Philippe Marchand

PFOS (perfluorooctanesulfonate) in serum is negatively associated with testosterone levels, but not with semen quality, in healthy men

STUDY QUESTIONnIs exposure to perfluorinated compounds (PFCs) associated with testicular function (reproductive hormone levels and semen quality) in healthy men?nnnSUMMARY ANSWERnPFOS levels were significantly negatively associated with serum testosterone (total and calculated free), but not with any other reproductive hormones or semen quality.nnnWHAT IS KNOWN ALREADYnIn animals, some PFCs have endocrine disrupting potential, but few studies have investigated PFCs in relation to human testicular function. Previously, we and others have observed a negative association between serum PFC levels and sperm morphology. The potential associations with reproductive hormones remain largely unresolved.nnnSTUDY DESIGN, SIZE, DURATIONnA cross-sectional study of 247 men was conducted during 2008-2009.nnnPARTICIPANTS/MATERIALS, SETTING, METHODSnHealthy men from the general population, median age of 19 years, gave serum and semen samples. Serum samples were analysed for total testosterone (T), estradiol (E), sex hormone-binding globulin (SHBG), luteinizing hormone (LH), follicle-stimulating hormone (FSH) and inhibin-B and 14 PFCs, including perfluorooctanesulfonate (PFOS). Semen samples were analysed according to the WHO criteria.nnnMAIN RESULTS AND THE ROLE OF CHANCEnPFOS levels were negatively associated with testosterone (T), calculated free testosterone (FT), free androgen index (FAI) and ratios of T/LH, FAI/LH and FT/LH. Other PFCs were found at lower levels than PFOS and did not exhibit the same associations. PFC levels were not significantly associated with semen quality. PFOS levels in these samples collected in 2008-2009 were lower than in our previous study of men participating in 2003.nnnLIMITATIONS, REASONS FOR CAUTIONnResults were robust to adjustment for relevant confounders; however, the possibility of chance associations due to multiple testing or effects of uncontrolled confounding cannot be ruled out.nnnWIDER IMPLICATIONS OF THE FINDINGSnOur previous findings of decreased sperm morphology in the most highly PFC exposed men were not replicated, possibly due to a lack of highly exposed individuals; however, a recent independent study also did corroborate such an inverse association. The negative association between serum PFOS and testosterone indicates that testosterone production may be compromised in individuals with high PFOS exposure.nnnSTUDY FUNDING/COMPETING INTEREST(S)nThe authors received financial support from the European Commission (DEER, FP7-2007-212844), the Danish Agency for Science, Technology and Innovation (grant nos. 27107068 and 09-067180), Rigshospitalet (grant no. 961506336), the University of Copenhagen, the Danish Ministry of Health and the Danish Environmental Protection Agency (MST-621-00013), and Kirsten and Freddy Johansen Foundation (grant no. 95-103-72087). The funding organizations played no role in the design and conduct of the study, in collection, management, analysis and interpretation of the data; or in the presentation, review or approval of the manuscript. The authors declare that they have no competing financial interests.

Dietary exposure to perfluoroalkyl acids of specific French adult sub-populations: high seafood consumers, high freshwater fish consumers and pregnant women.

Perfluoroalkyl acids (PFAAs) are globally found in various media, including food and especially fishery products. In the present study, the dietary exposure to 15 perfluoroalkyl acids was assessed for 3 French adult populations, namely high seafood consumers, high freshwater fish consumers, and pregnant women. Purified food extracts were analysed by LC-MS/MS and PFBA, PFPA, PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnA, PFTrDA, PFTeDA, PFBS, PFHxS, PFHpS, PFOS and PFDS were monitored and quantified according to the isotope dilution principle. Under lower bound (LB) hypothesis (i.e. contamination values<LOD considered as 0), high freshwater fish consumers appear as the most exposed to PFOS (7.5 ng.kg(-1) bw.d(-1)), PFUnA (1.3 ng.kg(-1) bw.d(-1)), PFDA (0.4 ng.kg(-1) bw.d(-1)) and PFHpS (0.03 ng.kg(-1) bw.d(-1)) while high seafood consumers appear as the most exposed to PFOA (1.2 ng.kg(-1) bw.d(-1)), PFNA (0.2 ng.kg(-1) bw.d(-1)) and PFHxS (0.06 ng.kg(-1) bw.d(-1)). For all considered populations, the major exposure contributors are fish, seafood and water under LB hypothesis, while dairy products, bread and crispbread are the main contributors under upper bound (UB) hypothesis. Besides this food exposure assessment, further studies are needed to assess the more global PFAA exposure, taking into account indoor and outdoor air, dust and cutaneous contact, which could be other important contributors for this particular class of chemicals.

Simultaneous determination of 16 brominated flame retardants in food and feed of animal origin by fast gas chromatography coupled to tandem mass spectrometry using atmospheric pressure chemical ionisation

A gas chromatography tandem mass spectrometry method using atmospheric pressure chemical ionisation was developed for the monitoring of 16 brominated flame retardants (7 usually monitored polybromodiphenylethers (PBDEs) and BDE #209 and 8 additional emerging and novel BFRs) in food and feed of animal origin. The developed analytical method has decreased the run time by three compared to conventional strategies, using a 2.5m column length (5% phenyl stationary phase, 0.1mm i.d., 0.1μmf.t.), a pulsed split injection (1:5) with carrier gas helium flow rate at 0.48mLmin(-1) in one run of 20u2009min. For most BFRs, analytical data were compared with the current analytical strategy relying on GC/EI/HRMS (double sector, R=10000 at 10% valley). Performances in terms of sensitivity were found to meet the Commission recommendation (118/2014/EC) for nBFRs. GC/APCI/MS/MS represents a promising alternative for multi-BFRs analysis in complex matrices, in that it allows the monitoring of a wider list of contaminants in a single injection and a shorter run time.

Effects of pan cooking on micropollutants in meat.

This work presents the effects of pan cooking on PCBs, PCDD/Fs, pesticides and trace elements in meat from a risk assessment perspective. Three different realistic cooking intensities were studied. A GC×GC-TOF/MS method was set up for the multiresidue analysis of 189 PCBs, 17 PCDD/Fs and 16 pesticides whereas Cd, As, Pb and Hg were assayed by ICP-MS. In terms of quantity, average PCB losses after cooking were 18±5% for rare, 30±3% for medium, and 48±2% for well-done meat. In contrast, average PCDD/F losses were not significant. For pesticides, no loss occurred for aldrin, lindane, DDE or DDD, whereas losses exceeding 80% were found for dieldrin, sulfotep or phorate. Losses close to the margin of error were observed for trace elements. These results are discussed in light of the physicochemical properties of the micropollutants as well as of water and fat losses into cooking juice.

Study on polychlorobiphenyl serum levels in French consumers of freshwater fish

INTRODUCTIONnPolychlorobiphenyls (PCBs) are persistent pollutants that are widespread in the environment and in foodstuffs, particularly in freshwater fish, which frequently exceed the maximum levels set by European regulations.nnnOBJECTIVESnFirst, we describe the consumption of freshwater fish and serum PCB levels in French anglers, a population expected to have the highest level of dietary PCB exposure. Second, we investigated whether there is a statistical relationship between serum PCB levels and the angler consumption of freshwater fish with high PCB bioaccumulation potential (PCB-BP(+) freshwater fish) in order to make recommendations with regard to safe consumption of freshwater fish.nnnMETHODSnWe conducted a survey of anglers from six sites with contrasting PCB contamination levels. The survey included a food consumption frequency questionnaire and blood samples were taken to assess serum PCB levels. We used a regression model to determine the main factors contributing to serum PCB levels.nnnRESULTSnConsumption of PCB-BP(+) freshwater fish was relatively infrequent. Serum PCB levels of the study population and of women of childbearing age were in the same range as those observed in the French population and in neighbouring European countries, but higher than in the North American population. The two factors with the highest positive association with serum PCB levels were age (R(2)=61%) and the consumption of PCB-BP(+) freshwater fish (R(2)=2%). Using the regression model, we calculated, for several scenarios depending on the age and gender of the population, the maximum annual frequencies for PCB-BP(+) freshwater fish consumption that do not exceed the critical body burden threshold.nnnCONCLUSIONnFollowing the results of this study, the French agency for food, environmental and occupational health and safety (ANSES) issued an opinion and recommended some specific maximum freshwater fish consumption frequencies to protect the French general population.

Field investigation to determine the environmental source of PCBs in a pig farm

During a national monitoring plan, a pork fat sample was declared non-compliant for the sum of dioxins and PCB-DL (EU regulation). The National Reference Laboratory together with competent authorities conducted extended investigations to determine rapidly the contamination source at the farm level. A range of samples (nu202f=u202f129), representative of potential contamination sources, was selected for further characterization (fat, feed, materials, dust, soil) and was analyzed for PCDD/Fs and DL-PCBs by GC-HRMS. A hot spot was localized in the farm, which corresponded to a pre-feed storage tank, the paints of which presented huge DL-PCB concentrations (>1u202f×u202f106u202fpgu202fg-1), responsible for the contamination. The present case report describes a new source of PCB contamination, previously undescribed.

Simultaneous analysis of historical, emerging and novel brominated flame retardants in food and feed using a common extraction and purification method

Brominated Flame Retardants (BFRs) are still widely used for industrial purposes. These contaminants may enter the food chain where they mainly occur in food of animal origin. The aim of our work was to provide a unique method able to quantify the widest range of BFRs in feed and food items. After freeze-drying and grinding, a pressurized liquid extraction was carried out. The extract was purified on acidified silica, Florisil® and carbon columns, the four separated fractions were analyzed by gas and liquid chromatography coupled to high resolution and tandem mass spectrometry. Isotopic dilution was preferentially used when commercial labelled compounds were available. Analytical sensitivity was in accordance with the expectations of Recommendation 2014/118/EU for PBDEs, HBCDDs, TBBPA, TBBPA-bME, EHTBB, BEHTEBP and TBBPA-bME. Additional BFRs were included in this analytical method with the same level of performances (LOQs below 0.01u202fngu202fg-1 ww). These are PBBs, pTBX, TBCT, PBBz, PBT, PBEB, HBBz, BTBPE, OBIND and T23BPIC. However, some of the BFRs listed in Recommendation 2014/118/EU are not yet covered by our analytical method, i.e. TBBPA-bOHEE, TBBPA-bAE, TBBPA-bGE, TBBPA-bDiBPrE, TBBPS, TBBPS-bME, TDBPP, EBTEBPI, HBCYD and DBNPG. The uncertainty measurement was fully calculated for 21 of the 31 analytes monitored in the method. Reproducibility uncertainty was below 23% in isotopic dilution. Certified reference materials are now required to better characterize the trueness of this method, which was applied in the French National Control Plans.

Past and recent state of sediment contamination by persistent organic pollutants (POPs) in the Rhône River: Overview of ecotoxicological implications

Twenty-one sediment samples were taken from five dated sediment cores collected along the Rhône River from 2008 to 2011. A total of 17 polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs), 7 polychlorinated biphenyls (PCBs), 8 polybrominated diphenyl ethers (PBDEs), 3 polybrominated biphenyls (PBBs), 3 hexabromocyclododecanes (HBCD) and 31 organochlorine pesticides (OCPs) were investigated to provide information on deposition dynamics in time and space, but also regarding the ecotoxicological risks associated with these contaminants. Median concentrations of total PBDEs are nine times lower than the levels of total PCBs along the entire studied stretch of the Rhône River. The results show that total PBDEs concentrations range from 0.06 to 239u202fμg·kg-1 DW with a median value of 3.81u202fμg·kg-1 DW and a maximum concentration measured in the years 2000s. These maximum concentrations are identical to those measured for total PCBs at the end of the 1990s, but show a different pattern of distribution. Abnormal dichlorodiphenyltrichloroethane (DDT) levels were also detected in the downstream section of the river, with a peak concentration of 147.5u202fμg·kg-1 DW measured at the GEC site from 2005 onwards. Analyses of the enantiomeric fractions reveal a fresh input resulting from a technical formulation. Sediments from the core sampled at the most downstream site (GEC) are found to be highly toxic to organisms living nearby, particularly because of the total PCDD/Fs, DDE and DDT levels. In addition, based on available sediment quality guidelines, there may be a potential bioaccumulation risk for humans not only for these three compounds of concern but also for total PCBs and 7 out of the 8 analysed PBDEs.

The challenging use and interpretation of circulating biomarkers of exposure to persistent organic pollutants in environmental health: Comparison of lipid adjustment approaches in a case study related to endometriosis

The gold-standard matrix for measuring the internal levels of persistent organic pollutants (POPs) is the adipose tissue, however in epidemiological studies the use of serum is preferred due to the low cost and higher accessibility. The interpretation of serum biomarkers is tightly related to the understanding of the underlying causal structure relating the POPs, serum lipids and the disease. Considering the extended benefits of using serum biomarkers we aimed to further examine if through statistical modelling we would be able to improve the use and interpretation of serum biomarkers in the study of endometriosis. Hence, we have conducted a systematic comparison of statistical approaches commonly used to lipid-adjust the circulating biomarkers of POPs based on existing methods, using data from a pilot case-control study focused on severe deep infiltrating endometriosis. The odds ratios (ORs) obtained from unconditional regression for those models with serum biomarkers were further compared to those obtained from adipose tissue. The results of this exploratory study did not support the use of blood biomarkers as proxy estimates of POPs in adipose tissue to implement in risk models for endometriosis with the available statistical approaches to correct for lipids. The current statistical approaches commonly used to lipid-adjust circulating POPs, do not fully represent the underlying biological complexity between POPs, lipids and disease (especially those directly or indirectly affecting or affected by lipid metabolism). Hence, further investigations are warranted to improve the use and interpretation of blood biomarkers under complex scenarios of lipid dynamics.

Occurence of legacy and novel brominated flame retardants in food and feed in France for the period 2014 to 2016

Determination of the occurrence levels of legacy and novel BFRs is today required to better understand the trends of BFRs contamination in food consecutive to the EU PBDEs restrictions and to proceed to a recent human food exposure in parallel. Therefore, concentrations of a large set of brominated flame retardants (BFRs) (nu202f=u202f27) including PBDEs, HBCDDs, TBBPA and novel flame retardants (nBFRs) have been determined in more than 600 food and feed samples collected between 2014 and 2016 in the context of French monitoring plans. Although legacy BFRs had already been studied in France, such a survey constituted the very first determination of nBFRs occurrence in foodstuffs at the national level. The concentration levels measured in fish and fish products were in general higher than in the other food categories. PBDEs were detected in 70% of the samples and were observed as the most abundant congeners (representing 80% of the sum of the monitored BFRs), while α-HBCDD could also be considered as a predominant congener (up to 26% of the sum of the monitored BFRs in fishes). nBFRs concentration levels were most of the time below the LOQ, except PBT, PBBz and HBBz which were more frequently detected at low levels. Also investigated in the study, BRPs exhibited high concentration levels in crustaceous (maximum valueu202f>u202f2700u202fpg/g ww).