Article pubs.acs.org/est
Temporal Changes of PBDE Levels in California House Cats and a Link to Cat Hyperthyroidism Weihong Guo,*,† Stephen Gardner,‡ Simon Yen,§ Myrto Petreas,† and June-Soo Park† †
California Department of Toxic Substances Control, California Environmental Protection Agency, 700 Heinz Avenue, Berkeley, California 94710, United States, ‡ VCA Albany Animal Hospital, 1550 Solano Avenue, Albany, California 94707, United States § Campus Veterinary Clinic, 1807 M.L.K. Jr Way, Berkeley, California 94709, United States S Supporting Information *
ABSTRACT: In this study, we measured serum PBDE levels in California (CA) house cats during two time periods: 2008−2010 and 2012−2013 to assess the impacts of the decline in use of these materials after the bans. The median ∑19PBDE level in CA household cats (age ≥10 yr) was 3479 ng/g lipid in 2008−2010 (1st time period, n = 21) and 1518 ng/g lipid in 2012− 2013 (2nd time period, n = 22), about 2 times lower than in the first time period (p = 0.006). In contrast, PCB and OCP levels showed no statistically significant changes. With better matched group size and age (HT = 11 vs non-HT = 11, age ≥10 yr) in the second time period, we found that ∑19PBDE level (mean ± SE ng/g lipid) was significantly higher in the HT group (3906 ± 1442) than those in the non-HT group (1125 ± 244) (p = 0.0030). Higher levels of PCBs and OCPs were also found in HT group. Despite the declines of PBDE levels, our findings indicate that the current levels of PBDEs, as well as PCBs and OCPs, may still pose health effects for house cats and, possibly, humans.
■
INTRODUCTION Largely manufactured since the 1970s, polybrominated diphenyl ethers (PBDEs) were used in various products and building materials for alleged fire safety.1 Accumulated scientific evidence has shown that PBDEs may cause health problems such as thyroid homeostasis disruption, neurodevelopmental deficits, reproductive changes, and even cancer in animals.2,3 Recent human studies have also reported that PBDEs are linked to lower birth weight, behavior problems in youth, and altered reproductive function in adults,4,5 confirming results from animal studies. Due to these health concerns, penta-BDE and octa-BDE mixtures were banned in 2004 in the E.U.6 and the United States.7 Deca-BDE was banned in 2008 in the E.U.,8 and voluntarily phased out in 2013 in the United States.9 In our earlier study,10 we reported extremely high PBDE levels (∑19PBDEs median = 2903 ng/g lipid, max = 22 537 ng/ g lipid) in the serum of California (CA) house cats (aged between 3 yr −20 yr) sampled between 2008 and 2010 (n = 26). Levels of major congeners (BDE-47, -99, -100, -153) in cats’ serum (∑4PBDEs median = 2647 ng/g lipid) were about 60 times higher than the levels in CA residents (∑4PBDEs median = 43.2 ng/g lipid) selected from NHANES (sampled in 2003−2004), who themselves were among the highest human levels in the world probably due to CA’s unique fire safety standards.11,12 In addition, our cats’ PBDE levels far exceeded levels of legacy contaminants−polychlorinated biphenyls (PCBs) and organochlorine pesticides (OCPs). We suggested © 2015 American Chemical Society
that house dust was a major source of PBDE exposures. As new policies started to be implemented in CA almost 10 years ago to ban and restrict PBDE commercial mixtures, our recent human population study reported that serum PBDE levels declined 39% from 2008−2009 to 2011−2012 in CA women.13 Similar PBDE declines were also observed in other studies from Australia, European, and Asian countries in humans and in fish.14−16 However, an uncertain trend (downward, then upward) was found in marine animals (shrimp, common carp, and yellow catfish) from East China in 2009−2012.17 Inconsistent changes were also reported in CA house dust studies.18,19 Cat hyperthyroidism (HT), first reported as a new disease in 1979, has become a major health problem among middle-aged and older household cats and is a leading cause of cat morbidity and mortality worldwide.20,21 HT is a multisystem disorder resulting from excessive circulating concentrations of thyroxine (T4) and triiodothyronine (T3). Causes of the disease are not well understood, but it is believed that multiple factors including nutritional imbalance (e.g., iodine or selenium deficiencies or excesses), indoor living (PBDE exposure from dust), canned food, and drinking water contaminated with Received: Revised: Accepted: Published: 1510
September 1, 2015 December 16, 2015 December 24, 2015 December 24, 2015 DOI: 10.1021/acs.est.5b04252 Environ. Sci. Technol. 2016, 50, 1510−1518
Article
Environmental Science & Technology Table 1. Cat Demographic Information Obtained from Questionnaires (mean ± S.D. and range) number total male female HT non-HT indoor outdoor
22 10 12 11 11 17 5
(45%) (55%) (50%) (50%) (77%) (23%)
age (yr) 13.0 12.8 13.1 13.1 12.8 12.9 13.1
± ± ± ± ± ± ±
2.0 1.8 2.3 1.9 2.2 2.1 1.9
weight (lbs)
(10.0−17.0) (10.0−15.0) (10.0−17.0) (10.5−16.0) (10.0−17.0) (10.0−17.0) (11.4−16.0)
10.4 11.5 9.5 9.5 11.4 10.9 8.9
± ± ± ± ± ± ±
2.6 2.5 2.5 2.4 2.7 2.7 2.1
(6.3−15.1) (6.7−15.1) (6.3−13.2) (6.3−12.6) (6.7−15.1) (6.4−15.1) (6.3−11.6)
canned food consumption (oz) 4.2 5.2 3.3 2.5 5.8 4.8 2.2
± ± ± ± ± ± ±
2.7 3.1 2.0 1.0 2.8 2.8 0.8
(1.0−9.0) (1.0−9.0) (1.0−7.0) (1.0−4.0) (1.0−9.0) (1.0−9.0) (1.0−3.0)
Sample Collection. Whole blood was collected by veterinarians in plastic serum separation tubes (SSTs) containing clot activator gel (Becton Dickinson, catalog# 367985). Blood samples were allowed to clot at room temperature for at least 30 min and then centrifuged at 2000 rpm for 15 min to achieve serum separation. The centrifuged SSTs were frozen and delivered to the Environmental Chemistry Laboratory (ECL) of the CA Department of Toxic Substances Control (DTSC) where they were stored at −20 °C until analysis.37 Serum Extraction and Instrumental Analysis. Sample extraction, cleanup, and instrumental analysis procedures were as reported in our previous study with a few modifications.10 Briefly, thawed cat serum samples (1−2 mL) were fortified with a panel of 13C12- labeled surrogate standards (Table S2) and mixed well. Then the mixture was treated with 6 M hydrochloric acid (1 mL) and 70% isopropanol (6 mL) to denature proteins. Analytes of interest were extracted with a 1:1 v/v mixture of hexane and methyl t-butyl ether. After phase separation, the neutrals (nonpolar compounds) were cleanedup through activated and acidified silica cartridges (500 °C prebaked, manually packed, 3 cm3) on a Biotage RapidTrace system (Biotage, Sweden). The collected final eluates in hexane/dichlomethane (1:1) were concentrated in TurboVap (Biotage, Sweden), and spiked with recovery standard (13C12 PCB-209). Standard reference material (NIST SRM1958) and bovine serum prespiked with known amounts of target analytes (in-house quality control samples) were used as QA/QC samples. The 19 PBDEs (BDE-17, -28, -47, -66, -85, -99, -100, -153, -154, -183, -196, -197, -201, -202, -203, -206, -207, -208, −209), 15 PCBs (PCB-66, -74, -99, -101, -105, -118, -138, -153, -156, -170, -180, -183, -187, -194, -203), and 7 OCPs (2,4′-DDT, 4,4′-DDE, 4,4′-DDT, hexachlorobenzene, oxychlordane, transnonachlor, and β-BHC) were measured by gas chromatography/high resolution mass spectrometry (GC-HRMS, DFS, ThermoFisher, Bremen, Germany) using isotope dilution. The same sample extracts were injected in two separate modes: for PCB and OCP analyses, 2 μL of extract were injected in splitless mode and separated using a HT8-PCB column (60 m × 0.25 mm I.D., 0.25 μm film thickness, SGE International Pty Ltd. Australia & Pacific Region) with helium carrier gas. For PBDE analysis, 2 μL of extract were injected by PTV injector and separated using a DB-5MS column (15 m × 0.25 mm I.D., 0.10 μm film thickness, Agilent J&W, U.S.A.) with helium carrier gas. The MS was operated in electron impact ionization mode using multiple ion detection. For both analyte groups, the source temperature was set to 260 °C, ionization energy was set to 42 V, and electron current was typically 0.7 mA, with a mass resolution power of 10,000. Perfluorokerosene (PFK) was used as the mass reference.
thyroid-disrupting compounds (e.g., bisphenol A from canned food) are the major players based on several cat studies.22−26 PBDEs are endocrine disruptors, affecting thyroid homeostasis (as are some PCBs and OCPs) due to their structural similarities to T4.27,28 PBDEs are one of the major groups of indoor contaminants29−31 and researchers have studied the links of PBDEs with cat HT based on the biological plausibility and the timing of the HT pandemic, which occurred shortly after PBDE use became widespread.32−35 However, most studies have found no strong associations between higher serum PBDE levels and cat HT due to various reasons (e.g., small samples size, high variation in cat groups). Only one recently published study from Sweden with relatively large sample size (n = 60, 2 yr −18 yr) showed significantly higher BDE-99, -153, and -183 levels in cats with thyroid disorders after adjusting for cat’s age.36 In the present study, we sampled 22 older cats with age ≥10 yr during 2012−2013 (referred to as the second time period) to compare their PBDE levels to those we reported in our earlier study (sampled in 2008−2010 and referred to as the first time period).10 We also compared PBDE as well as PCB and OCP levels based on their thyroid status (HT vs non-HT) to better understand the possible links between serum contaminant levels and cat HT.
■
MATERIALS AND METHODS Study Subjects. To follow-up on our earlier study,10 we recruited 22 new cats (age ≥10 yr) from 3 veterinary clinics in 2012−2013 in the CA San Francisco Bay region. HT diagnoses were made by veterinarians based upon physical appearance and total T4 values. The 22 cats were divided into 2 groups according to their thyroid status: non-HT cats (n = 11) and HT cats (n = 11). Blood samples from non-HT cats were collected during routine physical exams or during treatment for nonthyroid problems (e.g., dental disease, wounds, heart murmur, kidney disease, liver disease, etc.). A consent form and short questionnaire were filled out by the cat’s owner to provide basic information about cat’s demographic characteristics (Table 1 and Table S1). Table 1 shows the summary statistics of the demographic characteristics. Among the 22 cats, there were 10 (45%) males and 12 (55%) females from 10 yr to 17 yr. Their weight range was 6.3 lbs to 15.1 lbs with lower average weight in the HT group than in the non-HT group (9.5 lbs vs 11.4 lbs, p = 0.0915). All HT cats had T4 values, while only 5 (one indicated of within normal limit) out of the 11 non-HT had T4 values. Most cats were kept indoors except 4 from the HT group and 1 from the non-HT who spent 3−12 h per day outdoors. All cats ate various amounts of canned cat food daily supplemented with dried cat food. It appears that canned food intake is quite different between male and female, between HT and non-HT, and between indoor and outdoor cats. 1511
DOI: 10.1021/acs.est.5b04252 Environ. Sci. Technol. 2016, 50, 1510−1518
Article
Environmental Science & Technology
Table 2. Summary Statistics of Major PBDE, PCB and OCP Levels (ng/g lipid) in 2 Time Periods (2008−2010 and 2012− 2013). PBDE Levels Decreased Significantly while PCB and OCP Levels Remained the Same compound age (yrs) mean ± S.E median (min−max) ∑19PBDEs mean ± S.E median (min−max) BDE-47 mean ± S.E median (min−max) BDE-99 mean ± S.E median (min−max) BDE-100 mean ± S.E median (min−max) BDE-153 mean ± S.E median (min−max) BDE-154 mean ± S.E median (min−max) BDE-183 mean ± S.E median (min−max) BDE-197 mean ± S.E median (min−max) BDE-207 mean ± S.E median (min−max) BDE-209 mean ± S.E median (min−max)
first time perioda,d (n = 21)
second time periode (n = 22)
14.2 ± 0.498 14.0 (11.0−20.0)
13.0 ± 0.436 13.0 (10.0−17.0)
4510 ± 1038 3479 (631−22537)
2516 ± 775 1518 (231− 16568)
0.0060c
960 ± 258 622 (49.3−4941)
604 ± 238 289 (19.8−5321)
0.0307c
2232 ± 623 1674 (176−13 554)
1118 ± 396 626 (89.8−7880)
0.0080c
107 ± 42.6 34.1 (7.77−826)
39.8 ± 12.5 18.6 (1.11−257)
0.0172c
351 ± 81.0 317 (21.2−1695)
162 ± 59.8 72.5 (12.0−1277)
0.0063c
184 ± 48.8 136 (13.1−1046)
95.4 ± 35.4 45.3 (7.51−698)
0.0123c
23.9 ± 3.68 18.6 (4.53−74.4)
45.2 ± 30.0 10.8 (0.00−673)
0.2808
29.8 ± 7.92 18.1 (5.00−153)
50.3 ± 23.0 17.8 (0.00−520)
0.7900
78.4 ± 29.4 38.0 (13.2−627)
85.4 ± 16.2 52.4 (15.0−260)
0.8669
452 ± 153 227 (90.6−3373)
235 ± 42.1 170 (0.00−720)
0.0422c
p-valueb
compound ∑15PCBs mean ± S.E median (min−max) PCB-99 mean ± S.E median (min−max) PCB-101 mean ± S.E median (min−max) PCB-118 mean ± S.E median (min−max) PCB-138 Mean ± S.E median (min−max) PCB-153 mean ± S.E median (min−max) PCB-180 mean ± S.E median (min−max) PCB-187 mean ± S.E median (min−max) 4,4′-DDT mean ± S.E median (min−max) 4,4′-DDE mean ± S.E median (min−max) t-nonachlor mean ± S.E median (min−max)
first time periodd (n = 21)
second time periode (n = 22)
240 ± 31.4 227 (61.2−566)
293 ± 63.1 206 (56.3−1497)
0.7307
20.2 ± 2.22 18.8 (7.71−43.7)
23.2 ± 3.89 16.9 (5.10−75.8)
0.6084
19.9 ± 1.90 19.9 (6.70−41.2)
26.2 ± 4.59 20.6 (5.65−85.2)
0.7541
21.1 ± 2.58 20.2 (0.00−42.9)
30.5 ± 8.39 19.9 (6.63−198)
0.7490
31.4 ± 5.94 29.2 (0.00−91.3)
45.5 ± 13.0 28.0 (9.43−304)
0.7029
58.4 ± 9.25 51.7 (7.90−158)
63.5 ± 11.9 46.8 (12.7−273)
0.7354
28.4 ± 4.87 19.2 (0.00−78.4)
28.5 ± 5.72 19.1 (6.12−122)
0.3352
18.8 ± 3.16 14.6 (0.00−49.6)
18.1 ± 2.41 15.3 (4.59−46.1)
0.2695
64.6 ± 20.1 28.9 (0.00−365)
198 ± 83.6 64.1 (0.00−1721)
0.4275
618 ± 109 376 (106−1746)
454 ± 104 338 (84.3−1962)
0.1015
385 ± 297 65.0 (0.00−6315)
193 ± 83.6 26.5 (5.89−1556)
0.1074
p-valueb
Data (cats ≥10 yr) seltected from previous study (Guo et al). bp-values were determined from 1 tailed t-test. cp < 0.05. dDetection frequencies are 100%, except for PCB-118 (95%), PCB-138 (86%), PCB-180 (95%), PCB-187 (91%), and 4,4′-DDT (57%) in 1st time period. eDetection frequencies are 100%, except for BDE-183 (91%), BDE-197 (95%), and BDE-209 (95%), and 4,4′-DDT (95%) in 2nd time period. a
Lipid and T4 Measurements. Serum samples were sent to CERLab of the Department of Laboratory Medicine in Boston Children’s Hospital (Boston, MA, U.S.A.) for cholesterol (CHOL) and triglyceride (TG) measurements. The lipidnormalized concentrations (ng/g lipid) were calculated from CHOL and TG measurements using Phillip’s formula,38 which has been validated for human population studies. Serum samples were sent to IDEXX Reference Laboratories, Inc. for total T4 measurement using DRI Thyroxine (T4) Assay kit (Microgenics Corporation, Freemont, CA, U.S.A.). Data Analysis. We used the same statistical analysis procedures as we did for the first time period. We used “0” values to replace the concentrations below MDL. The lipidnormalized concentrations (ng/g lipid) were used throughout the data analysis. Due to their skewed distributions, all measurement data were log-transformed to achieve normal distributions, as shown by Chi-square goodness of fit test. The t-test and Pearson’s correlation test were performed on the logtransformed data using STATA version 11.0 (StataCorp LP) to assess: (1) different POP levels between the two time periods; (2) different POP levels between cat groups (HT vs non-HT;
male vs female; indoor vs outdoor); and (3) correlations between POP levels and cat’s age, weight, and T4 values. We set the level of significance as α < 0.05. We coded cats as “indoor cats” if they spent less than 3 h outdoors. For canned food consumption, we converted number of cans to ounces (oz). There are 2 sizes of canned cat food available on the market, 5.5 and 3 oz, respectively. If the canned food size was not specified on the questionnaire form, then we used the average size of 4 oz per can to determine the daily canned food intake.
■
RESULTS AND DISCUSSION PBDE Levels Decreased over Time in CA Cats. Summary statistics for PBDEs, PCBs, and OCPs measured from the two time periods are shown in Table 2. Major PBDEs (BDE-47, -99, -100, -153, -154, -183, -197, -207, and -209), PCBs (PCB-99, -101, -118, -138, -153, -180, and -187) and OCPs (4,4′-DDT, 4,4′-DDE, and t-nonachlor) are listed and detection frequencies are mostly 100% (Table 2, Table S3). Levels of BDE-47 (p = 0.0307), -99 (p = 0.0080), -100 (p = 0.0172), -153 (p = 0.0063), -154 (p = 0.0123), -209 (p = 1512
DOI: 10.1021/acs.est.5b04252 Environ. Sci. Technol. 2016, 50, 1510−1518
Article
Environmental Science & Technology
Figure 1. PBDE congener profiles (based on median values) in CA house dust, cats (1st and 2nd time periods), and humans (CA breast milk). Markedly different PBDE congener patterns in house cats with BDE-99 and BDE-209 as significant congeners, whereas the major congeners in humans were BDE-47 > BDE-153 > BDE-99 > BDE-100 > BDE-154.
0.0422), and ∑19PBDE (p = 0.0060) significantly decreased between the two time periods (Table 2). The median ∑19PBDE level in CA household cats (age ≥10 yr) were 3479 ng/g lipid with a range of 631 to 22 537 ng/g lipid in first time period and 1518 ng/g lipid with a range of 231 to 16 568 ng/g lipid in second time period. Since we had more HT cats in the first time period compared to the second time period (16 cats vs 11 cats), to eliminate possible bias of HT status on the temporal comparison, we also compared PBDE levels in each of the HT and non-HT cat groups between the two time periods. The median of the ∑19PBDE level in HT cats was 3,479 ng/g lipid with a range of 631 to 9491 ng/g lipid in the first time period and 2156 ng/g lipid with a range of 711 to 16 568 ng/g lipid in the second time period, i.e., we observed an approximate 38% decline. The median of the ∑19PBDE level in non-HT cats was 3742 ng/g lipid with a range of 874 to 22 537 ng/g lipid in the first time period) and 981 ng/g lipid with a range of 231 to 3276 ng/g lipid in the second time period), i.e., about a 74% decline. The consistent findings in both HT and non-HT cat groups as well as the combined cat group confirmed the declines reported in our previous human population study.13 These declines in human and house cats are probably due to the ban of penta-BDE and octa-BDE in 2004, as well as the voluntary ban/restricted production of deca-BDE in CA,39 suggesting the efficacy of regulatory policies. A similar decline was also reported in 12 pools of Australian human milk collected between 1993 and 2009.16 In that study, PBDE levels peaked in 2002−2004 and then decreased in 2007−2009. Two studies in South China and Finland found that PBDE levels decreased in fish, dropping back to levels found in the 1970s.14,15 Nonetheless, CA cat serum PBDE levels are still consistently higher than those in cats living elsewhere and sampled around the same time period.32 As we hypothesized in our earlier report,10 house dust may be the major source of PBDE exposure in household cats due to the time the cats spend on/ near the floor and grooming. CA house dust has among the highest PBDE levels in the world,40 resulting from CA’s unique flammability standard (Technical Bulletin 117).11 We expect that PBDE levels will continue to decrease in house dust18 and biological matrices13 because of the change of TB117 to TB117−201341 and other policy changes. However, exposures will continue from household products containing PBDEs still
in use. Therefore, like other legacy contaminants (e.g., PCBs), PBDEs may remain at detectable levels in environmental and biological systems for a long period of time. Moreover, the temporal trend in household cats may reflect the CA indoor environment exposure changes in humans and especially in small children, who share many similarities with house cats (e.g., small body size, higher dust ingestion through hand-tomouth contact, and time spend on the floor). In contrast to PBDEs, the PCB and OCP levels remained about the same between the two time periods (all p > 0.05) (Table 2). The median concentration of ∑15PCB was 206 ng/g lipid (2nd time period) vs 227 ng/g lipid (1st time period) and the median concentration of 4,4′-DDE was 338 ng/g lipid (2nd time period) vs 376 ng/g lipid (1st time period). Most of the PCB and OCP levels were strongly correlated whereas PBDE levels showed weaker correlations with OCP levels. Most of the PBDE levels (BDE-47, -99, -100, -153, and -154) were moderately or strongly correlated with PCB levels, which was not the case in the first time period (data not shown). This indicates that the role of other sources of exposure in addition to house dust (e.g., diet) may become greater as exposures from indoor sources of PBDE decrease due to the PBDE bans and as PBDEs get established in the food web. Cat PBDE, PCB, and OCP Levels and Congener Profiles in Comparison to Those of Humans. The median of ∑19PBDE level (1518 ng/g lipid) is about 7 times higher than the median of ∑15PCB (206 ng/g lipid) and 4 times that of 4,4′-DDE (338 ng/g lipid) even with decreased PBDE levels in the second time period. PBDE levels are much higher than those of PCBs in house dust18,19 and this suggests that house dust is still a major contributor to higher PBDE levels in house cats through dust ingestion (grooming, time spent on floor and furniture), confirming our earlier report.10 Comparing cats’ levels with CA first-time mothers sampled at similar time period42 (2009−2012, n = 67), we found that (1) the median of ∑5PBDE (BDE-47, -99, -100, -153, and -154) was more than 30 times higher in cats (1064 ng/g lipid vs 30.8 ng/g lipid); (2) the median of ∑15PCB (PCB-66, -74, -99, -101, -105, -118, -138, -153, -156, -170, -180, -183, -187, -194, and -203, was about 6 times higher in cats (206 ng/g lipid vs 34.0 ng/g lipid); (3) median of 4,4′-DDE (338 ng/g vs 97.5 ng/g lipid) was about 3 times higher in cats. This is similar to our report for the first time period. The higher PBDE levels in 1513
DOI: 10.1021/acs.est.5b04252 Environ. Sci. Technol. 2016, 50, 1510−1518
Article
Environmental Science & Technology Table 3. Major PBDE, PCB, and OCP Levels in HT and non-HT Cats in 2nd Time Period (ng/g lipid) compound age (yrs) mean ± S.E median (min−max) ∑19PBDEs mean ± S.E median (min−max) BDE-47 mean ± S.E median (min−max) BDE-99 mean ± S.E median (min−max) BDE-100 mean ± S.E median (min−max) BDE-153 mean ± S.E median (min−max) BDE-154 mean ± S.E median (min−max) BDE-183 mean ± S.E median (min−max) BDE-197 mean ± S.E median (min−max) BDE-207 mean ± S.E median (min−max) BDE-209 mean ± S.E median (min−max) a
HT (n = 11)
non-HT (n = 11)
p-valuea
13.1 ± 0.586 13.0 (10.5−16.0)
12.8 ± 0.672 13.0 (10.0−17.0)
3906 ± 1442 2156 (711−16568)
1125 ± 244 981 (231−3276)
0.0030b
933 ± 455 468 (126−5321)
276 ± 100 180 (19.8−1220)
0.0163b
1775 ± 747 1000 (202−7880)
461 ± 122 337 (89.8−1477)
0.0147b
57.3 ± 22.8 26.5 (11.6−257)
22.3 ± 8.60 14.8 (1.11−98.8)
0.0190b
261 ± 114 121 (27.9−1277)
62.1 ± 11.2 70.8 (12.0−126)
0.0103b
155 ± 67.1 59.2 (18.5−698)
36.4 ± 8.30 26.8 (7.51−95.0)
0.0096b
81.9 ± 59.3 15.9 (5.86−673)
8.52 ± 2.00 8.98 (0.00−21.1)
0.0244b
81.7 ± 44.6 38.7 (7.26−520)
18.8 ± 5.07 13.7 (0.00- 62.7)
0.0448b
122 ± 27.2 85.7 (25.1−260)
49.0 ± 9.80 45.1 (15.0−135)
0.0155b
319 ± 70.9 207 (87.9−720)
150 ± 31.5 128 (0.00−383)
0.0418b
compound ∑15PCBs mean ± S.E median (min−max) PCB-99 mean ± S.E median (min−max) PCB-101 mean ± S.E median (min−max) PCB-118 mean ± S.E median (min−max) PCB-138 mean ± S.E median (min−max) PCB-153 mean ± S.E median (min−max) PCB-180 mean ± S.E median (min−max) PCB-187 mean ± S.E median (min−max) 4,4′-DDT mean ± S.E median (min−max) 4,4′-DDE mean ± S.E median (min−max) t-nonachlor mean ± S.E median (min−max)
HT (n = 11)
non-HT (n = 11)
p-value
397 ± 118 254 (129−1497)
188 ± 25.7 189 (56.3−396)
0.0239b
31.7 ± 6.79 20.8 (10.7−75.8)
14.8 ± 1.74 15.8 (5.10−26.1)
0.0087b
35.4 ± 8.23 27.7 (5.65−85.2)
17.0 ± 1.94 16.4 (7.65−28.1)
0.0434b
42.8 ± 16.1 22.8 (14.1−198)
18.1 ± 2.50 18.3 (6.63−34.0)
0.0193b
64.3 ± 25.0 34.7 (17.9−304)
26.7 ± 3.32 27.6 (9.43−50.5)
0.0416b
83.9 ± 21.8 53.3 (23.1−273)
43.1 ± 5.95 43.8 (12.7−88.5)
0.0313b
35.6 ± 10.4 19.5 (8.28−122)
21.5 ± 4.42 17.7 (6.12−57.9)
0.1453
22.0 ± 3.97 19.9 (7.43−46.1)
14.3 ± 2.36 13.9 (4.59−32.8)
0.0559
336 ± 159 150 (0.00−1721)
60.0 ± 14.3 40.7 (9.81−166)
0.0452b
587 ± 196 349 (84.3−1962)
322 ± 54.8 257 (102−683)
0.2221
365 ± 153 79.3 (6.07−1556)
20.6 ± 3.14 20.9 (5.89−36.9)
0.0059b
p-values were determined from 1 tailed t-test. bp < 0.05.
and compared PBDE, PCB, and OCP levels based on their thyroid status. Lipid concentrations (mg/mL, mean ± S.D.) were 5.25 ± 1.44 and 5.24 ± 1.13, respectively, in HT and nonHT groups, and they were basically identical between the two cat groups (p = 0.9785), therefore, they are not subject to any differences to lipid normalized POP levels In general, the HT group showed significantly higher PBDE, PCB, and OCP levels (all p < 0.05, except for PCB-180, -187, and 4,4′-DDE) (Table 3). The median concentrations of ∑19PBDE, ∑15PCB, and 4,4′-DDE were 2156 ng/g lipid, 254 ng/g lipid, and 349 ng/g lipid, respectively in the HT group. The median concentrations of ∑19PBDE, ∑15PCB, and 4,4′DDE were 981 ng/g lipid, 189 ng/g lipid, and 257 ng/g lipid, respectively in non-HT group (Table 3). Better matched sample size for each group enhanced the power to observe differences between the two groups. This is consistent with the results found in Dye et al.,33 who first reported a link between PBDEs and cat HT. However, due to limited subjects and sample variations in each cat group, statistical significance was not established in that study. Clinical total T4 values were available from all 11 HT cats and 5 (one indicated of within normal limit) out of the 11 nonHT cats From these limited numbers, we found that most of the PBDE levels (but not PCBs and OCPs) were correlated with T4 values (corr. coef. ∑19PBDE = 0.6718, p = 0.0061,
cats than in humans probably result from greater amounts of house dust ingestion, and lack of hepatic glucuronidase that may slow PBDE detoxification and increase bioaccumulation.35,43 The higher PCB and OCP levels in cats may mainly be attributed to their limited diet through processed food (both canned and dried food) and lack of hepatic glucuronidase. While PCB congener patterns were similar between CA house cats and humans (PCB-153 predominating, Figure S2), PBDE congener patterns differed markedly (Figure 1). These congener patterns from CA cats remain unchanged from the first time period and are similar to those found in house dust, where BDE-99 and BDE-209 were major congeners,10,19 confirming that house dust is still the primary source of PBDE exposure to house cats. In contrast, humans had low BDE-209 detection frequencies (
