Phosphorous flame retardants in Swedish market basket food samples, and estimation of per capita intake

(1)

Report to the Swedish EPA (the Health-Related Environmental Monitoring Program) Överenskommelse nr 2215-15-010

Phosphorous flame retardants in Swedish market basket food samples, and estimation of per capita intake

Per Ola Darnerud and Anders Glynn, Livsmedelsverket

Adrian Covaci, Govindan Malarvannan and Giulia Poma, University of Antwerp

2016-08-11

(2)

NATIONELL MILJÖÖVERVAKNING

PÅUPPDRAGAV NATURVÅRDSVERKET

ÄRENDENNUMMER AVTALSNUMMER PROGRAMOMRÅDE

DELPROGRAM

NV-05468-15 2215-15-010 Hälsorelaterad MÖ Biologiska mätdata – organiska ämnen

Phosphorous flame retardants in Swedish market basket food samples, and estimation of per capita intake

Rapportförfattare

Per Ola Darnerud and Anders Glynn, Livsmedelsverket

Adrian Covaci, Govindan Malarvannan and Guilia Poma, University of Antwerp, Belgium

Utgivare Livsmedelsverket Postadress

Box 622, 751 26 Uppsala Telefon

018-175500

Rapporttitel

Phosphorous flame retardants in Swedish market basket food samples, and estimation of per capita intake

Beställare Naturvårdsverket 106 48 Stockholm Finansiering

Nationell hälsorelaterad miljöövervakning

Nyckelord för plats Uppsala

Nyckelord för ämne

Fosforinnehållande flamskyddsmedel, PFRs, TEHP, TNBP, TCEP, EHDPHP, TDCIPP, TCIPP

Tidpunkt för insamling av underlagsdata 2015

Sammanfattning

Analyser av åtta fosforinnehållande flamskyddsmedel (PFRs) har utförts i matkorgsprover från den senaste matkorgsundersökningen (Matkorgen 2015). De ämnen som analyseras är (i förkortning) TEHP, TNBP, TCEP, TBOEP, TPHP, EHDPHP, TDCIPP samt TCIPP. De undersökta

matkorgsproverna kommer från studier där matkassar från vanliga livsmedelskedjor i Sverige tas in för analys, och där de insamlade livsmedlen analyseras gruppvis (ex kött, fisk, mejeriprodukter, spannmål) för en mängd näringsämnen men även toxiska ämnen. Med hjälp av försäljningsstatistik kan

konsumtionen av olika livsmedelsgrupper beräknas, och tillsammans med haltdata kan per capita- intaget för dessa ämnen tas fram. Det är första gången som matkorgsprover analyseras för förekomst av PFRs.

Mätbara halter av ett flertal av undersökta PFRs kunde analyseras i proverna från de tretton olika livsmedelsgrupperna, och högst halter återfanns generellt i prover från cerealier, bakverk, fetter/oljor , vegetabilier, samt socker och sötsaker. Högst halter i livsmedel, och även det beräknade högsta per capita-intaget (49 ng/kg kroppsvikt/dag), observerades för EHDPHP (etylhexyldifenylfosfat). För fyra PFRs (TCEP, TPHP, TDCIPP och TCIPP) låg per capita-intaget på 6-12 ng/kg kroppsvikt/dag, medan inga beräkningar gjordes för de tre återstående pga att flertalet haltdata <LOQ. I jämförelse med hälsobaserade referensvärden, ligger de beräknade per capita-intagen lägre med en faktor mer än 2 000. Fastän denna undersökning visar en stor marginal mellan beräknat intag och nivåer där skadliga effekter kan förekomma i experimentella modeller, vet vi fortfarande litet om det totala intaget då inhalation och nedsväljning av damm kan vara de största exponeringsvägarna.

(3)

DELPROGRAM

Background

The need for fire protection devices and chemicals is linked to our current lifestyle and the increasing use of electronic equipment. The presence of flame retardants (FRs) in various products is considered to be lifesaving, and in UK alone estimations claim over 1 000 saved lives (plus decrease in injuries and material damage) during a 10-year period (Kucewicz, 2006). At the same time, environmental and human exposure to FRs has been related to a number of adverse outcomes, exemplified by the hexabromobenzene (HBB/FireMaster), accident in Michigan, USA, and (if flame retardant oils are included) polychlorinated biphenyls (PCBs) contaminations in Taiwan and Japan (the Yusho and Yusheng accidents). Therefore, FRs with adverse effects on man and wildlife have gradually been phased out and replaced with new ones. During the 1960s/70s, the increased use of brominated flame retardants resulted in subsequent exposure to these BFRs (e.g. polybrominated diphenyl ethers - PBDEs, hexabromocyclododecane - HBCD), affecting both wildlife and humans in the

following years (de Wit et al., 2010; Meironyté et al., 1999). Environmental persistence and reported adverse outcomes in individuals exposed to BFRs (e.g. Lyche et al., 2015) has resulted in restriction in use and phasing out, and again the search for new FR continues. Phosphorous flame retardants (PFRs) have gained an increasing interest, although they have been in use for over 150 years (van der Veen and de Boer, 2012).

By chemical structure, PFRs can be divided into three main groups: inorganic, organic (non-halogen), and halogenated PFRs. The phosphorous content could vary between 8 and 100% (100% in red phosphorous). Within these three groups, there are two basic types of compounds, depending on the chemical bonding to the manufactured product, namely reactive and additive PFRs. Whereas the reactive compounds are chemically bound to the product matrix, the additive compounds are not and could more easily leach out from the products, leading to decreased FR protection, and, as a consequence, an increased risk for human and wildlife exposure to these FRs. PFRs are responsible for 20% (ca. 90 000 tons) of the FR consumption in Europe (data from 2006; CEFIC, 2007), are easy to handle, and can be used both in a wide range of textile materials and in other industrial product processes. As an example, the use of PFRs in printed circuit boards will facilitate the recyclability of these products.

Numerous reports have shown that PFRs are found in environmental matrices as air, dust, surface water, sediment, and biota in a number of countries (den der Veen and de Boer, 2012). In Swedish studies, several PFRs were found in indoor air and dust (Björklund et al. 2004; Marklund et al. 2003;

Marklund et al., 2005a) and also in a kindergarten (Tollbäck et al., 2006). In a study on a Swedish sewage treatment plant, Marklund et al. (2005b) found that tris-2-chloroethyl phosphate (TCEP) passes through the plant without being eliminated. In biota, Sundkvist et al. (2010) analyzed herring, perch, mussels, eelpout, and salmon from Swedish waters and were able to quantify several PFRs, some of them in µg/g levels (tris-chloro-isopropyl phosphate (TCIPP) (max 1.3 µg/g, mussels from

(4)

DELPROGRAM

marine waters). From these results, it could be hypothesized that human PFR exposure from both air/dust and food is plausible, but data are scarce.

The relatively few toxicological studies on organic PFRs in mammals show a low acute toxicity, but the halogen-containing PFRs show, in many cases, carcinogenic potential in animals (WHO, 1998).

Some PFRs also induce skin and eye dermatitis and may have immunological effects (Björklund, 2004;

Saabori et al., 1991). In a US study, hormone levels and semen quality parameters in 50 men were suggested to be associated with levels of two PFRs, TCIPP and tris(1,3-dichloro-2-propyl) phosphate (TDCIPP) found in house dust (Meeker and Stapleton, 2010). The endocrine disruption potential was later studied in experimental models (cell lines and zebrafish), and results showed that organic PFRs could alter sex hormone balance through several mechanisms (Liu et al., 2012). Another potential effect is neurotoxicity, based on the PFR similarities with organophosphate (OP) pesticides that exert many neurodevelopmental effects through mechanisms that are unrelated to acute toxicity via cholinesterase inhibition (Dishaw et al., 2011). Indeed, neurotoxicity assay studies and reviews suggest that organic PFRs may affect neurodevelopment with at least similar potency compared to other known or suspected neurotoxicants, including some BFRs (Dishaw et al., 2012; Behl et al., 2015; Hendriks and Westerink, 2015).

In the present report, eight organic PFRs (Table 1), including halogenated PFRs, were analyzed in food homogenates from a recent Swedish market basket study (Market Basket 2015), and based on these results the per capita exposure from food was estimated and discussed.

Methods

Analytical methods

Food samples (ca 0.5 g) from the market basket survey, representing 13 different matrices, were extracted and cleaned up as described in Appendix 1. Analyses were performed by GC-MS in the electron-impact (EI) mode. Recoveries were 53-71%, except for tris(2-butoxyethyl) phosphate TBOEP-d6 (33%). LOQs were calculated as the “blank + 3*SD of the blank” and normalized by sample weight. Further analytical issues on PFRs are discussed by Brandsma et al. (2013). The analyses were carried out at the Toxicological Center of the University of Antwerp.

Abbreviations of PFR substances (see Table 1) follow the nomenclature review by Bergman et al.

(2012).

Estimation of per capita intake

The per capita exposure concept is based on the Swedish Board of Agriculture (SBA) data on food production and trade statistics. The calculation is based on the per capita consumption, which represents the calculated mean population consumption of various food groups derived from Swedish sales and production statistics by dividing the total volume (of a food item/category) by the number of inhabitants in Sweden. From these figures, the per capita intake can be derived by

multiplying the per capita consumption figure for a specific food category by the concentration of the

(5)

DELPROGRAM

actual compound found in the food homogenate. As we used a food homogenate that represents 1%

of the yearly per capita consumption, we have to multiply by a factor 100, and the daily consumption is obtained by dividing the figures per 365.

General formulas:

Conc. in food x homogenate weight x 100 / 365 = daily intake from specific food category, per person (A)

Addition of all separate intake from food categories = Total daily intake from food, per person (B) The above-estimated intake is given on a per-person basis. To present the data also on a body weight basis, we have chosen to use a calculated figure for the average body weight of the Swedish

population (67.2 kg), which was produced in the previous marked basket survey (Market Basket 2010; NFA 2012).

Results

Table 2 presents all analytical PFR results produced in the frame of this project. As shown (figures in yellow), many results are below LOQ, and the percentages of values below LOQ vary from 55% to 100%, depending on the substance. Because of this, tris(2-ethylhexyl)phosphat (TEHP), tri(n- butyl)phosphate (TNBP), and TBOEP (96-100% of the values were below LOQ) were not included in the subsequent calculations on these compounds.

Table 3 gives the compiled data on PFR levels in the different food categories, presented as lower, medium and upper bound (LB, MB, UB) figures. At an overview, the 2-ethylhexyl diphenyl phosphate (EHDPHP) levels are highest among the five PFRs, and within EHDPHP cereals, pastries, fats/oils, and sugar etc. are the major food categories. As could be seen by comparing LB and UB levels, this compound contains also relatively few LOQ levels.

The estimated per capita intakes of the five PFRs have been calculated in Table 4 (based on MB values). When the consumption figures are included, they generally give the largest weight to cereals, pastries, fats/oils, sugars etc., and beverages, but compound differences occur. The largest intake comes from EHDPHP, adding up to 3.3 µg/person and day, or 49 ng/kg bw/day. In Figure 1, the intake figures have been used to produce diagrams on EHDPHP and TDCIPP exposure and the

contribution from different food categories to the total estimated intake. For EHDPHP, the four most prominent food categories (cereals, pastries, sugars, and beverages) constitute 71%, and the

corresponding figure for TDCIPP is 57%. In addition, within these four categories, the relative contribution differs widely for the two compounds.

To get some information on how levels below LOQ will influence the intake estimations, the intake calculations including all food categories were presented based on LB, MB, of UB figures (Table 5). In this case, the quotient UB/LB, which indicate the influence of <LOQ levels, results in a factor above 5

(6)

DELPROGRAM

for TCEP and thereby gives a wide range within which the “true” levels should be found. TCEP is indeed the compound of the five that resulted in most analytical results below LOQ (83%). For the other compounds, the UB/LB factors are smaller (1.5-2.1) which decrease this uncertainty.

Discussion

The presented data on the estimated intake of PFRs via food should be considered valuable, as very few earlier food intake studies have been presented. In a US study (Gunderson, 1988), market basket results gave intake ranges for eight age groups of 3.5-99, 0.3-4.4 and 23-71 ng/kg bw/day for TBP, TPHP and TEHP, respectively (as cited in Wei et al., 2015). In a Belgian study, intake of PFRs was based on levels in eel, and median intake for high consumers of certain PFRs were at most 1 ng/kg bw/day (for TCIPP) (Malarvannan et al., 2015). In a Swedish study, based on the levels of the sum of eight PFRs in eelpout and a fish consumption of 375 g/week, the resulting consumption of sumPFRs was calculated to be 180 ng/kg bw/day. However, the Belgian study was only based on fish intake, whereas the present estimation is based on all relevant food categories, and our data show that many food groups contribute to the total intake, especially cereals, pastries, sugars, beverages, etc.

Food intake of five separate PFRs (6-49 ng/kg bw/day) are roughly in the same range as from these earlier studies, but differences in studied analytes between studies make comparisons difficult.

We found that some PFRs are present in levels above LOQ in many of the studied food categories, and PFRs are not distributed in food similarly to lipophilic POPs, such as PCBs and chloropesticides, which are present at highest levels in foods of animal origin. One possible answer could be the moderate lipophilicity, which makes them considerably less prone to accumulate in fat depots of food-producing animals compared to PBDEs (e.g. Malarvannan et al., 2015). The ubiquitous presence of PFRs in our environment (e.g. Wei et al., 2015) makes it easy for these compounds to enter in various food chains and reach our food items. The food categories with the highest levels (cereals, pastries, fats/oils, sugar etc.) are also all industrially processed to a higher degree compared to many other food categories, and contamination during food processing is therefore a possibility. In

addition, the presence of PFRs as plasticizers in food packages (which is the case for e.g. EHDPHP) may also play a role. As compounds within the PFR group have many different applications both as flame retardants and plasticizers, the release of compounds to the environments could take place as result from these various fields of application.

There are many potential exposure routes for PFRs in humans, due to the ubiquitous usage of these compounds. However, the most important exposure route for most people seems to be via indoor inhalation and ingestion of (inhaled) dust (Sundkvist et al., 2010; van den Eede et al., 2011; Cequier et al., 2015; Wei et al., 2015). In general, the ingestion of dust seems to be quantitatively more important than inhalation. In indoor environments there are several materials and products containing PFRs, e.g., various building materials, soft foams, paints, wallpapers, insulation and sealant foams (Wei et al., 2015). In a Swedish estimation, the mean exposure to TCEP and TCIPP was estimated to 4 and 3 µg/day, and the major exposure occurred during transportation and working

(7)

DELPROGRAM

hours (Staaf and Östman, 2005). These estimated intakes are more than 4-fold higher than the per capita intakes estimated from food.

The paper of Ali et al. (2012) reported reference doses for a number of PFRs, which were, obtained by dividing chronic NOAELs by a factor of 1 000. These reference doses were subsequently used in risk estimations by Malarvannan et al. (2015).For four of the analyzed PFR compounds in our study (i.e. TCEP, TPHP, TDCIPP and TCIPP), we could compare the calculated per capita intake with the reference doses given in the paper of Ali. Our calculated per capita intake figures of the four

compounds (6-12 ng/kg bw/day; MB) were much lower than the corresponding reference doses (15 000-80 000 ng/kg bw/day), i.e. by a factor of more than 2 000. To conclude, our data show that there is a large margin between the estimated per capita intakes and corresponding reference doses.

However, the present per capita intake estimates exposure from food only, and cannot be used to speculate what the total exposure to these compounds would be, as we did not study the other exposure routes in the present study.

The major strength of the presented paper is that it contributes to a better knowledge about food as a route of exposure to PFRs, as very few papers have earlier reported on PFR exposure from food. In our report, we have tried to monitor all relevant food categories and we have not made any

assumptions on which food type that is most interesting to study. A weakness of the study may be that many analytical results were below LOQ, which decreases precision of data. A reason for the low PFR levels is the comparably fast excretion of these compounds, compared to e.g. PCBs and chloropesticides. However, there were enough QA/QC measures to conclude that the analyses have been performed in optimal conditions. These QA/QC measures included an appropriate number of laboratory blanks and a fish oil material which was used in the 1^st interlaboratory test on PFRs (Brandsma et al., 2013). The choice of PFR compound to include in the study was done based on earlier analytical results and chemical-analytical capacity.

In conclusion, we have analyzed eight PFRs in market basket samples obtained in Swedish shops in 2015. Measurable levels were found in many of the 13 studied food groups and highest levels were generally found in samples from cereals, pastries, fats/oils, and sugar/sweets. The medium bound per capita intakes were estimated for five PFRs, ranging from 6 to 49 ng/kg bw/day. In comparison to health-based reference points, the estimated intake figures were lower by a factor of more than 2000. Although there is a large margin between the estimated food intakes and levels causing effects in animals, we still know little about the total intake of PFRs and its relation to health, as inhalation and ingestion of dust seem to be the major exposure pathways.

(8)

DELPROGRAM

References

Ali, N., Dirtu, A.C., Van den Eede, N., Goosey, E., Harrad, S., Neels, H., Mannetje, A.t., Coakley, J., Douwes, J., Covaci, A., 2012. Occurrence of alternative flame retardants in indoor dust from New Zealand indoor sources and human exposure assessment. Chemosphere 88, 1276-1282.

Behl, M., Hseih, J.-H., Shafer, T.J., Mundy, W.R., Rice, J.R., Boyd, W. A., Freedman, J.H., Hunter III, E.S., Jarema, K.A., Padilla, S., Tice, R.R., 2015. Use of alternative assays to identify and prioritize

organophosphorous flame retardants for potential developmental and neurotoxicity. Neurotoxicol.

Teratol. 52, 181-193.

Bergman, Å., Rydén, A., Law, R.J., de Boer, J., Covci, A., Alaee, M., Birnbaum, L., Petreas, M., Rose, M., Sakai, S., van den Eede, N., van der Ven, I., 2012. A novel abbreviation standard for organobromine, organochlorine and organophosphorous flame retardants and som characteristics of the chemicals.

Environ. Int. 49, 57-82.

Björklund, J., Isetun, S., Nilsson, U., 2004. Selective determination of organophosphate flame

retardants and plasticizers in indoor air by gas chromatography, positive-ion chemical ionization and collision-induced dissociation mass spectrometry. Rapid Commun. Mass Spectrom. 18, 3079-3083.

Brandsma, S.H., de Boer, J.,Cofino, W.P., Covaci, A., Leonards, P.E.G.,2013. Organophosphorus flame- retardant and plasticizer analysis, including recommendations from the first worldwide

interlaboratory study.TRAC Trends Anal. Chem. 43, 217–228.

CEFIC, 2007. European Flame Retardants Association (EFRA). What are FRs? Flame Retardant Market Statistics. <www.cefic-efra.eu> (accessed 30.03.07).

Cequier, E., Sakhi, A.K., Marcé, R.M., Becker, G., Thomsen, C., 2015. Human exposure pathways to organophosphate triesters – A biomonitoring studiy of mother-child pairs. Environ. Int. 75, 159-165.

De Wit, C.A., Herzke, D., Vorkamp, K., 2010. Brominated flame retardants in the Arctic environment

— trends and new candidates. Sci. Tot. Environ. 408, 2885-2918.

Dishaw, L.V., Powers, C.M., Ryde, I.T., Roberts, S.C., Seidler, F.J., Slotkin, T.A., Stapleton, H.M., 2011.

Is the PentaBDE replacement, tris (1,3-dichloropropyl) phosphate (TDCPP), a developmental neurotoxicant? Studies in PC12 cells. Toxicol. Appl. Pharmacol. 256, 281-289.

Gunderson, E.L., 1988. J. Assoc. Off. Anal. Chem. 71, 1200.

Hendriks, H.S., Westerink, R.H.S., 2015. Neurotoxicity and risk assessment of brominated and alternative flame retardants. Neurotoxicol. Teratol. 52, 248-269.

Kucewicz, W.P., 2006. Brominated flame retardants – a burning issue. American Council on Science and Health. New York, USA, August 2006 (16 pp).

(9)

DELPROGRAM

Liu, X., Ji,K., Choi, K., 2012. Endocrine disruption potentials of organophosphate flame retardants and related mechanisms in H295R and MVLN cell lines and in zebrafish. Aquat. Toxicol.114–115,173–181.

Lyche, J.L., Rosseland, C., Berge, G., Polder, A., 2015. Human health risk associated with brominated flame-retardants (BFRs). Environ. Int. 74, 170-180.

Malarvannan, G., Belpaire, C., Geeraerts, C., Eulaers, I., Neels, H., Covaci, A., 2015.

Organophosphorous flame retardants in the European eel in Flanders, Belgium: Occurrence, fate and human health risk. Environ Res. 140, 604-610.

Marklund, A., Andersson, B., Haglund, P., 2003. Screening of organophosphorus compounds and their distribution in various indoor environments. Chemosphere 53, 1137-1146.

Marklund, A., Andersson, B., Haglund, P., 2005a. Organophosphorus flame retardants and plasticizers in air from various indoor environments. J. Environ. Monit. 7, 814-819.

Marklund, A., Andersson, B., Haglund, P., 2005b. Organophosphorus flame retardants and plasticizers in Swedish sewage treatment plants. Environ. Sci. Technol. 39, 7423-7429.

Meeker, J.D., Stapleton, H.M., 2010. House dust concentrations of organophosphate flame retardants in relation to hormone levels and semen quality parameters. Environ. Health Perspect.

118, 318-323.

Meironyté, D., Norén, K., Bergman, Å., 1999. Analysis of polybrominated diphenyl ethers in Swedish human milk. A time related trend study, 1972-1997. J. Toxicol. Environ. Health 58, 329-341.

NFA, 2012. Market Basket 2010 – chemical analysis, exposure estimation and health-related assessment of nutrients and toxic compounds in Swedish food baskets. The National Food Agency, Uppsala, Sweden (Report no. 7- 2012)

Saboori, A.M., Lang, D.M., Newcombe, D.S., 1991. Structural requirements for the inhibition of human monocyte carboxylesterase by organophosphorus compounds. Chem.-Biol. Interact. 80, 327–

338.

Staaf, T., Östman, C., 2005. Organophosphate triesters in indoor environments. J. Environ. Monit. 7, 883-887.

Sundkvist, A.M., Olofssona, U., Haglunda, P., 2010. Organophosphorus flame retardants and plasticizers in marine and fresh water biota and in human milk. J. Environ. Monit. 12, 943-951.

Tollbäck, J., Tamburro, D., Crescenzi, C., Carlsson, H., 2006. Air sampling with Empore solid phase extraction membranes and online single-channel desorption/liquid chromatography/mass spectrometry analysis: determination of volatile and semi-volatile organophosphate esters. J.

Chromatogr. A 1129, 1–8.

(10)

DELPROGRAM

Van de Eede, N., Dirtu, A., Neels, H., Covaci, A., 2011. Analytical developments and preliminary assessment of human exposure to organophosphate flame retardants from indoor dust. Environ. Int.

37, 454-461.

Van der Veen, I., De Boer, J., 2012. Phosphorus flame retardants: properties, production, environmental occurrence, toxicity and analysis. Chemosphere 88, 1119-1153.

Wei, G.-L., Li, D.-Q., Zhou, M.-N., Liao, Y.-S., Xie, Z.-Y., Guo, T.-L., Li, J.-J., Zhang, S-Y., Liang, Z.-Q., 2015. Organophosphorous flame retardants and plasticizers: Sources, occurrence, toxicity and human exposure. Environ Pollut.

World Health Organization, 1998. Environmental Health Criteria 209, Flame Retardants:

Tris(chloropropyl) Phosphate and Tris(2-chloroethyl) Phosphate. World Health Organization, Geneva, Switzerland.

(11)

DELPROGRAM

Table 1. Abbreviations of the eight PFRs studied in the present project

TDCIPP Tris(1,3-dichloro-2-propyl) phosphate TCIPP Tris(1-chloro-2-propyl) phosphate TCEP Tris(2-chloroethyl) phosphate TNBP Tri-n-butyl phosphate

TEHP Tris(2-ethylhexyl) phosphate TPHP Triphenyl phosphate

TBOEP Tris(2-butoxyethyl) phosphate EHDPHP 2-Ethylhexyl diphenyl phosphate

(12)

DELPROGRAM

Table 2. Total data set of PFR levels in Swedish market basket samples from 2015 (pg/g fresh wt.).

Levels in yellow are beneath LOD, showing specific LOD levels for different food categories

No. Sample cat. TEHP TNBP TCEP TBOEP TPHP EHDPHP TDCIPP TCIPP 1+2

1 Cereals < 2150 < 3000 <500 <3000 673 4236 <500 2803

2 " < 2150 <3000 <500 <3000 <500 <3000 <500 2100

3 " < 2150 < 3000 <500 <3000 <500 9248 893 <400

4 " < 2150 < 3000 <500 <3000 <500 1188 <500 470 5 " < 2150 < 3000 <500 <3000 <500 4681 <500 589 6 Pastries < 2150 < 3000 <500 <3000 1240 8443 <500 914 7 " < 2150 < 3000 <500 <3000 <500 10057 <500 701 8 Meat <800 <1000 < 200 <1000 <200 <1000 <200 <150 9 " <800 <1000 < 200 <1000 <200 <1000 <200 <150

10 " <800 <1000 < 200 <1000 324 <1000 522 <150

11 " <800 <1000 < 200 <1000 1539 1215 <200 <150

12 " <800 <1000 < 200 <1000 228 <1000 <200 <150 13 Fish <800 <1000 < 200 <1000 <200 <1000 <200 <150

14 " <800 <1000 < 200 <1000 434 1700 1051 <150

15 " <800 <1000 < 200 <1000 950 5802 <200 <150

16 " <800 <1000 < 200 <1000 1561 2552 <200 <150

17 " <800 <1000 < 200 <1000 <200 1753 <200 <150 18 Dairy, fluid^# <800 <1000 218 <1000 <200 <1000 500 <150 19 " <600 <800 <100 <700 <100 <700 <100 <100 20 " <800 <1000 192 <1000 <200 <1000 <200 <150 21 " <600 <800 <100 <700 <100 <700 <100 <100 22 Dairy, solid <1000 <1500 <300 <1500 <300 <2000 <300 <200 23 " <1000 <1500 <300 <1500 <300 <2000 <300 <200 24 " <1000 <1500 <300 <1500 <300 <2000 <300 <200 25 " <1000 <1500 <300 <1500 <300 <2000 <300 <200 26 " <1000 <1500 <300 <1500 <300 <2000 <300 <200

27 Eggs^# <800 <1000 <150 <700 <150 584 <150 <150

28 " <600 <1000 <150 <700 <150 1263 173 153

29 " <600 <1000 <150 <700 <150 901 <150 231

30 " <800 <1000 < 200 <1000 <200 876 393 <150

31 Fats, oils^# < 5000 <8000 < 2000 <6000 1754 <6000 <2000 <1500 32 " < 5000 <8000 < 2000 <6000 12370 6706 <2000 <1500 33 " < 5000 <8000 < 2000 <6000 3489 7613 <2000 <1500 34 " < 5000 <8000 < 2000 <6000 1356 <6000 <2000 <1500

35 Vegetables <200 <300 316 <300 131 240 1061 333

36 " <200 <300 445 <300 58 386 358 167

37 " <200 466 356 <300 <50 288 211 316

38 " <200 <300 506 <300 94 394 <50 <50

(13)

DELPROGRAM

39 " <200 <300 445 <300 <50 <200 177 70

40 Fruit <800 <1000 <150 <700 <150 946 574 <150

41 " <800 <1000 <150 <700 <150 <700 237 <150

42 " <800 <1000 <150 <700 <150 <700 233 <150

43 " <800 <1000 161 <700 <150 <700 <150 241

44 " <800 <1000 <150 <700 <150 <700 339 <150

45 Potatoes <800 <1000 255 <700 <150 <700 293 233

46 " <800 1015 <150 <700 476 <700 204 <150

47 " <800 <1000 <150 <700 <150 <700 177 176

48 " <800 <1000 <150 <700 <150 <700 485 278

49 " <800 <1000 <150 <700 215 <700 290 <150

50 Sugar, sweets < 2150 < 3000 <450 <3000 <500 <3000 1228 <400 51 " < 2150 <3000 <450 <3000 <500 5923 <500 <400 52 Beverages < 2150 < 3000 <450 <3000 <500 <3000 1069 <400 53 " < 2150 < 3000 <450 <3000 <500 <3000 642 <400

<LOQ/all 53/53 51/53 44/53 53/53 36/53 29/53 31/53 37/53

Detection freqv.(%) 0 4 17 0 32 45 42 30

# One out of five samples missing in these groups

(14)

DELPROGRAM

Table 3. Levels of PFRs in the 2015 market basket food categories, purchased on the Swedish market (pg/g fresh wt; mean of 2-5 analyses)

Sample cat. (mean v.) TCEP TPHP EHDPHP TDCIPP TCIPP 1+2

Cereals LB 0 135 3871 179 1192

MB 250 335 4171 379 1232

UB 500 535 4471 579 1272

Pastries LB 0 620 9250 0 807

MB 250 745 9250 250 807

UB 500 870 9250 500 807

Meat LB 0 418 243 104 0

MB 100 458 643 184 75

UB 200 498 1043 264 150

Fish LB 0 589 2362 210 0

MB 100 629 2462 290 75

UB 200 669 2562 370 150

Dairy, fluid LB 102 0 0 125 0

MB 127 75 425 175 63

UB 152 150 850 225 125

Dairy, solid LB 0 0 0 0 0

MB 150 150 1000 150 100

UB 300 300 2000 300 200

Eggs LB 0 0 906 142 96

MB 81 81 906 179 133

UB 163 163 906 217 171

Fats, oils LB 0 4742 3580 0 0

MB 1000 4742 5080 1000 750

UB 2000 4742 6580 2000 1500

Vegetables LB 414 57 262 361 177

MB 414 67 282 366 182

UB 414 77 302 371 187

Fruit LB 32 0 189 277 48

MB 92 75 469 292 108

UB 152 150 749 307 168

Potatoes LB 51 138 0 290 137

MB 111 183 350 290 167

UB 171 228 700 290 197

(15)

DELPROGRAM

Sugar etc LB 0 0 2961 614 0

MB 225 250 3711 739 200

UB 450 500 4461 864 400

Beverages LB 0 0 0 855 0

MB 225 250 1500 855 200

UB 450 500 3000 855 400

(16)

DELPROGRAM

Table 4. The Market basket 2015 estimated per capita intake of PFRs from the analysed food categories, and summarized as the total PRF intake, based on medium bound levels (values in ng/person/day or *ng/kg b.w./day (hypothetic population mean weight 67.2 kg)).

Sample cat. TCEP TPHP EHDPHP TDCIPP TCIPP 1+2

Cereals 57 77 955 87 282

Pastries 12 36 448 12 39

Meat 21 97 136 39 15

Fish 4 28 112 13 3

Dairy, fluid 41 24 137 56 20

Dairy, solid 11 11 79 11 7

Eggs 2 2 25 4 3

Fats, oils 44 213 228 44 33

Vegetables 81 13 55 72 36

Fruit 21 17 109 67 25

Potatoes 14 23 44 36 21

Sugar etc 28 31 466 92 25

Beverages 70 78 472 269 63

TOTAL (ng/day) 406 650 3266 802 572

TOTAL* (ng/kg bw/day) 6.0 9.7 48.6 11.9 8.5

(17)

DELPROGRAM

Table 5. Based on the Market basket 2015 study, the estimated total per capita intake of PFRs from all analysed food categories presented as lower, medium and upper bound (LB, MB, and UB) levels (values in ng/person/day). The quotient UB/LB indicate the influence of <LOD values.

TCEP TPHP EHDPHP TDCIPP TCIPP 1+2

LB 127.0 415.8 2145.5 631.9 377.1

MB 406.3 649.6 3266.2 801.7 572.3

UB 688.5 883.5 4388.9 974.5 768.4

UB/LB 5.4 2.1 2.0 1.5 2.0

(18)

DELPROGRAM

Figure 1. Estimated intake of EHDPHP and EDCPP divided into separate food categories (MB values).

Differences in relative importance of food categories between the two compounds indicate (at least partial) differences in contamination routes

Cereals Pastries Meat Fish Dairy, fluid Dairy, solid Eggs Fats, oils Vegetables Fruit Potatoes Sugar etc Beverages

EHDPHP

Cereals Pastries Meat Fish Dairy, fluid Dairy, solid Eggs Fats, oils Vegetables Fruit Potatoes Sugar etc Beverages