) and 1668 (US Environmental Protection Agency 1999). For blood samples, congeners were extracted from serum, cleaned following a multi-column protocol and quantified using high resolution mass spectrometry (HRMS). Final analyte concentrations were provided on both a whole weight and lipid-adjusted weight basis, expressed in parts per trillion (ppt) (equivalent to picograms gram1). Household dust samples were collected using high volume small surface samplers (HVS3s) manufactured by CS-3, Inc. (Sandpoint, ID, USA), and equipped with a cyclone and fine particle filter capable of capturing 99.95% of particles above 0.3 lm aerodynamic mean diameter. One 10 g composite sample was taken in each household from frequently used areas, from both hard and soft surfaces with carpets and area rugs being preferred, according to the American Society for Testing and Materials (ASTM) method ‘‘Standard Practice for Collection of Floor Dust for Chemical Analysis” (ASTM 2000). The soil samples were archived in 4 oz dioxin-grade amber glass containers to avoid photolytic degradation reactions, and stored in dedicated 4 °C cold rooms prior to analysis. For all residences, the top 1 in. composited sample from the perimeter was submitted for analysis using HRGC/HRMS. The results for dioxin concentrations that fell below the limit of p detection (LOD) were estimated by LOD/ 2 (Hornung and Reed, 1990). Statistical analyses, which accounted for the complex survey design, were performed using SAS version 9.1.3 (SAS Institute, Inc., Cary, NC, 2002–2003) and R package ‘‘survey” (R Release 2.6.1, 2007).
3. Results Among those 946 participants who donated blood, 93% were white and 57% were female, with a mean age of 51 years and a mean BMI of 29 kg m2. Table 2 presents the comparisons between TEQ–WHO98 and TEQ–WHO05 for serum, household dust, and soil samples, respectively, with descriptive statistics giving the mean (and 95% confidence interval), 95th percentile (and its 95% confidence interval), and the percent changes in the mean and 95th percentile TEQs. The 95th percentile serum TEQs are also compared with the values reported by Patterson et al. (2008) based on the NHANES 2001–2002 serum dioxin concentrations. The mean total TEQ–WHO05 levels in serum, household dust, and soil in the UMDES data were reduced by 26%, 12% and 14%, respectively, com-
729
B. Hong et al. / Chemosphere 76 (2009) 727–733 Table 2 Comparison between TEQ–WHO98 and TEQ–WHO05 for samples in the UMDES study and in NHANES 2001–2002 (unit: ppt). N
UMDES serum data Total PCDDs PCDFs Non-ortho PCBs Mono-ortho PCBs
946 946 946 946 946
TEQ based on WHO 1998 TEFs
UMDES household dust data Total 764 PCDDs 764 PCDFs 764 Non-ortho PCBs 764 Mono-ortho PCBs 764
b
95% CI for P95
Mean
95% CI for mean
75.3 35.6 10.6 11.9 24.6
68.3–104.0 32.4–42.7 9.6–12.3 8.9–16.9 21.1–29.9
23.9 15.3 3.6 4.0 1.1
22.1–25.7 14.2–16.4 3.4–3.8 3.4–4.5 1.0–1.2
68.9 34.8 12.3 11.0 18.0
62.9–80.8 28.7–43.3 11.0–14.4 9.7–12.0 15.1–20.4
Mean
95% CI for mean
P95
32.4 15.2 4.9 3.4 8.9
30.0–34.8 14.1–16.3 4.7–5.2 2.9–3.9 8.0–9.7
NHANES 2001–2002 serum datac Total 1194 PCDDs 1194 PCDFs 1194 Non-ortho PCBs 1194 Mono-ortho PCBs 1194
TEQ based on WHO 2005 TEFs a
% Change 95% CI for P95
In mean
In P95a
58.6 35.8 7.5 13.1 3.0
50.8–73.3 32.5–43.1 6.9–8.3 10.2–18.1 2.6–4.0
26.0 0.7 27.0 18.0 88.0
22.0 0.6 29.0 10.0 88.0
56.1 34.8 8.9 11.9 2.6
47.6–65.4 28.7–43.4 7.7–10.2 10.8–12.9 2.3–3.0
P95
a
b
18.6 0.0 27.6 8.2 85.6
41.9 22.3 4.9 7.0 7.7
32.0–51.9 16.4–28.2 4.1–5.8 3.5–10.5 4.0–11.5
142.5 70.2 13.5 10.4 19.1
95.4–275.7 46.2–99.9 12.6–18.7 7.3–167.6 10.1–81.1
36.7 24.0 4.2 7.1 1.5
28.4–45.1 17.6–30.3 3.5–5.0 3.6–10.6 0.85–2.1
126.4 74.3 12.7 10.5 3.8
90.6–220.8 49.1–110.7 10.3–17.5 7.4–168.6 2.3–30.6
12.4 7.6 14.3 1.4 80.5
11.3 5.8 5.9 1.0 80.1
UMDES house perimeter soil data Total 766 16.6 PCDDs 766 6.2 PCDFs 766 8.2 Non-ortho PCBs 766 1.4 Mono-ortho PCBs 766 0.8
12.7–20.5 5.0–7.4 5.2–11.2 0.8–2.0 0.2–1.5
73.3 25.4 23.9 3.7 1.2
51.1–85.5 15.5–48.7 13.5–63.9 2.3–4.9 0.7–2.6
14.3 6.4 6.3 1.4 0.2
11.2–17.4 5.1–7.6 4.1–8.6 0.8–2.0 0.03–0.3
59.4 25.8 18.5 3.7 0.3
50.1–74.7 16.2–49.3 11.3–48.9 2.3–5.0 0.1–0.5
13.9 3.2 23.2 0.0 75.0
19.0 1.6 22.6 0.0 75.0
Note: All results accounted for complex survey design. a P95 is the 95th percentile of the distribution. b The 95% CI (Confidence Interval) for P95 is calculated by R using method of Woodruff (1952). c From Patterson et al. (2008) Table 1.
pared to the mean total TEQ–WHO98 levels. The 95th percentile for serum total TEQ was reduced by 22% in UMDES data, compared to 19% in NHANES 2001–2002 data (Patterson et al., 2008). We also examined the TEQ for the PCDD congeners alone, for the PCDF congeners, the non-ortho PCBs, and the mono–ortho PCBs. The PCDD TEQ showed only modest changes when comparing the 1998 and 2005 TEFs at both the mean and at the 95th percentile (the mean PCDD TEQ increased 0.7% in serum, 7.6% in household dust, and 3.2% in soil). In contrast, the mean PCDF TEQ decreased substantially at both the mean and at the 95th percentile in serum and soil, with less pronounced changes in the household dust (the mean PCDF TEQ decreased 27% in serum, 14% in household dust, and 23% in soil). For the non-ortho PCBs, the mean serum TEQ increased by 18%, but showed almost no change in household dust or soil. The most dramatic effect of the 2005 TEFs was seen for the mono–ortho PCB TEQ, for which the mean serum level decreased by 88%, household dust decreased by 81%, and soil decreased by 75%. Similar changes were seen at the 95th percentile. Comparison of the UMDES and NHANES serum results showed virtually identical changes for the 2005 TEFs at the 95th percentile TEQ. We also evaluated the serum total TEQs by age groups (Table 3). The serum TEQ levels increased with age using either the 1998 TEFs or the 2005 TEFs. The mean serum TEQ for the oldest age group (1998 TEFs: 54.4 ppt for age 60+; 2005 TEFs: 39.6 ppt for age 60+) is about five times larger than the mean serum TEQ in the youngest age group (1998 TEFs: 9.5 ppt for age 18–29; 2005 TEFs: 8 ppt for age 18–29). The trends by age are similar for the 95th percentile for serum total TEQ. Patterson’s results (Patterson et al., 2008) on serum TEQs by age group show a similar pattern. Table 3 also shows the serum mean TEQs decreased in all age groups when based on the 2005 TEFs compared to the 1998 TEFs (16% in age 18–29, 22% in age 30–44, 28% in age 45–59, and 27% in age 60+). The pattern of decreases was similar for the serum 95th percentile (P95) TEQ across age groups, except for in the
age 60+ group which showed only 18% decrease after the implementation of the 2005 TEFs. Table 4 presents the mean percent contribution of PCDDs, PCDFs, non-ortho PCBs, and mono–ortho PCBs to total TEQ– WHO98 and TEQ–WHO05 for serum, household dust, and soil samples in the study. The results for serum samples are also compared with the values reported by Patterson et al. (2008) based on the NHANES 2001–2002 serum dioxin concentrations. Table 5 provides the mean percent contribution to the total TEQ by each specific congener for serum, household dust and soil samples. Fig. 1 provides the percent contribution of the seven important congeners (i.e., 2,3,7,8-TCDD, 1,2,3,7,8-PeCDD, 1,2,3,6,7,8-HxCDD, 2,3,4,7,8PeCDF, PCB 126, PCB 118, and PCB 156) to the serum total TEQ– WHO98 and TEQ–WHO05 in UMDES data, shown using box plots. For serum TEQ, the mean percent contribution of PCDDs increased 16% in the TEQ–WHO05 compared to the TEQ–WHO98 (Table 4), while the contribution of the mono–ortho PCBs decreased 21%. There were small changes in the contribution of the PCDFs (0.4%) and non-ortho PCBs (5.3%). The Patterson et al. (2008) results showed similar results (for PCDDs: there was a 14% increase; for PCDFs: there was a 1% decrease; for non-ortho PCBs: there was a 6% increase, and for mono–ortho PCBs there was a 19% decrease). The reduction of the contribution in mono–ortho PCBs was due to the down-weighting of the TEFs for the majority of mono–ortho PCBs. For example, PCB 156 contributed 1.2% to the serum TEQ– WHO05 compared to 14% to the serum TEQ–WHO98 (Table 5). These changes in the TEFs for the mono–ortho PCBs had a dramatic impact on which congeners were the major contributors to the TEQ. Seven congeners accounted for more than 80% of the serum TEQ–WHO98 (i.e., 2,3,7,8-TCDD, 1,2,3,7,8-PeCDD, 1,2,3,6,7,8HxCDD, 2,3,4,7,8-PeCDF, PCB 126, PCB 118, and PCB 156) in our study. These results are in agreement with the NHANES 2001– 2002 data (Patterson et al., 2008). With the new 2005 TEFs, both PCB 118 and PCB 156 dropped from the top seven contributors
730
B. Hong et al. / Chemosphere 76 (2009) 727–733
Table 3 Comparison between serum total TEQ–WHO98 and TEQ–WHO05 by age groups in the UMDES study and in NHANES 2001–2002 (unit: ppt). N
UMDES serum data Age 18–29 55 Age 30–44 200 Age 45–59 378 Age 60+ 313
Total TEQ based on WHO 1998 TEFs a
Total TEQ based on WHO 2005 TEFs 95% CI for P95
Mean
95% CI for mean
P95
9.5 18.6 32.2 54.4
8.5–10.4 17.0–20.1 29.1–35.2 49.6–59.3
13.6 28.2 66.5 106.7
12.1–21.1 26.0–47.1 48.2–96.1 97.2–128.1
34.7 56.7 102
29.9–40.7 46.4–82.6 86.9–114
NHANES 2001–2002 serum datac Age 20–39 443 Age 40–59 362 Age 60+ 389
b
% Change 95% CI for P95
In mean
In P95a
11.2 21.3 48.8 87.2
10.4–17.8 20.6–40.5 36.1–70.6 70.6–103.8
15.8 22.0 28.0 27.2
17.6 24.5 26.6 18.3
26.2 46.9 79.7
23.7–32.5 36.4–66.1 68.2–96.3
Mean
95% CI for mean
P95
8.0 14.5 23.2 39.6
7.2–8.7 13.3–15.8 21.1–25.4 35.8–43.4
a
b
24 17 22
Note: All results accounted for complex survey design. a P95 is the 95th percentile of the distribution. b The 95% CI (confidence interval) for P95 is calculated by R using method of Woodruff (1952). c From Patterson et al. (2008) Table 8.
Table 4 Mean percent contribution to the total TEQ for PCDDs, PCDFs, non-ortho PCBs and mono–ortho PCBs (unit: %). Percent contribution to the TEQ (%)
Serum PCDDs PCDFs Non-ortho PCBs Mono-ortho PCBs Household dust PCDDs PCDFs Non-ortho PCBs Mono-ortho PCBs House perimeter soil PCDDs PCDFs Non-ortho PCBs Mono-ortho PCBs
NHANES 2001–2002 dataa
UMDES data WHO 1998
WHO 2005
Changes
WHO 1998
WHO 2005
Changes
48.7 16.8 9.4 25.1
64.7 16.4 14.7 4.1
16.1 0.4 5.3 21.0
48.0 15.0 14.0 23.0
62.0 14.0 20.0 4.0
14.0 1.0 6.0 19.0
59.5 17.0 12.4 11.2
66.8 15.6 14.6 3.0
7.3 1.3 2.2 8.2
49.4 36.2 11.1 3.2
54.4 32.1 12.8 0.8
4.9 4.1 1.7 2.4
Note: All results accounted for complex survey design. a These results are from Patterson et al. (2008) Table 7.
and only contributed about 3% to the serum TEQ–WHO05 (Table 5 and Fig. 1). However, the TEQ–WHO05 was still dominated by 2,3,7,8-TCDD, 1,2,3,7,8-PeCDD, 1,2,3,6,7,8-HxCDD, 2,3,4,7,8-PeCDF, and PCB 126, which contributed 9.6%, 27%, 20%, 8.7% and 11% to the serum TEQ, respectively. The new two congeners among the top seven contributors to the TEQ–WHO05 were PCB 169 and 1,2,3,7,8,9-HxCDD, which contributed 3.3% and 3.2% to the serum TEQ, respectively. For household dust TEQ, the mean percent contribution of PCDDs increased about 7.3% in the TEQ–WHO05 compared to the TEQ–WHO98, while the contributions of mono–ortho PCBs decreased about 8.2% (Table 4). There were small changes in the contribution of the PCDFs (1.3%) and non-ortho PCBs (2.2%). For each specific congener, the percent contribution changed slightly within a range of 2.9 to 4.5% (Table 5). The largest impact was on OCDD, which contributed 4.5% more for household dust in the TEQ– WHO05 compared to the TEQ–WHO98 (Table 5), due to the 3-fold increase in the TEF from 0.0001 to 0.0003. The major contributors for the household dust TEQ–WHO05 were 1,2,3,7,8-PeCDD (9.8%), 1,2,3,6,7,8-HxCDD (9.8%), 1,2,3,4,6,7,8-HpCDD (31%), and PCB 126 (14%). For the house perimeter soil samples, the mean percent contribution of PCDDs increased 4.9%, while the contributions of PCDFs decreased 4.1% (Table 4). There were small changes in the contribution of the non-ortho PCBs (1.7%) and mono–ortho PCBs
(2.4%). The reduction of the contribution in PCDFs was mainly due to the down-weighting of 2,3,4,7,8-PeCDF (1998 TEF = 0.5, 2005 TEF = 0.3), which accounted for 18% of the soil TEQ–WHO98 but 12% of TEQ–WHO05 (Table 5). The major contributors for soil TEQ–WHO05 were 2,3,7,8-TCDD (15%), 1,2,3,7,8-PeCDD (15%), 1,2,3,4,6,7,8-HpCDD (10%), 2,3,4,7,8-PeCDF (12%), and PCB 126 (12%).
4. Discussion The mean total TEQ–WHO05 levels in serum, household dust, and soil samples in our study were reduced, by 26%, 12% and 14%, respectively, after the implementation of the 2005 TEFs compared to the 1998 TEFs (Table 2). The change in the serum TEQ was at the upper end of the range of 10–25% for the general population predicted by Van den Berg et al. (2006). Although these changes are appreciable, it should be kept in mind that laboratory variation in congener measurements may vary by roughly the same amount. We performed duplicate GC/MS analyses of our serum samples in cooperation with the National Center for Environmental Health laboratory and found that variation was within about two standard deviations (mean variation was 11% and maximum variation was 27%). Our results for the change in serum TEQ 95th percentile were quite similar to the
731
B. Hong et al. / Chemosphere 76 (2009) 727–733 Table 5 Mean percent contribution to the total TEQ for each WHO dioxin congener in UMDES study (unit: %). Compound
Serum
Household dust
WHO 1998 2378_TCDD 12378_PeCDD 123478_HxCDD 123678_HxCDD 123789_HxCDD 1234678_HpCDD OCDD 2378_TCDF 12378_PeCDF 23478_PeCDF 123478_HxCDF 123678_HxCDF 123789_HxCDF 234678_HxCDF 1234678_HpCDF 1234789_HpCDF OCDF PCB_77 PCB_81 PCB_126 PCB_169 PCB_105 PCB_114 PCB_118 PCB_123 PCB_156 PCB_157 PCB_167 PCB_189
WHO 2005
c
c
7.3 20.6c 2.0 14.6c 2.5 1.6 0.1 0.2 0.1 10.9c 2.2 2.1 0.4 0.5 0.4 0b 0b 0b 0b 8.6c 0.8 0.9 2.0 4.2c 0.1 14.4c 3.3 0.1 0.2
9.6 27.3c 2.6 19.5c 3.2c 2.1 0.4 0.3 0.1 8.7c 2.9 2.8 0.5 0.6 0.5 0.1 0b 0b 0b 11.4c 3.3c 0.4 0.2 1.7 0b 1.2 0.3 0.2 0.1
Changes
a
WHO 1998
WHO 2005
c
2.3 6.7 0.6 4.9 0.8 0.5 0.3 0.1 0b 2.2 0.7 0.7 0.1 0.1 0.1 0b 0b 0b 0b 2.8 2.5 0.5 1.9 2.5 0b 13.2 3.0 0.2 0.1
House perimeter soil
c
4.2 9.2c 1.5 9.3c 3.6 29.7c 2.1 1.3 0.5 6.4c 1.9 1.9 0.4 1.8 2.6 0.1 0.1 0.2 0b 12.1c 0.1 2.0 0.6 4.7c 0.1 3.1 0.7 0b 0b
4.5 9.8c 1.6 9.8c 3.8 30.7c 6.6c 1.4 0.3 4.2c 2.1 2.1 0.5 2.0 2.7 0.1 0.2 0.2 0b 14.1c 0.3 0.8 0.1 1.8 0b 0.2 0.1 0.1 0b
Changes
a
WHO 1998 c
0.3 0.6 0.1 0.5 0.2 1.1 4.5 0.1 0.2 2.2 0.2 0.2 0b 0.1 0.2 0b 0.2 0b 0b 2.0 0.2 1.3 0.6 2.9 0b 2.9 0.6 0b 0b
14.6 14.4c 1.8 4.5c 3.8c 9.6c 0.8 3.4 1.1 18.0c 3.1 3.5 0.7 2.7 3.3 0.2 0.1 0b 0b 10.9c 0.2 0.6 0.1 1.1 0b 1.1 0.3 0b 0b
WHO 2005 c
15.4 15.4c 1.9 4.8c 4.0c 10.2c 2.6 4.0 0.8 12.3c 3.5 3.8 0.8 2.9 3.6 0.2 0.2 0.1 0b 12.2c 0.5 0.2 0b 0.4 0b 0.1 0b 0b 0b
Changesa 0.8 1.0 0.1 0.3 0.3 0.6 1.8 0.6 0.4 5.7 0.4 0.3 0.1 0.3 0.3 0b 0.1 0b 0b 1.3 0.4 0.4 0b 0.7 0b 1.0 0b 0b 0b
Note: All results accounted for complex survey design. a Percent contribution to TEQ–WHO05 – percent contribution to TEQ–WHO98 (the italicised values indicate the largest change due to the 2005 TEFs). b Values less than 0.05. c Indicate the top seven contributed congeners to the TEQ.
UMDES serum data
Percent contribution to Total TEQ
60
40
20
Percent contribution to TEQ-WHO98
PCB_156
PCB_118
PCB_126
23478_PeCDF
123678_HxCDD
12378_PeCDD
2378_TCDD
PCB_156
PCB_118
PCB_126
23478_PeCDF
123678_HxCDD
12378_PeCDD
2378_TCDD
0
Percent contribution to TEQ-WHO05
Fig. 1. The percent contribution of seven important congeners to the serum total TEQ–WHO98 versus TEQ–WHO05 in UMDES data, shown using box plots. The plus sign indicates the arithmetic mean, the horizontal line across the box indicates the 50th percentile; the lower and upper margins of the box indicate the 25th and 75th percentiles, respectively; the lower and upper ticked lines extend to the 1st and 99th percentile, respectively; the stars show values below the 1st or above 99th percentile.
results reported by Patterson et al. (2008) based on the NHANES 2001–2002 data. The changes in TEFs will make comparisons be-
tween older publications (Flesch-Janys et al., 1995; DeVito et al., 1995; Wittsiepe et al., 2000; Millette, 2000) and newer publica-
732
B. Hong et al. / Chemosphere 76 (2009) 727–733
tions (based on TEQ–WHO05) (Wittsiepe et al. 2007) more challenging. In our study, the serum TEQs were calculated from both sets of TEFs (WHO 1998 TEFs and 2005 TEFs), and the full results for each of our study populations by age group are provided on the study website (www.umdioxin.org) in an effort to facilitate comparisons to past and future studies. The contribution of mono–ortho PCBs to the serum TEQ was reduced when the TEQ–WHO05 was compared to the TEQ–WHO98 (21%) (Table 4), due to the down-weighting of the majority of mono–ortho PCBs. The largest decrease in the serum TEQ was due to PCB 156, which alone accounted for 14% of the TEQ– WHO98 versus 1.2% of the TEQ–WHO05 (Table 5). The serum TEQ–WHO05 was still dominated by the five major congeners (2,3,7,8-TCDD, 1,2,3,7,8-PeCDD, 1,2,3,6,7,8-HxCDD, 2,3,4,7,8-PeCDF, and PCB 126), with the highest contribution coming from 1,2,3,7,8-PeCDD (27%). The impact of the changes in TEFs on total TEQ was not as substantial for soil as it was for serum in the present study. The 4.1% reduction in the contribution of PCDFs in soil (Table 4) was mainly due to the down-weighting of 2,3,4,7,8-PeCDF (the TEF decreased from 0.5 to 0.3), which was the largest contributor to TEQ–WHO98 (accounting for 18%). The contribution to the TEQ–WHO98 or TEQ–WHO05 of the other two PCDFs for which the TEFs changed (1,2,3,7,8-PeCDF and OCDF) was small (less than 1.5%). The impact of the changes in TEFs on total TEQ was small for household dust. The TEQ approach is intended to estimate exposure to dioxin mixtures via dietary intake. Interpreting TEQs from environmental matrices such as soil and household dust in settings where there is no ingestion may have no relevance to risk assessment unless the bioavailability and environmental fate and transport of specific congeners are taken into account. Because of this concern, we have provided the congener-specific data that underlies the TEQs and recommend that for purposes of risk assessment the congener specific concentrations be used. In the instance of household dust, where toddlers may ingest material, the TEQ may have direct relevance to risk assessment. It is also worthwhile to examine the percent contributions of the different groups of congeners (PCDDs, PCDFs, non-ortho PCBs or mono–ortho PCBs) to the total TEQs (Table 4). The mean percent contribution of PCDDs increased for all three types of samples for the TEQ–WHO05, compared to the TEQ–WHO98 (16% for serum, 7.3% for household dust, and 4.9% for soil). This was due to the proportional decrease in the contribution by the PCDFs and mono–ortho PCBs. For the PCDDs, the only change in TEFs in 2005 was for OCDD, which increased from 0.0001 in 1998 to 0.0003 in 2005. The contribution of the PCDFs was slightly reduced for serum (0.4%) and household dust (1.3%), but somewhat greater (4.1%) for soil. The changes in TEQ–WHO05 compared to the TEQ–WHO98 may have an impact in locations where regulations of soil contamination are triggered by specific TEQ levels. For example, the Michigan Department of Environmental Quality (MDEQ) residential soil direct contact criterion is 90 ppt TEQ (
ppt, 75th percentile = 8.5 ppt, 95th percentile = 23.8 ppt, maximum = 329.7 ppt), compared to 0.3% using the TEQ–WHO05 (mean = 6.9 ppt, median = 3.6 ppt, 75th percentile = 7.6 ppt, 95th percentile = 22.6 ppt, maximum = 186.2 ppt) (Demond et al., 2008). The change in TEFs had a substantial impact (1.8% versus 0.3%) on the proportion of properties above the regulatory limit in the latter population. In contrast, it had a small effect on the proportion of properties above this regulatory limit in the contaminated area (38.1% versus 35.8%). The changes in mean TEQ–WHO05 compared to the TEQ– WHO98 seen in serum were quite similar between the populations in Midland/Saginaw and the Jackson/Calhoun referent population (total TEQ: 26% to 23% for Midland/Saginaw populations and 28% for Jackson/Calhoun population; for PCDD TEQ: 0%–0.7% for Midland/Saginaw populations and 0.8% for Jackson/Calhoun population; for PCDF TEQ: 28% to 29% for Midland/Saginaw populations and 26% for Jackson/Calhoun population; for non-ortho PCB TEQ: 14% to 22% for Midland/Saginaw populations and 13% for Jackson/Calhoun population; for mono–ortho PCB TEQ: 88% to 87% for Midland/Saginaw populations and 87% for Jackson/Calhoun population), suggesting that the effect of the new TEFs on the TEQ is fairly stable across populations. This pattern may not be the same in settings where there is excessive exposure to a specific pattern of congeners that is different than what is commonly seen in the general population. The change in TEFs in 2005 was principally a reflection of updated information on toxicity of specific congeners. Acknowledgments Financial support came from The Dow Chemical Company through an unrestricted grant to the University of Michigan. The authors acknowledge Ms. Sharyn Vantine for her assistance, and Drs. Linda Birnbaum, Ronald Hites, Paolo Boffetta and Marie Haring Sweeney for their guidance as members of our Scientific Advisory Board. Drs. Garabrant and Franzblau declare they have competing financial interests. Drs. Garabrant and Franzblau have at times been retained as consultants and served as expert witnesses for the Dow Chemical Company. Appendix A WHO TEFs TEQ PeCDF TEQ–WHO98 TEQ–WHO05 TCDD OCDD OCDF UMDES PCDDs PCDFs PCBs HRMS ppt HVS3s ASTM LOD MDEQ NHANES
World Health Organization toxic equivalency factors toxic equivalent pentachlorodibenzofuran TEQ based on the WHO 1998 TEFs TEQ based on the WHO 2005 TEFs tetrachlorodibenzo-p-dioxin octachlorodibenzo-p-dioxin octachlorodibenzofuran University of Michigan Dioxin Exposure Study polychlorinated dibenzo-p-dioxins polychlorinated dibenzofurans polychlorinated biphenyls high resolution mass spectrometry parts per trillion high volume small surface samplers American Society for Testing and Materials limit of detection Michigan Department of Environmental Quality the National Health and Nutrition Examination Survey
B. Hong et al. / Chemosphere 76 (2009) 727–733
References ASTM, 2000. Standard Practice for Collection of Floor Dust for Chemical Analysis. D 5438-00, 1–7. American Society for Testing and Materials, West Conshohocken, PA. (ref. type: report). Demond, A., Adriaens, P., Towey, T., Chang, S.-C., Hong, B., Chen, Q., Chang, C.-W., Franzblau, A., Garabrant, D., Gillespie, B., Hedgeman, E., Knutson, K., Lee, C.-Y., Lepkowski, J., Olson, K., Ward, B., Zwica, L., Luksemburg, W., Maier, M., 2008. Statistical comparison of residential soil concentrations of PCDDs, PCDFs, and PCBs from two communities in Michigan. Environmental Science and Technology 42 (15), 5441–5448. doi:10.1021/es702554. DeVito, M., Birnbaum, L., Farland, W., Gasiewicz, T., 1995. Comparisons of estimated human body burdens of dioxinlike chemicals and TCDD body burdens in experimentally exposed animals. Environmental Health Perspectives 103, 820– 831. EPA, 1994. Method 8290, Polychlorinated Dibenzodioxins (PCDDs) and Polychlorinated Dibenzofurans (PCDFs) by High-Resolution Gas Chromatography/High-Resolution Mass Spectrometry (HRGC/HRMS). US Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington, DC. EPA, 1999. Method 1668, Revision A: Chlorinated Biphenyl Congeners in Water, Soil, Sediment, and Tissue by HRGC/HRMS. US Environmental Protection Agency, Office of Water, Washington, DC. Flesch-Janys, D., Berger, J., Gurn, P., Manz, A., Nagel, S., Waltsgott, H., Dwyer, J., 1995. Exposure to polychlorinated dioxins and furans (PCDD/F) and mortality in a cohort of workers from a herbicide-producing plant in Hamburg, Federal Republic of Germany. The American Journal of Epidemiology 142, 1165–1175. Garabrant, D.H., Franzblau, A., lepkowski, J., Gillespie, B.W., Adriaens, P., Demond, A., Ward, B., LaDronka, K., Hedgeman, E., Knutson, K., Zwica, L., Olson, K., Towey, T.,
733
Chen, Hong, B., 2008. The University of Michigan dioxin exposure study: methods for an environmental exposure study of polychlorinated dioxins, furans and biphenyls. Environmental Health Perspectives. doi:10.1289/ ehp.11777.