ARTICLE pubs.acs.org/est
Analysis of a Homologous Series of Perfluorocarboxylates from American Red Cross Adult Blood Donors, 20002001 and 2006 Geary W. Olsen,*,† Mark E. Ellefson,‡ David C. Mair,§ Timothy R. Church,|| Corinne L. Goldberg,^ Ross M. Herron,# Zahra Medhdizadehkashi,$ John B. Nobiletti,& Jorge A. Rios,@ William K. Reagen,‡ and Larry R. Zobel† †
Medical Department, 3M Company, St. Paul, Minnesota, United States Environmental Laboratory, 3M Company, St. Paul, Minnesota, United States § American Red Cross, North Central Region, St. Paul, Minnesota, United States Division of Environmental Health Sciences, University of Minnesota, Minneapolis, Minnesota, United States ^ American Red Cross, Carolinas Region, Charlotte, North Carolina, United States # American Red Cross, Southern California Region, Los Angeles, California, United States $ American Red Cross, Pacific Northwest Region, Portland, Oregon, United States & American Red Cross, Greater Alleghenies Region, Johnstown, Pennsylvania, United States @ American Red Cross, New England Region, Dedham, Massachusetts, United States
)
‡
bS Supporting Information ABSTRACT: The purpose of this study was to determine the concentration trends of a nine-target-analyte homologous series of perfluorocarboxylates from six American Red Cross adult blood donor centers. A total of 645 serum and 600 plasma samples were obtained in 20002001 and 2006, respectively, with samples stratified for each 10-year (2069) age- and sex-group per each location. Samples were extracted by protein precipitation and quantified by using tandem mass spectrometry. The nine perfluorocarboxylates were perfluorobutanoate (PFBA, C3F7CO2), perfluoropentanoate (PFPeA, C4F9CO2), perfluorohexanoate (PFHxA, C5F11CO2), perfluoroheptanoate (PFHpA, C6F13CO2), perfluorooctanoate (PFOA, C7F15CO2), perfluorononanoate (PFNA, C8F17CO2), perfluorodecanoate (PFDA, C9F19CO2), perfluoroundecanoate (PFUnA,C10F21CO2), and perfluorododecanoate (PFDoA, C11F23CO2). The majority of measurements were less than the lower limit of quantitation for PFPeA, PFHxA, and PFDoA. For the remaining targeted analytes, the geometric mean serum and plasma concentrations (ng/mL) for 20002001 and 2006 were, respectively, as follows: PFBA 2.61 vs 0.33, PFHpA 0.13 vs 0.09, PFOA 4.70 vs 3.44, PFNA 0.57 vs 0.97, PFDA 0.16 vs 0.34, and PFUnA 0.10 vs 0.18. Estimates of the 95th percent tolerance limits (ng/mL) were as follows: PFBA 5.3 vs 1.4, PFHpA 0.4 vs 0.4, PFOA 12.3 vs 7.7, PFNA 1.4 vs 2.2, PFDA 0.4 vs 0.8, and PFUnA 0.3 vs 0.5. Important observations were the decline in PFBA and increase in PFNA, PFDA, and PFUnA concentrations between 20002001 and 2006. The longer chain length perfluorocarboxylates were also highly correlated with each other.
’ INTRODUCTION Two major series of cross-sectional biomonitoring studies have been published pertaining to perfluoroalkyl levels in populations in the United States. The largest is the Centers for Disease Control and Prevention (CDC) National Health and Nutrition Examination Survey (NHANES) study.1 Although designed as a nutritional and health survey of a representative sample of the U.S. noninstitutionalized population age 12 and older, a subset of collected serum samples is used for periodic biomonitoring investigations. Although neither representative sample nor as large of a study population, the American Red Cross adult blood donor study, nevertheless, has been at the forefront of published data presenting the earliest indication of individual serum perfluoroalkyl concentrations in the general r 2011 American Chemical Society
population,2 as well as reporting the initial decline in concentrations of perfluorooctanesulfonate (PFOS, C8F17SO3) after the voluntary phase-out by the primary manufacturer, 3M Company, of PFOSbased materials.3 There have been similar conclusions reached by CDC NHANES and the American Red Cross adult blood donor study teams regarding perfluoroalkyls in the general population.1,4 Special Issue: Perfluoroalkyl Acid Received: December 27, 2010 Accepted: April 8, 2011 Revised: March 22, 2011 Published: April 29, 2011 8022
dx.doi.org/10.1021/es1043535 | Environ. Sci. Technol. 2011, 45, 8022–8029
Environmental Science & Technology The purpose of the present study was to compare age-, sex-, and location-specific concentrations of a homologous series of nine perfluorocarboxylates in 600 individual plasma samples collected in mid-2006 from six American Red Cross adult blood donor centers to 645 nonpaired serum samples collected from the same locations in 20002001. Both direct (e.g., manufacturing) and indirect (impurities and precursors that degrade) sources of perfluorocarboxylates occur in the environment.5 Associating chain length patterns in environmental samples can be complicated because of partitioning and uptake/clearance rates in biota for the various perfluorocarboxylates. Unique to the present study is the examination of perfluorobutanoate (PFBA, C3F7CO2). PFBA has been found in a variety of environmental matrices including as a contaminant in groundwater used as a source of community drinking water.6,7
’ MATERIAL AND METHODS Sample Collection. Through collaboration with six American Red Cross blood donor centers, a total of 645 serum samples (332 male, 313 female) were obtained in 20002001 and a nonidentical 600 (301 male, 299 female) plasma samples in 2006 from adult blood donors 2069 years of age. Each blood donor center was requested to provide 100 serum (20002001) or plasma (2006) samples: 10 samples per every 10-year age interval (2029, 3039, 4049, 5059, and 6069) for each sex. The six American Red Cross blood donor centers represented the following areas: Boston, Massachusetts; Charlotte, North Carolina; Hagerstown, Maryland; Los Angeles, California; Minneapolis-St. Paul, Minnesota; and Portland, Oregon. Prior results have been reported for PFOS, perfluorohexanesulfonate, perfluorobutanesulfonate, N-methyl perfluorooctanesulfonamidoacetate, N-ethyl perfluorooctanesulfonamidoacetate, and perfluorooctanoate (PFOA, C7F15CO2) but not the homologous series of perfluorocarboxylates.2,4 Upon completion of the initial studies, the remaining samples were stored at 80 C in plastic tubes until laboratory analysis for the present study. Samples were void of personal identifiers. The only available demographic factors were age, sex, and location. The study was approved by the American Red Cross Biomedical Services Institutional Review Board. Reference Materials. Methods and results for PFOA have been previously published.2,4 The eight additional perfluorocarboxylate target analytes quantified specifically in this present study were PFBA, perfluoropentanoate (PFPeA, C4F9CO2), perfluorohexanoate (PFHxA, C5F11CO2), perfluoroheptanoate (PFHpA, C6F13CO2), perfluorononanoate (PFNA, C8F17CO2), perfluorodecanoate (PFDA, C9F19CO2), perfluoroundecanoate (PFUnA,C10F21CO2), and perfluorododecanoate (PFDoA, C11F23CO2). The reference materials used to make the calibration standards for these 8 perfluorocarboxylates were of 98.7%, 98.5%, 97.7%, 98.2%, 98.0%, 96.4%, 99.7%, and 98.3% purity, respectively. Reference materials were purchased from the following (in parentheses): PFBA and PFHpA (Aldrich, St. Louis, MO); PFPeA (Alfa Aesar, Ward Hill, MA); and PFHxA, PFNA, PFDA, PFUnA, and PFDoA (Oakwood Products, West Columbia, SC). Isotopically labeled reference standards were purchased from Wellington Laboratories (Guelph, Ontario, Canada). To account for matrix effects and differences in extraction efficiency, surrogate matrix-matched internal standard calibration curves were constructed using nonmass-labeled reference standards for each of the target analytes in bovine serum. Mass-labeled internal standards MPFBA, MPFHxA, and
ARTICLE
MPFOA were added at a nominal concentration of 1.0 ng/mL serum to the surrogate matrix-matched bovine calibration curve, each sample, quality control samples, and one-half of the method and matrix blanks. MPFBA was used as an internal standard to quantify PFBA. MPFHxA was used as an internal standard to quantify PFPeA, PFHxA, and PFHpA. MPFOA was used as an internal standard to quantify PFNA, PFDA, PFUnA, PFDoA, and the surrogate recovery standards MPFDA and MPFDoA. MPFDA and MPFDoA surrogate recovery standards were added to the matrix-matched bovine serum calibration curve and at a nominal concentration of 1.0 ng/mL to each sample prior to extraction. These surrogate recovery standards were used as surrogates to demonstrate method accuracy of 100% ( 30% for individual samples. Surrogate recovery standard recoveries in all samples ranged from 73.4% to 118% for MPFDA and 71.5% to 129% for MPFDoA. Recoveries for individual samples were determined by dividing the measured value by the theoretical spike value, and multiplied by 100. Sample Preparation. The 3M Environmental Laboratory prepared and analyzed the samples. Target analytes were extracted from serum or plasma by protein precipitation in acetonitrile via a MultiPROBE II HT EX robotic liquid handling system (PerkinElmer, Wellesley, MA).8 Additional details are found in the Supporting Information. HPLC/MS/MS Analysis. Quantitation of sample extracts was accomplished by high-performance liquid chromatography tandem mass spectrometry (HPLC/MS/MS) using an Agilent Series 1200 Liquid Chromatograph system (Agilent Technologies, Santa Clara, CA) attached to an Applied Biosystems MDS Sciex API 5000 triple-quadrupole mass spectrometer equipped with a TurboIon Spray Source (PE Sciex, Concord, Ontario, Canada) maintained at 450 C in the negative ion mode. The target analytes, internal standards, and surrogate recovery standards were monitored in two separate runs recording 13 transitions for each ion to improve sensitivity.9 Details of this analysis along with the ions monitored to quantitate the target analytes (Table S1) are provided in the Supporting Information. Calibration, Accuracy, and Precision. PFBA was quantified using MPFBA as an internal standard. PFPeA, PFHxA, PFHpA, and MPFBA were quantified using MPFHxA as an internal standard. PFNA, PFDA, PFUnA, PFDoA, MPFDA, and MPFDoA were quantitated using MPFOA as an internal standard. All calibration curves had coefficients of determination of 0.990 or greater. MPFDA and MPFDoA were spiked into all samples and the surrogate matrix-matched bovine serum calibration curve prior to extraction, and were used as surrogate recovery standards to monitor method accuracy. As a partial method validation, PFBA, PFPeA, PFHxA, PFHpA, PFNA, PFDA, PFUnA, PFDoA, MPFBA, MPFDA, and MPFDoA were matrix spiked into control human plasma and control bovine serum to demonstrate analytical method accuracy and precision.10,11 Data accuracy and precision of 99.9% ( 24% (for PFBA), 98.8% ( 14% (for PFPeA), 101% ( 5.9% (for PFHxA), 97.0% ( 13% (for PFHpA), 100% ( 12% (for PFNA), 98.4% ( 15% (for PFDA), 103% ( 16% (for PFUnA), 105% ( 15% (for PFDoA), 99.8% ( 26% (for MPFBA), 98.1% ( 14% (for MPFDA), and 105% ( 14% (for MPFDoA) were determined by analyzing two levels of QC samples in replicates of three in control human plasma for each analytical batch, and were demonstrated intrabatch and between batch over a 0.25 18 ng/mL range. Bovine serum was demonstrated to be an equivalent surrogate test matrix for quantitation of the target 8023
dx.doi.org/10.1021/es1043535 |Environ. Sci. Technol. 2011, 45, 8022–8029
Environmental Science & Technology
ARTICLE
Table 1. Measures of Central Tendency and Distribution for Nine Perfluorocarboxylate Concentrations (ng/mL), American Red Cross Adult Blood Donors, 20002001 and 2006 selected percentiles range
N of samples
