Article pubs.acs.org/est
Perfluorinated Alkyl Acids in Blood Serum from Primiparous Women in Sweden: Serial Sampling during Pregnancy and Nursing, And Temporal Trends 1996−2010 Anders Glynn,†,* Urs Berger,‡ Anders Bignert,§ Shahid Ullah,‡ Marie Aune,† Sanna Lignell,† and Per Ola Darnerud† †
Department of Research and Development, National Food Agency, P.O. Box 622, SE-751 26 Uppsala, Sweden, Department of Applied Environmental Science (ITM), Stockholm University, SE-106 91 Stockholm, Sweden, and § Department of Contaminant Research, Swedish Museum of Natural History, SE-104 05 Stockholm, Sweden ‡
S Supporting Information *
ABSTRACT: We investigated temporal trends of blood serum levels of 13 perfluorinated alkyl acids (PFAAs) and perfluorooctane sulfonamide (FOSA) in primiparous women (N = 413) from Uppsala County, Sweden, sampled 3 weeks after delivery 1996−2010. Levels of the short-chain perfluorobutane sulfonate (PFBS) and perfluorohexane sulfonate (PFHxS) increased 11%/y and 8.3%/y, respectively, and levels of the long-chain perfluorononanoate (PFNA) and perfluorodecanoate (PFDA) increased 4.3%/y and 3.8%/y, respectively. Concomitantly, levels of FOSA (22%/y), perfluorooctane sulfonate (PFOS, 8.4%/y), perfluorodecane sulfonate (PFDS, 10%/y), and perfluorooctanoate (PFOA, 3.1%/y) decreased. Thus, one or several sources of exposure to the latter compounds have been reduced or eliminated, whereas exposure to the former compounds has recently increased. We explored if maternal levels of PFOS, PFOA, and PFNA during the early nursing period are representative for the fetal development period, using serial maternal serum samples, including cord blood (N = 19). PFAA levels in maternal serum sampled during pregnancy and the nursing period as well as in cord blood were strongly correlated. Strongest correlations between cord blood levels and maternal levels were observed for maternal serum sampled shortly before or after the delivery (r = 0.70−0.89 for PFOS and PFOA). A similar pattern was observed for PFNA, although the correlations were less strong due to levels close to the method detection limit in cord blood.
■
under investigation.7 Intake of contaminated fish may be a significant source of exposure to certain PFAAs such as perfluorooctane sulfonate (PFOS).8 Apart from food, also drinking water and dust may contribute to the exposure to PFOS and perfluorooctanoate (PFOA).9,10 In 2000 the chemical manufacturer 3M announced that it voluntarily was going to phase out the production of PFOSrelated chemicals and PFOA.11 PFOS had been found to be very persistent in the environment, to have a strong tendency to accumulate in human and animal tissues, and to pose a risk
INTRODUCTION Perfluorinated alkyl acids (PFAAs) are fully fluorinated organic acids, which are environmentally persistent and possess a strong surface tension lowering potential.1 Manufacturing of certain PFAAs has spanned over six decades. PFAAs are used in industrial processes, for instance production of fluoropolymers, and in products such as water and stain proofing agents, lubricants, paints, and fire-fighting foams.1 PFAAs bind to serum proteins and accumulate in blood and in protein-rich tissues of exposed organisms.2−4 An increased fluorinated carbon chain length of a PFAA generally results in increased bioaccumulation.5 Human exposure to PFAAs occurs globally, reflected by detectable levels in human blood in most of the studied areas of the world.6 The relative importance of different pathways for human exposure to PFAAs is currently © 2012 American Chemical Society
Received: Revised: Accepted: Published: 9071
April 12, 2012 July 2, 2012 July 6, 2012 July 6, 2012 dx.doi.org/10.1021/es301168c | Environ. Sci. Technol. 2012, 46, 9071−9079
Environmental Science & Technology
Article
to human health.11 The phase-out of PFOS-related products by 3M was completed in 2002 and the short-chain PFAA perfluorobutane sulfonate (PFBS) was launched as a replacement for PFOS.12,13 Likely as a result of the phase-out by 3M, studies of temporal trends of PFAAs in humans have shown declining levels of PFOS in blood in several areas of the world since the turn of the century.14−17 Moreover, levels of PFOA also seem to decline in some parts of the world.14−17 This could be due to a combination of the phase-out of PFOA production by 3M and, in more recent years, the launch of the “PFOA Stewardship Program” in the U.S. in 2006. In this program eight of the major PFOA-producing companies committed to reduce global emissions and residual product content of PFOA and related chemicals by 95% by 2010.18 However, less is known about temporal trends of human serum concentrations of other PFAAs than PFOS and PFOA. We investigated the temporal trends of 13 perfluoroalkyl carboxylates and sulfonates with varying carbon chain lengths, as well as perfluorooctane sulfonamide (FOSA) in human blood. The aim was to determine how the phase-out of PFOSrelated chemicals and the expected emission reductions for PFOA have influenced the serum concentrations of PFAAs in humans in Sweden. Temporal trends (1996−2010) of PFAA levels were studied in pooled samples of blood serum sampled 3 weeks after delivery from nursing primiparous women living in the Uppsala area. Uppsala County is situated in the densely populated area in the vicinity of the Swedish capital Stockholm on the central east coast of Sweden. In order to investigate if the PFAA levels observed during pregnancy and the nursing period are representative for the period of fetal development, PFOS, PFOA, and perfluorononanoate (PFNA) were measured in serial serum samples of pregnant and nursing women, including cord blood. A few studies have found strong positive correlations between maternal levels of PFAAs at delivery, such as PFOS and PFOA, and fetal cord blood levels.19,20 We wanted to take these observations further by determining if the timing of maternal sampling affected the representativeness of maternal PFAA levels as a marker for fetal cord blood levels. Moreover, we also tested the hypothesis that serum albumin levels are determinants of PFAA levels in maternal serum during pregnancy and after delivery. 21
samples were produced by adding equal amounts of serum from each woman to the pool. Details about the samples and the pooling scheme are given in Table S1 in the Supporting Information (SI). Serial Blood Samples during Pregnancy and Nursing Period. Serial maternal blood serum samples during pregnancy and after delivery as well as cord whole blood samples were analyzed for PFOS, PFOA, and PFNA. The aim was to study if PFAA concentrations in the maternal serum were indicative of the PFAA levels in cord blood of the newborn. Mothers were randomly selected among the primiparous women having donated blood during the period 1996−1999. This subgroup of women (N = 19, mean age 28 years, range 23−34) donated blood in the first and third trimester of pregnancy (weeks 6−11 and 32−34, respectively), at delivery (cord blood), and 3 weeks as well as 3 months after delivery. Chemical Analysis. Standards and Reagents. Abbreviations of perfluoroalkyl substances used in this study are according to Buck et al.23 and summarized in Table S2 in the SI. The target analytes for the temporal trend study were PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnDA, PFDoDA, PFTrDA, PFTeDA, PFBS, PFHxS, PFOS, PFDS, and FOSA, while in the serial blood sample study PFOA, PFNA, and PFOS were analyzed. All employed analytical standard compounds including mass-labeled standards were purchased from Wellington (Guelph, ON, Canada), with the exception of technical PFOS from Fluka (Buchs, Switzerland) used in the serial blood sample study. All solvents and reagents were of highest commercial purity and employed as received. Water obtained from a Milli-Q water purification unit (Millipore AB, Solna, Sweden) or, alternatively, HPLC grade water (Chromanorm, VWR Int., Stockholm, Sweden) was used. Extraction and Cleanup. An aliquot of 0.5 g of human serum or whole blood was weighed into a 13 mL polypropylene centrifuge tube and spiked with 25 μL of a MPFAC-MXA (Table S2 in the SI) solution in methanol containing 0.2 ng/μL each of eight mass-labeled internal standards (for the temporal trend study) or with 25 μL of a 0.2 ng/μL solution in methanol containing each of the mass-labeled internal standards 13C4− PFOA and 13C4−PFOS (for the serial blood sample study). The target analytes were extracted (and proteins precipitated) with 4 mL of acetonitrile in an ultrasonic bath at room temperature for 15 min. Following centrifugation at 2000 rpm for 5 min, the supernatant extract was removed and concentrated to 1 mL at 30 °C under nitrogen. The concentrated extract underwent dispersive cleanup on 25 mg graphitized carbon (Supelclean ENVI-Carb 120/400, Supelco, Stockholm, Sweden) and 50 μL glacial acetic acid by vortex mixing for 10 s in a 1.7 mL Eppendorf centrifuge tube. The sample was centrifuged at 10 000 rpm for 10 min and a volume of 0.5 mL of the clear supernatant extract was added to 0.5 mL of 4 mM aqueous ammonium acetate in an autoinjector vial. Before instrumental analysis 25 μL of a 0.1 ng/μL solution of the volumetric standards 13C8−PFOA and 13C8−PFOS (Table S2 in the SI) in methanol were added. The volumetric standards (sometimes also called injection standards or recovery (internal) standards) were used in the calculation of recoveries of the internal standards for correction of differences in final extract volumes and injection volumes. Instrumental Analysis and Quantification. Aliquots of the final extracts were injected automatically on a high performance liquid chromatography system coupled to high resolution mass spectrometry (HPLC/HRMS for the temporal trend study,
■
MATERIALS AND METHODS Study Design and samples. Temporal Trends of PFAAs 1996−2010. Between 1996 and 2010, primiparous women living in Uppsala County, Sweden, were recruited in a study of temporal trends of serum and breast milk concentrations of persistent halogenated organic compounds in pregnant and nursing women, the POPUP study (Persistent Organic Pollutants in Uppsala Primiparas).22 Blood samples from 413 mothers (ages 19−41 years) were taken 3 weeks (more precisely within the fourth week) after delivery. No samples were taken during 2003 and 2005 due to lack of funds. The study was approved by the regional ethical vetting board in Uppsala, and the participating women gave informed consent before donating the blood samples. Blood serum levels were determined for 13 PFAAs and FOSA. The aim was to investigate the temporal trends in human serum during the period of industrial changes of production and emission patterns of PFAAs. In general three pooled serum samples were produced for each sampling year, resulting in a total of 36 pooled samples for analysis. Pooled 9072
dx.doi.org/10.1021/es301168c | Environ. Sci. Technol. 2012, 46, 9071−9079
Environmental Science & Technology
Article
5 μL injections) or tandem mass spectrometry (HPLC/MS/ MS for the serial samples study, 25 μL injections). The HPLC/ HRMS system consisted of an Acquity Ultra Performance LC (Waters, Milford, MA) coupled to a Premier quadrupole timeof-flight instrument (Micromass, Manchester, UK). The instrumental setup for the HPLC/MS/MS method was an Alliance 2695 HPLC pump (Waters; degasser replaced with helium degassing) coupled to a Quattro II triple quadrupole MS (Micromass). The instrumental methods and parameters are described in detail by Ullah et al.9 (HPLC/HRMS) and Verreault et al.24 (HPLC/MS/MS). Optimized instrumental parameters of both mass spectrometric systems are further given in Table S2 in the SI. Quantification was performed in extracted high resolution mass chromatograms (HRMS) or in selected reaction monitoring chromatograms (MS/MS) using the internal standard method. An external solvent-based linear calibration curve was applied for calculation of relative response factors. The calibration included concentrations from the individual quantification limits to 5 ng/mL, thus bracketing the quantifiable concentrations of the PFAAs in the sample extracts. The linear regression (r2 values) and residuals for all analytes are given in Table S3 in the SI. In the temporal trend study 13C4−PFOA was used as internal standard for PFHpA, 13 C2−PFDoDA for PFTrDA and PFTeDA, 18O2−PFHxS for PFBS and 13C4−PFOS was used as internal standard for PFDS and FOSA. 13C4−PFOA was employed as internal standard for PFNA in the serial blood sample study. For PFHxS, PFOS, PFDS, and FOSA, which showed quantifiable signals of branched isomers, one area for the sum of all isomers was used for calculation of the (sum) concentration. The response factors of relevant branched PFOS isomers were shown to be similar to the response factor of linear PFOS in single stage MS.25 Thus the approach of quantifying the sum of all isomers using a purely linear calibration standard was assumed to be accurate in the temporal trend study using HRMS. In MS/MS this approach may lead to an overestimation of the true concentration of PFOS when using the product ion m/z 80 and a purely linear calibration standard.26 Therefore, a technical PFOS standard with a similar isomer pattern as encountered in the serum samples was used as calibration standard in the serial blood sample study (Table S2 in the SI). Analytical Quality Control. Sample collection tube (polyethylene or glass) blanks and instrumental blank experiments (solvent injections) did not show any contamination with the target analytes. To monitor the background contamination from the complete method, a procedural blank extraction was performed with every batch of samples, resulting in four (time trend study analyzed by HRMS) and seven (serial blood sample study analyzed by MS/MS) procedural blank experiments. The procedural blanks were treated in exactly the same way as the samples but without serum or whole blood matrix present. None of the target analytes was detected in any of the procedural blank extracts from the serial blood sample study. A minor procedural blank contamination was consistently found for PFOA, PFNA, and PFUnDA in the time trend study, due to the better sensitivity of the HRMS method compared to the MS/MS method. For these three analytes method detection limits (MDLs) were defined as mean plus three standard deviations of the quantified background contamination signals and method quantification limits (MQLs) were defined as three times the corresponding MDLs. In the absence of procedural blank contamination MDLs and MQLs were defined as the
lowest concentration in a serum sample giving a chromatographic signal with a signal-to-noise ratio of 3 and 10, respectively. A complete list of compound specific MDLs and MQLs for both instrumental systems is given in Table S4 in the SI. Matrix (serum and whole blood) spike extraction experiments were performed for all analytes and individual absolute recoveries ranged between 68 and 93% (mean of triplicate determination after correction for endogenous PFAA concentrations). Individual recoveries are given in Table S4 in the SI. Recoveries of the stable isotope mass-labeled internal standards in the temporal trend study (N = 36) were on an average between 78 and 92% and are given in Table S5 in the SI. All absolute recoveries were calculated using a solvent-based calibration standard. The consistency of quantified recoveries from spiked serum and whole blood samples, as well as the consistency of recoveries between different PFAA homologues (SI Table S4) showed that there was no significant matrix effect associated with the analytical method. Furthermore, quantification using the internal standard method corrected for recovery losses. The largest deviation of calculated recoveries between a quantified analyte and its internal standard was 7% for PFBS and 18O2−PFHxS, respectively (Tables S4 and S5 in the SI). A human serum sample previously used in an international interlaboratory comparison study in 200627 was analyzed six times along with different batches of samples. Precision expressed as coefficient of variation for the six replicates was ≤11% for all quantified analytes. The mean quantified concentrations deviated by
