Document K6967ZyB7J3oj9gMVLZrXJ49X

ORIGINAL D O E S N O T C O N TA IN C B I Taft/ Taft Stettinius & Hollister LLP 425 Walnut Street, Suite 1800 / Cincinnati, OH 45202-3957 / Tel: 513.381 2838 / Fax. 513.381 0205 / www.taftlaw.com Cincinnati / Cleveland / Columbus / Dayton / Indianapolis / Northern Kentucky / Phoenix Robert A. B ilott 513.357.9638 bllott@taftlaw com 8EHQ-1 3-18966 January 9, 2013 TSCA Confidential Business Information Center (7407M) EPA East - Room 6428, Attn: Section 8(e)/FYI U.S. Environmental Protection Agency 1200 Pennsylvania Avenue, NW Washington, DC 20460-0001 Re: TSCA Section 8(e) And FYI Submission Re: PFOA - Human Health Effects To TSCA 8(e)/FY! Database: W e hereby submit the following information to USEPA, pursuant to its TSCA Section 8(e) and "FYI" submission procedures, providing information relating to human health effects linked to exposure to perfluorooctanoate acid ("PFOA," a/k/a "C-8"): 1. Javins, B., Pregnancy etal., in the "CC8ircHuelaatlitnhgSMtuadteyr,"nEalnvPierorfnlu. oSrcoia. l&kylTeScuhb,s(tdaoni:ces during 10.1021 /es3028082) (on-line Jan. 7, 2013); etal.,2. Dong, G-H., "Serum Polyfluoroalkyl Concentrations, Asthma Outcomes, Environ. Health Perspec.and Immunological Markers in a Case-Control Study of Taiwanese Children," (dx.doi.org/10.1289/ehp.1205351) (on-line Jan. 8, 2013); and 3. et a!.,Vieira, V.M., "Perfluorooctanoic Acid Exposure Contaminated Community: A Geographic Analysis," EanndvirCona.ncHeeraOlthutPcoemrspeseci.n a (dx.doi.org/10.1289/ehp. 1205829) (on-line Jan. 8, 2013). Very truly yours, RAB:mdm Enclosures 13235539 1 ' R o b e r t A. Bilott 1 13 0 0 0 124 CONTAINS NO CBI Circulating Maternal Perfluoroalkly Substances during Pregnancy in the C8 Health Study Beth Javins, Gerald Hobbs, Alan M. Ducatman, Courtney Pilkerton, Danyel Tacker, and Sarah S Knox Environ. Sci. Techno!., Just Accepted Manuscript DOI: 10.1021/es3028082 * Publication Date (Web): 28 Dec 2012 Downloaded from http://pubs.acs.org on January 7,2013 Just Accepted I "Just Accepted" manuscripts have been peer-reviewed and accepted for publication. They are posted j online prior to technical editing, formatting for publication and author proofing. The American Chemical j Society provides 'Just Accepted" as a free service to the research community to expedite the I dissemination of scientific material as soon as possible after acceptance. "Just Accepted" manuscripts i appear in full in PDF format accompanied by an HTML abstract. "Just Accepted" manuscripts have been j fully peer reviewed, but should not be considered the official version of record. They are accessible to all j readers and citable by the Digital Object Identifier (DOI). "Just Accepted" is an optional service offered j to authors. Therefore, the "Just Accepted" Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the "Just Accepted" Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and ail legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors | dr consequences arising froth the use of information contained in these "Just Accepted" manuscripts. ACS Publications ?K.-v/tnopvgy p {.ythv' V*i-KvmChe 11,\fj, [>" 200`>i Published by Am erican C hem ical Society Copyright > Am erican Chem ical Society How ever, no copyright claim is m ade to original U .S G overnm ent works, or works produced by em ployees o f any Com m onwealth realm C rown governm ent in the course o f their duties. Page 1 of 28 Environmental Science & Technology 1 2 3 4 5 6 7 Circulating Maternal Perfluoroalkyl Substances 8 9 10 11 12 during Pregnancy in the C8 Health Study 13 14 15 16 17 Beth Javins, Gerald Hobbs, Alan M. Ducatman, Courtney Filkerton, Darnel Tacker, Sarah S. 18 19 Knox* 20 21 22 23 West Virginia University, School of Public Health, Department of Occupational and 24 25 Environmental Sciences, P.O. Box 9190, Morgantown, W.V. 26506-9190, U.S.A,, Phone: 26 27 (304)293-1058, Fax: (304)293-6685, sknox@hsc.wvu.edu 28 29 30 31 Keywords: Pregnancy, Perfluoroalkyl substances, Fetal Exposure, Fetal Origins of Disease 32 33 34 35 ABSTRACT 36 37 38 39 Perfluoroalkyl substances (PFAFs) are man-made chemicals used in many consumer products and 40 41 have become ubiquitous in the environment. Animal studies and a limited number of human 42 43 studies have demonstrated developmental effects in offspring exposed to PFAFs in utero but the 44 45 46 implications o f timing of in utero exposure have not been systematically investigated. The 47 48 current study investigated variation in periluorocarbon levels of 9,952 women of childbearing 49 50 51 age who had been exposed to perfluorooctanoicacid (PFOA) in drinking water contaminated by 52 53 industrial waste. An analysis of variance with contrast was performed to compare the levels of 54 55 PFOA and perfluorooctanesulfonicacid (PFOS) in pregnant and non-pregnant women overall 56 57 58 1 59 60 ACS Paragon Plus Environm ent Environmental Science & Technology Page 2 of 28 and during each trimester o f pregnancy. We found that pregnant women had lower circulating PFOA and PFOS concentrations in peripheral blood than non-pregnant women and that PFOA levels were consistently lower throughout all trimesters for pregnancy, suggesting transfer to the fetus at an early stage of gestation. These results are discussed in the context of the endocrine disrupting properties o f perfluoroalkyl substances that have been characterized in animal and human studies. Our conclusion is that further, systematic study of the potential implications o f intrauterine perfluorocarbon exposure during critical periods of fetal development is urgently needed. INTRODUCTION Perfluorooctanesulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) are man-made perfluoroalkyl substances (PFASs) historically used in many consumer and industrial products. PFOA has been used primarily in the linings of cookware and food containers, and PFOS in stain resistant clothing and textiles, cleaning agents (waxes and floor polishes), and paint and varnish [1J. Because of their widespread use, PFASs have become ubiquitous in the environment and can be found in soil, water, wildlife, and in measurable quantities in most humans [2], even those living in remote areas [3], Data from a representative random sample (NHANES) in the United States indicate that 98% o f participants had measurable quantities of PFOA and PFOS [4]. Several animal studies have addressed the issue of reproductive toxicity. Leubker [5] and Grasty [6j found that pup PFASs were proportional to serum concentrations in the mothers, indicating that females transferred PFASs to their offspring. It has also been shown that daily PFAS exposure in pregnant rats, mice, and rabbits resulted in complications varying from altered feed consumption, to hepatomegaly, decreased litter size, and increased litter reabsorption [6 1Oj. Additional studies have found complications in offspring exposed to PFASs in utero that included a decreased rate of survival [ i l |, decreased fetal weight [10], and developmental and birth defects (e.g., 2 ACS Paragon Plus Environm ent Page 3 of 28 Environmental Science & Technology 1 2 3 delayed ossification of bones, enlarged right atrium, cleft palate, and inhibition o f lung maturation) [8- 4 5 6 N]. 7 8 Results from these studies have raised concerns about the health consequences of PFAS 9 10 exposure in pregnant women. Although human research is limited, PFOA was documented in all 11 12 13 100 umbilical cord blood samples of one study [12] and an association between maternal and fetal 14 15 exposure to PFASs has been documented through umbilical cord blood concentrations at birth that were 16 17 proportional to maternal serum concentrations [13-18]. The Danish Cohort Study 18 19 20 reported that maternal serum levels o f PFOA and PFOS decreased in the second trimester compared to the 21 22 first [16] and a small study of 105 babies detected both PFOS and PFOA in umbilical cord blood. That 23 24 25 same study showed that mid-pregnancy maternal serum PFOS levels were higher than the maternal values 26 27 at delivery [17|. Together these data lend strong support for transfer PFAFs to the fetus in the latter 28 29 30 stages of pregnancy. Further human research has found an inverse association between gestational age and 31 32 circulating PFASs in pregnant mothers [16, 17] and data from a small sample of women in the National 33 34 Health and Nutritional Examination Survey (NHANES) indicate that pregnant women may have lower 35 36 37 PFAS levels than non-pregnant women [ 19J. 38 39 Thus, accumulating human and animal research raises concerns about in utero exposure to PFASs. 40 41 42 The purpose o f the present study was to investigate variation in PFAS levels in pregnant women during 43 44 gestational trimester and compare them with those o f non-pregnant women in a large geographic area where 45 46 exposure to PFOA in drinking water had resulted from toxic waste disposal. Based on earlier animal and 47 48 49 human data, we hypothesized that redistribution of PFASs to the fetus during gestation would result in 50 51 lower circulating serum PFAS levels in the peripheral blood o f mothers during all trimesters of 52 53 pregnancy compared with levels found in the non-pregnant women. 54 55 56 MATERIALS AND METHODS 57 58 3 59 60 ACS Paragon Plus Environment Environmental Science & Technology Page 4 of 28 1 2 3 Participants 4 5 6 Participants were 9,952 women between the ages of 18 and 42 years enrolled in the C8 Health Project, 7 8 which resulted from a class action suit. The "Class" was defined as individuals in West Virginia or Ohio 9 10 who had consumed contaminated drinking water at a residence, place of employment, or school within a 11 12 13 specific area of West Virginia and Ohio for at least a year 14 15 between 1950 and December 3, 2004. Details have been described elsewhere [20J. Because o f the 16 17 18 circumstances of the study, the average level o f exposure in these participants was higher than in 19 20 NHANES f 191. Eligibility requirements for inclusion o f women in the study were: that their residential 21 22 water level exposure was classified as consistent with respect to geographic area prior to measurement (as 23 24 25 described below), that they were o f childbearing age (defined as 18-42 years), and that they had 26 27 responded to the question on pregnancy status. There were 498 pregnant women and 9,454 non 28 29 30 pregnant women who met the inclusion criteria. Eligible enrolled participants filled out surveys with 31 32 demographic, medical and other information and submitted a voluntary blood sample between August 1, 33 34 2005 and August 31, 2006. A detailed description of the consent, surveys, blood processing, and blood 35 36 37 sample storage is available online at http://www.hsc.wvu.edu/som/cmed/c8/. Month of pregnancy was 38 39 measured by self- report. The study was cross-sectional, meaning that perfluorocarbon measurements 40 41 were done once in each woman. 42 43 44 Residential History of Exposure 45 46 Because this was an exposure study, a complete residential history was obtained from each participant 47 48 49 to provide an indication o f consistency and amount o f exposures. Participants were eligible for the study if 50 51 they had regularly consumed drinking-water from one of the six water districts of PFOA exposure based on 52 53 residence, or having regularly worked or gone to school there. Interpreting the exposure o f participants who 54 55 56 had never lived in the C8 Health study water districts but were included due to regular exposure in their 57 58 4 ` 59 60 ACS Paragon Plus Environm ent Page 5 of 28 Environmental Science & Technology 1 2 3 workplaces or schools would have been difficult. Also, including women who had moved between districts 4 5 6 with high, medium or low levels of exposure would also have made interpretation difficult. Therefore, 7 8 non-residents and inconsistent residents were excluded from the analyses. There were 1,604 women 9 10 excluded 11 12 13 due to not having resided in any of the C8 Health Study water districts, and 1,750 excluded due to 14 15 16 inconsistent residential history. This left 498 pregnant women and 9,454 non-pregnant women whose 17 18 data were analyzed. 19 20 Blood Sample Processing and Laboratory Methods 21 22 23 Participants voluntarily submitted up to 26mL of blood for analysis in the C8 Health Study, and the 24 25 resulting serum was banked during the collection period (2005-2006). PFAS concentrations were quantified 26 27 at Exygen Research Inc., State College, PA, USA. The analytic protocol was a modification of a previously 28 29 30 described procedure combining protein precipitation extraction and reverse-phase high-performance liquid 31 32 chromatography-tandem mass spectrometry. Spectrometric detection was performed using a triple 33 34 quadrupole mass spectrometer in selected reaction monitoring mode, monitoring for the individual m/z 35 36 37 transitions 38 39 for each of the 10 PFASs and the u C-PFOA internal standard. A description of the 40 41 perfluoroalkyl acid analytic techniques and quality assurance protocols for the C8 Health Project has been 42 43 44 published elsewhere [20], 45 46 Statistical Analyses 47 48 49 Statistical analyses were performed using JMP/PRO 10 Visualization Software (SAS Institute Inc., Cary 50 51 NC). Because month of gestation was self-report and calculating exactly when the pregnancy started was 52 53 not exact, months one and two were combined in the analyses. An analysis of variance and covariance 54 55 56 with a contrast statement comparing pregnant and non- pregnant women was calculated to investigate the 57 58 5 59 60 ACS Paragon Plus Environm ent Environmental Science & Technology Page 6 of 28 1 2 3 differences in PFOA and PFOS concentrations between pregnant and non-pregnant women. Separate 4 5 6 analyses were also calculated to compare non-pregnant women with those in each trimester of pregnancy 7 8 (defined as months 1-3, 4-6, and 9 10 7-9). The natural logarithm for PFOA and PFOS was used in all analyses to mitigate the effects 11 12 of outliers. All analyses were controlled for level of education and income, (proxies of socioeconomic status 13 14 that could influence simultaneous exposure to other toxicants based on living area within the water 15 16 17 district); age (older women may have longer exposure times); parity (increased number of pregnancies may 18 19 decrease perfluorocarbon levels [16]), smoking and alcohol consumption (which can impair blood flow 20 21 to the placenta [21, 22]), as well as a surrogate for plasma volume, which changes during pregnancy 22 23 24 123] and could potentially affect interpretation of the PFAS measurements. Direct measures of plasma 25 26 volume such as Evan's Blue Dye [24] were not available in the dataset, so we performed plasma volume 27 28 adjustments using two different surrogate measures: hematocrit and the hemoglobin/hematocrit ratio. 29 30 31 Hematocrit is a crude estimate of overall plasma volume, i.e., the percentage of blood volume consisting 32 33 of packed red blood cells, and the hemoglobin-hematocrit ratio adds additional information concerning 34 35 whether red blood morphology might be contributing to a hemoglobin value. If the hemoglobin-hematocrit 36 37 38 ratio decreases and there is no other reason for a reduced red blood cell mass, the reduction is attributed 39 40 to increased plasma volume. Equations were calculated with both versions with similar results. The proxy 41 42 43 for plasma volume described in the results section is hemoglobin-hematocrit ratio. Covariates were allowed 44 45 to enter the equation in order of importance. 46 47 RESULTS 48 49 50 Descriptive characteristics of pregnant and non-pregnant women are listed in Table 1. Their PFOS and 51 52 PFOA concentrations are listed in Table 2. Both PFOS and PFOA are lower in pregnant women than in 53 54 55 non-pregnant women and this is reflected also in the covariates in Table 56 57 1. 58 6 59 60 ACS Paragon Plus Environm ent Page 7 of 28 Environmental Science & Technology 1 2 3 Table 1. Select characteristics of the study population. Unless indicated otherwise, results are 4 5 6 presented as: median, mean standard deviation. Percentages are rounded to the nearest decimal. 7 8 g Non-Pregnant Women Pregnant Women 10 11 (n=9,454*) (n=498*) 12 13 14 15 Median Mean Std. Dev. Median Mean Std. Dev. 16 17 Age (Years) 30.74 30.44 7.00 26.26 26.93 5.14 18 19 Hematocrit (%) 40.90 40.79 3.04 36.90 37.06 3.30 20 21 Hemoglobin (g/di) 13.80 13.74 1.05 12.50 12.54 3.30 22 23 Hemoglobin/Hematocrit Ratio . 0.34 0.34 . 03H 0.34 0.34 0.01 24 25 Parity (Number of Times Pregnant) 1.00 1.57 1.48 2.00 1.75 1,44 26 27 % Non- % 28 Pregnant PFOAf PFOSt Pregnant PFOA* PFGS* 29 Women Women 30 31 Highest Level of Education Obtained 32 33 Less than a high school diploma 8.4% 32.29 15.00 11.3% 22.01 11.08 34 35 High school diploma or GED 34.1% 44.56 16.88 29.8% 22.03 12.93 36 37 Some college 38 41.5% 44.18 17.60 41.0% 22.11 16.41 39 Bachelor's degree or higher 16.0% 40.21 17.60 17.9% 35.84 15.70 40 41 Health Behaviors 42 43 Currently smokes 44 45 46 Currently drinks 31.6% 53.2% 42.73 45.73 15.11 17.20 22.4% 32.2% 19.54 25.62 12.62 14.93 47 48 Average Yearly Household Income 49 50 <S 10,000 14.6% 34.21 15.58 17.8% 20.76 12.76 51 52 SI 0,000-19,999 16.8% 39.28 16.10 20.0% 19.26 14.10 53 54 $20,000-29,999 15.8% 40.28 16.80 15.2% 22.18 16.30 55 56 57 58 7 59 60 ACS Paragon Plus Environm ent Environmental Science & Technology Page 8 of 28 1 2 3 $30,000-39,999 4 J 13.7% 41.67 17.36 9.7% 33.04 15.34 5 $40,000-49,999 6 :i ii.o% 43.86 17.74 13.3% 26.83 15.29 7 $50,000-59,999 8 8.8% 55.67 . 17.85 8.3% 26.25 14.56 9 $60,000-69,999 10 7.0% 48.28 18.03 5.7% 23.22 17.66 11 > $70,000 12 12.4% 51.31 18.68 10.0% 40.06 13.95 13 Residential History of PFAS 14 15 Exposure 16 17 High 41.0% 77.14 17.83 37.0% 45.61 15.58 18 19 Medium 30.8% 27.56 16.52 34.5% 15.87 13.91 20 21 Low 24.2% 11.18 16.80 28.5% 8.54 14.57 22 23 *This number reflects the total number o f participants for the respective group. However, it varies slightly 24 25 across cells due to missing values. 26 27 28 'Mean PFOA and PFOS levels are in units of ng/ml. 29 30 31 32 33 34 Table 2. PFOA and PFOS levels in pregnant and non-pregnant women. 35 36 37 38 39 PFOA (ng/ml) PFOS (ng/ml) 40 41 Mean Median Min. Max. Mean Median Min. Max. 42 43 44 All women (n=9,952) 42.26 17.50 0.25 8,162.80 17.01 15.20 0.25 106.60 45 46 Non-Pregnant Women 43.19 17.80 0.25 8,162.80 17.13 15.30 0.25 106.60 47 48 (n=9,454) 49 50 51 Pregnant Women (n=498) 24.49 12.20 1.10 295.60 14.71 12.90 0.25 56.00 52 53 54 55 56 57 58 8 59 60 ACS Paragon Pius Environment Page 9 of 28 Environmental Science & Technology 1 2 3 Women in First Trimester 25.42 12.40 1.50 227.30 16.32 15.00 1.80 56.00 4 5 6 (n=128) 7 8 Women in Second 24.92 12.50 1.10 295.60 14.46 12.40 0.25 46.50 9 10 Trimester (n=193) 11 12 13 Women in Third Trimester 23.69 12.00 1.40 194.50 13.63 12.10 1.30 41.50 14 15 (0=166) 16 17 18 19 20 Comparison of PFOA and PFOS in Pregnant vs. Non-Pregnant Women 21 22 23 Assumptions for normality in this population were met. After controlling for all 24 25 covariates (age, alcohol consumption, smoking, educational level, hemoglobin-hematocrit 26 27 ratio, parity, residential exposure, and income), the results of the ANOVA using month as a 28 29 30 predictor (months 31 32 l and 2 were combined) showed significant differences between pregnant and non-pregnant 33 34 35 women in PFOA (F= 17.33; p< 0.001) AND PFOS (F=9.78; p=0.0018). Pregnant women in the 36 37 study population had lower circulating levels of PFOA [15] than non-pregnant women. 38 39 Mean values for pregnant and non-pregnant women are shown in Figure 1. In the figures, 40 41 42 natural log values have been converted to analog values to make interpretation easier. 43 44 45 46 Insert Figure 1 about here 47 48 49 50 51 52 53 Perfluorocarbon Concentrations by Trimester of Pregnancy vs. Non-Pregnancy 54 55 To examine PFAS concentrations during different stages of fetal development, we also 56 57 58 9 59 60 ACS Paragon Pius Environm ent Environmental Science & Technology Page 10 of 28 1 2 3 analyzed the data by trimester. After controlling for the same covariates, comparison of non 4 5 6 pregnant women to women in each trimester of pregnancy showed significantly lower levels of 7 8 PFOA in pregnant women in the second and third trimesters of pregnancy (second trimester 9 10 T=2.8I, p=0.005 and third trimester T=2.67, p=0.008, than in non-pregnant women and a strong 11 12 13 trend towards lower concentrations in the first trimester (T=1.85 p=0.06). However, PFOS 14 15 concentrations in pregnant women were significantly lower than those in non-pregnant women 16 17 only during the third trimester T-3.05, p=Q.0G2, although the tendency in the second semester 18 19 20 was strong (T=1.79, p=0.07). The PFAS concentrations by trimester of pregnancy compared 21 22 with non-pregnant women are shown in Figure 2. 23 24 25 26 27 Insert Figure 2 about here 28 29 30 31 32 33 DISCUSSION 34 35 This data from a population of women exposed to PFOA in drinking water substantiate the 36 37 earlier indications from NHANES that circulating levels do indeed drop in pregnant women 38 39 40 | I9j. Pregnant women in our study had consistently lower levels of circulating PFOA in all 41 42 trimesters o f pregnancy (strong tendency in the first trimester and significant in the second and 43 44 third); as well as lower concentrations o f PFOS that reached significance during the 3rd 45 46 47 trimester o f pregnancy. Since pregnant women do not menstruate, they might have been 48 49 expected, based on earlier reported hysterectomy data [25], to exhibit higher levels of 50 51 circulating PFASs than non-pregnant women. Not only were their PFOA levels not higher, 52 53 54 they were lower. One plausible explanation for the lower peripheral blood levels in pregnant 55 56 women is redistribution to the fetus. This explanation is consistent with existing animal data, 57 58 10 59 60 ACS Paragon Plus Environment Page 11 of 28 Environmental Science & Technology 1 2 3 which have demonstrated transfer of PFOS to the fetus [5, 6] and with human data showing 4 5 6 the presence of both PFOS and PFOA in the cord blood of neonates from exposed mothers 7 8 [14-17], Furthermore, a recent cross-sectional study in humans reported steadily increasing 9 10 PFOS in amniotic fluid by gestational week [26], The results of these studies further support 11 12 our hypothesis that maternal PFASs are offloaded into the fetus throughout the course of 13 14 15 pregnancy. 16 17 There are a number of potential consequences of fetal exposure to PFASs that could have 18 19 major consequences for subsequent child development. Our data show that the first trimester is 20 21 22 the beginning of a decline in maternal circulating PFOA levels, indicating that redistribution of 23 2.4 this chemical occurs during a critical period of gestational development. This is a period during 25 26 27 which structures o f the nervous system, cardiovascular system, digestive system, respiratory 28 29 system, endocrine system, and kidneys are being formed [27], and when the potential for 30 31 epigenetic influences is particularly critical. Extensive data indicate that epigenetic influences 32 33 34 during critical periods o f fetal development can cause permanent changes in metabolism and 35 36 chronic disease susceptibility [28]; and an increasing body of research shows that a complex 37 38 combination of adult health-related disorders can originate from developmental events that occur 39 40 41 in iitero even without direct effects on pregnancy or birth weight [29]. A great many epigenetic 42 43 changes that occur in utero do not manifest until later in child- or even adulthood and may 44 45 therefore not be immediately identifiable as birth outcomes. Because animal research has shown 46 47 48 that perfiuorocarbon exposure to the fetus is, in fact, associated with epigenetic changes, this 49 50 issue is important. It has been demonstrated in rat L02 liver cells [30], that there is a dose-related 51 52 53 increase in methylation (a process that turns genes off) o f the glutationine-S-transferase (GSTP) 54 55 promoter, a gene that encodes an enzyme involved with detoxification metabolism and 56 57 58 11 59 60 ACS Paragon Plus Environment Environmental Science & Technology Page 12 of 28 1 2 3 susceptibility to cancer) 31]. DNA methylation in human cord serum has also been 4 5 6 demonstrated to be inversely associated with serum PFOA [32]. The implications of these 7 8 methylation changes have not been investigated but given the increasing evidence of maternal - 9 10 fetal transfer of these chemicals, the need for such research is urgently needed. Animal research 11 12 13 on other toxicants (e.g. Vinclozolin) has demonstrated that embryonic exposure was associated 14 15 with tissue abnormalities including prostate, kidney, immune system, and testis, as well as tumor 16 17 development in the adult FI generation as well as in subsequent generations (F2-F4) [33]. 18 19 20 The relevance of Vinclozolin is that it is an endocrine disrupter and endocrine disruption has 21 22 been an important focus of research related to perfluoroalkyi substances [34, 35], Although 23 24 25 endocrine disrupting consequences o f fetal exposure to PFASs, have to our knowledge not yet 26 27 been investigated, the fact that exposure to other endocrine disrupting chemicals in utero has 28 29 been associated with the above mentioned abnormalities as well as with human urogenital 30 31 32 malformations and cancer [36-38], impaired reproductive function and infertility [38], increased 33 34 risk of breast cancer 139], and intellectual impairment and neurodevelopmental changes 35 36 manifesting in children [40], raises cause for concern. Whether or not these effects also 37 38 39 generalize to PFOA and PFOS remains to be seen, however, accumulating data suggest that this 40 41 is an important area for future investigation. 42 43 44 The C8 Health Study has reported that in adults, PFOS and PFOA are associated with 45 46 endocrine disruption related to thyroid dysfunction [41], In analyses stratified by age and 47 48 gender, both PFOA and PFOS were shown to be associated with significant elevations in serum 49 50 51 thyroxin and a significant reduction of T3 uptake in all participants. These effects were 52 53 significantly stronger in women. The pattern found in those data were interpreted as being 54 55 56 consistent with what occurs with the use of exogenous estrogens in patients, namely an increase 57 58 12 59 60 ACS Paragon Plus Environm ent Page 13 of 28 Environmental Science & Technology 1 2 3 in thyroid binding globulin (TBG) but not thyroid stimulating hormone (TSH), an increase in 4 5 6 total thyroxin and a decrease in T3 uptake. The limitation of that study was that the only binding 7 8 protein actually measured was albumin, which binds a much smaller amount of thyroxin than 9 10 TBG. However, albumin showed the same positive association with serum PFOA and PFOS that 11 12 13 would have been expected from TBG. 14 15 The fact that PFOS and PFOA are associated with thyroid dysfunction in adults has 16 17 implications for pregnant mothers. It has been demonstrated in clinical studies that thyroid 18 19 20 hormone reaches the fetus and affects gene expression in the fetal brain [42], In fact, thyroid 21 22 hormones control neuronal and glial proliferation in certain brain regions, and contribute to the 23 24 regulation of neuronal migration and differentiation [43], Thus, factors that disrupt thyroid 25 26 27 function in the mother have the potential to affect brain development in the fetus [42], In fetal 28 29 fluids, a major proportion of T4 is not protein-bound (i.e., it is `free') and is correlated to that in 30 31 32 maternal serum [44]. The primary research focus with respect to maternal thyroid function and 33 34 subsequent fetal and child development has been on the damaging effects of hypothyroidism to 35 36 the central nervous system and cognitive development [45-49J. Hypothyroidism has not been 37 38 39 observed in the C8 dataset. Rather the effect of perfluoroalkyl substances on thyroid function 40 41 was an increase in T4 and circulating T3 (based on a reduction in T3 uptake). However, also 42 43 increases in maternal thyroid hormones are associated with biochemical disturbances in the fetus, 44 45 46 including effects on the neurotransmitters acetylcholine, dopamine and serotonins [50]. Again, 47 48 the accumulating data indicate the need for more systematic research on fetal exposure to PFOS 49 50 and PFOA. 51 52 53 Endocrine disruption is not the only threat from exposure to PFOS and PFOA. These 54 55 56 chemicals have also been shown to be associated with increased total and low-density lipoprotein 57 58 13 59 60 ACS Paragon Plus Environm ent Environmental Science & Technology Page 14 of 28 1 2 3 (LDL) cholesterol in children and adults [51, 52]. The significance o f this is that maternal 4 5 6 hypercholesterolemia is associated with the increased formation of fatty streaks in human fetal 7 8 arteries [53] as well as accelerated atherosclerosis progression in childhood [54]. -9 10 Overall, studies investigating the effects of intrauterine PFAS exposure on child development 11 12 13 in humans are limited and most have not investigated outcomes associated with endocrine 14 15 disruption. Studies of infant allergies and infectious diseases from mothers exposed to PFASs 16 17 dining pregnancy have found an inverse association between maternal PFOA levels during 18 19 20 pregnancy and IgE levels in cord blood of infant girls [55]; as well as negative associations 21 22 between PFOS and PFOA and antidiphtheria prebooster antibody concentrations [56]. 23 24 Consistent with the Developmental Origins o f Disease hypothesis, PFOA concentrations in 25 26 27 pregnant mothers have also been found to be positively associated with BMI and waist 28 29 circumference among their 20 year old daughters [57] and with the daughters' biomarkers of 30 31 32 adiposity (e.g. insulin, leptin, and leptin-adiponectin ratio) while being inversely associated with 33 34 their adiponectin [57], 35 36 The comparison of maternal PFOS and PFOA concentrations during pregnancy with those of 37 38 39 non-pregnant women is complex because of the significant changes in plasma volume that occur. 40 41 One weakness of our study is that we were forced to use a proxy for plasma volume in our 42 43 analyses because the dataset did not contain a direct measure. Interestingly, previously published 44 45 46 data from this same cohort reported that levels o f PFASs were higher in women who had had 47 48 hysterectomies (i.e., were not menstruating) [25J, the implication is that our comparison may 49 50 51 actually be conservative. If non-menstruating women accumulate more perfluoroakly substances, 52 53 then pregnant women who also do not menstruate may actually be transferring higher 54 55 56 concentrations to the fetus than would be expected from comparing concentrations between 57 58 14 59 60 ACS Paragon Plus Environm ent Page 15 of 28 Environmental Science & Technology 1 2 3 pregnant and non-pregnant women as a whole. 4 5 6 A second limitation to our study is that the month o f pregnancy is measured by self-report. 7 8 Because women use more than one way to measure pregnancy onset (e.g., date of last menstrual 9 10 11 period, date of conception), our designation of trimesters is not exact. Despite this fact, there is 12 13 no reason to assume any systematic bias with respect to this variable. Another study limitation is 14 15 that we were not able to adjust for thyroid binding globulin, which varies greatly during 16 17 pregnancy, because it was not part of the dataset. Problems of this type with covariates are often 18 19 20 a trade-off in large population databases. These databases provide a lot of power for 21 22 investigating population effects but do not always contain the covariates one would desire. We 23 24 25 nevertheless, believe these analyses are informative. The fact that the results were stronger for 26 27 PFOA than for PFOS may reflect the results of a Chinese study which reported a higher partition 28 29 ratio of PFOA through placental barrier and lactation than for PFOS. There is an additional 30 31 32 covariate we would have liked to analy/.e but couldn't, namely history of lactation. Breast 33 34 feeding could potentially influence the level of maternal PFAFs by off-loading some of them into 35 36 the baby. This is a limitation of the study. Fortunately there is no reason to believe that history of 37 3389 lactation would vary by water district and thus introduce systematic bias. The major strength of 40 41 this study is that the population was large enough to compare perfluorocarbon levels at different 42 43 trimesters of pregnancy with those of non-pregnant women. 44 45 46 Although there has been some human research with regard to PFASs and developmental 47 48 outcomes, it has not been systematic. Study designs have tended to measure exposure at more 49 50 51 advanced stages of gestation, and reports of clinical outcomes have not, for the most part, been 52 53 theor>' driven. Because of the tremendous strides that have been made in the field of epigenetics, 54 55 56 a whole new avenue for investigating fetal origins o f adult disease has emerged. Given the 57 58 15 59 60 ACS Paragon Plus Environm ent Environmental Science & Technology Page 16 of 28 1 2 3 accumulating data on PFASs and endocrine disruption and the indications that PFOA is entering 4 5 6 the fetus in the first trimester of development, more systematic research needs to be applied to 7 8 the epigenetic effects of these chemicals on the developing fetus. 9 10 11 12 AUTHOR INFORMATION 13 14 15 Corresponding Author 16 17 TSarah S. Knox, West Virginia University, School of Public Health, P.O. Box 9190, 18 19 Morgantown, W.V., 26506-9190, U.S.A., sknox@hsc.wvu.edu. Phone: (304)293-1058, Fax: 20 21 22 (304)293-6685 23 24 25 Author Contributions 26 27 All authors have given approval to the final version of the manuscript. Contributions of 28 29 individual authors area as follows: Beth Javins contributed to the statistical analyses and wrote 30 31 32 the first draft of the manuscript, Gerald Hobbs was the statistician, Alan M. Ducatman 33 34 contributed to editing the document, interpretation of results and was the technical consultant on 35 36 PFOS and PFOA, and Courtney Pilkerton contributed to discussions concerning implications of 37 38 39 the results for fetal development. Sarah S. Knox designed the study, contributed to the analyses, 40 41 interpreted the results and did a lot of the writing and editing. 42 43 44 Funding Sources 45 46 The authors are grateful to the C8 Health Project for partial funding of this project through 47 48 Brookmar contract number 10009179.4.1003778R to Alan M. Ducatman. All authors work at 49 50 51 West Virginia University and have no competing financial interest. 52 53 54 55 ACKNOWLEDGMENT . . . 56 57 We would like to thank William Holls, MD, Professor, Department of Obstetrics and 58 16 59 60 ACS Paragon Pius Environm ent Page 17 of 28 Environmental Science & Technology 1 2 3 Gynecology at the West Virginia University School of Medicine, for taking the time to discuss 4 5 6 with us the issue of plasma volume during pregnancy. 7 8 9 ABBREVIATIONS 10 11 PFAS, Perfluorinated Chemicals; PFOS, Perfluorooctanesulfonic acid; PFOA, Perfluorooctanoic 12 13 acid 14 15 16 REFERENCES 17 18 19 20 1. Poulsen, P. B.; Jensen, A. A.; Wallstrom, E.; Aps, E. More environmentally friendly 21 22 alternatives to PFOS-compounds and PFOA. In Environmental Project, Danish Ministry of the 23 24 Environment: 2005; Vol. 1013, p 2005. 25 26 27 28 2. Houde, M.; Martin, J. W.; Letcher, R. J.; Solomon, K. R.; Muir, D. C. G. Biological 29 30 monitoring of polyfluoroalkyl substances: a review. Environ. Sci. Technol. 2006, 40 (11), 3463 31 32 3473. 33 34 35 36 3. Martin, J. W.; Smilhwick, M. M.; Braune, B. M.; Hoekstra, P. F,; Muir, D. C. G.; 37 38 Mabury, S. A. Identification o f long-chain perfluorinated acids in biota from the Canadian 39 40 Arctic. Environ. Sci. Technol 2003, 38 (2), 373-380. 41 42 43 4. Calafai, A. M.; Wong, L.; Kuklenyik, Z.; Reidy, J. A.; Needham, L. L. Polyfluoroalkyl 44 45 chemicals in the U.S. population: data from the National Health and Nutrition Examination 46 47 48 Survey (NHANES) 2003-2004 and comparisons with NHANES 1999-2000. Environ. Health 49 50 Perspecl. 2007,115 (11), 1596-1602. 51 52 53 5. Luebker, D. J.; York, R. G.; Hansen, K. J.; Moore, J. A.; BulenhofT, J. L. Neonatal 54 55 56 mortality from in tero exposure to perfluorooctanesulfonate (PFOS) in Sprague-Dawley rats: 57 58 17 59 60 ACS Paragon Plus Environm ent Environmental Science & Technology Page 18 of 28 1 2 3 dose-response, and biochemical and pharamacokinetic parameters. Toxicology 2005, 215 (1-2), 4 5 6 149-169. 7 8 9 6. Grasty, R C.; Grey, B. E.; Lau, C. S.; Rogers, J. M. Prenatal window of susceptibility to 10 11 perfluorooctane sulfonate induced neonatal mortality in the Sprague Dawley rat. Birth Defects 12 13 14 Res. B Dev. Reprod. Toxicol. 2003, 68 (6), 465-471. 15 16 17 7. Lau, C.; Thibodeaux, J. R.; Hanson, R. G.; Narotsky, M. G.; Rogers, J. M.; Lindsirom, A. 18 19 B.; Strynar, M. J. Effects o f perfluorooctanoic acid exposure during pregnancy in the mouse. 20 21 22 Toxicol. Sci. 2006, 90 (2), 510-518. 23 24 25 8. Thibodeaux, J. R.; Hanson, R. G.; Rogers, J. M.; Grey, B. E.; Barbee, B. D.; Richards, J. 26 27 IT; Butenhoff, J. L.; Stevenson, L. A.; Lau, C. Exposure to perfluorooctane sulfonate during 28 29 30 pregnancy in rat and mouse. I; maternal and prenatal evaluations. Toxicol. Sci. 2003, 74 (2), 369 31 32 381. 33 34 35 9. Yahia, D.; Tsukuba, C.; Yoshida, M.; Sato, I.; Tsuda, S. Neonatal death of mice treated 36 37 with perfluorooctane sulfonate. J. Toxicol. Sci. 2008, 33,(2), 219-226. 38 39 40 10. Case, M. T.; York, R. G.; Christian, M. S. Rat and rabbit oral developmental toxicology 41 42 43 studies with two perfluorinated compounds. Int. J. Toxicol. 2001, 20 (2), 101. 44 45 46 11. Grasty, R. C\; Bjork, J.; Wallace, K.; Lau, C.; Rogers, J. Effects of prenatal 47 48 perfluorooctane sulfonate (PFOS) exposure on lung maturation in the perinatal rat. Birth Defects 49 50 51 Res. B Dev. Reprod. Toxicol. 2005, 74 (5), 405-416. 52 53 54 12. Arbuckle, T. E.; Kubwabo, C.; Walker, M.; Davis, K.; Lalonde, K.; Kosarac, I.; Wen, S. 55 56 W.; Arnold, D. L. Umbilical cord blood levels of perfluoroalkyl acids and polybrominated flame 57 58 18 59 60 ACS Paragon Plus Environm ent Page 19 of 28 Environmental Science & Technology 1 2 3 retardants. Int. J. Hyg. Envir. Heal. 2012, Article In Press. 4 5 6 7 13. Inoue, K.; Okada, F.; Ito, R.; Kato, S.; Sasaki, S.; Nakajima, S.; Uno, A.; Saijo, Y.; Sata, 8 9 F.; Yoshimura, Y. Periluorooctane sulfonate (PFOS) and related perfluorinated compounds in 10 11 human maternal and cord blood samples: assessment of PFOS exposure in a susceptible 12 13 14 population during pregnancy. Environ. Health Ferspect. 2004,112 (11), 1204-1207. 15 16 17 14. Fromme, H.; Mosch, C.; Morovitz, M.; Alba-Alejandre, I.; Boehmer, S.; Kiranoglu, M.; 18 19 Faber, F.; Hannibal, I.; Genzel-Borovic/.ny, O.; Koletzko, B.; Volkel, W., Pre- and postnatal 20 21 22 exposure to perfluorinated compounds (PFASs). Environ. Sci. Technol. 2010, 44 (18), 7123 23 24 7129. 25 26 27 15. Apelberg, B. J.: Witter, F. R.; Herbstman, J. B.; Calafat, A. M.; Halden, R. U.; Needham, 28 29 30 L. L.; Goldman, L. R. Cord serum concentrations of perfluorooctane sulfonate (PFOS) and 31 32 perfluorooctanoale (PFOA) in relation to weight and size at birth. Environ. Health Ferspect. 33 34 2007, //5 (1 1 ), 1670 35 36 16. Fei, .; McLaughlin, J. K.; Tarone, R. E,; Olsen, J. Perfluorinated chemicals and fetal 37 38 growth: a study within the Danish National Birth Cohort. Environ. Health Ferspect. 2007, 115 39 40 41 (II), 1677-1682. 42 43 44 17. Monroy, R ; Morrison, K.; Teo, K.; Atkinson, S.; Kubwabo, C.; Stewart, B.; Foster, W. 45 46 G. Serum levels o f perfluoroalkyl compounds in human maternal and umbilical cord blood 47 48 49 samples. Environ. Res. 2008,108(1), 56-62. 50 51 52 18. Giitzkow, K.. B.; Haug, L. S.; Thomsen, C.; Sabaredzovic, A.; Becher, G.; Brunborg, G. 53 54 Placental transfer of perfluorinated compounds is selective - a Norwegian mother and child sub 55 56 57 cohort study. Int. ./. Hyg. Envir. Heal 2012,215 (2), 216-219. 58 19 59 60 ACS Paragon Plus Environm ent Environmental Science & Technology Page 20 of 28 1 2 3 4 19. Woodruff; T. J.; Zota, A. R.; Schwartz, J. M. Environmental chemicals in pregnant 5 6 women in the United States: NHANES 2003-2004. Environ. Health Perspect. 2011, 119 (6), 7 8 9 878-885. 10 11 20. Frisbee, S. J.; Brooks, A. P.; Maher, A.; Flensborg, P.; Arnold, S.; Fletcher, T.; 12 13 14 Steenland, K.; Shankar, A.; Knox, S. S.; Pollard, C.; Halverson, J. A.; Vieira, V. M.; Jin, C. F.; 15 16 Leyden, K. M.; Ducatman, A. M. The C8 Health Project: design, methods, and participants. 17 18 Environ. Health Perspect. 2009, 117 ( 12), 1873-1882. 19 20 21 21. Burd, L.; Roberts, D.; Olson, M.; Odendaal, H. Ethanol and the placenta: a review. J. 22 23 Matern.-Fetal Neo. M. 2007, 20 (5), 361-375. 24 25 22. Habek, D.; Habek, J. C.; Ivanisevic, M.; Djelmis, J. Fetal tobacco syndrome and perinatal 26 27 28 outcome, l-'etal Diagn. 1'her. 2002, 17 (6), 367-371. 29 30 23. Faupel-Badger, J. M.; Hsieh, C, C.; Troisi, R.; Lagiou, P.; Potischman, N. Plasma volume 31 32 33 expansion in pregnancy: implications for biomarkers in population studies. Cancer Epidemiol. 34 35 Biomarkers Prev. 2007,16(9), 1720-1723. 36 37 24. Faijanei, J.; Denis, C.; Chatard, J. C.; Geyssant, A. An accurate method of plasma 38 39 40 volume measurement by direct analysis of Evans blue spectra in plasma without dye extraction: 41 42 origins of albu min-space variations during maximal exercise. Eur. J. Appli. Physiol. Occup. 43 44 45 Physiol. 1996, 75 (1), 75-82. 46 47 25. Knox, S. S.; Jackson, T.; Javins, B.; Frisbee, S. J.; Shankar, A.; Ducatman, A. M. 48 49 Implications of early menopause in women exposed to perfluorocarbons. J. Clin. Endocrinol. 50 51 52 Metab. 2011,96(6), 1747-1753. 53 54 26. Jensen, M. S.; Norgaard-Pedersen, B.; Toft, G.; Hougaard, D. M.; Bonde, J. P.; Cohen, 55 56 A.; Thulstrup, A. M.: Ivell, R.: Anand-lvell, R.; Lindh, C. H.; Jnsson, B. A. G. Phthalates and 57 58 20 59 60 ACS Paragon Plus Environment Page 21 of 28 Environmental Science & Technology 1 2 3 perfluorooctanesulfonic acid in human amniotic fluid: temporal trends and timing of 4 5 6 amniocentesis in pregnancy. Environ. Health Perspect. 2012, 20 (6), 897-903. 7 8 27. Yamada, S.; Samlani, R. R.; Lee, E. S.; Lockett, E.; Uwabe, C.; Shiota, K.; Anderson, S. 9 10 A.; Lo, C. W. Developmental atlas of the early first trimester human embryo. Dev. Dynam. 2010, 11 12 13 239(61 1585-1595. 14 15 28. Waterland, R. A.; Michels, K. B.,Epigenetic epidemiology of the developmental origins 16 17 hypothesis. Anna. Rev. Nutr. 2007, 27, 363-388. 18 19 20 29. Sinclair, K. D.; Allegrucci, C.; Singh, R.; Gardner, D. S.; Sebastian, S.; Bispham, J.; 21 22 Thurston, A.; Huntley, J. F.; Rees, W. D.; Maloney, C. A. DNA mthylation, insulin resistance, 23 24 25 and blood pressure in offspring determined by maternal perconceptional B vitamin and 26 27 methionine status. Proc. Natl. Acad. Sci. USA 2007,104 (49), 19351-19356. 28 29 30 30. Tian, M.; Peng, S.; Martin, F. L.; Zhang, J.; Liu, L.; Wang, Z.; Dong, S.; Shen, H. 31 32 Perfluorooctanoic acid induces gene promoter hypermthylation of glutathione-S-transferase Pi 33 34 in human liver L02 cells. Toxicology 2012,2 9 6 ( 1-3), 48-55. 35 36 37 31. Harries, L. W.; Stubbins, M. J.; Forman, D.; Howard, G.; Wolf, C. R. Identification of 38 39 genetic polymorphisms at the glutathione-S-transferase Pi locus and association with 40 41 susceptibility to bladder, testicular and prostate cancer. Carcinogenesis 1997,1 8 (4), 641-644. 42 43 44 32. Guerrero-Preston, R.; Goldman, L. R.; Brebi-Mieville, P.; Ili-Gangas, C.; LeBron, C ; 45 46 Witter, F. R.; Apelberg, B. J.; Hemndez-Roystacher, M.; Jaffe, A.; Halden, R. U. Global DNA 47 48 hypomethylation is associated with in utero exposure to cotinine and perfluorinated alkyl 49 50 51 compounds. Epigenetics 2010, 5, (6), 539-546. 52 53 33. Anway, M. D.; Leathers, C.; Skinner, M. K. Endocrine disruptor vinclozolin induced 54 55 56 epigenetic transgenerational adult-onset disease. Endocrinology 2006, 147 (12), 5515-5523. 57 58 21 59 60 ACS Paragon Plus Environment Environmental Science & Technology Page 22 of 28 1 2 3 34. Casals-Casas, C.; Desvergne, B. Endocrine disruptors: from endocrine to metabolic 4 5 6 disruption. Anna. Rev. Physiol. 2011, 73, 135-162. 7 8 35. White, S. S.; Fenton, S. E.; Hines, E. P. Endocrine disrupting properties of 9 10 perfluorooctanoic acid. J. SteroidBiochem. Mol. Biol. 2011,127 (1-2), 16-26. 11 12 13 36. Fernandez, M. F.; Olmos, B.; Granada, A.; Lopez-Espinosa, M. J.; Moiina-Molina, J. M.; 14 15 Fernandez, J. M.; Cruz, M.; Olea-Serrano, F.; Olea, N. Human exposure to endocrine-disrupting 16 17 18 chemicals and prenatal risk factors for cryptorchidism and hypospadias: a nested case-control 19 20 study. Environ. Health Perspect. 2007,115 (S-l), 8-14. 21 22 23 37. Strohsnitter, W. C.; Noller, K L.; Hoover, R. K ; Robboy, S. J.; Palmer, J. R.; Tilus- 24 25 26 Ernstoff, L.; Kaufman, R. H.; Adam, E.; Herbst, A. L.; Hatch, E. E. Cancer Risk in Men Exposed 27 28 In Utero to Diethylstilbestrol../. Natl. Cancer Inst. 2001, 93 (7), 545-551. 29 30 31 38. Swan, S. H. Intrauterine exposure to diethylstilbestrol: Long-term effects in humans. 32 33 34 A P M IS im i, 109 (SI03), S210-S222. 35 36 37 39. Palmer, J. R.: Wise, L. A.; Hatch, E. E.; Troisi, R.; Titus-Ernstoff, L.; Strohsnitter, W,,; 38 39 Kaufman, R.; Herbst, A. L.; Noller, K. L.; Hyer, M.: Hoover, R. N. Prenatal Diethylstilbestrol 40 41 42 Exposure and Risk o f Breast Cancer. Cancer Epidemiol. Biomarkers Prev. 2006, 15 (8), 1509 43 44 1514. 45 46 47 40. Jacobson, J. L.; Jacobson, S. W. Intellectual Impairment in Children Exposed to 48 49 50 Polychlorinated Biphenyls in Utero. New Engl. J. Med. 1996, 335 (11), 783-789. 51 52 53 41. Knox, S. S.; Jackson, T.; Frisbee, S. J.; Javins, B.; Ducatman, A. M. Perfluorocarbon 54 55 exposure, gender and thyroid function in the C8 Health Project. J. Toxicol. Sci. 2011, 36 (4), 56 57 58 22 59 60 ACS Paragon Plus Environment Page 23 of 28 Environmental Science & Technology 1 2 3 403-410. 4 5 6 42. Zoeller, R. T.; Crofton, K. M. Thyroid hormone action in fetal brain development and 7 8 9 potential for disruption by environmental chemicals. Neuroloxicology 2000, 21 (6), 935-946. 10 11 12 43. Porterfield, S. P. Thyroidal dysfunction and environmental chemicals-potential impact 13 14 on brain development. Environ. Health Persped. 2000, JOS (Suppl 3), 433-438. 15 16 17 44. Escobar, G. M.; Obregon, M. J.; Del Rey, F. E. Role of thyroid hormone during early 18 19 brain development. Eur. J. Endocrinol. 2004,151 (Suppl 3), U25-U37. 20 21 22 23 45. Escobar, G. M.; Obregon, M. J.; Rey, F. E. Maternal thyroid hormones early in 24 25 pregnancy and fetal brain development. Best. Israel. Res. Clin. Endocrinol. Metab. 2004, 18 (2), 26 27 225-248. 28 29 30 31 46. Henrichs, J.; Bongers-Schokking, J. J.; Schenk, J. J.; Ghassabian, A.; Schmidt, H. G.; 32 33 Visser, T. J.; Hooijkaas, H.; de Muinck Keizer-Schrama, S. M. P. F.; Hofman, A.; Jaddoe, V. V. 34 35 W. Maternal thyroid function during early pregnancy and cognitive functioning in early 36 37 38 childhood: the generation R study../. Clin. Endocrinol. Metab. 2010,95 (9), 4227-4234. 39 40 41 47. Pop, V. J.; Kuijpens, J. L.; van Baar, A. L.; Verkerk, G.; van Son, M. M.; de Vijlder, J. J.; 42 43 Vulsma, T.; Wiersinga, W. M.; Drexhage, H. A.; Vader, H. L. Low maternal free thyroxine 44 45 46 concentrations during early pregnancy are associated with impaired psychomotor development in 47 48 infancy. Clin, endocrinol. 2001, JO (2), 149-155. 49 50 51 48. Smil, B.; Kok, J.; Vulsma, T.; Briet, J.; Boer, K_.; Wiersinga, W. Neurologic development 52 53 54 of the newborn and young child in relation to maternal thyroid function. Acta. Paediatrica 2012, 55 56 89 (3), 291-295. 57 58 23 59 60 AS Paragon Plus Environm ent Environmental Science & Technology Page 24 of 28 1 2 3 4 5 49. Abalovich, M.; Amino, N.; Barbour, L. A.; Cobin, R. H.; De Groot, L. J.; Glinoer, D.; 6 7 8 Mandel, S. J.; Stagnaro-Green. A. Management o f thyroid dysfunction during pregnancy and 9 10 postpartum: an Endocrine Society Clinical Practice Guideline.,/. Clin. Endocrinol. Metab. 2007, 11 12 92 (8 suppl), sl-s47. 13 14 50. Ahmed, O.; Abd El-Tawab, S.; Ahmed, R. Effects of experimentally induced maternal 15 16 hypothyroidism and hyperthyroidism on the development of rat offspring: 1. The development of 17 18 19 the thyroid hormones-neurotransmitters and adenosinergie system interactions. Ini. J. Devel. 20 21 Neurosci. 2010, 28 (6), 437-454. 22 23 24 51. Frisbee, S. J.; Shankar, A.; Knox, S. S.; Steenland, K.; Savitz, D. A.; Fletcher, T.; 25 26 27 Ducalmau, A. M. Perfluorooctanoic Acid, Perfluorooctanesulfonate, and Serum Lipids in 28 29 Children and Adolescents: Results From the C8 Health Project. Arch. Pediatr. Adolesc. Med. 30 31 2010,164 (9), 860-869. 32 33 34 35 52. Steenland, K.; Tinker, S.; Frisbee, S.; Ducatman, A.; Vaccarino, V. Association of 36 37 perfluorooctanoic acid and perfluorooctane sulfonate with serum lipids among adults living near 38 39 a chemical plant. Am. J. Epidemiol. 2009,170 (10), 1268-1278. 40 41 42 43 53. Napoli, C.; D'armiento, F.; Mancini, F.; Postiglione, A.; Witztum, J.; Palumbo, G.; 44 .45 Palinski, W. Fatty streak formation occurs in human fetal aortas and is greatly enhanced by 46 47 maternal hypercholesterolemia. Intimal accumulation o f low density lipoprotein and its oxidation 48 49 50 precede monocyte recruitment into early atherosclerotic lesions../. Clin. Invest. 1997, 100 (11), 51 52 2680-2690. 53 54 55 54. Palinski, W.; Napoli, C. The fetal origins of atherosclerosis: maternal 56 57 58 24 59 60 ACS Paragon Plus Environment Page 25 of 28 Environmental Science & Technology 1 2 3 hypercholesterolemia, and cholesterol-lowering or antioxidant treatment during pregnancy 4 5 6 influence in utero programming and postnatal susceptibility to atherogenesis. FASEB J. 2002, 16 7 8 (11), 1348-1360. 9 10 11 55. Okada, E.; Sasaki, S.; Saijo, Y.; Washino, N.; Miyashita, C.; Kobayashi, S.; Konishi, K.; 12 13 Ito, Y. M ; Ito, R.; Nakata, A.; Iwasaki, Y.; Saito, K.; Nakazawa, H.; Kishi, R. Prenatal exposure 14 15 to periluorinated chemicals and relationship with allergies and infectious diseases in infants. 16 17 18 Environ, lies. 2012, 1 1 2 ,118-125. 19 20 21 56. Grandjean, P.; Andersen, E. W.; Budtz-Jorgensen, E.; Nielsen, F.; Molbak, K.; Weihe, P.: 22 23 Heilmann, C. Serum vaccine antibody concentrations in children exposed to perfluorinated 24 25 26 compounds. JAMA 2012, 307 (4), 391-397. 27 28 29 57. Halldorsson, T. I.; Rytter, D.; Haug, L. S.; Bech, B. H.; Danielsen, I.; Becher, G.; 30 31 Henriksen, T. B.; Olsen, S. F. Prenatal exposure to perfluorooctanoate and risk of overweight at 32 33 34 20 years of age: a prospective cohort study. Environ. Health Perspect. 2012,120 (5), 668-673. 35 36 37 58. Liu, J.; Li, J.; Liu, Y.; Chan, H. M.; Zhao, Y.; Cai, Z.; Wu, Y. Comparison of gestation 38 39 and lactation exposure of perfluorinated compouds for newborns. Environ. Int. 2012, 37, 1206 40 41 42 1212. 43 44 45 46 Figure Legends 47 48 49 Finnre 1. Average PFOS and PFOA Concentrations in Pregnant and Non-Pregnant Women 50 51 52 Figure 2. Cross-Sectional Maternal Perfluorocarbon Concentrations by Trimester 53 54 55 56 57 58 25 59 60 ACS Paragon Pius Environm ent Environmental Science & Technology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Figure 1. Perfluorocarbon Averages in Pregnant and Non-Pregnant Women PFOA and PFOS Averages in Pregnant and Non-Pregnant Women 2S - - . . .. _ - - ... ................ - -- - ... .- ----------- PFOA ft?OS Mean Values and Standard Errors PFOA in non-pregnant women 20.25+1.13ng/ml, PFOA in pregnant women 14.611.20ng/ml, PFOS in non-pregnant women 14.371.09ng/ml, and PFOS in pregnant women 13.73+1.14ng/ml. ACS Paragon Plus Environm ent Page 26 of 28 Page 27 o f 28 Environmental Science & Technology 1 2 3 Figure 2. Cross-Sectional Maternal Perfluorocarbon Concentrations by Trimester 4 5 6 7 8 Cross-Sectional PFOA and PFOS Averages in each 9 10 Trimester 11 12 25 13 14 15 16 17 18 19 MPFOA 20 21 a PFOS 22 23 24 25 26 27 28 Non-Pregnant First Trimester Second Trimester Third Trimester 29 30 Mean Values and Standard Errors 31 32 Mean PFOA PFOA Standard Error Mean PFOS PFOS Standard Error 33 34 Non-Pregnant 20.25 1.13 14.37 1.09 35 First Trimester 15.66 36 Second Trimester 14.19 37 38 Third Trimester 14.35 +1.20 1.19 1.21 14.22 12.51 11.55 1.14 1.13 1.14 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Environmental Science & Technology Page 28 of 28 1 2 3 | 4 5 PFOA and PFOS Averages in Pregnant and Non 6 7 ; Pregnant Women 8 i 25 r * -- - ---------------------------------------------------- ------------ ---------------------9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Mean Values and Standard Errors 29 PFOA in non-pregnant women 20.25+1.13ng/ml, PFOA in pregnant women 14.61ll.20ng/m l, PFOS in 30 non-pregnant women 14.371.09ng/ml, and PFOS in pregnant women 13.731.14ng/ml. 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environm ent i ehponnne.org I IfIs I i ENVIRONMENTAL HEALTH PERSPECTIVES I I Serum Polyfluoroalkyl Concentrations, Asthma I Outcomes, and Immunological Markers in a Case-Control Study o f Taiwanese Children > Guang-Hui Dong, Kuan-Yen Tung, Ching-Hui Tsai, j Miao-Miao Liu, Da Wang, Wei Liu, Yi-He Jin, I Wu-Shiun Hsieh, Yungiing Leo Lee, and Pau-Chung Chen I I http://dx.doi.Org/10.1289/ehp.1205351 J Online 8 January 2013 National Institute of Environmental Health Sciences National Institutes o f Health U.S. Departm ent o f Health and Human Services 2 .i Page 1 of 32 Serum Polyfluoroalkyl Concentrations, Asthma Outcomes, and Immunological M arkers in a Case-Control Study of Taiwanese Children Guang-Hui Dong1-2, Kuan-Yen Tung', Ching-Hui Tsai3, Miao-Miao Liu1, Da Wang1, Wei Liu4, Yi-He Jin4, Wu-Shiun Hsieh5, Yungling Leo Lee3'6'7*, and Pau-Chung Chen6'8-9* `Department of Biostatistics and Epidemiology, and Department of Occupational and Environmental Health, School of Public Health, China Medical University, Shenyang, China departm ent of Epidemiology, School o f Public Health, Saint Louis University, Saint Louis, Missouri, USA 'Institute of Epidemiology and Preventive Medicine, College o f Public Health, National Taiwan University, Taipei, Taiwan 4School of Environmental Science and Technology, Dalian University of Technology, Dalian, China departm ent of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan departm ent of Public Health, College of Public Health, National Taiwan University, Taipei, Taiwan 'institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan "institute of Occupational Medicine and Industrial Hygiene, College o f Public Health, National Taiwan University, Taipei, Taiwan 9Department o f Environmental and Occupational Medicine, National Taiwan University College of Medicine and National Taiwan University Hospital, Taipei, Taiwan *Contributed equally to this work i Page 2 of 32 Corresponding Author: Yungling Leo Lee, MD, PhD, Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, No.17 Xuzhou Road, Taipei 100, Taiwan. E-mail: leolee(u)ntu.edu.tw. Telephone: +886-2-33668016 Running Head: PFCs and Asthma Outcomes in Children Competing Interests: The authors declare that they have no competing interests. Acknowledgments: This study was supported by Grant # 98-2314-B-002-138-MY3 and #101-2621 -M-002-005 from the National Science Council in Taiwan. The views expressed in this article are those of the authors and do not necessarily represent those o f the funding source. The funding source had no role in the design or analysis of the study publication. Key Words: Asthma, AEC, ECP, IgE, Perfluorinated compounds Abbreviations: ACT: Asthma Control Test AEC: absolute eosinophil counts ECP: eosinophilic cationic protein GBCA: Genetic and Biomarkers study for Childhood Asthma Page 3 of 32 IQR: interquartile range IgE: immunoglobulin E ORs: odds ratio 95%CI: 95% confidence intervals PFBS: perfluorobutane sulfonate PFCs: periluorinated compounds PFDA: perfluorodecanoic acid PFDoA: perfluorododecanoic acid PFHpA: perfluoroheptanoic acid PFHxA: perfluorohexane acid PFHxS: pertluorohexane sulfonate PFNA: perlluorononanoic acid (PFNA) PFOA: perfluorooctanic acid PFOS: perfluorooctane sulfonate PFTA: perfluorotetradecanoic acid Page 4 of 32 ABSTRACT Background: Perfluorinated compounds (PFCs) are ubiquitous pollutants. Experimental data suggest that they may be associated with adverse health outcomes, including asthma. However, there is little supporting epidemiological evidence. Methods: A total of 231 asthmatic children and 225 non-asthmatic controls, all from Northern Taiwan, were recruited in the Genetic and Biomarker study for Childhood Asthma. Structure questionnaires were administered by face-to-face interview. Serum concentrations of 11 PFCs and levels of immunological markers were also measured. Associations of PFC quartiles with concentrations of immunological markers and asthma outcomes were estimated using multivariable regression models. Results: Nine PFCs were detectable in most children (>84.4%), of which periluorooclane sulfonate (PFOS) was the most abundant (median serum concentration o f 33.9 ng/mL in asthmatics and 28.9 ng/mL in controls). Adjusted odds ratios for asthma among those with the highest versus lowest quartile of PFC exposure ranged from 1.81 (95%CI: 1.02, 3.23) for the perfluorododecanoic acid (PFDoA) to 4.05 (95%CI: 2.21, 7.42) for perfluorooctanic acid (PFOA). PFOS, PFOA, and subsets of the other PFCs were positively associated with serum IgE concentrations, absolute eosinophil counts (AEC), eosinophilic cationic protein (ECP) concentrations, and asthma severity scores among asthmatics. Conclusions: This study suggests an association between PFC exposure and juvenile asthma. Due to widespread exposure to these chemicals, these findings may be of potential public health concern. 4 Page 5 of 32 INTRODUCTION Perfluorinated compounds (PFCs) include a class of man-made organic chemicals comprised o f a fluorinated carbon backbone of varying length, terminated by a carboxylate or sulfonate functional group. Such PFCs are extremely stable, thermally, biologically, and chemically, and additionally possess hydrophobic and lipophobic characteristics that enable products coated in them to repel both oil and water, and resist staining (Conder et al. 2008; Hoffman et al. 2010). Accordingly, PFCs are widely used, for example in surfactants, emulsifiers, food packaging, nonstick pan coatings, fire-fighting foams, paper and textile coatings, and personal care products (Lau et al. 2007; Lindstorm et al. 2011; Renner 2001). This combination o f extreme resistance to degradation and environmental ubiquity has raised concerns in recent years (Giesy and Kannan 2001; Lau et al. 2007). Furthermore, studies have shown that PFCs accumulate among the higher trophic level of the food chain, such as predators and human beings (Conder et al. 2008; Hbude et al. 2006; Noorlander el al. 2011). Although data from the National Health and Nutrition Examination Survey have indicated a decrease in serum PFC concentrations in the general U.S, population since the production of some PFCs has been phased out (for example, PFOS decreased from 30.4 ng/mL in 1999 to 13.2 ng/inL in 2008) (Kato et al. 2011), PFCs are still manufactured abroad (Paul et al. 2009). PFCs bioaccumulate by binding to proteins in the liver and serum, in contrast with many other persistent organic pollutants that persist primarily in adipose tissue (Conder et al. 2008), and they are slowly eliminated without biotransformation (Lau et al. 2007). Estimated serum half-life estimates in an occupationally exposed cohort ranged from 5.4 years for perfluorooctane sulfonate (PFOS) to 8.5 years for perfiuorohexane sulfonic acid (PFHxS) (Olsen et al. 2007). Several attempts have been made to understand the toxicological hazards that may be 5 Page 6 of 32 associated with exposure. Early animal studies focused almost exclusively on exposure to PFOS and perfiuoroocianic acid (PFOA), two o f the most common PFCs. Evidence of several potential effects has been reported based on experimental studies, including hepatotoxieity, immunotoxicity, developmental toxicity, reproductive toxicity, neurotoxicity, endocrine toxicity, and tumors o f the liver, thyroid, and mammary glands (Corsini et al. 2011; Lau et al. 2004,2007; Olsen et al. 2009). Preliminary data suggest that PFCs have the potential to exacerbate atopic diseases such as asthma. In a murine model of asthma, Fairley et al. (2007) found that PFOA increased serum levels of immunoglobulin E (IgE) and enhanced the hypersensitivity response to ovalbumin, suggesting that PFOA exposure may augment the IgE response to environmental allergens. Another recent study reported that PFOS exposure decreased baseline airway resistance but significantly increased airway responsiveness in an allergic murine model (Basu et al. 2011). Furthermore, in our experimental studies, in vivo exposure to PFOS was associated with decreased secretion of ThI -type cytokines (1L-2 and IFN-y) and increased secretion of TH2-type cytokines (IL-4 and IL-10) and IgE, which suggested that PFOS exposure might shift the host's immune state toward a more T ^-lik e state (Dong et al. 2011). Th1/Th2 polarization (towards Th2 response) is a hallmark of atopy diseases (Colavita el al. 2000), and IgE is known to play a role in mediated type 1 hypersensitivity reactions, including asthma (Platts-Mills 2001). Accordingly, we hypothesized that exposure to PFCs may have a role asthma development in humans. Asthma is the most common respiratory disease in young children, and although recent studies indicated that asthma prevalence has plateaued or may be declining (Montefort et al. 2011; Pearce et al. 2007), it is still a major public health problem among young people. The 6 Page 7 of 32 case-control design of the Genetic and Biomarkers study for Childhood Asthma (GBCA) provided an opportunity to explore the association between PFCs exposure and asthma in children, in addition, two validated questionnaires [asthma severity score questionnaire and Asthma Control Test (ACT)j were used to examine the association between PFCs and asthma severity and control in asthmatic children. METHODS Study population The Genetic and Biomarkers study for Childhood Asthma (GBCA) was conducted between 2009 and 2010. A total of 231 ten to fifteen-year-old children with physician-diagnosed asthma in the previous year were recruited from two hospitals in Northern Taiwan. Controls were selected from our previous cohort study population in seven public schools of Northern Taiwan (Tsai et al. 2010). These schools had diverse geographical and socioeconomic settings, being located in city, rural, and high altitude communities, respectively; In each targeted school, children of the same age range and without a personal or family history of asthma were invited to participate, and 225 non-asthmatic controls enrolled in the study (response rate 72% among those contacted by phone). Information pertaining to demographic variables, environmental exposures and asthma outcomes were collected from questionnaires. We also collected urine and serum samples for each child after 8 hours of fasting. A trained field worker measured each child's height, weight, waist circumference and blood pressure. All participants and their parents signed written informed consents. The study protocol was approved by the Institutional Review Board (National Taiwan University Hospital Research Ethics Committee), and complied with the principles outlined in the Helsinki Declaration (Declaration of Helsinki., 1990). 7 Page 8 of 32 Serum IgE, AEC. ECP level detection Venous blood was placed into EDTA tubes, a peroxydase coloration was performed, and the absolute eosinophil counts (AEC) were calculated using automatic analyzer (>100()0 cells/mL; Sysntex, XE2100, Japan). Serum samples were stored at -80 C until total IgE and eosinophilic cationic protein (ECP) levels were analyzed. Serum total IgE levels were determined using a Pharmacia UniCap assay lest system (Pharmacia Diagnostics, Uppsala, Sweden). Total IgE concentrations below 0.35 kU/L were defined as absent or undetectable. ELISA kits were used to measure ECP levels in serum samples according to the manufacturer's instructions (R&D Systems Europe, Abingdon, UK). The limit o f quantitation for ECP concentrations was 0.125 ng/niL. Asthma Control Test and asthma Severity evaluation The Asthma Control Test (ACT), a five-item questionnaire used to assess asthma control in the previous 4 weeks (Nathan et al. 2004), was administered to the asthmatic children. Questions pertaining to asthmatic symptoms, use of rescue medication, and limitation of daily activities are used to ascertain asthma management. The reliability, empirical validity, and discriminative properties in assessing the control of asthma by Chinese children are good (Chen et al. 2008). The sum of the scores o f the five questions gives the total ACT score (range 5-25); the higher the score, the better controlled the disease. We also used a 13-item asthma severity score questionnaire to evaluate four overall components of asthma severity in the asthmatic children, including frequency of current asthma symptoms, use of systemic corticosteroids, use of other medications (besides systemic corticosteroids), and history of hospitalizations and intubations 8 Page 9 of 32 (Eisner et al. 1998), Possible total scores range from 0 to 28, with higher scores reflecting more severe asthma. PFC concentrations PFCs were measured from 0.5 mL of serum using Agilent high-performance liquid chromatography (HPLC) - in tandem with an Agilent 64 i0 Triple Quadruple (QQQ) mass spectrometer (MS/MS) (Agilent, Palo Alto and Santa Clara, CA), Detailed information about standards and reagents, sample preparation and extraction, instrumental analysis, quality assurance and quality control, and recovery experiments in the present study is provided in Supplementary Material (see Supplementary Material, p.2-p.4) and is described elsewhere (Bao el al. 2011). Ten PFCs were analyzed in serum samples: PFOS, PFOA, perfluorobutane sulfonate (PFBS), perfluorodecanoic acid (PFDA), perfluorododecanoic acid (PFDoA), perfluoroheptanoic acid (PFHpA), perfluorohexane acid (PFHxA), perfiuorohexane sulfonate (PFHxS), perlluorononanoic acid (PFNA), and perfluorotetradecanoic acid (PFTA). The limit of quantification (LOQ) for PFOS, PFOA and PFNA was 0.03 ng/mL, for PFBS and PFHxS was 0.07 ng/mL, for PFDA and PFDoA was 0.1 ng/mL, for PFHpA and PFHxA was 0.05 ng/mL, and for PFTA was 0.02 ng/mL. All tests were duplicated and average of the two measures Was calculated as the concentrations of PFC. Statistical analysis Statistical analysis was performed using SAS software (Version 9.2; SAS Institute Inc., Cary, North Carolina, USA). Concentrations of PFCs and biomarkers below the LOQ were assigned a value equal to the LOQ divided by the square root o f 2 for statistical analyses. We calculated 9 Page 10 of 32 univariate statistics, including medians, interquartile ranges (IQR), and ranges for each PFC. Because PFC concentrations were highly skewed, we utilized the Wilcoxon rank-sum test to compare PFC concentrations between children with and without asthma. We used logistic regression to estimate odds ratio (ORs) and 95% confidence intervals (CIs) for each PFC quartile relative to the lowest quartiie, with a priori adjustment for child age and gender. To determine the magnitude o f other potential confounding, we examined the following variables using a backward deletion strategy: parental education, body mass index (BM1), environmental tobacco smoke (ETS) exposure and month of survey. If the estimated PFC effect changed by at least 10% when a covariate was included in the base model, the covariate was included in the final model. Multiple general linear models were used to estimate associations with continuous outcomes (IgE, AEC, and ECP) in PFCs quartiles, with the lowest PFCs quartile as reference group and adjusted for identified covariates. These models were applied separately for cases and controls. We modeled an ordinal variable assigned the median value for each corresponding quartile to estimate p-values for trend. A p-value of < 0.05 was considered statistically significant. RESULTS Compared to children without asthma, asthmatic children tended to be younger and less likely to report ETS exposure (Table 1). In addition, asthmatic children had significantly higher median plasma concentrations of IgE, AEC, and ECP. Nearly all study participants had detectable serum concentrations o f all PFCs (>94% of PFCs) except for PFDoA (84.4% in children with and without asthma) and PFHpA (53.3% in unasthmatic children, and 70.6% in asthmatic children) (Table 2). Because of the large numbers to Page 11 of 32 of samples <LOQ, we did not conduct further analyses of PFHpA. Serum concentrations of PFCs were significantly higher in asthmatic children than in non-asthmatic children (p<0.05), except for PFHxA concentrations, which were similar in both groups, and PFTA concentrations, which were significantly higher in non-asthmatic children. Crude and adjusted ORs for asthma in association with the highest versus lowest quartile of exposure were significantly elevated for all PFCs except for PFHxA and PFTA (Table 3). In general, the data suggest increasing odds of asthma with increasing PFCs, with the strongest associations for exposures in the fourth quartile. Specifically, adjusted ORs for the highest versus lowest quartile were 2.63 (95%CI: 1.48, 4.69) for PFOS, 4.05 (95%CI: 2.21, 7.42) forPFOA, 1.90 (95%CJ: 1.08, 3.37) forPFBS, 3.22 (95%CI: 1.75, 5.94) for PFDA, 1.81 (95%CI: 1.02, 3.23) for PFDoA, 3.83 (95%CI: 2.11, 6.93) for PFHxS, and 2.56 (95%CI: 1.41, 4.65) for PFNA. None of the PFCs were significantly associated with serum levels of IgE or absolute eosinophil counts (AEC) among children without asthma, but serum ECP concentration was positively associated with PFDA and PFDoA (Table 4). In contrast, among children with asthma all three biomarkers were positively associated with PFOS and PFOA, with significant monotonic trends with increasing exposure (Table 5). For example, asthmatic children in the highest of PFOS quarti le had mean IgE levels of 877.3 IU/dL (95%CI: 695.2, 1059.5), compared with 517.9 IU/dL (95% Cl: 336.7, 699.2) for in the lowest quartile (Figure 1). In addition, with the exception of PFHxA, which was not associated with any of the biomarkers, all the remaining PFCs were associated with 2 of the 3 biomarkers evaluated. Among asthmatic children, positive trends for associations with asthma severity scores were significant for PFOS, PFDA, PFDoA and PFTA, but none of the PFCs was associated with ACT scores (Table 6). u Page 12 of 32 DISCUSSION Serum concentrations o f PFCs were significantly higher in asthmatic children compared with controls, and among children with asthma, all but one of the PFCs evaluated were positively associated with at least two of the three immunological biomarkers (IgE, AEC, and ECP). Although temporality cannot be determined due to the cross-sectional nature of the data, and non-causal associations due to uncontrolled confounding or other sources of bias cannot be ruled out, the robust associations of PFCs with asthma and asthma related biomarkers in children suggest that a causal relationship may be present. However, it should also be noted that concentrations of individual PFCs were positively correlated (see Supplemental Material, Table S l ), and therefore it is not possible to determine if associations apply to multiple PFCs or to only a subset of individual PFCs. There is little information in the literature about associations between environmental PFCs and asthma or asthma-related biomarkers in children. In a systematic Medline search, we identified only two studies of PFCs and atopic disease in humans. The first was a cross-sectional study of 566 residents with prolonged exposure to PFOA in their drinking water (Anderson-Mahoney et al. 2008). In that study, respiratory illness was evaluated by questionnaire, and standardized prevalence ratios (SPR) using national Health and Examination Survey (NHANES) data for comparison rates suggested an increased prevalence of asthma among the exposed participants than in the general U.S. population (SPR=1.82, 95%CI: 1.47, 2.25). The second was a cohort study o f prenatal exposure to PFCs in association with IgE levels and atopic dermatitis in 244 newborns (Wang et al. 2011). In that study, prenatal PFOA and PFOS exposures were positively correlated with cord blood IgE levels, but were not significantly associated with atopic dermatitis. 12 Page 13 of 32 In a murine model of asthma, Fairley et al. (2007) evaluated the effects of PFOA dermal exposure on the hypersensitivity response to ovalbumin (OVA). The authors reported that IgE increased to a greater extent in animals exposed to PFOA and OVA, and that the severity of the OVA-specific airway hyperreactivity response, and a pleiotropic cell response characterized by eosinophilia and mucin production, increased with increasing concentrations of PFOA. Grasty et al. (2005) evaluated the effect of prenatal PFOS exposure on maturation of the terminal airway epithelium in rais based on histological and morphometric examination, and reported that there were significant histologic and morphometric differences between control and PFOS-lreated lungs in newborns, suggesting that PFOS may inhibit or delay perinatal lung development. Also, a recent study using an allergic murine model to evaluate effects of PFOS exposure on pulmonary function and airway responsiveness reported that PFOS exposure decreased baseline airway resistance, but significantly increased airway responsiveness in allergic mice (Basu et al. 2 0 1J). In our experimental studies, in vivo exposure to PFOS was linked decreased secretion of Ini-type cytokines (IL-2 and IFN-y), and increased secretion of TfI2-type cytokines (IL-4 and IL-I0) and serum IgE, which suggested that PFOS exposure might shift immune responses toward a more T|[2-like state (Dong et al. 2011). Th1/Th2 polarization towards Th2 responses is a hallmark of atopy diseases (Coiavita et al. 2000), and IgE plays a role in mediating type 1 hypersensitivity reactions, including asthma (Platts-Mills 2001). Therefore, we hypothesize that exposure to PFCs may augment the TM2 response, resulting in airway hyper-reactivity to environmental allergens. Potential mechanisms for effects of PFCs on immune response and asthma development in humans are uncertain. One possibility is an effect of PFCs on regulatory T cells that influence the development of immune-related diseases including asthma and allergy (Akbari el al. 2003). 13 Page 14 of 32 Another possible mechanism for effects of PFCs on immune responses pertains to the peroxisome proliferator activated receptors (PPARs) signaling pathway. PFCs are known as agonists for peroxisome proliferator activated receptors (PPARs) (Vanden huevel et al. 2006). Both PPAR a and y are expressed on cells of the monocyte / macrophage lineage, suggesting a possible role in immune function (Braissant and Wahli 1998). Although PPAR y activation has been implicated as an important contributor to the pathogenesis o f the toluene diisocyanate-induced asthma phenotype in a female BALB/c mice model (Lee et al. 2006), PFOC and PFOA have been shown to significantly increase activation of mouse or human PPAR a and [L but not of PPAR y, in vitro (Takacs and Abbott, 2006). Lovett-Racke et al. (2004) reported that PPAR a agonists, including gemfibrozil, ciprofibrate, and fenofibrate, can increase the production of the Th2 cytokine IL-4, and suppress MBP A c l-ll induced proliferation by TCR transgenic T cells. In addition, gemfibrozil shifted cytokine secretion by inhibiting interferon-y and promoting IL-4 secretion in human T-cell lines. These results suggest that PFCs may potentially augment the Th2 response and subsequent airway hyperreactivity to environmental allergens through a PPAR-mediated mechanism (Fairley et al. 2007). In the present study, nearly all study participants had detectable serum concentrations o f all PFCs, including both asthmatic children and controls. PFOS was the most abundant PFC in the serum measured in 2009 - 2010 in these Taiwanese children, with median concentrations (28.9 ng/rnL in controls and 33.9 ng/mL in asthmatics) that were similar to levels reported for children age 12 - 19 years in the general U.S. population in 1999 - 2000 (29.4 ng/mL), but higher than the median concentration reported for U.S. children in 2007 - 2008 (11.3 ng/mL) (Kalo el al. 2011) and concentrations reported for other populations of children sampled during the mid- to late- 2000s (see Supplemental Material, Table S2). In contrast, median concentrations of PFOA 14 Page 15 of 32 in the present study population (0.2 ng/mL in controls and 1.2 ng/mL in asthmatics) were lower than reported for other populations o f children during the late 2000s (e.g., 4.0 ng/mL based on NHANES data for 2007 - 2008), and substantially lower than median concentrations reported for children living in a U.S. community near a manufacturing facility (26.3 - 32.6 ng/mL) (Frisbee et al. 2010). Differences in PFC concentrations among populations may reflect changes in exposures over time, as well as differences in diet and other sources of exposure, and individual differences in rates or patterns of metabolism or excretion. The limitations o f these analyses should be noted. We based the PFCs measures on a single serum sample, and although PFCs have a half-life o f 5.4 to 8.5 years (Olsen et al. 2007), samples taken at several time points might be more accurate than a single sample for classifying exposure. As previously noted, this is a cross-sectional study, and temporal relationships between exposures and outcomes cannot be established. In addition, associations with individual PFCs may be biased due to correlations with other PFCs. Finally, cases were recruited from two hospitals in Northern Taiwan, whereas controls were recruited from schools in the same region. Therefore, estimates also may have been influenced by selection bias or uncontrolled confounding. In conclusion, we observed positive associations between serum PFCs and asthma, and positive associations between PFCs and IgE, AEC and ECP levels, and (to a lesser extent) asthma severity scores, in asthmatic children. These findings suggest that exposure to PFCs may not only be related to asthma outcomes but also to asthma severity._Although the production of some PFCs has been phased out in the United States and Europe, PFCs are still manufactured in Asia and elsewhere, and exposure may also result from the breakdown o f similar compounds to PFCs in the environment (Organisation for Economic Co-operation and Development 2002; U.S. 15 Page 16 of 32 EPA 2006). Therefore, continued exposure is of public health concern, and additional research on potential immunoloxic effects of PFCs is warranted. 16 Page 17 of 32 REFERENCES Akbari 0 , Stock P, DeKruyffRH, Umetsu DT. 2003. Role of regulatory T cells in allergy and asthma. Curr Opin Immunol 15:627-633. Anderson-Mahoney P, Kotlerman J, Takhar H, Gray D, Dahlgren J. 2008. Self-reported health effects among community residents exposed to perfluorooctanoate. New Solut 18:129-143. Bao J, Liu W, Liu L, Jin Y, Dai J, Ran X, et al. 2011. Perfluorinated compounds in the environment and the blood of residents living near fluorochemical plants in Fuxin, China. Environ Sci Technol 45:8075-8080. Basu S. Becker A, Bondy G, Haiayko A, Kozyrskyj A, Loewen M. 2011. The Impact O f A Perfluorinated Compound (PFC) On Airway Function In An Allergic Murine Model. ATS. Available: http://myats201 Lzerista.com/event/member?itern_id=996058. [accessed 17 February 2012) Braissant O, Wahli W. 1998. Differential expression of peroxisome proliferator-activated receptor-alpha, -beta, and -gamma during rat embryonic development. Endocrinology 139:2748-2754. C8 Science Panel (Tony Fletcher, Kyle Steenland, David Savitz). 2009. Status Report: PFOA and immune biomarkers in adults exposed to PFOA in drinking water in the mid Ohio valley. Available: http://www.c8sciencepanel.org/pdfs/Status_Report_C8_and_Imm une_markers_ M arch2009.pdf [accessed 24 January 2012] Chen HH, Wang JY, Jan RL, Liu YH, Liu LF. 2008. Reliability and validity of childhood asthma control test in a population of Chinese asthmatic children. Qual Life Res 17:585-593. Corsini E, Avogadro A, Galbiati V, deH'Agli M, Marinovich M, Galli CL, et al. 2011. In vitro evaluation of the immunotoxic potential of perfluorinated compounds (PFCs). Toxicol Appl Pharmacol 250:108-116. Colavita AM, Reinach AJ, Peters SP. 2000. Contributing factors to the pathobiology of asthma. The Thl/Th2 paradigm. Clin Chest Med 21:263-277. 17 Page 18 of 32 Conder JM, Hoke RA, De Wolf W, Russell MH, Buck RC. 2008. Are PFCAs bioaccumulative? A critical review and comparison with regulatory criteria and persistent lipophilic compounds. Environ Sci Technol 42.995-1003. Declaration o f Helsinki. 41st World Medical Assembly. 1990. Declaration of Helsinki: recommendations guiding physicians in biomedical research involving human subjects. Bull Pan Am Health Organization 24: 606-609. Dong GH, Liu MM, Wang D, Zheng L, Liang ZF, Jin YH. 2011. Sub-chronic effect of perfluorooctanesulfonate (PFOS) on the balance of type 1 and type 2 cytokine in adult C57BL6 mice. Arch Toxicol 85:1235-1244. Eisner MD, Katz PP, Yelin EH, Henke J, Smith S, Blanc PD. 1998. Assessment of asthma severity in adults with asthma treated by family practitioners, allergists, and pulmonologists. Med Care 36:1567-1577. Fairley KJ, Purdy R, Kearns S, Anderson SE, Meade BJ. 2007. Exposure to the immunosuppressant, perfluorooctanoic acid, enhances the murine IgE and airway hyperreactivity response to ovalbumin. Toxicol Sci 97:375-383. Fei C, Olsen J. 2011. Prenatal exposure to perfluorinated chemicals and behavioral or coordination problems at age 7 years. Environ Health Perspect 119:573-578. Frisbee SJ, Shankar A, Knox SS, Steenland K, Savitz DA, Fletcher T, et al. 2010. Perfluorooctanoic acid, perfluorooctanesulfonate, and serum lipids in children and adolescenLresults from the C8 Health Project. Arch Pediatr Adolesc Med 164:860-869. Giesy JP, Kannan K. 2001. Global distribution of perfluorooctane sulfonate in wildlife. Environ Sci Technol 35:1339-1342. Grasty RC, Bjork JA, Wallace KB, Wolf DC, Lau CS, Rogers RJ. 2005. Effects of prenatal perfluorooctane sulfonate (PFOS) exposure on lung maturation in the perinatal rat. Birth Defects Res B Dev Reprod Toxicol 74:405-416. Haug LS, Thomsen C, Becher G 2009. Time trends and the influence o f age and gender on serum concentrations of perfluorinated compounds in archived human samples. Environ Sci is Page 19 of 32 Technol 43: 2131-2136. Hemat H, Wilhelm M, Volkel W, Mosch C, Fromme H, Witlsiepe J. 2010. Low serum levels of perfluorooctanoic acid (PFOA), perfluorooctane sulfonate (PFOS) and perfluorohexane sulfonate (PFHxS) in children and adults from Afghanistan. Sci Total Environ 408:3493-3495. Hoffman K, Webster TF, Weisskopf MG, Weinberg J, Vieira VM. 2010. Exposure to polyfluoroalkyl chemicals and attention deficit/hyperactivity disorder in U.S. children 12-15 years o f age. Environ Health Perspect 118:1762-1767. Houde M, Bujas TA, Small J, Wells RS, Fair PA, Bossart GD, el al. 2006. Biomagnification of perfluoroalkyl compounds in the bottlenose dolphin (Tursiops truncatus) food web. Environ Sci Technol 40:4138-4144. Kato K, Wong L-Y, Jia LT, Kuklenyik Z, Calafat AM. 2011. Trends in exposure to polyfluoroaikyl chemicalS'in the U.S. population: 1999-2008. Environ Sci Technol 45: 8037-8045. Kudo N, Kawashima Y. 2003. Toxicity and toxicokinetics of perfluorooctanoic acid in humans and animals. J Toxicol Sci 28:49-57, Lau C, Anitole K, Hodes C, Lai D, Pfahles-Hutchens A, Seed J. 2007. Perfluoroalkyl acids: a review of monitoring and toxicological findings. Toxicol Sci 99:366-394. Lau C, ButenhoiTJL, Rogers JM. 2004. The developmental toxicity of perfluoroalkyl acids and their derivatives. Toxicol Appl Pharmacol 198:231-241. Lee KS, Park SJ, Kim SR, Min KH, Jin SM, Lee HK, et al. 2006. Modulation of airway remodeling and airway inflammation by peroxisome proliferator-activated receptor gamma in a murine model of toluene diisocyanate-induced asthma. J Immunol 177:5248-5257. Lindstrom AB, Strynar MJ, Libelo EL. 2011. Polyfluorinated compounds: past, present, and future. Environ Sci Technol 45: 7954-7961. 9 Page 20 of 32 Lovett-Racke AE, Hussain RZ, Northrop S, Choy J, Rocchini A, Matthes L, et al. 2004. Peroxisome proliferator-activated receptor alpha agonists as therapy for autoimmune disease. J Immunol 172:5790-5798. Montefort S. Ellul P, Monteforl M, Camana S, Agius Muscat H. 2011. A decrease in the prevalence and improved control of allergic conditions in 13- to 15-yr-old Maltese children (ISAAC). Pediatr Allergy Immunol 22:eI07-el 11. Nathan RA, Sorkness CA, Kosinski M, Schatz M, Li JT, Marcus P, et al. 2004. Development of the asthma control test: a survey for assessing asthma control. J Allergy Clin Immunol 113: 59-65. Noorlander CW, van Leeuwen SP, Te Biesebeek ID, Mengelers MJ, Zeilinaker MJ. 2011. Levels of perfluorinated compounds in food and dietary intake of PFOS and PFOA in the Netherlands. J Agric Food Chem 59:7496-7505. Olsen GW, Burris JM, Ehresman DJ, Froehlich JW, Seacat AM, Butenhoff JL, et al. 2007. Half-life of serum elimination of perfluorooctanesulfonate,perfluorohexanesulfonate, and perfiuorooctanoate in retired fluorochemical production workers. Environ Health Perspect 115:1298-1305. Olsen GW, Butenhoff JL, Zobel LR. 2009. Perfluoroalkyl chemicals and human fetal development: an epidemiologic review with clinical and toxicological perspectives. Reprod Toxicol 27:212-230. OECD (Organisation for Economic Co-operation and Development).2002. Hazard Assessment of Perfluorooctane Sulfonate (PFOS) and Its Salts. U.S. EPA Docket AR-226-1140. 2002. Available: http://www.oecd.Org/dataoecd/23/l 8/2382880.pdf [accessed 9 January 2012). Paul AG, Jones KC, Sweetman AJ. 2009. A first global production, emission, and environmental inventory for perfluorooctane sulfonate. Environ Sei Technol 43: 386-392. Pearce N, At-Khaled N, Beasley R, Mallol J, Keil U, Mitchell E, et al, 2007. Worldwide trends in the prevalence of asthma symptoms: phase III of the International Study ofAsthma and Allergies in Childhood (ISAAC). Thorax 62:758-766. 20 Page 21 of 32 Peden-Adams MM, Keller JM, EuDaly JG 2008. Suppression of humoral immunity in mice following exposure to perfluorooctane sulfonate. Toxicol Sci 104:144-154. Platts-Mills TA. 2001. The role o f immunoglobulin E in allergy and asthma. Am J Respir Crit Care Med 164:S1-S5. Renner R. 2001. Growing concern over perfluorinated chemicals. Environ Sci Technol 35:154A-160A. Takacs ML, Abbott BD. 2007. Activation of mouse and human peroxisome proliferator-activated receptors (alpha, beta/delta, gamma) by perfluorooctanoic acid and perfluorooctane sulfonate. Toxicol Sci 95:108-117. Toms LM, Calafat AM, Kato K, Thompson J, Harden F, Hobson P, et al. 2009. Polyfluoroalkyl chemicals in pooled blood serum from infants, children, and adults in Australia. Environ Sci Technol 43:4194-4199. Tsai CH, Huang JH, Hwang BF, Lee YL. 2010. Household Environmental Tobacco Smoke and Risks of Asthma, Wheeze and Bronchitic Symptoms among Children in Taiwan. Repir Res 11:11. U.S. EPA (U.S. Environmental Protection Agency). Basic Information on PFOA. 2006. Available: http://www.epa.gov/oppt-intr/pfoa/pubs/pfoainfo.htm [accessed 9 January 2012]. Vanden Heuvel JP, Thompson JT, Frame SR, Gillies PJ. 2006. Differential-activation o f nuclear receptors by perfluorinated fatty acid analogs and natural fatty acids: a comparison of human, mouse, and rat peroxisome proliferator-activated receptor-alpha, -beta, and -gamma, liver X receptor-beta, and retinoid X receptor-alpha. Toxicol Sci 92:476-489. Wang U, Hsieh WS, Chen CY, Fletcher T, Lien GW, Chiang HL, et al. 2011. The effect of prenatal perfluorinated chemicals exposures on pediatric atopy. Environ Res 111:785-791. 21 Page 22 of 32 Table 1. Characteristics of children with and without asthma in the study population Characteristic Children without asthma (n=225) Children with asthma (n=231) Age (year)" 13.6+/-0.7 12.9+/-1.7 Height (cm)" 159.8+/-7.0 156.6+/-10.4 Weight (kg)a 52.5+/-13.2 49.8+/-13.3 BMI (kg/m2)" 20.4+/-4.1 20.1+/-3.9 Gender, n(%) Male 102 (45.3) 158(68.4) Female 123 (54.7) 73 (31.6) Parental education, n (%) <High School 86 (38.2) 90 (39.0) >High School 139 (61.8) 141 (61.0) ETS exposure, n (%) No 93 (41.3) 138 (59,7) Ever 22 (9.8) 23(10.0) Current 110 (48.9) 70 (30.3) Month o f survey 7-9 156(69.3) 106 (45.9) 11-12 69 (30.7) 125(54.1) IgE (lU/dL)a 331.4+/-486.6 684.6+/-679.2 AEC (x Io'Vl)" 152.3+/-150.3 395.0+/-280.9 ECP (pg/L)a 28.4+/-41.1 42.2+/-57.8 "Mean +/'- SD. AEC, absolute eosinophil count; ECP, eosinophilic cationic protein P value <0.001 <0.001 0.033 0.379 <0.001 0.871 <0.001 <0.001 <0.001 <0.001 0.004 22 Page 23 of 32 Table 2. Serum PFC concentrations (ng/mL) in children with and without asthma rrt PFOS PFOA PFBS PFDA PFDoA PFHpA PFHxA PFFlxS PFNA PFTA Children without asthma (n=225) Mean +/- SD 33.4+/- 26.4 1.0-/-1.1 0.5 +/- 0.2 1.0+/-0.5 4.5 +/- 6.0 0.2 +/- 0.3 0.2 +/- 0.2 2.1 +/- 2.2 0.9 +/- 0.3 28.9+/- 81.6 Median (IQR) 28.9 (14.1, 43.0) 0.5 (0.4, 1.3) 0.5 (0.4, 0.5) 1.0 (0.8, 1.2) 2.7 (0.8-6.0) 0.2 (LOQ, 0.2) 0.2 (0.1, 0.3) 1.3 (0.6, 2.8) 0.8 (0.6, 1.1) 5.2 (0.4, 23.3) Range LOQ-148.1 LOQ-1I.3 IOQ-2.7 LOQ-5.0 LOQ-43.1 LOQ-4.3 LOQ-2.4 LOQ-11.8 0.26-2.5 LOQ-793.6 % above LOQ 97.3 95.1 94.2 95.1 84.4 53.3 98.7 99.6 100.0 99.6 Children with asthma (n=231) ____________ _ P Mean +/- SD Median (1QR) Range % above LOQ value* 45.5 +/- 37.3 33.9 (19.6, 61.1) LOQ-149.6 1.5+/- 1.3 1.2 (0.5, 2.2) LOQ-9.0 973) 0.002 99.6 <0.001 0.5 +/- 0.2 0.5 (0.4, 0.6) LOQ-2.7 99.1 0.022 1.2 +/- 0.5 1.1 (0.9, 1.5) LOQ-3.5 97.4 <0.001 5.8+/- 6.0 3.8 (1.1, 8,4) LOQ-36.1 84.4 0.014 0.3 +/- 0.5 0.2 (LOQ, 0.3) LOQ-5.0 70.6 <0.001 0.3 +/- 0.3 0.2 (0.1, 0.3) LOQ-3.9 97.0 0.765 3.9+/- 9.0 1.1 +/- 0.5 2.5 (1.3, 4.3) LOQ-129.1 1.0 (0.7, 1.3) 0.28-3.6 . 98.3 <0.001 100.0 <0.001 54.6+/- 1.01.3 4,1 (0.2, 31.7) LQQ-429.1 99.6 0.003 " ,,ua,u ucv'auuu' inierquarme range; LUQ, Limit o f quantitation; Wtlcoxon rank-sum test to compare the difference of PFCs levels between children without asthma and children with asthma. 23 Page 24 of 32 Table 3. Association between PFCs and asthma among 456 participants in the Genetic and Biomarkers study for Childhood Asthma, Taiwan, 2009-2010________________________ Exposure No. No. Controls Cases Crude OR (95%CI) Adjusted OR (95%CI)3 PFOS Quartile 1(lowest)6 67 47 1.00 1.00 Quartile 2 54 60 1.59(0.94,2.67) 1.96(1.11,3.47) Quartile 3 64 50 1.11 (0.66, 1.88) 1.32 (0.75, 2.32) Quartile 4 (highest) 40 74 2.64(1.54,4.51) 2.63(1.48,4.69) l ` for trend1 0.003 0.003 PFOA Quartile 1(lowest)6 71 43 1.00 1.00 Quartile 2 64 50 1.29(0.76,2.19) 1.58 (0.89,2.80) Quartile 3 53 61 1.90(1.12,3.22) 2.67(1.49,4.79) Quartile 4 (highest) 37 77 3.43 (1.99, 5.93) 4.05 (2.21, 7.42) P for trend1' <0.001 <0.001 PFBS Quartile 1(lowest)6 63 51 1.00 1.00 Quartile 2 56 58 1.28 (0.76,2.15) 1.31 (0.74,2.31) Quartile 3 58 56 1.19(0.71,2.01) 1.24 (0.70, 2.20) Quartile 4 (highest) 48 66 1.70(1.01,2.87) 1.90(1.08,3.37) P for trend' 0.072 0.021 PFDA Quartile I(lowest)6 70 44 1.00 1.00 Quartile 2 68 46 1.08 (0.64, 1.83) 1.02 (0.58, 1.80) Quartile 3 53 61 1.83 (1.08,3.10) 1.30 (0.72, 2.33) Quartile 4 (highest) 34 80 3.74 (2.16,6.49) 3.22(1.75,5.94) P for trendc <0,001 <0.001 PFDoA Quartile I(lowest)6 60 54 1.00 1.00 Quartile. 2 61 53 0.97 (0.57,. 1.'62) 0,81 (0.46, 1.43) Quartile 3 63 51 0.90(0.53, 1.52) 0.71 (0.40, 1.26) Quartile 4 (highest) 41 73 1.97(1.16,3.36) 1.81 (1.02,3.23) P for trend' 0.021 0.044 PFHxA Quartile 1(lowest)6 54 60 1.00 1.00 Quartile 2 56 58 0.93 (0.55, 1.57) 1.21 (0.69,2.12) Quartile 3 68 46 0.61 (0.36, 1.03) 0.90 (0.51,1.60) Quartile 4 (highest) 47 67 1.28(0.76, 2.17) 1.60 (0.90,2.86) P for trend' 0.706 0.168 PFHxS Quartile l(lowcst)6 72 42 1.00 1.00 Quartile 2 69 45 1.12(0.66, 1.91) 1.54 (0.85,2.77) Quartile 3 45 69 2.63 (1.54, 4.49) 2.94(1.65,5.25) Quartile 4 (highest) 39 75 3.30(1.92,5.67) 3.83 (2.11,6.93) P for trend' <0.001 <0.001 PFNA Quartile 1(lowest)6 69 45 1.00 1.00 Quartile 2 64 50 1.20 (0.71,2.03) 1.19(0.68,2.09) Quartile 3 53 61 1.76(1.04, 2.99) 1.54 (0.86,2.76) Quartile 4 (highest) 39 75 2.95(1.72,5.06) 2.56(1.4U 4.65) P for trend' <0.001 0.001 PFTA Quartile I(lowest)6 52 62 1.00 1.00 Quartile 2 56 58 0.69(0.41, 1.16) 0.62(0.35,1.09) Quartile 3 63 51 0.61 (0.36, 1.02) 0.65(0.37, 1.14) Quartile 4 (highest) 54 60 0.84 (0.50, 1.41) 0.96(0.55, 1.67) P for trend' 0.410 0.899 24 Page 25 of 32 3Adjusted for age, gender, BMI, parental education, ETS exposure, and month of survey. bRcference category. ''P values were calculated using categories representing the median value of corresponding quartiie. 25 Page 26 of 32 Table 4. Estimated mean values (95% Cl) serum IgE, AEC, and serum ECP according to serum PFCs concentration among children without asthma (n=225)a________________________ Exposure IgE (IU/dL) AEC (x]09/L) ECP (pg/L) PFOS Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trcndb PFOA Quartile I (lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trend" PFBS Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trend1' PFDA Quartile 1(lowest) Quartile 2 286.3(157.0,415.6) 298.3 (164.6. 432.1) 403.5(274.1,532.9) 336.3 (208.3,464.2) 0.404 223.1 (76.8,369.5) 298.9(170.9,427.0) 406.2 (274.6, 537.9) 393.9 (258.0, 529.8) 0.123 360.1 (229.6, 490.7) 345.0(214.4,475.7) 329.4(198.2,460.6) 291.8 (161.9,421.8) 0.447 248.6(122.2,375.0) 305.4(179.9,430.8) 138.9(100.1, 177.8) 141.6(102.2, 181.1) 167.8 (128.9, 206.6) 160.9(122.2, 199.7) 0.445 118.5 (78.6, 158.5) 110.1(71.3,148.8) 198.3(160.7, 235.9) 182.0(139.3,224.8) 0.224 156.0(117.6, 194.3) 108.2 (70.5, 145.9) 151.7(113.2, 190.3) 194.1 (155.6, 232.6) 0.070 118.2 (79.7, 156.7) 148.0(109.7, 186.3) 29.4(18.5, 40.3) 24.2(13.0,35.4) 33.5 (22.5, 44.5) 26.6(15.8,37.4) 0.972 15.4 (3.2,27.7) 28.3 (17.6, 39.0) 38.3 (27.3, 49.3) 31.2 (19,8, 42.6) 0.133 23.9(13.0,34.9) 32.1 (21.2,43.0) 28.9(17.9,40.0) 28.7(17.7, 39.8) 0.648 18.1 (7.2,28.9) 26.7(15.8, 37.5) Quartile 3 Quartile 4 (highest) P for trend" PFDoA Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trend" PFHxA Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trend" PFHxS Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) 395.6 (267.4. 523.9) 379.4 (253.4, 505.5) 0.092 358.4 (230.9, 485.8) 423.7 (293.2, 554.2) 281.8(151.8,411.8) 261.9 (132.4, 391.4) 0.145 215.2(83.7,346.7) 386.7 (257.7,515.9) 427.9 (299.6, 556.2) 296.7(163.5,429.9) 0.330 257.1 (125.3,389.0) 390.6 (259.3, 521.9) 363.6(233.5,493.8) 316.0(180.4,451.7) 178.7(140.3,217.1) 164.6(125.6, 203.6) 0.073 89.9 (52.7, 127.1) 190.6(152,4, 228.7) 172.0(134.5,209.6) 158.6(120.8, 196.4) 0.067 127.4 (87.5, 167.4) 158.3 (119.4,197.0) 178.7 (140.1, 217.2) 144.6(104.8, 184.3) 0.104 182.3 (141.9, 222.8) 119.3 (80.5, 158.1) 147.3(108.1, 186.4) 159.0(120.0, 198.0) 28.1 (17.4, 38.8) 40.7(30.0,51.4)* 0.004 19.2 (8.6, 29.7) 25.9(15.1,36.7) 21.7(11.0, 32.4) 46.5(36.1,57.0)' 0.001 25.2 (13.9,36.5) 26.0(15.1,37.0) 32.1 (21.2, 43.0) 30.4(19.0,41.7) 0.429 24.7(13.4,36.1) 39.6 (28.7, 50.5) 25.4 (14.3, 36.5) 24.2(13,2,35.2) P for trend" 0.581 PFNA Quartile 1(lowest) 278.4(150.9, 405.9) 0.321 138.6(100.5, 176.7) 0.537 28.2(17.4, 39.0) Quartile 2 331.3 (202.8,459.9) 122.7 (83.7,161.6) 20.1 (9.2,31.1) Quartile 3 237.5 (108.4,366.6) 172.2 (134.1, 210.3) 28.7(17.7, 39.7) Quartile 4 (highest) P for trend" PFTA Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trend" 474.1 (347.8, 600.4) 0.084 275.6(142.6, 408.6) 330.5 (199.6,461.5) 344.3(212.6, 476.0) 375.0 (245.6, 504.4) 0.293 175.5 (136.7,214.3) 0.086 156.7(117.9, 195.5) 133.1 (93.6, 172.6) 161.0(121.8, 200.3) 158.5 (118.9, 198.0) 0.954 36.4(25.7, 47.1) 0.167 29.7(18.9,40.6) 34.9 (23.9, 46.0) 28.6(17.6, 39.6) 20.4 (9.3,31.6) 0.196 26 Page 27 of 32 "Models were adjusted for age, gender, BMI, parental education, ETS exposure, and month of survey. bP values were calculated using categories representing the median value of corresponding quartile. Compared with the lowest of quartile: p<0.05. AEC, absolute eosinophil count; ECP, eosinophilic cationic protein Page 28 of 32 Table 5. Estimated mean values (95% Cl) serum IgE, AEC, and serumECP according to serum PFCs concentration among children with asthma (n=231)a IgE (lU/dL) PFOS Quartile 1(lowest) 517.9 (336.7, 699.2) Quartile 2 686.2(501.3,871.1) Quartile 3 658.1 (475.2,841.1) Quartile 4 (highest) 877.3 (695.2, P for trend1' 0.008 PFOA Quartile 1(lowest) 512.1 (329.4,694.8) Quartile 2 604.6 (422.1,787.1) Quartile 3 788.2 (607.1, 969.2) Quartile 4 (highest) 836.4 (652.0, 1020.8)* P for trend1* 0.005 PFBS Quartile I(lowest) 683.6 (497.0, 870.2) Quartile 2 601.2 (416.7, 785.7) Quartile 3 Quartile 4 (highest) P for trend1' 671.3 (485.9,856.8) 780.6 (598.4, 962.7) 0.496 PFDA Quartile 1(lowest) 470.6 (289.7, 651.6) Quartile 2 615.6(436.0, 795.2) Quartile 3 789.8(608.2,971.4) Quartile 4 (highest) 862.2 (682.7, 1041.7)* P for trend1' 0.001 PFDoA Quartile 1(lowest) 533.0(348.1,717.9) Quartile 2 653.4 (470.7, 836.0) Quartile 3 726.7 (546.6, 906.8) Quartile 4 (highest) 823.5 (640)0, 1006.9) P for lrendb 0.016 PFHxA Quartile 1(lowest) 539.7 (355.0, 724.4) Quartile 2 744.7 (561.1,928.2) Quartile 3 661.7(480.3,843.1) Quartile 4 (highest) 794.9(610.9, 978.8) P for trcndb 0.075 PFHxS Quartile 1(lowest) 682.4 (495.0, 869.7) Quartile 2 643.0 (456.1,830.0) Quartile 3 679.9 (495.4, 864.5) Quartile 4 (highest) 734.2 (549.1,919;4) P for frendb 0.632 PFNA Quartile 1(lowest) 410.9 (230.6, 591.2) Quartile 2 704.5 (524.1,884.9) Quartile 3 828.8(651.6, 1006.0)* Quartile 4 (highest) 790.9(610.1,971.6)* /' for trendb 0.001 PFTA Quartile l(lowest) 541.9(356.4,727.5) Quartile 2 659.9 (493.4, 826.5) Quartile 3 709.6 (507.5, 911.8) Quartile 4 (highest) 833.1 (650.7, 1015.5) P for trcndb 0.011 AEC (*107U 329.4 (255.8,403.0) 368.6(293.9, 443.3) 431.3(358.1,504.6) 453.4 (379.4, 527.3) 0.009 325.9 (253.7, 398.1) 339.7 (266.8,412.6) 422.1 (349.9,494.2) 498.0 (423.7, 572.3) <0.001 343.0 (268.8,417.2) 374.0 (301.6,446.5) 380.4 (307.2,453.5) 487.4 (413.4,561.4)* 0.009 333.6 (256.0, 407.2) 351.7(277.7,425.8) 422.2 (349.3, 495.0) 470.8 (398.6, 542.9)* 0.004 344,1 (270.9,417.2) 385.2 (313.4,457.0) 356.3 (282.9, 429.8) 496.9 (423.8,570.0)* 0.011 397.5(323.2,471.7) 369.7 (294.7, 444.7) 371.9(298.4, 445.4) 443.4(368.9,517.9) 0.407 331.7 (256.3,407.2) 379.1 (305.2,453.1) 430.5 (356.8,504.2) 439.7(365.5,513.6) 0.029 309.7 (236.4,383.0) 353.1 (280.3,425.8) 431.7(359.5, 503.9) 482.5 (411.1,553.9)' <0.001 328.8 (262.9, 394.8) 351.4 (270.5,432.3) 405.0 (332.7, 477.2) 502.3 (429.4,575.1)* <0.001 ECP (ug/L) 25.9(10.4,41.3) 37.4 (21.9, 52.8) 43.5 (27.5, 59.4) 62.4 (46.3, 78.4)' 0.001 30.3 (14.3,46.3) 34.8(18.9, 50.7) 44.3 (28.4, 60.2) 57.8 (42.2, 73.4) 0.010 32.6(16.3,48.9) 44.8(29.1,60.5) 42.9(27.0, 58.8) 47.3(31.1,63.6) 0.210 19.0(3.6, 34.3) 45.3 (29.9, 60.7) 44.7 (28.9,60.6) 59.7 (44.0, 75.3)* 0.001 28.7(13.3, 44.1) 36.3 (20.7, 52.0) 42.8 (27.3, 58.2) 62.0 (45.8, 78.2)* 0.003 36.9(20.8,53.0) 31.5(15.3,47.8) 49.5 (33.8, 65.2) 49.0 (33.3,64.6) 0.148 25.8 (9.5,42.0) 39.6 (24.0, 55.2) 41.0 (25.3, 56.6) 61.0(45.4, 76.6)* 0.004 28.8(13.1,44.4) 34.8(19.3, 50.3) 43.5 (27,6, 59.5) 61.0(45.3,76.6)' 0.003 36.8 (20.9, 52.8) 39.1 (23.3, 54.9) 49.9 (31.9,67.9) 43.6 (28.9, 58.2) 0.409 28 Page 29 of 32 '`Models were adjusted for age, gender, BMI, parental education, ETS exposure, and month of survey. bP values were calculated using categories representing the median value of corresponding quartilc. Compared with the lowest of quartilc: p<0.05. AEC, absolute eosinophil count; ECP, eosinophilic cationic protein 29 Page 30 of 32 Table 6. Estimated mean scores (95% Cl) for the relationship between the PFCs levels and asthma severity, asthma control test (ACT) among children with asthma (n=231)a Exposure PFOS Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trcndb PFOA Quartile (lowest) Asthma severity score 3.33 (2.36,4.31) 4.18(3.19,5.17) 4.49 (3.52, 5.45) 4.57(3.61,5.54) 0.045 3.63 (2.66,4.60) ACT score 22.51 (21.71,23.32) 22.72 (21.92, 23.52) 22.13(21.30,22.94) 22.21 (21.41, 23.02) 0.450 22.02 (21.22, 22.82) Quartile 2 3.99 (3.02,4.96) 22.14 (21.33, 22.96) Quartile 3 Quartile 4 (highest) P for trend1' PFBS Quartile I (lowest) 4.39 (3.40, 5.38) 4.57 (3.59, 5.55) 0.119 3.48 (2.50,4.47) 22.76 (21.96, 23.56) 22.65 (21.84,23.45) 0.168 ' 22.23 (21.41,23.04) Quartile 2 Quartile 3 Quartile 4 (highest) P for trend11 PFDA Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trcndb PFDoA Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trendb PFHxA Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trendb PFHxS Quartile 1(lowest) 4.42 (3.46,5.38) 3.82 (2.85,4.79) 4.85 (3.88, 5.82) 0.092 3.43 (2.48, 4.37) 3.79 (2.83,4.74) 4.07(3.10, 5.04) 5.32 (4.36,6.29)' 0.005 3.68.(2.71,4.65) 3.50 (2.55,4.45) 4.50 (3,53, 5.46) 4.91 (3.94, 5.88) 0.024 4.34 (3.36, 5.32) 4.10(3.13,5.07) 4.06 (3.08,5.04) 4.07 (3.09, 5.05) 0.854 3.96 (2.99,4.94) 22.39(21.60,23.19) 22.73 (21.92, 23.53) 22.23 (21.42, 23.03) 0.836 22.35 (21.54,23.16) 22.55 (21.76, 23.35) 22.33 (21.53, 23.13) 22.34 (21.53,23.15) 0.857 22.57(21.77,23.38) 22.54(21.75, 23.33) 21.90 (21.10, 22.70) 22.57(21.77,23.38) 0.709 22.21 (21.40,23.01) 22.41 (21.62,23.20) 21.92 (21.12, 22.72) 23.03 (22.23, 23.83) 0.284 21.97(21.17, 22.78) Quartile 2 Quartile 3 Quartile 4 (highest) P for trendb PFNA Quartile 1(lowest) Quartile 2 Quartile 3 Quartile 4 (highest) P for trend0 PFTA Quartile 1(lowest) Quartile 2 4.17(3.19, 5.14) 4.44 (3.46, 5.42) 4.01 (3.02, 5.00) 0.722 4.05 (3.08, 5.02) 3.62 (2.64, 4.60) 4.25 (3.29, 5.21) 4.65 (3.68, 5.63) 0.217 3.40 (2.51,4.28) 4.45 (3.39, 5.52) 22.39(21.59,23.19) 22.48 (21.68,23.29) 22.72 (21.91,23.54) 0.251 22.35 (21.55, 23.15) 22.79 (21.98, 23.60) 22.09(21.30,22.88) 22.35 (21.55,23.16) 0.695 22.57(21.84,23.30) 22.33 (21.44, 23.22) Quartile 3 Quartile 4 (highest) P for trendb 4.05 (3 08,5.01) 4.89 (3.92,5.86) 0.050 22.01 (21.21,22.81) 22.63(21.82, 23.44) 0.917 30 Page 31 of 32 3Models were adjusted for age, gender, BMI, parental education, ETS exposure, and month of survey. bCompared with the lowest of quartile: p<0.05. . 'P values were calculated using categories representing the median value of corresponding quartile. 31 Page 32 of 32 FIGURE LEGEND . Figure 1. Immunological markers (A) IgE, (B) AEC and (C) ECP among asthmatic children according to quartilcs of PFOS exposure. The data are expressed as estimated mean with 95%CI adjusted for age, gender, BMI, parental education, ETS exposure, and month of survey. P values of trend are calculated using categories representing the median value of corresponding quartile (Quartilc 1: < 19.64 ng/mL; Quartile 2: 19.64-33.85 ng/mL; Quartile 3: 33.85-61.08 ng/mL; Quartile 4: > 61.08 ng/mL). ""Compared with the lowest of quartile (Quartile 1): P < 0.05. Cvartito i MUilt) Qumli-X ? Qocmi ( rP H V vfklinin* tjwmi* t Q j& Q t.' t f c a K i QmWW 4 **FOSIoclO Environmental Health Perspectives Supplemental Material h ttp M lx doi.org/10 1289/ehp. 1205351 Serum Polyfluoroalky! Concentrations, Asthma Outcomes, and Immunological Markers in a Case-Control Study of Taiwanese Children Authors Guang-Hui Dong, Kuan-Yen Tung, Ching-Hui Tsai, Miao-Miao Liu, Da Wang, Wei Liu, Yi-He Jin, Wu-Shiun Hsieh, Yungling Leo Lee, Pau-Chung Chen Table of Contents Standards and Reagents: page 2 Sample Preparation and Extraction: page 2 Instrumental Analysis: page 3 Quality Assurance and Quality Control: page 3-page 4 Supplementary Material, Table SI: page 5 Supplementary Material, Table S2: page 6 References: page 7-page 8 l Environmental Health Perspectives h ttp M x dot org/10. 1289/ehp 1205351 STANDARDS AND REAGENTS The potassium salt of heptadecafluorooctane sulfonate (PFOS, 98%) was acquired from Fluka (Steinheim, Germany). The potassium salts o f perfluorohexane sulfonate (PFHxS, 98%), periluoroheptanoic acid (PFHpA, 98%), perfluorohexane acid (PFHxA, 98%) were acquired from interchim (Montlucon, France). The potassium salt of nonafluoro-1-butanesulfonate (PFBS, 98%) and hepatadecafluoropelargonic acid (PFNA, 95%) were acquired from Tokyo Chemical industry (Tokyo, Japan). Pentadecafluorooctanoic acid (PFOA, 95%) was purchased from Wako Pure Chemical Industries (Osaka, Japan). Nonadecafluorodecanoic acid (PFDA, 96%) and perfluorododecanoic acid (PFDoA, 97%) were purchased from Acres Organics (Geel, Belgium). Perfluorotetradecanoic acid (PFTA, 97%) was purchased from Aldrich (Steinheim, Germany). Tetrabutylammonium hydrogensulfate (TBAHS) of HPLC grade and anhydrous extra pure sodium carbonate (NuiCO.t, 99.5%) were obtained from Acros Organics (Geel, Belgium). HPLC grade ammonium acetate was obtained from Dikma Technology (Richmond, VA). HPLC grade methyl tert-butyl ether (MTBE), methanol, and acetonitrile were obtained from Tedia (Fairfield, OH). Milli-Q water was cleaned using Waters Oasis HLB Plus (225 mg) cartridges (Milford, MA) to remove the potential residue of PFCs. Mixed stock PFC standard solution was prepared in methanol. All reagents were used as received. SAMPLE PREPARATION AND EXTRACTION Serum samples were extracted following a method developed by Hansen et al. (2001). Two millilitres of 0.25 M Na^COi and one liter o f 0.5 M TBAHS were added to 0.5 mL o f serum and then extracted twice with MTBE. The combined MTBE extracts were brought to dryness under a gentle stream o f high purity nitrogen, and reconstituted in 1 mL mixture o f methanol and 10 mM ammonium acetate (2:3, v/v) before final filtration with a 0.22 pm nylon filter. Environmental Health Perspectives hltp.Hdx doi.org/I0.1289Jehp 1205351 INSTRUMENTAL ANALYSIS Extracts of serum samples were analyzed via high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). Chromatography was performed by an Agilent 1200 HPLC system (Palo Alto, CA). A 25 pL aliquot o f extract was injected onto a 2.1 x 100 mm (3.5 pm) Agilent Eclipse Plus C18 column (Palo Alto, CA) with 10 mM ammonium acetate and acetonitrile as mobile phases starting with 40% acetonitrile at a flow rate of 250 pL/min and column temperature o f 40 C. The gradient was increased to 90% acetonitrile at 9 min and then held for 2 min. In addition, an 8 min re-equilibration interval was run before each following sample. The HPLC system was interfaced to an Agilent 6410 Triple Quadrupole (QQQ) mass spectrometer (Santa Clara, CA) operated with electrospray ionization (ESI) in negative mode. Instrumental parameters were optimized to transmit the [M-K.]' ion before fragmentation to one or more product ions. Declustering potential and collision energies were optimized for each analyte and ranged from 35 to 90V and 10 to 35eV, respectively. Data were acquired by tandem mass spectrometry using multiple reaction monitoring (MRM) at transitions, 499 > 99 for PFOS, 413 > 369 for PFOA, 299 > 99 for PFBS, 399 > 99 for PFHxS, 363 > 3 19 for PFHpA, 313 > 269 for PFHxA, 463 > 419 forPFNA, 513 > 469 for PFDA, 613 > 569 for PFDoA, 713 > 669 for PFTA. Moreover, multiple daughter ions were monitored for confirmation, but quantitation was based on a single production. In all cases, the capillary was held at -4 kV and the desolvation temperature was kept at 350 C. QUALITY ASSURANCE AND QUALITY CONTROL Procedural blanks were prepared at an interval o f every ten samples to check if contamination had occurred during the extraction of samples. Solvent blanks containing acetonitrile and Milli-Q water (2:3, v/v) were run after every twenty samples to monitor for background contamination. Duplicate injections and calibration check standards were run after every twenty samples to assure the precision and accuracy o f each run. Matrix spike recoveries were tested by spiking native 3 Environmental Health Perspectives http:f/dx doi.org/10.1289/ehp. 1205351 standards of all 10 target compounds into 12 randomly selected samples, at levels o f 10 ng and 20 ng for each of the target compounds. All the matrix spike samples were analyzed in duplicate. Recoveries of native standards spiked in serum matrix were 98 5%, 101 6%, 95 3%, and 96 5%, for PFOS, PFOA, PFHxS, and PFNA, respectively. The average recoveries for other perfluorochemicals ranged from 82% to 96%. The relative standard deviations (RSD) of duplicate analyses were less than 5% for PFOS, PFOA, PFBS, PFDA, PFHpA, PFHxA, PFHxS, PFNA, PFTA.and less than 10% for PFDoA. The concentrations of serum extracts were quantified via nine-point matrix-matched calibration curves ranging from 0.01 to 100 ng/mL, which were performed by adding mixed PFC standard solution into blank and newborn bovine serum, respectively. The regression coefficients ( r ) o f calibration curves for all the target analytes in different matrixes were higher than 0.99. The limit of detection (LOD) was defined as the peak of analyte that needed to yield a signal-to-noise (S/N) ratio o f 3:1, and the limit of quantification (LOQ) was defined as the lowest point on the standard curve, above the LOD, with a RSD less than 10%. 4 Environmental Health Perspectives http://dx.doi.org/10.I28P/ehp.i205351 Supplemental Material, Table SI. Spearman's rank correlation coefficients (rho) among different PFCs in blood samples (n=456)______________________________________________________________ PFC PFOS PFOA PFBS PFDA PFDoA PFHpA PFHxA PFHxS PFNA PFTA PFOS 1.00 **0.64 **0.27 **0.34 **0.53 **0.27 **0.21 **0.37 **0.35 **0.24 PFOA 1.00 **0.42 **0.41 **0.34 **0.43 **0.26 **0.59 **0.46 *0.11 PFBS 1.00 **0.30 **0.28 **0.29 **0.19 **0.21 **0.18 **0.22 PFDA 1.00 **0.53 **0.31 0.09 **0.42 **0.79 **0.17 PFDoA 1.00 **0.33 *0.10 **0.28 **0.41 **0.71 PFHpA 1.00 **0.29 **0.27 **0.22 **0.24 PFHxA 1.00 **0.33 0.06 0.02 PFHxS 1.00 **0.50 0.05 PFNA 1.00 *0.16 PFTA **p<().0|; *p<0.05 1.00 5 Environmental [feuIth Perspectives http://dxdoi.org/10.1289/ehp. 1205351 Supplemental Material, Table S2. Median serum PFOS and PFOA concentrations (ng/mL) reported for other populations_______________________________________________________________________ Reference Studies/Country Time Age Sample (years) Median PFOS PFOA Kato ctal. (2011) NHANES (USA) 1999-2000 543 12-19 29.4 5.60 NHANES (USA) NHANES (USA) 2003-2004 2005-2006 640 640 12-19 12-19 19.9 14.9 4.00 3.80 NHANES (USA) 2007-2008 357 12-19 11.3 4.00 Frisbee et al. (2010) C8 Health Project (USA) 2005-2006 6536 1-11.9 20.7 32.6 C8 Health Project (USA) 2005-2006 5934 12-17.9 19.3 26.3 OECD (2002) USA 1995 599 2-12 37.5a -- Holzer et al. (2008) Srcgcn, Gennari 2006 80 5-6 5.2* 5.2" Amsberg, German 2006 90 5-6 5.4a 24.6a Turgeon et al. (2012) Canada 2006-2008 86 1-4.5 3.4* 1.74 Tomset al. (2009) Australia 2006-2007 -- 6-9 18.3 8.2 -- 9-12 17.7 7.0 Zhang ct al. (2010) China 2009 85 5-10 5.6 2.2 Fei and Olsen(2011) Danish 1998-2002 787 Pregnancy 34.4 5.4 OECD (2002) USA 2000 645 20-69 34.9" -- Sagamihara, Japan 1999 32 Adults 40.3 b -- Tokyo,Japan 1999 30 Adults 52,3 b -- ' Geometric mean; "Arithmetric mean; 6 Environmental Health Perspectives http://dx.doi.org/JO. 1289/ehp. 1205351 References Fei C, Olsen J, 2011. Prenatal exposure to perfluorinated chemicals and behavioral or coordination problems at age 7 years. Environ Health Perspect i 19:573-578. Frisbee SJ, Shankar A, Knox SS, Steenland K, Savitz DA, Fletcher T, et al. 2010. Perfluorooctanoic acid, perfluorooctanesulfonate, and serum lipids in children and adolescents: results from the C8 Health Project. Arch Pediatr Adolesc Med 164:860-869. Hansen KJ, Clemen LA, Ellefson ME, Johnson HO. 2001. Compound-specific, quantitative characterization of organic fluorochemicals in biological matrices. Environ Sei Technol 35: 766-770. Hang LS, Thomsen C, Becher G 2009. Time trends and the influence of age and gender on serum concentrations of perfluorinated compounds in archived human samples. Environ Sei Technol 43: 2131-2136. Hlzer J, Midasch O, Rauchfuss K, Kraft M, Reupert R, Angerer J, et al. 2008. Biomonitoring of perfluorinated compounds in children and adults exposed to periluorooctanoate-contaminated drinking water. Environ Health Perspect 116:651-657. Kato K, Wong L-Y, Jia LT, Kuklenyik Z, Calafat AM. 2011. Trends in exposure to polyfluoroalkyl chemicals in the U.S. population: 1999-2008. Environ Sei Technol 45: 8037-8045. OECD (Organization for Economic Co-operation and Development) Hazard assessment of perXuorooctane sulfonate (PFOS) and its salts. 2002. Available at: http://www.oecd.org/dataoecd/23/18/2382880.pdf [accessed 17 June 2012] Toms LM, Calafat AM, Kato K, Thompson J, Harden F, Hobson P, et al. 2009. Polyfluoroalkyl chemicals in pooled blood serum from infants, children, and adults in Australia. Environ Sei Technol 43:4194-4199. Turgeon O'Brien H, Blanchet R, Gagn D, Lauzire J, Vzina C, Vaissire E, et al. 2012. Exposure to toxic metals and persistent organic pollutants in Inuit children attending childcare centers in Nunavik, Canada. Environ Sei Technol 46:4614-4623. 7 pnviromiieiitul Health Perspectives http://dx doi. org/10.1289/ehp. 1205351 Zhang T, Wu Q, Sun HW, Zhang XZ, Yun SH, Kannan K. 2010. Periluorinated compounds in whole blood samples from infants, children, and adults in China. Environ Sci Technol 44:4341-4347. 8 ENVIRONMENTAL HEALTH PERSPECTIVES Perfluorooctanoic Acid Exposure and Cancer Outcomes in a Contaminated Community: A Geographic Analysis Vernica M. Vieira, Kate Hoffman, Hyeong-Moo Shin, Janice M. Weinberg, Thomas F. Webster, Tony Fletcher http://dx.doi.org/10.1289/ehp.1205829 Online 8 January 2013 tesm Page 1 of 26 Perfluorooctanoic Acid Exposure and Cancer Outcomes in a Contaminated Community: A Geographic Analysis Vernica M. Vieira1,2,K ate Hoffman1'3, Hyeong-Moo Shin4,5, Janice M. Weinberg6, Thomas F. W ebster1, Tony Fletcher7 'Department o f Environmental Health, Boston University School of Public Health, Boston, Massachusetts, USA Program m Public Health, Chao Family Comprehensive Cancer Center, University of California, Irvine, California, USA 3University o f North Carolina, Gillings School o f Global Public Health, Chapel Hill, North Carolina, USA, 4School of Social Ecology, University of California, Irvine, California, USA 5 Department o f Public Health Sciences, University of California, Davis, California, USA d ep artm en t of Biostatistics, Boston University School of Public Health, Boston, Massachusetts, USA Department o f Social and Environmental Health Research, London School o f Hygiene and Tropical Medicine, London, United Kingdom Corresponding author: Vernica M. Vieira, Program in Public Health, AIRB 2042, University of California, Irvine, CA 92697; vvieira@uci.edu: 949.824.7017 (tel); 949.824.0527 (fax) Running title: PFOA exposure and cancer 1 Page 2 of 26 Key Words: C8, GIS, kidney cancer, PFOA, testicular cancer Acknowledgements: We acknowledge the staff at the West Virginia Cancer Registry (Office of Epidemiology and Prevention Services, Bureau for Public Health, West Virginia Department of Health and Human Resources, 350 Capitol Street, Room 125, Charleston, WV 25301-3715) and the Ohio Cancer Incidence Surveillance System (OCISS), Ohio Department o f Health (ODH) for their assistance with the cancer data. OCISS is a registry participating in the National Program of Cancer Registries o f the Centers for Disease Control and Prevention (CDC). Information about OCISS can be obtained at http://www.healthyohioprogram.org/cancer/ocisshs/ci_survl.aspx. Use of these data does not imply ODH or CDC either agrees or disagrees with any presentations, analyses, interpretations or conclusions. The project was supported by the C8 Class Action Settlement Agreement (Circuit Court o f Wood County, WV, USA) between DuPont and plaintiffs, which resulted from releases into drinking water o f the chemical perfluorooctanoic acid (PFOA, or C8). Funds were administered by the Garden City Group (Melville, NY) that reports to the court. Our work and conclusions are independent o f either party to the lawsuit. Competing interests: The authors declare they have no competing interests. Abbreviations: Cl: Confidence Interval GIS: Geographic Information Systems LACS: Locatable Address Conversion System OCISS: Ohio Cancer Incidence Surveillance System 2 Page 3 of 26 OH: Ohio OR: Odds Ratio PFOA, C8: Perfluorooctanoic acid SES: Socioeconomic Status SIR: Standardized Incidence Ratio WD: Water District WV: West Virginia 3 Page 4 of 26 A bstract B ackground: Perfluorooctanoic acid (PFOA) has been linked to cancer in occupational mortality studies and animal toxicological research. Objective: We investigated the relationship between PFOA exposure and cancer among residents living near the DuPont plant in Parkersburg, West Virginia (WV). M ethods: Our analyses included incident cases o f 18 cancers diagnosed from 1996-2005 in five Ohio (OH) counties and eight WV counties. For analyses o f each cancer outcome, controls comprised all other cancers in the study dataset except kidney, pancreatic, testicular, and liver cancers, which have been associated with PFOA in animal or human studies. We applied logistic regression models to individual-level data to calculate odds ratios (OR) and confidence intervals (Cl). For the combined analysis of WV and OH data, the exposure of interest was resident water district. Within OH, geocoded addresses were integrated with a PFOA exposure model to examine the relationship between cancer odds and categories of estimated PFOA serum. Results: Our final dataset included 7,869 OH cases and 17,238 WV cases. There was a positive association between kidney cancer and the very high and high serum exposure categories (OR: 2.0, 95% Cl: 1.0, 3.9; n=9 and OR: 2.0, 95% Cl: 1.3, 3.2; n=22, respectively) and a null association with the other exposure categories compared to the unexposed. The largest OR was for testicular cancer with the very high exposure category (OR: 2.8, 95% Cl: 0.8, 9.2; n=6) but there was an inverse association with the lower exposure groups, and all estimates are imprecise because o f small case numbers. Conclusions: This study suggests that higher PFOA serum levels may be associated with testicular, kidney, prostate, and ovarian cancers and non-Hodgkin's lymphoma. Strengths of this 4 Page 5 of 26 study include near-complete case ascertainment for state residents, and well characterized contrasts in predicted PFOA serum levels from 6 contaminated water supplies. 5 Page 6 of 26 Introduction The current study uses geographic methods to investigate the relationship between exposure to PFOA (perfluorooctanoic acid, C8) and patterns o f cancer risk in the mid-Ohio River Valley using data from the West Virginia (WV) and Ohio (OH) cancer registries. This work is one o f a series of studies investigating health effects o f PFOA exposure among residents living near the Washington Works DuPont Teflon-manufacturing plant in Parkersburg, WV (Steenland et al. 2009). PFOA was released into the environment via aerial emissions and discharged into the Ohio River since the 1950s, resulting in the contamination o f the local drinking water (Paustenbach et al. 2007; Shin et al. 201 la). In addition to hundreds o f impacted private drinking water wells, six nearby public water districts in WV and OH were also contaminated (Figure 1), and monitoring data show that even after a drastic reduction in releases, PFOA contamination o f drinking water persisted and continued to increase in some water districts near the plant (Shin et al. 2012). As part o f a settlement from a large class action lawsuit against DuPont, the C8 Science Panel (2012) was established to investigate potential health effects resulting from PFOA exposure, and a one-year cross sectional survey (2005-2006), known as the C8 Health Project, was conducted among over 69,000 residents with a minimum o f one year residency in public water districts contaminated by PFOA (Frisbee et al. 2009). Measured mean PFOA levels in public drinking water supplies at the time o f the survey ranged from 0.03 pg/L in Mason, WV to 3.49 pg/L in Little Hocking, OH, and private drinking water was measured at levels as high as 22.1 pg/L (Shin et al. 201 la). The median serum PFOA level in this cross-sectional study population was 28.2 pg/L with a range o f 0.2 to 22,412 pg/L (Steenland et al. 2009). PFOA is 6 Page 7 of 26 also detected in the serum o f the general U.S. population, albeit at a much lower median level o f 3.9 pg/L (Calafat et al. 2007). The stain-resistant and water-repellant properties o f PFOA make it widely used, and given its persistence it is ubiquitous in many indoor environments, including homes and work places (Fraser et al. 2012; Haug et al. 2011). Animal toxicologic data links PFOA to pancreatic cancer (acinar cells), testicular cancer (Leydig cells), and liver cancer (Lau et al. 2007). A recent review o f the epidemiologic data concluded that more studies were needed to determine if any potential health effects exist, and, specifically, that the evidence for cancer is not conclusive (Steenland et al. 2010). Fluman data for cancer from two occupational cohorts are limited to mortality and are based on small numbers. One of the two cohorts showed excesses o f kidney cancer (Leonard et al. 2008), while the other showed positive exposure-response trends for pancreatic and prostate cancer which were not statistically significant (Lundin et al. 2009). In a prospective Danish cohort study, plasma concentrations of background PFOA exposures were not associated with prostate, bladder, pancreatic, or liver cancer (Eriksen et al. 2009). A casecontrol study o f Greenland Inuit women found a positive but not statistically significant association between PFOA exposure and breast cancer (Bonefeld-Jorgensen et al. 2011). The positive associations were generally not consistent among cancer sites between studies and for the remaining cancer sites reported, no associations were observed. The objective o f this study is to investigate the association between PFOA exposure and the odds o f cancer using data from the WV and OH cancer registries. The current study includes residents o f exposed water districts and unexposed geographic areas outside o f the C8 Health Project area, enabling a comparison between populations exposed to varying degrees and unexposed populations while controlling for individual-level risk factors. Results o f this 7 Page 8 of 26 geographic study complement other studies being done within the C8 Health Project population by including a more complete ascertainment o f cases, including those who died prior to the 2005-6 survey. The weight o f evidence from the combination o f studies within this population was heavily considered in the determination by the C8 Science Panel that there was a probable link between PFOA exposure and testicular and kidney cancers. Methods Study Population and Data The study area encompasses 6 contaminated public water districts (WD) and 13 counties in WV and OH that surround the Washington Works DuPont facility (Figure 1). Incident cancer cases diagnosed from 1996-2005 in the OH counties o f Athens, Meigs, Gallia, Washington, and Morgan and the WV counties o f Wood, Mason, Wirt, Putnam, Jackson, Pleasants, Ritchie, and Cabell, were obtained from the OH Cancer Incidence Surveillance System (OCISS) and from the WV Cancer Registry (WVCR), respectively. The WVCR provided an incident cancer dataset of all cancer types with a total of 19,716 individual cases. There were 10,044 (51%) male cases and 9,673 (49%) female cases. The OCISS provided us with registry data for 9,402 incident cases o f all cancer types. For our analyses, we were able to geocode 8,650 (92%) o f the OH addresses at diagnosis to the street level and the remaining 752 (8%) at the ZIP code level, with little variation in these proportions by cancer type. There were 4,396 (51%) male cases and 4,254 (49%) female cases. The median age for both datasets was 67 years. We excluded 745 OH cases and 2,411 WV cases o f cancer types including oral cavity, pharynx, esophagus, larynx, stomach, and Hodgkin's lymphoma with too few cases (<100 OH cases, the smaller o f the two analyses) for meaningful analysis, or that have not been previously investigated in relation to PFOA in 8 Page 9 of 26 animal toxicologic studies or occupational mortality studies (Lau et al. 2007; Leonard et al. 2008; Steenland and Woskie 2012). We also excluded 36 OH cases and 67 WV cases that were diagnosed at less than 15 years o f age. Our final dataset included 7,869 geocoded OH cases and 17,238 WV cases o f 18 cancer categories (bladder, brain, female breast, cervix, colon/rectum, kidney, leukemia, liver, lung, melanoma o f the skin, multiple myeloma, non-Hodgkin's lymphoma, ovary, pancreas, prostate, testis, thyroid, and uterus). Based on 2010 U.S. census population estimates, the population o f the study area is over 500,000, with one third in OH and two thirds in WV. Using a PFOA exposure model and data collected from the C8 Health Project, the corresponding 1995 median PFOA serum concentrations in the 6 public WDs were previously estimated as follows: Little Hocking (Washington and Athens Counties, OH) =125 pg/L; Lubeck (Wood County, WV) = 65.8 pg/L; Tupper Plains (Athens and Meigs Counties, OH) = 23.9 pg/L; Belpre (Washington County, OH) = 18.7 pg/L; Pomeroy (Meigs County, OH) = 10.7 pg/L; and Mason (Mason County, WV) = 5.3 pg/L (Shin et al. 201 lb). The Institutional Review Boards at the Boston University Medical Campus, the London School of Hygiene and Tropical Medicine, the OH Department o f Health, and the WV Bureau o f Health Statistics have approved the research. This study was granted a waiver o f informed consent. Overview o f Analyses The final dataset included information for study area residents diagnosed with 18 different categories o f cancer. We applied logistic regression to individual-level data using registry-based cancer controls to calculate adjusted odds ratios (ORs) and confidence intervals (CIs) for each cancer category, with the other cancer categories excluding kidney, pancreatic, 9 Page 10 of 26 testicular, and liver cancers (which have been linked to PFOA exposure in animal and human studies) serving as controls. As a sensitivity analysis, we also performed analyses using a control group consisting o f those with all other cancer diagnoses included in the dataset, without exclusions. We adjusted for age, gender, diagnosis year, smoking status (current, past, unknown, with never smoker as the reference) and insurance provider (government insured Medicaid, uninsured, unknown, with privately insured as the reference). We ran additional analyses stratified by gender for cancers with >100 cases o f each gender; this included cancers o f the bladder, colon-rectum, kidney, and lung, as well as melanoma o f the skin and non-Hodgkin's lymphoma. To test the sensitivity o f results to missing smoking and health insurance data, we generated 10 datasets with imputation o f missing values using default predictive mean matching and logistic regression imputation via the `mice' library in R (Van Buuren and Oudshoorn 2007). We obtained parameter estimates by averaging over all 10 datasets o f parameter estimates, and variance estimates by combining the between- and within-imputation variances. For exposure assessment purposes, OCISS provided addresses at diagnosis that we geocoded, while the WVCR provided an identifier for geographic unit, which allowed us to assign case addresses to contaminated water district areas or to the unexposed group. We conducted two different analyses to compare the robustness of our results across different exposure metrics. The first analysis used water district of residence as the exposure of interest and included both OH and WV data. The second analysis was restricted to OH and took advantage o f the availability of geocoded OH addresses at time o f diagnosis. We used an existing PFOA exposure model (Shin et al. 201 la; 201 lb) to estimate serum levels at a finer geographic resolution for different latency assumptions. For OH-only analyses, we also adjusted for race, modeled as a binary variable for white or non-white, which was provided by OCISS but not 10 Page 11 of 26 WVCR due to confidentiality concerns. The two analyses are described in detail below. All statistical analyses were conducted using R 2.10.1 (Vienna, Austria). Water District Analysis in WV and OH For the combined WV and OH data, we used residency within a contaminated WD area as our exposure of interest. In OH, we assigned cases to WDs using geocoding, the process by which measures of longitude and latitude are calculated for street addresses using reference street files. We first cleaned and standardized addresses using ZP4 address correction software with the LACS database (version expiring April 1 2011; www.semaphorecorp.com) and converted additional rural route boxes to street addresses using Enhanced 911 address conversion tables (Vieira et al. 2010). Geocoding was then performed using a geographic information system (GIS), ESRI ArcView version 9.3 (Redlands, CA) with the ESRI StreetMap Premium North America NAVTEQ 2010 enhanced street dataset as the reference address locator. Using geocoding, we were able to identify cases living within a contaminated water district area. Cases not in contaminated water districts were assigned to the unexposed group. For WV cases, data release restrictions prohibited identifiable geographic location from being included with the cancer data. Instead, a variable was provided to indicate whether cases were located in Lubeck WD, Mason County WD, or unexposed areas. Only addresses in Wood County were geocoded to the WD distribution system at the WVCR to determine if the case was living at a street address serviced by the Lubeck WD. Wood County cases not on the Lubeck WD distribution system were considered unexposed. All cases in Mason County were assigned exposure to the Mason County WD. Mason County addresses were not geocoded because the median PFOA serum levels were close to background. 11 Page 12 of 26 We calculated adjusted ORs and CIs for each o f the 18 cancer categories in association with one o f the six contaminated WDs versus an unexposed WD. We also calculated the adjusted ORs for living in any exposed WD relative to unexposed WDs. Estimated PFOA Serum Level Analysis in OH To take advantage o f the availability o f geocoded street addresses in the OH data, we also used modeled serum PFOA concentration as an exposure metric. All OH addresses at time o f diagnosis were geocoded to determine if the case was serviced by one o f the contaminated public WDs, a contaminated private well, or unexposed. This geocoding allowed us to be even more specific about exposure as cases living within a water district area, but not on a street serviced by a distribution pipe (or before the year o f pipe installation), would likely be accessing drinking water from a private residential well. The methods for estimating individual serum PFOA levels from linked environmental, exposure, and pharmacokinetic models are described in detail elsewhere (Shin et al. 201 la; 201 lb). Briefly, the environmental models integrate facility emissions data, fate and transport characteristics o f PFOA, and hydrogeological properties o f the study area to estimate PFOA air and water concentrations from 1951-2008. Using GIS, we were also able to determine what year the pipe that serviced the case was installed. For each case, annual PFOA serum levels were.calculated from 1951 to date o f diagnosis by linking historical air and groundwater concentrations to residential information at time o f diagnosis and applying standard assumptions about water intake, body weights, and a PFOA half-life in the exposure and pharmacokinetic models (Shin et al. 201 lb). Because only the residence at diagnosis was available, annual serum levels were estimated assuming cases lived at that address for 10 years. As a sensitivity analysis, we also estimated serum levels with and without a 10-year latency 12 Page 13 of 26 period prior to date o f diagnosis assuming a lifetime residency at that address. We then extracted two exposure metrics for each latency and residency assumption: the estimated annual serum level corresponding to the year o f diagnosis or 10 years prior for latency analyses and a cumulative measure summed over the corresponding years of exposure. The estimated annual serum level is equivalent to what would be measured in a serum sample taken during that year whereas the cumulative measure is the area under the serum level profile curve. We first categorized individual-level exposure as very high, high, medium, low, and unexposed using cutoffs based on the distribution of the annual PFOA serum concentrations among the exposed study population assuming a 10-year residency. The distribution o f estimated annual PFOA serum levels among the exposed study population ranged from 3.7-655 pg/L for 10-yr residency assuming 10-yr latency (see Supplemental Material, Figure SI). The tertile breaks o f the distribution defined the cutoffs for low, medium, and high. We used the tertile breaks o f 12.9 and 30.8 to define high (30.8-109 pg/L), medium (12.9-30.7 pg/L), and low exposure categories (3.7-12.9), with unexposed serving as the reference category. There is a large break in the distribution at 110 pg/L so a very high group was created based on this break value which included the upper 10% of our exposed population (see Supplemental Material, Figure SI). Cumulative exposure categories were based on the distribution among the exposed cases and were divided into the following groups: very high=600-4,679 pg/L-years; high= 198 599 pg/L-years; medium=89-197 pg/L-years; and low=3.9-88 pg/L-years. We analyzed the individual-level OH data using logistic regression to calculate adjusted ORs and CIs for exposure categories with unexposed serving as the referent. For comparison, separate analyses were conducted for the annual and cumulative exposure measures calculated for the different latency and residency assumptions. 13 Page 14 of 26 Results WV and OH Water District Analyses Table 1 shows the distribution o f cases in the contaminated WD areas and surrounding unexposed geographic area. The Little Hocking WD is the highest exposed district, followed by Lubeck, Tuppers Plains, Belpre, Pomeroy, and Mason County. The odds o f testicular cancer was increased in Little Hocking (OR: 5.1,95% Cl: 1.6, 15.6; n=8), and the odds o f kidney cancer was elevated in Little Hocking (OR: 1.7, 95% Cl: 0.4, 3.3; n=10) and Tuppers Plains (OR: 2.0, 95% Cl: 1.3, 3.1; n=23). Residents o f Little Hocking also had increased odds o f non-Hodgkin lymphoma (OR: 1.6, 95% Cl: 0.9, 2.8; n=14) and prostate cancer (OR: 1.4, 95% Cl: 0.9, 2.3; n=36). OH Serum Level Analyses Adjusted odds ratios suggested associations between the very high PFOA exposure category and several cancers, but ORs for lower exposure categories generally did not support a positive dose-response relation (Table 2). Kidney cancer was positively associated with very high and high exposure categories (OR: 2.0, 95% Cl: 1.0, 3.9; n=9 and OR: 2.0, 95% Cl: 1.3, 3.2; n=22, respectively) while ORs for medium and low exposure categories were close to the null compared to the unexposed. The largest OR was for testicular cancer with the very high exposure category (OR: 2.8, 95% Cl: 0.8, 9.2; n=6) but the estimate was imprecise due to small numbers, and ORs for high, medium or low exposure categories, which were based on only 1,3 and 1 cases, respectively, were all <1.0. Ovarian cancer was also positively associated with the very high exposure category (OR: 2.1, 95% Cl: 0.8, 5.5; n=5) but again imprecise due to small numbers, and weaker associations for the high and medium exposure categories with a negative 14 Page 15 of 26 association in the lowest exposure category. ORs for the association between the very high and medium exposure categories and non-Hodgkin's lymphoma were moderate (OR: 1.8, 95% Cl: 1.0, 3.4; n=l 1 and OR: 1.5, 95% Cl: 1.0, 2.2; n=28, respectively), while ORs for high and low exposure categories were close to the null compared to the unexposed. Prostate cancer showed a weak but relatively precise positive association with very high exposure (OR: 1.5, 95% Cl: 0.9, 2.5; n= 31) and no association with lower levels o f exposure. Results were very similar for associations with the cumulative exposure measure (see Supplemental Material, Table SI), and for exposure estimates that did not account for latency (see Supplemental Material, Table S2) which were highly correlated with estimated exposures that assumed a 10-year latency (Spearman's rank correlation p=0.997, p-value<0.001). In addition, associations were similar when the alternative control group (that included kidney, liver, pancreas and testis cancer cases) was used (see Supplemental Material, Table S3). To test the sensitivity o f our analyses to missing smoking (n=2,452) and health insurance data (n=1824), we ran multiple imputations for cancers of the bladder, colon/rectum, female breast, kidney, lung, prostate, uterus, melanoma o f the skin, and non-Hodgkin's lymphoma with sufficient numbers (>100 cases with complete covariate data) and we observed similar results (see Supplemental Material, Table S4). For cancers o f the bladder, colon/rectum, kidney, and lung, melanoma o f the skin, and non-Hodgkin's lymphoma with sufficient numbers to stratify by gender (>100 cases in men and women, respectively), we observed generally similar results with regards to PFOA exposure categories (data not shown). An exception was kidney cancer, which was positively associated with very high exposure in women (n=108) (OR: 3.5, 95% Cl: 1.4, 8.3; n=6) but not men (n=138) (OR: 1.0, 95% Cl: 0.3, 3.4; n=3) relative to the unexposed. 15 Page 16 of 26 Discussion , Testicular cancer was positively associated with the highest PFOA water district (OR: 5.1, 95% Cl: 1.6, 15.6; n=8) and the highest serum exposure category (OR: 2.8, 95% Cl: 0.8, 9.2; n=6) compared to cases living in unexposed areas. However, we also observed an inverse association between testicular cancer and the lower exposure groups, and all o f the estimates are imprecise because o f small numbers of cases. Evidence of effects of PFOA on testicular Leydig cell tumors in animal models has been reported (Lau et al. 2007). Kidney cancer was increased in association with both high and very high PFOA exposure, based on larger numbers of cases. We also observed elevated adjusted odds ratios for very high PFOA exposure and ovarian (OR: 2.0, 95% Cl: 0.8, 5.1; n=5) and prostate (OR: 1.5, 95% Cl: 1.0, 2.5; n=31) cancers and non-Hodgkin lymphoma (OR: 1.8, 95% Cl: 0.9, 3.3; n=l 1). A limitation o f our study is that we used other types o f cancer as our controls (referents). Our analysis assumes that referent cancers are not associated with exposure to PFOA. For our main analyses we excluded kidney, pancreatic, testicular, and liver cancers from controls as these cancers have been linked to PFOA exposure in animal or human studies, but analyses using all other cancers as referents were comparable. We further assume that different types of cancer are ascertained by the registry in the same way, and that they are sampled from the same source population. Despite the large overall sample size o f our study, the water district analyses and the analyses of the very high exposure group were limited by small numbers of many individual cancers. There is also little consistency in the results across exposure categories, with no evidence of a positive dose response. We were further limited by the covariates we could adjust 16 Page 17 of 26 for, which included only age, gender, race (white or non-white, OH only), smoking status, and health insurance provider. We were therefore unable to adjust for other risk factors o f potential interest-- e.g., prenatal exposure to xenoestrogens in relation to testicular cancer-- although such factors would also have to be associated with exposure to cause confounding. Chance is also a concern as we are investigating multiple cancer sites. As expected under the assumption that a positive association is truly present, we observed similar but weaker associations for most outcomes when no latency was assumed. Because exposure in our study is dependent on location and the ranking o f exposure between districts generally remained stable over time, there is very little movement o f cases across exposure categories with respect to latency assumptions. As a sensitivity analysis, we also modeled exposures assuming the cases lived at their residence their entire lifetime and observed similar results (see Supplemental Materials, Table S4). However both the latency and residential history measures were highly correlated (Spearman's rank correlations: p>0.99, p-value<0.001). Moving within the same public water district would also have little or no impact on the estimated serum values, but moving across water districts, or especially from more distant locations, could lead to exposure misclassification. Based on data from the C8 Health Project for residents older than 50 years o f age, the median residency duration for their current residence in 2005-6 was 17 years. Therefore, we felt 10-year residency duration with a 10-year latency was a reasonable assumption. Any resulting exposure misclassification is likely non-differential, so the bias in the highest exposure category should on average be towards the null. Strengths of our study include a relatively large overall sample size, ascertainment of cases from cancer registries using controls from the same population as the cases, good success in geocoding o f OH residences, and the use o f a validated exposure model for predicting serum 17 Page 18 of 26 levels. Previous work has shown that the correlation between measured and predicted serum in 2005-6 using this exposure model was 0.82 (Shin et al. 201 lb). Water and serum concentrations in the more exposed areas are well above background, providing a larger exposure contrast compared to other cancer studies o f PFOA in general populations. We found qualitatively similar results for testicular, kidney, ovarian, prostate cancers, and non-Hodgkin lymphoma using the two different analyses; the robustness o f these results is another strength. Both analyses used individual-level outcome and risk factor information, but the first analysis used a group-level water district exposure measure so that both WV and OH data could be analyzed together. The second analysis used estimates of serum PFOA as the exposure of interest, allowing us to use geocoded residences to estimate exposure metrics based on points in time or cumulative measures, but for OH cases only. The second analysis has the advantages of being a fully individual-level design, eliminating the possible semi-ecologic bias in the other analysis (Webster 2007), and using an exposure model that has been validated as a predictor of serum levels in this context. However, there is still a potential for exposure misclassification using residence at diagnosis. A disadvantage is that we were only able to analyze OH data because geocoded residences were not available for WV. Associations between PFOA exposures and the same cancers have been reported in other unpublished C8 Science Panel (2012) studies o f the same community. Interviews o f 32,254 adult community residents and DuPont workers were conducted in 2008-11, and medical records were sought. Cox regression o f hazard ratios o f medically validated cancers in relation to modeled cumulative PFOA serum levels at dates o f diagnoses indicated increasing kidney cancer risk with increasing exposure when latency was not considered. When 10-year latency was included in the exposure metric, the association was less evident. The relative risks (RR) for testicular cancer in 18 Page 19 of 26 relation to increasing exposure quartiles with 10-year latency were 1.0, 1.2, 1.7, and 3.0. Cross sectional analysis o f prevalent cancers among 49,082 adult community members interviewed in 2005-6 in relation to measured PFOA indicated increased RRs with increased serum PFOA quartiles compared to the lowest quartile (RRs =1.0, 1.5, 1.7, and 1.7, respectively). Conclusions The geographic analyses of cancer registry data provide some evidence that higher PFOA serum levels may be associated with certain cancers. The association in the highest PFOA exposure group was largest but very imprecise for testicular cancer, and smaller but more precise for kidney cancer. Non-Hodgkin's lymphoma, ovarian and prostate cancers were associated with very high exposure based on some models, but there was little or no evidence o f associations with other cancers. Analyses were limited by a case-only design with minimal control of confounders and small case numbers, despite having ten years o f data. In addition, residential history information was not available to account for latency, migration, and other issues regarding timing o f exposure relative to cancer. However, the registries cover all residents in the study area, which comprises water districts with large and known contrasts in contamination. Thus geographic analyses using cancer registry data contributed to the evidence for the C8 Science Panel (2012) conclusion that there is a probable link between PFOA exposure and testicular and kidney cancers. 19 Page 20 of 26 References Bonefeld-Jorgensen EC, Long M, Bossi R, Ayotte P, Asmund G, Krger T, et al. 2011. Perfluorinated compounds are related to breast cancer risk in Greenlandic Inuit: a case control study. Environ Health 10:88. Calafat AM, Wong LY, Kuklenyik Z, Reidy JA, Needham LL. 2007.Polyfluoroalkyl chemicals in the U.S. population: data from the National Health and Nutrition Examination Survey (NHANES) 2003-2004 and comparisons with NHANES 1999-2000. Environ Health Perspect 115:1596-1602. C8 Science Panel. 2012. Probable Link Evaluation o f Cancer. Available: http://www.c8sciencepanel.org/pdfs/Probable_Link_C8_Cancer_16ApriI2012_v2.pdf [accessed 18 Dec 2012], Eriksen KT, Sorensen M, McLaughlin JK, Lipworth L, Tjonneland A, Overvad K, et al. 2009. Perfluorooctanoate and perfluorooctanesulfonate plasma levels and risk o f cancer in the general Danish population. J Natl Cancer Inst 101:605-609. Fraser AJ, Webster TF, Watkins DJ, Nelson JW, Stapleton HM, Calafat AM, et al. 2012. Polyfluorinated compounds in serum linked to indoor air in office environments. Environ Sci Technol 46:1209-1215. Frisbee S, Brooks A, Maher A, Flensborg P, Arnold S, Fletcher T, et al. 2009. The C8 Health Project: Design, methods, and participants. Environ Health Perspect 117:1873-1882. Huag LS, Huber S, Schlabach M, Becher G, Thomsen C. 2011. Investigation on per- and polyfluorinated compounds in paired samples o f house dust and indoor air from Norwegian homes. Environ Sci Technol 45:7991-7998. Lau C, Anitole K, Hodes C, Lai D, Pfahles-Hutchens A, Seed J. 2007. Perfluoroalkyl acids: a review o f monitoring and toxicological findings. Toxicol Sci 99:366-394. Leonard RC, Kreckmann KH, Sakr CJ, Symons JM. 2008. Retrospective cohort mortality study o f workers in a polymer production plant including a reference population o f regional workers. Ann Epidemiol 18:15-22. Lundin JI, Alexander BH, Olsen GW, Church TR. 2009. Ammonium perfluorooctanoate production and occupational mortality. Epidemiology 20:921-928. 20 Page 21 of 26 Paustenbach DJ, Panko JM, Scott PK, Unice KM. 2007. A methodology for estimating human exposure to perfluorooctanoic acid (PFOA): a retrospective exposure assessment of a community (1951-2003). J Toxicol Environ Health A 70:28-57. Shin HM, Vieira VM, Ryan PB, Detwiler R, Sanders B, Steenland K, et al. 2011a. Environmental fate and transport modeling for perfluorooctanoic acid emitted from the Washington Works Facility in West Virginia. Environ Sei Technol 45:1435-1442. Shin HM, Vieira VM, Ryan PB, Steenland K, Bartell SM. 2011b. Retrospective exposure estimation and predicted versus observed serum perfluorooctanoic acid concentrations for participants in the C8 Health Project. Environ Health Perspect 119:1760-1765. Shin HM, Ryan PB, Vieira VM, Bartell SM. 2012. Modeling the air-soil transport pathway of perfluorooctanoic acid in the mid-Ohio Valley using linked air dispersion and vadose zone models. Atmos Environ 51:67-74. Steenland K, Jin C, MacNeil J, Lally C, Ducatman A, Vieira V, et al. 2009. Predictors o f PFOA levels in a community surrounding a chemical plant. Environ Health Perspect 117:1083 1088. Steenland K, Fletcher T, Savitz DA. 2010. Epidemiologic evidence on the health effects o f perfluorooctanoic acid (PFOA). Environ Health Perspect 118:1100-1108. Steenland K, Woskie S. 2012. Cohort Mortality Study o f Workers Exposed to Perfluorooctanoic Acid. Am J Epidemiol; doi: 10.1093/aje/kwsl71 [Online 18 October 2012] Van Buuren S, Oudshoorn C. 2007. Mice: multivariate imputation by chained equations. R package version 1.16. Vieira VM, Howard GJ, Gallagher LG, Fletcher T. 2010. Geocoding rural addresses in a community contaminated by PFOA: a comparison o f methods. Environ Health 9:18. Webster TF. 2007. Boas magnification in ecologic studies: a methodological investigation. Environ Health 6:17. 21 Page 22 of 26 Table 1 w v and 0H water district results: number (n), adjusted odds ratios3(AOR), and 95% confidence intervals (Cl) for exposure to contaminated water districts6 Outcome Total Total Exposed Little Hocking Lubeck Tuppers Plains Belpre Pomeroy Mason n n AOR (Cl) n AOR (Cl) n AOR (Cl) n AOR (Cl) n AOR (Cl) n AOR (Cl) n AOR (Cl) All 18 Cancer Types 25107 2932 208 430 405 454 100 1326 Bladder 1350 137 0.8 (0.7, 1.0) 7 0.6 (0.3, 1.4) 24 1.0 (0.6, 1.5) 20 0.9 (0.6, 1.5) 24 1.1 (0.7, 1.6) 4 0.8 (0.3, 2.1) 58 0.7 (0.6, 1.0) Brain 506 60 1.0 (0.8, 1.3) 1 0.2 (0.0, 1.5) 7 0.8 (0.4, 1.8) 9 1.1 (0.5, 2.1) 11 1.2 (0.6, 2.2) 3 1.7 (0.5, 5.4) 29 1.1 (0.7, 1.6) Female Breast 4057 436 1.0 (0.9, 1.1) 33 1.2 (0.8, 2.0) 69 1.2 (0.9, 1.7) 50 0.7 (0.5, 1.1) 73 1.1 (0.8, 1.5) 18 0.8 (0.5, 1.5) 193 1.0 (0.8, 1.2) Cervix 338 35 0.8 (0.6, 1.2) 4 0.9 (0.3, 2.9) 5 0.7 (0.3, 1.7) 8 1.8 (0.8, 3.8) 5 0.6 (0.2, 1.6) 2 0.9 (0.2, 4.1) 11 0.7 (0.4, 1.3) Colon/Rectum 3543 383 0.9 (0.8, 1.0) 20 0.7 (0.5, 1.2) 44 0.7 (0.5, 1.0) 66 1.2 (0.9, 1.6) 55 0.9 (0.7, 1.2) 18 1.2 (0.7, 2.1) 180 0.9 (0.8, 1.1) Kidney 751 94 1.1 (0.9, 1.4) 10 1.7 (0.9, 3.3) 9 0.7 (0.4, 1.3) 23 2.0 (1.3,3.1) 17 1.4 (0.8, 2.3) 0 35. 0.9 (0.6, 1.3) Leukemia 674 72 0.9 (0.7, 1.1) 5 1.0 (0.4, 2.3) 11 0.9 (0.5, 1.6) 9 0.8 (0.4, 1.7) 12 1.0 (0.6, 1.9) 1 0.4 (0.1, 2.8) 34 0.9 (0.6, 1.3) Liver 179 23 1.1 (0.7, 1.6) 1 0.8 (0.1, 5.6) 4 1.3 (0.5, 3.5) 3 1.0 (0.3, 3.3) 3 1.0 (0.3, 3.1) 1 1.4 (0.2, 10.5) 11 1.0 (0.5, 1.9) Lung 4926 632 1.2 (1.1, 1.3) 37 1.0 (0.7, 1.5) 85 1.1 (0.8, 1.4) 84 1.3 (1.0, 1.7) 90 1.1 (0.9, 1.4) 23 1.1 (0.7, 1.8) 313 1.3 (1.1, 1.5) Melanoma of the Skin 1428 168 0.9 (0.8, 1.1) 12 1.0 (0.6, 1.9) 32 1.2 (0.8, 1.7) 21 0.9 (0.6, 1.4) 38 1.4 (1.0, 2.0) 4 0.9 (0.3,2.5) 61 0.7 (0.5, 0.9) Multiple Myeloma 285 36 1.1 (0.8, 1.6) 1 0.5 (0.1, 3.6) 4 0.9 (0.3, 2.3) 3 0.7 (0.2, 2.2) 7 1.5 (0.7, 3.2) 1 0.9 (0.1,6.6) 20 1.4 (0.9, 2.2) Non-Hodgkin'sLymphoma 1124 152 1.2 (1.0, 1.5) 14 1.6 (0.9, 2.8) 20 1.1 (0.7, 1.7) 21 1.2 (0.8, 1.9) 24 1.3 (0.9, 2.0) 5 1.1 (0.4, 2.7) 68 1.2 (0.9, 1.5) Ovary 417 48 1.0 (0.8, 1.4) 5 1.8 (0.7, 4.4) 5 0.7 (0.3, 1.7) 6 1.1 (0.5, 2.4) 11 1.6 (0.9, 3.0) 2 1.1 (0.3, 4.4) 19 0.9 (0.5, 1.4) Pancreas 495 58 1.0 (0.8, 1.3) 4 1.1 (0.4, 3.0) 9 1.1 (0.6, 2.1) 10 1.3 (0.7, 2.5) 8 0.9 (0.4, 1.8) 2 1.0 (0.2, 4.1) 25 0.9 (0.6, 1.4) Prostate 3678 434 0.9 (0.8, 1.1) 36 1.4 (0.9, 2.3) 78 1.2 (0.9, 1.6) 56 0.8 (0.6, 1.1) 56 0.8 (0.6, 1.1) 12 1.3 (0.6, 2.6) 196 0.9 (0.7, 1.0) Testis 134 18 1.0 (0.6, 1.8) 8 5.1 (1.6, 15.6) 2 0.9 (0.2, 4.5) 2 0.4 (0.1, 2.0) 1 0.6 (0.1, 5.0) 0 ... 5 0.5 (0.2, 1.5) Thyroid 343 40 1.1 (0.7, 1.5) 3 0.8 (0.3, 2.7) 7 1.2 (0.6, 2.6) 2 0.3 (0.1, 1.4) 5 0.9 (0.4, 2.2) 0 ... 23 1.4 (0.9, 2.2) Uterus 879 97 1.0 (0.8, 1.3) 7 1.1 (0.5,2.4) 15 1.1 (0.6, 1.9) 12 0.9 (0.5, 1.6) 14 0.9 (0.5, 1.6) 4 0.9 (0.3, 2.4) 45 1.1 (0.8, 1.5) a. j - o ------>-- o*iiv/i\ux5 ottttuo, ^uuuuia we uuici adieu cAciuuing Muncy, iiver, pancreas ana testis cancers. b. The estimated 1995 median PFOA serum concentrations in the WDs are: Little Hocking =125 pg/L; Lubeck = 65.8 pg/L; Tupper Plains = 23.9 pg/L; Belpre =18.7 pg/L; Pomeroy = 10.7 pg/L; and Mason = 5.3 pg/L. Unexposed was the reference. 22 Page 23 of 26 23 Page 24 of 26 Table 2. OH serum-level results: number (n), adjusted odds ratios3(AOR), and 95% confidence intervals (Cl) for individual-level annual PFOA serumexposure categories' assuming 10-yr residency and latency Outcome All 18 Cancer Types Bladder Brain Female Breast Cervix Colon/Rectum Kidney Leukemia Liver Lung Melanoma of the Skin Multiple Myeloma Non-Hodgkin'sLymphoma Ovary Pancreas Prostate Testis Thyroid Uterus Total n 7869 395 150 1260 144 1149 246 191 61 1526 429 83 347 128 162 1155 61 94 288 Total Exposed n 1496 69 32 223 25 212 59 36 11 293 95 18 76 27 33 214 11 15 47 Very High n 159 4 0 29 2 13 9 2 0 29 9 1 11 5 2r 31 6 2 4 AOR (Cl) 0.6 (0.2, 1.5) ... 1.4 (0.9, 2.3) 0.6 (0.1, 2.6) 0.6 (0.3, 1.0) 2.0 (1.0, 3.9) 0.6 (0.1, 2.3) ... 1.0 (0.7, 1.6) 0.9 (0.5, 1.9) 0.6 (0.1, 4.7) 1.8 (1.0, 3.4) 2.1 (0.8, 5.5) 0.6 (0.1,2.5) 1.5 (0.9, 2.5) 2.8 (0.8, 9.2) 0.8 (0.2, 3.5) 0.7 (0.3, 1.5) High n AOR (Cl) 374 21 1.2 (0.8, 2.0) 4 0.6 (0.2, 1.6) 45 0.7 (0.5, 1.0) 8 1.7 (0.8, 3.8) 63 1.3 (1.0, 1.7) 22 2.0(13,3.2) 8 0.9 (0.4, 1.8) 3 1.0(03,3.1) 78 1.2 (0.9, 1.6) 21 1.0 (0.6, 1.5) 4 1.0(03,2.7) 17 1.1 (0.7, 1.9) 8 1.4 (0.7, 2.9) 9 1.1 (0.6,23) 47 0.8 (0.5, 1.1) 1 0.3 (0.0, 2.7) 3 0.7 (0.2, 2.1) 12 1.7 (1.2, 2.5) Medium n AOR (Cl) 489 21 0.9 (0.6, 1.4) 16 1.8 (1.1, 3.2) 77 1.1 (0.8, 1.5) 4 0.5 (0.2, 1.5) 64 0.9 (0.7, 1.2) 17 1.2 (0.7, 2.0) 12 1.0 (0.6, 1.9) 4 0.9 (03,2.5) 95 1.0 (0.8, 13) 38 1.3 (0.9, 1.8) 6 1.1 (0.5, 2.6) 28 1.5 (1.0, 2.2) 10 1.4 (0.7, 2.7) 10 0.9 (0.5, 1.7) 65 0.8 (0.6, 1.0) 3 0.6 (0.2, 2.2) 5 0.9 (0.4,23) 14 0.9 (0.6, 1.3) Low n 474 23 12 72 11 72 11 14 4 91 27 7 20 r 4 12 71 1 5 17 AOR (Cl) 0.9 (0.6, 1.4) 1.5 (0.8, 2.7) 0.9 (0.7, 1.2) 1.1 (0.6, 2.2) 1.0 (0.8, 1.3) 0.8 (0.4, 1.5) 1.2 (0.7, 2.1) 1.1 (0.4, 3.1) 1.0 (0.7, 1.2) 1.2 (0.8, 1.8) 1.4 (0.7, 3.2) 1.0 (0.6, 1.6) 0.5 (0.2, 1.4) 13 (0.7,23) 1.1 (0.8, 1.5) 0.2 (0.0, 1.6) 0.9 (0.4, 23) 1.2 (0.8, 1.7) a. Adjusted for age, race, gender, diagnosis year, insurance provider, and smoking status. Controls were other listed cancers excluding kidney, liver, pancreas and testis cancers b. Categories of modeled PFOA serumconcentrations (pg/L): Very High=l 10-655 pg/L; High=30.8-109 pg/L; Medium=12,9-30.7 pg/L; Low=3 7-12 8 pg/L- reference = unexposed. ' 24 Page 25 of 26 Figure Legend Figure 1. Study area o f 13 counties encompassing 6 contaminated water districts 25 Environmental Health Perspectives Supplemental Material http://dx.doi.org/10.1289/ehp. 1205829 Perfluorooctanoic Acid Exposure and Cancer Outcomes in a Contaminated Community: A Geographic Analysis Vernica M. Vieira, Kate Hoffman Hyeong-Moo Shin, Janice M. Weinberg, Thomas F. Webster, Tony Fletcher Table o f Contents: Supplemental Material, Table SI, page 2. OH serum-level results: adjusted odds ratios and 95% confidence intervals for cumulative PFOA serum exposure assuming 10-yr residency and latency Supplemental Material, Table S2, page 3. OH serum-level results: adjusted odds ratios and 95% confidence intervals for annual PFOA serum exposure assuming 10-yr residency and no latency Supplemental Material, Table S3, page 4. OH serum-level results: adjusted odds ratios and 95% confidence intervals for annual PFOA serum exposure assuming 10-yr residency and latency with alternative control group Supplemental Material, Table S4, page 5. OH serum-level results: adjusted odds ratios and 95% confidence intervals for annual PFOA serum exposure assuming 10-yr residency and latency with multiple imputations for missing data Supplemental Material, Figure S I, page 6. Histogram o f serum estimates assuming a 10-yr residency and latency among exposed OH study population 1 Environmental Health Perspectives http://dx.doi.org/!0.1289/ehp. 1205829 Supplemental Material, Table SI, OH serum-level results: adjusted odds ratios3and 95% confidence intervals for cumulative PFOA serum exposure15assuming 10-yr residency and latency Category Bladder Brain Female Breast Cervix Colon/Rectum Kidney Leukemia Liver Lung Melanoma o f the skin Multiple Myeloma Non-Hodgkin's Lymphoma Ovary Pancreas Prostate Testis Thyroid Uterus Very High 0.6 (0.2, 1.7) C 1.4 (0.8, 2.3) 0.6 (0.1, 2.7) 0.6 (0.3, 1.1) 2.1 (1.1, 4.2) 0.6 (0.1, 2.4) C 0.9 (0.6, 1.5) 0.9 (0.4, 1.8) 0.7 (0.1, 5.0) 2.0(1.0,3.7) 2.2 (0.9, 5.7) 0.6 (0.2, 2.7) 1.5 (0.9, 2.5) 2.8 (0.8, 9.6) 0.9 (0.2, 3.7) 0.7 (0.3, 1.7) High 1.0 (0.6, 1.7) 0.7 (0.3, 1.8) 0.7 (0.5, 0.9) 1.7 (0.7, 3.9) 1.2 (0.9, 1.6) 2.0(1.3,3.2) 1.4 (0.8, 2.6) 1.0 (0.3, 3.2) 1.2 (0.9, 1.6) 1.2 (0.8, 1.9) 0.7 (0.2, 2.3) 1.0 (0.6, 1.7) 1.7 (0.9, 3.4) 1.3 (0.7, 2.5) 0.8 (0.6, 1.1) 0.4 (0.0, 2.9) 0.9 (0.3, 2.5) 1.6 (1.1, 2.3) Medium 1.1 (0.7, 1.6) 1.7 (1.0, 2.9) 1.1 (0.8, 1.5) 0.6 (0.2, 1.6) 1.0 (0.7, 1.3) 1.2 (0.7, 2.0) 0.7 (0.4, 1.5) 1.2 (0.5, 3.0) 1.0 (0.8, 1.3) 1.2 (0.8, 1.7) 1.1 (0.5, 2.6) 1.5 (1.0, 2.2) 0.9 (0.4, 2.1) 0.7 (0.3, 1.4) 0.8 (0.6, 1.0) 0.4 (0.1, 1.8) 0.7 (0.3, 2.0) 0.9 (0.6, 1.3) Low 0.9 (0.6, 1.4) 1.5 (0.8, 2.7) 0.9 (0.7, 1.2) 1.1 (0.6, 2.2) 1.0 (0.8, 1.3) 0.8 (0.4, 1.5) 1.1 (0.6, 2.0) 0.9 (0.3, 2.8) 1.0 (0.7, 1.2) 1.1 (0.8, 1.7) 1.7 (0.8, 3.6) 1.0 (0.6, 1.6) 0.7 (0.3, 1.6) 1.4 (0.8, 2.4) 1.1 (0.8, 1.5) 0.4 (0.1, 1.9) 0.9 (0.4, 2.3) 1.2 (0.8, 1.7) a. Adjusted for age, race, gender, insurance provider, and smoking status. Controls were other listed cancers excluding kidney, liver, pancreas and testis cancers. a. Categories o f modeled PFOA serum concentrations (pg/L-year): Very High=600-4679 pg/Lyear; High: 198-599 pg/L-year; Medium=89-197 pg/L-year; Low=3.8-88 pg/L-year; reference: unexposed b. No cases in exposure category 2 Environmental Health Perspectives http://dx.doi.org/lO. 1289/ehp. 1205829 Supplemental Material, Table S2. OH serum-level results: adjusted odds ratios3and 95% confidence intervals for annual PFOA serum exposureb assuming 10 year residency and no latency Category Bladder Brain Female Breast Cervix Colon/Rectum Kidney Leukemia Liver Lung Melanoma o f the skin Multiple Myeloma Non-Hodgkin's Lymphoma Ovary Pancreas Prostate Testis Thyroid Uterus Very High 0.7 (0.3, 1.6) C 1.2 (0.7, 1.9) 1.0 (0.3, 2.8) 0.7 (0.4, 1.1) 1.7 (0.9, 3.3) 1.1 (0.4, 2.7) C 1.0 (0.7, 1.4) 1.0 (0.5, 1.9) 0.5 (0.1, 3.7) 1.8 (1.0, 3.1) 1.7 (0.7, 4.2) 0.9 (0.3, 2.6) 1.3 (0.8, 2.0) 2.2 (0.7, 6.6) 1.0 (0.3, 3.4) 0.7 (0.3, 1.5) High 1.1 (0.7, 1.7) 1.1 (0.6, 2.1) 0.7 (0.5, 0.9) 1.8 (0.9, 3.5) 1.2 (0.9, 1.5) 1.8 (1.2, 2.8) 0.9 (0.5, 1.7) 0.9 (0.3, 2.6) 1.2 (0.9, 1.5) 1.0 (0.6, 1.4) 0.9 (0.4, 2.3) 1.3 (0.9, 2.0) 1.7 (0.9, 3.1) 1.1 (0.6, 2.1) 0.8 (0.6, 1.1) 0.6 (0.2, 2.0) 0.5 (0.2, 1.6) 1.6(1.1,2.3) Medium 1.0 (0.6, 1.6) 1.7 (0.9, 3.1) 1.1 (0.9, 1.5) 0.5 (0.2, 1.3) 1.0 (0.7, 1.3) 1.1 (0.6, 1.9) 0.9 (0.5, 1.7) 0.8 (0.2, 2.5) 1.1 (0.8, 1.4) 1.3 (0.9, 1.9) 1.1 (0.5, 2.6) 1.3 (0.8, 2.0) 0.9 (0.4, 2.1) 0.9 (0.4, 1.7) 0.7 (0.5, 0.9) 0.3 (0.0, 2.9) 0.8 (0.3, 2.3) 0.9 (0.6, 1.4) Low 0.8 (0.5, 1.4) 1.3 (0.6, 2.7) 1.0 (0.7, 1.4) 0.9 (0.4, 2.2) 1.0 (0.7, 1.3) 0.9 (0.5, 1.8) 1.2 (0.6, 2.4) 1.3 (0.4, 4.2) 0.9 (0.6, 1.2) 1.2 (0.8, 2.0) 1.9 (0.8, 4.3) 0.9 (0.5, 1.6) 0.6 (0.2, 1.8) 1.2 (0.6, 2.5) 1.3 (0.9, 1.8) C 1.2 (0.5, 3.0) 1.1 (0.7, 1.7) b. Adjusted for age, race, gender, diagnosis year, insurance provider, and smoking status. Controls were other listed cancers excluding kidney, liver, pancreas and testis cancers. a. Categories o f modeled PFOA serum concentrations (pg/L): Very High=l 10-655 pg/L; High=30.8-109 pg/L; Medium=12.9-30.7 pg/L; Low=3.7-12.8 pg/L; reference = unexposed b. No cases in exposure category 3 Environmental Health Perspectives http://dx.doi.org/lO 1289/ehp. 1205829 Supplemental Material, Table S3. OH serum-level results: adjusted odds ratios3and 95% confidence intervals for annual PFOA serum exposure15assuming 10-yr residency and latency with alternative control group Category Bladder Brain Female Breast Cervix Colon/Rectum Kidney Leukemia Liver Li ig __________ Melanoma o f the skin Multiple Myeloma Non-Hodgkin's Lymphoma Ovary Pancreas Prostate Testis Thyroid Uterus Very High 0.6 (0.2, 1.5) C 1.3 (0.8, 2.1) 0.6 (0.1, 2.6) 0.6 (0.3, 1.0) 1.9 (1.0, 3.9) 0.5 (0.1, 2.2) C 1.0 (0.6, 1.5) 0.8 (0.4, 1.6) 0.6 (0.1, 4.6) 1.8 (0.9, 3.3) 2.0 (0.8, 5.1) 0.6 (0.1, 2.3) 1.5 (1.0, 2.5) 2.4 (0.7, 7.5) 0.8 (0.2, 3.3) 0.6 (0.3, 1.5) High 1.2 (0.8, 1.9) 0.6 (0.2, 1.6) 0.7 (0.5, 0.9) 1.7 (0.8,3 7) 1.2 (0.9, 1.6) 2.0 (1.3, 3.2) 0.9 (0.4, 1.8) 0.9 (0.3, 3.1) 1.2 (0.9, 1.5) 0.9 (0.6, 1.5) 0.9 (0.3, 2.6) 1.1 (0.7, 1.8) 1.3 (0.6, 2.8) 1.1 (0.6, 2.2) 0.7 (0.5, 1.0) 0.3 (0.0, 2.9) 0.7 (0.2, 2.1) 1.6 (1.1, 2.4) Medium 0.9 (0.6, 1.4) 1.8 (1.1, 3.2) 1.1 (0.8, 1.5) 0.5 (0.2, 1.5) 0.9 (0.7, 1.2) 1.2 (0.7, 2.0) 1.0 (0.6, 1.9) 0.9 (0.3, 2.6) 1.0 (0.8, 1.2) 1.3 (0.9, 1.8) 1.1 (0.5, 2.6) 1.5 (1.0, 2.2) 1.4 (0.7, 2.7) 0.9 (0.5, 1.7) 0.8 (0.6, 1.0) 0.6 (0.2, 2.3) 0.9 (0.4, 2.3) 0.9 (0.6, 1.3) Low 0.9 (0.6, 1.4) 1.5 (0.8, 2.7) 0.9 (0.7, 1.2) 1.1 (0.6, 2.2) 1.0 (0.8, 1.3) 0.8 (0.4, 1.5) 1.2 (0.7, 2.1) 1.2 (0.4, 3.3) 1.0 (0.8, 1.3) 1.2 (0.8, 1.8) 1.5 (0.7, 3.2) 1.0 (0.6, 1.6) 0.5 (0.2, 1.4) 1.3 (0.7, 2.3) 1.1 (0.8, 1.5) 0.2 (0.0, 1.5) 0.9 (0.4, 2.3) 1.2 (0.8, 1.7) a. Adjusted for age, gender, race, diagnosis year, insurance provider, and smoking status. Cases o f the other 17 cancer types were used as controls. b. Categories o f modeled PFOA serum concentrations (pg/L): Very High=l 10-655 pg/L; High=30.8-109 pg/L; Medium=12.9-30.7 pg/L; Low=3.7-12.8 pg/L; reference = unexposed c. No cases in exposure category 4 Environmental Health Perspectives http://dx.d0i.0r g /l0 .1289/ehp. 1205829 Supplemental Material, Table S4. OH serum-level results: adjusted odds ratios3and 95% confidence intervals for annual PFOA serum exposure15assuming 10-yr residency and latency with multiple imputations0 for missing data Category Bladder Female Breast Colon/Rectum Kidney Lung Melanoma of the Skin Non-Hodgkin's Lymphoma Prostate Uterus Very High 0.4 (0.1, 1.7) 1.2 (0.6, 2.3) 0.7 (0.3, 1.3) 2.0 (0.3, 3.3) 1.2 (0.6, 2.2) 0.9 (0.6, 1.3) High 1.3 (0.7, 2.4) 0.9 (0.5, 1.4) 1.1 (0.8, 1.7) 2.3 (1.3, 4.1) 1.6 (1.1, 2.4) 1.0 (0.8, 1.3) 2.1 (1.0, 4.5) 1.9 (1.0, 3.6) 0.9 (0.3, 3.0) 0.8 (0.4, 1.8) 0.6 (0.4, 1.1) 1.0 (0.5, 2.1) Medium 0.8 (0.4, 1.5) 0.9 (0.6, 1.4) 0.6 (0.4, 1.0) 0.9 (0.4, 2.0) 1.3 (0.9, 1.9) 1.5 (1.3, 1.8) 1.6 (0.9, 2.8) 1.0 (0.7, 1.7) 0.6 (0.2, 1.6) Low 0.9 (0.5, 1.5) 0.9 (0.6, 1.3) 1.1 (0.8, 1.5) 0.6 (0.3, 1.4) 0.9 (0.7, 1.3) 1.1 (0.9, 1.4) 1.1 (0.7, 1.9) 1.3 (0.9, 1.9) 0.9 (0.5, 1.7) c. Adjusted for age, race, gender, diagnosis year, insurance provider, and smoking status. Controls were other listed cancers excluding kidney, liver, pancreas and testis cancers. d. Categories o f modeled PFOA serum concentrations (pig/L): Very High=l 10-655 pg/L; High=30.8-109 pg/L; Medium^l 2.9-30.7 pg/L; Low=3.7-12.8 pg/L; reference = unexposed e. Multiple imputations were only run for cancers with sufficient number o f cases (> 100 cases with complete covariate information) 5 Environmental Health Perspectives http://dx.doi.org/lO.1289/ehp. 1205829 6 Taft/ Taft Stettinius & Hollister LLP 425 W alnut Street, Suite 1800 Cincinnati, OH 45202-3957 FCF 10015293 FCF 10015293