- Email: [email protected]
Exposure assessment to mycotoxins in gluten-free diet for celiac patients
Food and Chemical Toxicology 69 (2014) 13–17
Contents lists available at ScienceDirect
Food and Chemical Toxicology journal homepage: www.elsevier.com/locate/foodchemtox
Exposure assessment to mycotoxins in gluten-free diet for celiac patients C. Brera a,⇑, F. Debegnach a, B. De Santis a, S. Di Ianni a, E. Gregori a, S. Neuhold b, F. Valitutti c a
GMO and Mycotoxin Unit, Department of Veterinary Public Health and Food Safety, Italian National Institute of Health, Italy AIC – Italian Society for Celiac Disease, Genoa, Italy c Pediatric Gastroenterology and Liver Unit, Department of Pediatrics, Policlinico Umberto I, Sapienza University of Rome, Italy b
a r t i c l e
i n f o
Article history: Received 9 January 2014 Accepted 23 March 2014 Available online 30 March 2014 Keywords: Mycotoxins Fumonisins Zearalenone Gluten-free Exposure assessment Celiac disease
a b s t r a c t Mycotoxins are low molecular weight secondary metabolites produced by certain strains of filamentous fungi such as Aspergillus, Penicillium and Fusarium, which attack crops in the field, and grow on foods also during storage under favorable conditions of temperature and humidity. Foods mainly contributing to the intake of mycotoxins with diet are cereals, maize being the most risky commodity due to the potential cooccurrence of more than one mycotoxin, this can be of particular concern especially for vulnerable group of population such as celiac patients that show increased maize-based products consumption. In this study the exposure of celiac patients to fumonisins (FBs) and zearalenone (ZON) has been assessed. The higher exposures, for all the matrices and for both the selected mycotoxins, were for children age group. The lower and upper bound exposure ranged between 348–582 ng/kg bw/day for FBs and 22– 83 ng/kg bw/day for ZON; these values result well below the TDI for the selected mycotoxins, representing the 17–29% and 9–33% of the TDI set for FBs and ZON, respectively. Even considering the worst scenario the exposure values reported for children were lower, namely 1385 ng/kg bw/day for FBs and 237 ng/kg bw/day for ZON, than the corresponding toxicological thresholds. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Mycotoxins are low molecular weight secondary metabolites produced by certain strains of filamentous fungi of genera Aspergillus, Penicillium and Fusarium, which, under favorable conditions of temperature and humidity, attack crops in the field and grow on food commodities also during storage. Consequently, consumers can be exposed to mycotoxins either directly by ingesting contaminated foods or indirectly by consuming animal origin products derived from animals exposed to mycotoxins by feed. Thus, the metabolism of ingested mycotoxins can result in their accumulation in different organs and tissues, potentially affecting human health since mycotoxins may enter into the whole human food chain through a wide spectrum of foodstuffs such as cereals, meat, milk, wine, beer, dried fruits and spices. Mycotoxins are regularly implicated in toxic syndromes in animals and humans (Smith et al., 1995; Berry, 1998). The toxic properties of mycotoxins associated with animals and human include genotoxicity, carcinogenicity, teratogenicity, mutagenicity, nephrotoxicity and immunotoxicity. In particular, the specific ⇑ Corresponding author. Address: GMO and Mycotoxin Unit, Department of Veterinary Public Health and Food Safety, Italian National Institute of Health, Viale Regina Elena, 299-00161 Rome, Italy. Tel.: +39 06 49902377; fax: +39 06 49902363. E-mail address: [email protected] (C. Brera). http://dx.doi.org/10.1016/j.fct.2014.03.030 0278-6915/Ó 2014 Elsevier Ltd. All rights reserved.
immunotoxic action leads to depression of the lymphocytes activity, suppression of antibody production and damage to the functionality of macrophages and neutrophyles (Milic´evic´ et al., 2010). The most relevant mycotoxins found in foods are aflatoxins (AFs), ochratoxin A (OTA), trichothecenes (type A: T-2 and HT-2 toxin, and type B: deoxynivalenol), zearalenone (ZON) and fumonisins B1 (FB1) and B2 (FB2). Mycotoxins are estimated to affect as much as 25% of the world’s crops each year (Lawlor and Lynch, 2005). Foods that mainly contribute to the intake of mycotoxins with diet are cereals, accounting for 50%, followed by alcoholic beverages, dried fruits, cocoa and coffee (SCOOP Task, 2002, 2006). Among cereals, maize is the most risky commodity and the potential co-occurrence of more than one mycotoxin can be of particular concern especially for vulnerable groups of population that result high consumers of maize-based products due to regional diet habits or to specific pathologies like celiac disease. Many countries have adopted regulations to limit the mycotoxin exposure. According to the annual report of the Rapid Alert System for Food and Feed (RASFF), in 2012 mycotoxins were the main hazard in border rejection notification in the European Union (RASFF, 2012). A full review by Marín et al. (2013) reported several dietary exposure studies to mycotoxins. The estimate of population exposure to mycotoxins is traditionally based on food consumption patterns and mycotoxin
14
C. Brera et al. / Food and Chemical Toxicology 69 (2014) 13–17
contamination levels. However, for estimates of individual exposure such approaches are less useful, particularly in light of heterogeneity in contamination and variations in food processing and cooking (Wild and Gong, 2010). The use of biomarkers in risk assessment studies provides different information related to the biological responses to the intake of a mycotoxin (biomarkers of effect), the quantitative detection of the parent toxin and its metabolites (biomarkers of exposure) or the indication of the variabilities of susceptibility of an organism to the effects of an exposure to a toxic compound such as mycotoxins (biomarkers of susceptibility). Basically, the biomarker measurement is an additional approach with clear advantages in reducing uncertainties in risk assessment. Conversely, the not yet ascertained metabolic pathways for some mycotoxins, the unknown bioavailability of the parent compound, the lack of availability of commercial reference standards of metabolites constitute real disadvantages in using this approach. In addition, epidemiological studies, bioinformatics and advanced statistical methods are needed to substantiate the findings. Anyway, biomarkers will not replace traditional approaches used in risk assessment, but should be considered as an additional approach. Recently the estimation of dietary intake has been also focused on vulnerable groups and particular attention was devoted to children, vegetarian people and sub-groups of population, warning about possible higher exposure due to unfavorable body weight/ intake ratio, dietary habits or pathologies (Cano-Sancho et al., 2012; Dall’Asta et al., 2012; Ostry´ and Ruprich, 1998). In this study, the exposure of celiac patients to the main maize contaminating mycotoxins was evaluated; based on the occurrence data obtained, exposure calculations were conducted for FBs and ZON. This is the first study that assess the celiac patients exposure to ZON, while few works, focused on the exposure assessment to fumonisins of celiac patients, are present in literature (Cano-Sancho et al., 2012; Dall’Asta et al., 2012; Ostry´ and Ruprich, 1998). The present study was performed in the framework of an ongoing three years project in cooperation with, and supported by the Italian Celiac Association.
2. Materials and methods 2.1. Sample Gluten-free (GF) products were purchased from the Italian market. All products were collected from retail shops and pharmacies specialized in dietetic foods. Almost all products were foods specifically formulated for coeliacs of the Italian National Register of gluten-free products. In addition, also simple maize flour samples were collected. A total amount of 376 gluten-free samples were purchased, namely 133 GF pasta of different shapes, 109 GF savory snacks including crackers, nachos and breadstick, 87 GF bread and 47 GF flour to be used in recipes (polenta, pizza, cakes, breadcrumbs). Most of the collected samples were maize-based products. All the information about the sample composition was obtained from the labels. Samples were randomly selected collecting as many as possible leader and minor brands available on the market.
2.2. Determination of mycotoxins All the collected samples were analyzed for the determination of aflatoxin B1 (AFB1), ochratoxin A, fumonisins B1 e B2 (FB1, FB2), deoxynivalenol (DON), zearalenone and toxins T-2 and HT-2. The selected mycotoxins were simultaneously determined according to Brera et al. method (submitted for publication). Briefly, the ground samples were extracted with AcCN:MeOH:H2O (20:20:60, v:v:v), the extract was then centrifuged, filtered and injected in the UPLC–ESI MS/MS system without any further purification step. The method was in-house validated on GF pasta and GF bread matrices. The method performances, including precision and accuracy, were evaluated by analyzing 6 replicates for two spiked levels. The in-house validation results were compliant with the criteria reported in the European Commission Regulation 401/2006 (EC, 2006a).
2.3. Exposure assessment The exposure was calculated by a deterministic approach combining the mycotoxin contamination value with the food consumption divided by the body weight (EFSA, 2011a; WHO, 2009). The exposure assessment calculations were performed according to four age categories: children (3–9.9), teenagers (10–17.9), adults (18– 64.5) and elderly (P65); with the exception of children age group, also the gender was evaluated. For each category lower and upper bound exposures were assessed. Exposure calculations at the 95th percentile (P95) of consumption were also performed with the aim of evaluating a worst-case scenario. 2.3.1. Consumption data In literature, few consumption data expressly assessed for celiac patient exist (Dall’Asta et al., 2012; Gibert et al., 2006), but they do not really fit with the purpose of this work; therefore the Authors preferred to refer to the food consumption survey published by the Italian Institute for Nutrition (INRAN) (Leclercq et al., 2009), which is highly representative of the Italian population in terms of geographical area considered and of the number of households involved in the survey (3323 respondents involved). The basic assumption of this choice is that the celiac patient has the same diet habits of the population involved in the survey, with the substitution of the gluten-based foods with GF products specifically formulated for celiac patients (pasta, bread, etc.). This assumption is also supported by the consumption data reported, for Italian celiac patients, by Gibert et al. (2006), which are in line with the assumed substitution. 2.3.2. Concentration data The concentration data used for exposure calculation were derived from the analyses conducted within this study. As regards the management of left-censored data, a substitution method was applied (EFSA, 2010; GEMS/Food-EURO, 1995) generating lower and upper bound contamination values by substituting with zero or LOQ (Limit Of Quantification) value the results reported to be below the LOQ. As a consequence lower bound and upper bound exposures were assessed.
3. Results and discussion 3.1. Mycotoxin occurrence The results obtained from the analysis of the collected samples, are reported in Table 1. The mean contamination reported in Table 1 is calculated attributing LOQ/2 value to all the samples reported to be lower than LOQ for all considered mycotoxins. As reported in Table 1, none of the analyzed samples was contaminated by AFB1, despite it often occurs in maize and maize-based foods. Only few samples were found positive for OTA, DON and T-2+HT-2 content. As regards the OTA contamination, some of the positive samples exceeded the maximum level set by the European legislation (EC, 2006b). However, the positive samples ranged between the 2% and 6% of the analyzed matrices and the mean value reported was well below the legal limit set for OTA. Also for DON contamination, one snack sample was found to contain about the double of the reported legal limit, but similarly to the OTA situation, the mean value for the snack samples was well below the legal limit. As regards the FBs content, the 29% of the analyzed samples resulted contaminated, the presence of FBs was detected in all the selected food categories. However, the number of positive samples was below the occurrence reported in other studies (Dall’Asta et al., 2009, 2012; Lo Magro et al., 2011) where the percentage of contaminated samples ranged between 54% and 88%. Zearalenone contamination was found in all the investigated food categories with the exception of GF pasta. The positive samples represented the 11% of the overall collected samples. 3.2. Exposure assessment Due to the poor amount of positive samples for AFB1, OTA, DON and T-2 and HT-2 contamination, the exposure assessment was performed only for FBs and ZON. The FBs exposure calculation was made for all food categories (i.e. pasta, bread, flour and snack), while for ZON, pasta was not included since none of the analyzed samples resulted contaminated. The exposure assessment for FBs and ZON is reported in Tables 2 and 3, respectively. To stress the
15
C. Brera et al. / Food and Chemical Toxicology 69 (2014) 13–17
Table 1 Results obtained from the analysis of the collected samples. The reported mean value is obtained attributing LOQ/2 value to all the samples reported to be lower than LOQ.
LOQ (lg/kg) GF pasta (N = 133)
GF bread (N = 87)
GF flour (N = 47)
GF snack (N = 109)
Positive samples Contamination range Mean (lg/kg) Positive samples Contamination range Mean (lg/kg) Positive samples Contamination range Mean (lg/kg) Positive samples Contamination range Mean (lg/kg)
AFB1 0.88
OTA 1.25
FBs 38 (FB1); 12 (FB2)
DON 173
ZON 19
T2 + HT2 46 (T2); 47 (HT2)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
4 (5%) 1.45–4.37 0.75 3 (6%) 2.48–3.42 0.78 2 (2%) 1.79–5.32 0.68
79 (59%) 55–421 113 4 (5%) 73–205 30 16 (34%) 52–544 116 10 (9%) 42–761 48
1 (1%) 296 89 0 (0%)
15 (17%) 19–31 13 8 (17%) 19–61 14 20 (18%) 19–82 17
1 (1%) 115 48 0 (0%)
(lg/kg)
(lg/kg) 0 (0%) (lg/kg) 0 (0%) (lg/kg)
2 (2%) 290–1120 183
0 (0%)
Table 2 Exposure assessment to FBs for total population. The results are reported per age/sex category; single matrix and overall exposures are reported.
*
Matrix
Age/sex (years)
Body weight (kg)*
Mean consumption (g)*
Consumption P95 (g)*
Contamination LB–UB (lg/kg)
Exposure LB–UB (ng/kg bw/day)
Exposure P95 LB–UB (ng/kg bw/day)
Pasta
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
58.2 63.6 56.6 60.3 47.4 61.1 50.7
104.9 128.0 105.3 118.4 100.0 109.6 100.6
101–125
225–279 112–139 116–144 78–96 77–95 79–98 79–98
406–502 226–280 217–268 153–189 162–201 142–175 156–193
Bread
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
73.6 121.4 85.1 127.2 86.1 137.6 102.1
180.0 280.0 206.5 285.0 210.0 298.9 240.0
6–54
17–152 13–115 10–94 10–88 8–75 11–95 9–85
41–372 29–265 25–227 22–196 20–182 23–207 22–199
Snack
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
5.9 7.4 10.6 5.7 9.2 10.3 11.7
30.0 40.0 45.7 32.0 38.3 40.0 43.3
25–71
6–16 3–9 5–15 2–5 4–11 3–9 5–13
29–82 18–50 23–66 10–29 15–44 13–36 17–47
Flour
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
26.5 50.6 43.9 44.1 36.5 27.2 18.9
84.1 149.3 124.3 120.1 97.5 79.3 61.6
98–133
100–135 87–118 88–119 55–75 58–78 34–46 28–39
316–429 256–348 248–337 150–204 154–208 100–135 93–126
Total
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
348–582 215–381 219–372 145–264 147–259 127–248 121–235
792–1385 529–943 513–898 335–618 351–635 278–553 288–565
Data reported by Leclercq et al. (2009).
contribution of the single food category, the calculations are reported for single matrix in each age/sex group considered. Also the total exposure, combining each single matrix contribution together, was assessed. As a worst case scenario, exposure calculation at the 95th percentile of consumption was evaluated. The higher exposure values, for all the matrices and for both the selected mycotoxins, were for children age group. Also focusing on the worst case, represented by the exposure calculated with the consumption value at the 95th percentile, the exposure values reported for children were the highest, namely 1385 ng/kg bw/ day for FBs and 237 ng/kg bw/day for ZON. Matching the results
with the TDI values, which are set at 2000 ng/kg bw for FBs (SCF, 2003) and at 250 ng/kg bw for ZON (EFSA, 2011b), it comes out that the resulting exposures represent the 69% and the 95% of the TDI, respectively. However, it has to be highlighted that in none of the case studied, even the 95th percentile exposures, the TDI value has been exceeded. Moreover, looking at the lower and upper bound exposure range, contrarily to the worst case considered above, the reported exposure values ranged between 348– 582 ng/kg bw/day for FBs and 22–83 ng/kg bw/day for ZON (Tables 2 and 3), that are well below the TDI for both the selected mycotoxins, representing the 17–29% and 9–33% of the TDI set for FBs
16
C. Brera et al. / Food and Chemical Toxicology 69 (2014) 13–17
Table 3 Exposure assessment to ZON for total population. The results are reported per age/sex category; single matrix and overall exposures are reported.
*
Matrix
Age/sex (years)
Body weight (kg)*
Mean consumption (g)*
Consumption P95 (g)*
Contamination LB–UB (lg/kg)
Exposure LB–UB (ng/kg bw/day)
Exposure P95 LB–UB (ng/kg bw/day)
Bread
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
73.6 121.4 85.1 127.2 86.1 137.6 102.1
180.0 280.0 206.5 285.0 210.0 298.9 240.0
5–20
14–56 11–43 9–35 8–32 7–28 9–35 8–31
34–138 25–98 21–84 18–73 17–68 19–77 18–74
Snack
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
5.9 7.4 10.6 5.7 9.2 10.3 11.7
30.0 40.0 45.7 32.0 38.3 40.0 43.3
9–24
2–5 1–3 2–5 1–2 1–4 1–3 2–4
10–28 6–17 8–22 4–10 6–15 5–12 6–16
Flour
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
26.5 50.6 43.9 44.1 36.5 27.2 18.9
84.1 149.3 124.3 120.1 97.5 79.3 61.6
6–22
6–22 5–19 5–20 3–12 4–13 2–8 2–6
19–71 16–58 15–56 9–34 9–34 6–22 6–21
Total
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
22–83 17–65 16–60 12–46 12–45 12–46 12–41
63–237 47–173 44–162 31–117 32–117 30–111 30–111
Data reported by Leclercq et al. (2009).
and ZON, respectively. The different width of the exposure ranges, always reported for FBs and ZON is due to the difference in positive samples percentage, the higher are the positive samples, the lower is the influence of substituting with zero or LOQ value the samples reported to be lower than LOQ. The higher value of exposure assessed for children age group can be explained considering that the consumption value influences the exposure assessment less than the body weight variation. In fact, looking more deeply in the data reported in Tables 2 and 3, it can be noted that the children body weight is 2 or 3 times lower than the body weight of other categories, while the children consumption are only in some case, higher than the ones reported for the other investigated age/sex groups. To better evaluate the children scenario, their intake was compared with the exposure calculated for the total population (all ages); this group covers all the ages/ sex groups, including children. The results for FBs and ZON are shown in Figs. 1 and 2, respectively. The histograms show that children exposures are always about the double of the exposure of the
total population (all ages), confirming the influence of the body weight in the calculation. In conclusion, the exposure assessed for celiac patients to FBs and ZON results below the TDI value. In
ZON 237
Exposure P95 UB
122 63
Exposure P95 LB
34 83
Exposure UB
46 Children (3-9.9) Total populaon (all ages)
22
Exposure LB
12 0
50
100
150
1385 661 792
Exposure P95 LB
360 582
Exposure UB
Children (3-9.9)
271
Total populaon (all ages) 348
Exposure LB
149
0
200
400
600
250
Fig. 2. Comparison between the exposures (ng/kg bw/day) to zearalenone assessed for children and total population (all ages).
FBs Exposure P95 UB
200
ng/kg bw/day
800
1000
1200
1400
1600
1800
2000
ng/kg bw/day Fig. 1. Comparison between the exposures (ng/kg bw/day) to fumonisins assessed for children and total population (all ages).
C. Brera et al. / Food and Chemical Toxicology 69 (2014) 13–17
particular, the FBs exposure assessment scenario outlined in this work, is in agreement with the evaluation reported by Dall’Asta et al. (2012) although they accounted for higher FBs contamination level and occurrence. In fact, their higher FBs contamination is counterbalanced with our higher consumption values since we assumed that celiac patients have the same diet habits of the population of the consumption survey used for the exposure calculation. The possible overestimation, which could result under this assumption, will be clarified when the consumption data of the celiac population, scheduled for the end of the project, will be assessed. Anyway, also the variation in mycotoxins content, caused by high influence of climatic conditions, should be taken into consideration when comparing different set of results for exposure assessment purpose. As regards the ZON exposure assessment this study represents the first evaluation of celiac patient exposure. The absence of previous study suggested to the Authors to perform the exposure calculations even if the number of positive samples was quite low, 43 out of 376, representing only the 11% of the overall collected samples. The results obtained from this study were compared with the exposure assessment reported in the EFSA Scientific Opinion (2011b). Even if the EFSA evaluations were not focused on celiac patient, the exposure assessment obtained in this study is in agreement with lower and upper bound exposure range reported by EFSA (2011b). Conflict of Interest The authors declare that there are no conflicts of interest. Transparency Document The Transparency document associated with this article can be found in the online version.
Acknowledgement Authors wish to thank Mrs. Maria Cristina Barea Toscan for the technical assistance for editing the manuscript. References Berry, C.L., 1998. The pathology of mycotoxin. J. Pathol. 154, 301–311. Cano-Sancho, G., Ramos, A.J., Marín, S., Sanchis, V., 2012. Occurrence of fumonisins in Catalonia (Spain) and exposure assessment of specific population groups. Food Addit. Contam. A 29, 799–808. Dall’Asta, C., Galaverna, G., Mangia, M., Sforza, S., Dossena, A., Marchelli, R., 2009. Free and bound fumonisins in gluten-free food products. Mol. Nutr. Res. 53, 492–499.
17
Dall’Asta, C., Scarlato, A.P., Galaverna, G., Brighenti, F., Pellegrini, N., 2012. Dietary exposure to fumonisins and evaluation of nutrient intake in a group of adult celiac patients on a gluten-free diet. Mol. Nutr. Res. 56 (4), 632–640. EC, 2006a. European Commission Regulation (EC) No 401/2006 of 19 December 2006. Off. J. Eur. Union L70/12-34. EC, 2006b. European Commission Regulation (EC) No 1881/2006 of 23 February 2006. Off. J. Eur. Union L364/5-24. European Food Safety Authority (EFSA), 2010. Management of left-censored data in dietary exposure assessment of chemical substances. EFSA J. 8, 1557. European Food Safety Authority (EFSA), 2011a. Overview of the procedures currently used at EFSA for the assessment of dietary exposure to different chemical substances. EFSA J. 9, 2490. European Food Safety Authority (EFSA), 2011b. Scientific opinion on the risks for public health related to the presence of zearelenone in food. EFSA J. 9 (6), 2197. GEMS/Food-EURO, 1995. Reliable evaluation of low-level contamination of food. Report on a workshop in the frame of GEMS/Food-EURO 26-27 May 2005, Kulmbach, Germany. World Health Organization Regional Office for Europe, Geneva, 8 pp.. Gibert, A., Espadaler, M., Canela, M.A., Sánchez, A., Vaqué, C., Rafecas, M., 2006. Consumption of gluten-free products: should the threshold value for trace amounts of gluten be at 20, 100 or 200 ppm? Eur. J. Gastroenterol. Hepatol. 18, 1187–1195. Lawlor, P.G., Lynch, P.B., 2005. Mycotoxin management. Afr. Farming Food Process. 46, 12–13. Leclercq, C., Arcella, D., Piccinelli, R., Sette, S., Le Donne, C., Turrini, I., 2009. The Italian food consumption survey INRAN-SCAI 2005–06: main results in terms of food consumption. Public Health Nutr. 12 (12), 2504–2532. Lo Magro, S., Campaniello, M., Nardiello, D., Muscarella, M., 2011. Assessment of fumonisins B1 and B2 levels in commercial maize-based food products by liquid chromatography with fluorimetric detection and postcolumn chemical derivatization. J. Food Sci. 76 (1), T1–T4. Marín, S., Ramos, A.J., Cano-Sancho, G., Sanchis, V., 2013. Mycotoxins: occurrence, toxicology, and exposure assessment. Food Chem. Toxicol. 60, 218–237. Milic´evic´, D.R., Škrinjar, M., Baltic´, T., 2010. Real and perceived risks for mycotoxin contamination in foods and feeds: challenges for food safety control. Toxins 2, 572–592. Ostry´, V., Ruprich, J., 1998. Determination of the mycotoxin fumonisins in glutenfree diet (corn-based commodities) in the Czech Republic. Central Eur. J. Publ. Health 6, 57–60. Rapid Alert System for Food and Feed (RASFF), 2012. Annual Reports 2012. European Commission. . Scientific Committee on Food (SCF), 2003. Updated opinion of the Scientific Committee on Food on Fumonisin B1, B2 and B3. 4 April 2003. . SCOOP Task 3.2.7, 2002. Assessment of dietary intake of ochratoxin A by the population of EU Member States. Reports on tasks for scientific cooperation. Directorate General Health and Consumer Protection. . SCOOP Task 3.2.10, 2006. Collection of occurrence data of Fusarium toxins in food and assessment of dietary intake by the population of EU Member States. Reports on tasks for scientific cooperation. Directorate General Health and Consumer Protection. . Smith, J.E., Solomon, G., Lewis, C., Anderson, J.G., 1995. The role of mycotoxins in human and animal nutrition and health. Nat. Toxins 3, 187–192. World Health Organization (WHO), 2009. Dietary exposure assessment of chemicals in food (Chapter 6). Principles and methods for the risk assessment of chemicals in food. Environmental Health Criteria 240. FAO/WHO. International Programme on Chemical Safety (IPCS). Geneva. . Wild, C., Gong, Y.Y., 2010. Mycotoxins and human disease: alargely ignored global health issue. Carcinogenesis 31 (1), 71–82.
Contents lists available at ScienceDirect
Food and Chemical Toxicology journal homepage: www.elsevier.com/locate/foodchemtox
Exposure assessment to mycotoxins in gluten-free diet for celiac patients C. Brera a,⇑, F. Debegnach a, B. De Santis a, S. Di Ianni a, E. Gregori a, S. Neuhold b, F. Valitutti c a
GMO and Mycotoxin Unit, Department of Veterinary Public Health and Food Safety, Italian National Institute of Health, Italy AIC – Italian Society for Celiac Disease, Genoa, Italy c Pediatric Gastroenterology and Liver Unit, Department of Pediatrics, Policlinico Umberto I, Sapienza University of Rome, Italy b
a r t i c l e
i n f o
Article history: Received 9 January 2014 Accepted 23 March 2014 Available online 30 March 2014 Keywords: Mycotoxins Fumonisins Zearalenone Gluten-free Exposure assessment Celiac disease
a b s t r a c t Mycotoxins are low molecular weight secondary metabolites produced by certain strains of filamentous fungi such as Aspergillus, Penicillium and Fusarium, which attack crops in the field, and grow on foods also during storage under favorable conditions of temperature and humidity. Foods mainly contributing to the intake of mycotoxins with diet are cereals, maize being the most risky commodity due to the potential cooccurrence of more than one mycotoxin, this can be of particular concern especially for vulnerable group of population such as celiac patients that show increased maize-based products consumption. In this study the exposure of celiac patients to fumonisins (FBs) and zearalenone (ZON) has been assessed. The higher exposures, for all the matrices and for both the selected mycotoxins, were for children age group. The lower and upper bound exposure ranged between 348–582 ng/kg bw/day for FBs and 22– 83 ng/kg bw/day for ZON; these values result well below the TDI for the selected mycotoxins, representing the 17–29% and 9–33% of the TDI set for FBs and ZON, respectively. Even considering the worst scenario the exposure values reported for children were lower, namely 1385 ng/kg bw/day for FBs and 237 ng/kg bw/day for ZON, than the corresponding toxicological thresholds. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Mycotoxins are low molecular weight secondary metabolites produced by certain strains of filamentous fungi of genera Aspergillus, Penicillium and Fusarium, which, under favorable conditions of temperature and humidity, attack crops in the field and grow on food commodities also during storage. Consequently, consumers can be exposed to mycotoxins either directly by ingesting contaminated foods or indirectly by consuming animal origin products derived from animals exposed to mycotoxins by feed. Thus, the metabolism of ingested mycotoxins can result in their accumulation in different organs and tissues, potentially affecting human health since mycotoxins may enter into the whole human food chain through a wide spectrum of foodstuffs such as cereals, meat, milk, wine, beer, dried fruits and spices. Mycotoxins are regularly implicated in toxic syndromes in animals and humans (Smith et al., 1995; Berry, 1998). The toxic properties of mycotoxins associated with animals and human include genotoxicity, carcinogenicity, teratogenicity, mutagenicity, nephrotoxicity and immunotoxicity. In particular, the specific ⇑ Corresponding author. Address: GMO and Mycotoxin Unit, Department of Veterinary Public Health and Food Safety, Italian National Institute of Health, Viale Regina Elena, 299-00161 Rome, Italy. Tel.: +39 06 49902377; fax: +39 06 49902363. E-mail address: [email protected] (C. Brera). http://dx.doi.org/10.1016/j.fct.2014.03.030 0278-6915/Ó 2014 Elsevier Ltd. All rights reserved.
immunotoxic action leads to depression of the lymphocytes activity, suppression of antibody production and damage to the functionality of macrophages and neutrophyles (Milic´evic´ et al., 2010). The most relevant mycotoxins found in foods are aflatoxins (AFs), ochratoxin A (OTA), trichothecenes (type A: T-2 and HT-2 toxin, and type B: deoxynivalenol), zearalenone (ZON) and fumonisins B1 (FB1) and B2 (FB2). Mycotoxins are estimated to affect as much as 25% of the world’s crops each year (Lawlor and Lynch, 2005). Foods that mainly contribute to the intake of mycotoxins with diet are cereals, accounting for 50%, followed by alcoholic beverages, dried fruits, cocoa and coffee (SCOOP Task, 2002, 2006). Among cereals, maize is the most risky commodity and the potential co-occurrence of more than one mycotoxin can be of particular concern especially for vulnerable groups of population that result high consumers of maize-based products due to regional diet habits or to specific pathologies like celiac disease. Many countries have adopted regulations to limit the mycotoxin exposure. According to the annual report of the Rapid Alert System for Food and Feed (RASFF), in 2012 mycotoxins were the main hazard in border rejection notification in the European Union (RASFF, 2012). A full review by Marín et al. (2013) reported several dietary exposure studies to mycotoxins. The estimate of population exposure to mycotoxins is traditionally based on food consumption patterns and mycotoxin
14
C. Brera et al. / Food and Chemical Toxicology 69 (2014) 13–17
contamination levels. However, for estimates of individual exposure such approaches are less useful, particularly in light of heterogeneity in contamination and variations in food processing and cooking (Wild and Gong, 2010). The use of biomarkers in risk assessment studies provides different information related to the biological responses to the intake of a mycotoxin (biomarkers of effect), the quantitative detection of the parent toxin and its metabolites (biomarkers of exposure) or the indication of the variabilities of susceptibility of an organism to the effects of an exposure to a toxic compound such as mycotoxins (biomarkers of susceptibility). Basically, the biomarker measurement is an additional approach with clear advantages in reducing uncertainties in risk assessment. Conversely, the not yet ascertained metabolic pathways for some mycotoxins, the unknown bioavailability of the parent compound, the lack of availability of commercial reference standards of metabolites constitute real disadvantages in using this approach. In addition, epidemiological studies, bioinformatics and advanced statistical methods are needed to substantiate the findings. Anyway, biomarkers will not replace traditional approaches used in risk assessment, but should be considered as an additional approach. Recently the estimation of dietary intake has been also focused on vulnerable groups and particular attention was devoted to children, vegetarian people and sub-groups of population, warning about possible higher exposure due to unfavorable body weight/ intake ratio, dietary habits or pathologies (Cano-Sancho et al., 2012; Dall’Asta et al., 2012; Ostry´ and Ruprich, 1998). In this study, the exposure of celiac patients to the main maize contaminating mycotoxins was evaluated; based on the occurrence data obtained, exposure calculations were conducted for FBs and ZON. This is the first study that assess the celiac patients exposure to ZON, while few works, focused on the exposure assessment to fumonisins of celiac patients, are present in literature (Cano-Sancho et al., 2012; Dall’Asta et al., 2012; Ostry´ and Ruprich, 1998). The present study was performed in the framework of an ongoing three years project in cooperation with, and supported by the Italian Celiac Association.
2. Materials and methods 2.1. Sample Gluten-free (GF) products were purchased from the Italian market. All products were collected from retail shops and pharmacies specialized in dietetic foods. Almost all products were foods specifically formulated for coeliacs of the Italian National Register of gluten-free products. In addition, also simple maize flour samples were collected. A total amount of 376 gluten-free samples were purchased, namely 133 GF pasta of different shapes, 109 GF savory snacks including crackers, nachos and breadstick, 87 GF bread and 47 GF flour to be used in recipes (polenta, pizza, cakes, breadcrumbs). Most of the collected samples were maize-based products. All the information about the sample composition was obtained from the labels. Samples were randomly selected collecting as many as possible leader and minor brands available on the market.
2.2. Determination of mycotoxins All the collected samples were analyzed for the determination of aflatoxin B1 (AFB1), ochratoxin A, fumonisins B1 e B2 (FB1, FB2), deoxynivalenol (DON), zearalenone and toxins T-2 and HT-2. The selected mycotoxins were simultaneously determined according to Brera et al. method (submitted for publication). Briefly, the ground samples were extracted with AcCN:MeOH:H2O (20:20:60, v:v:v), the extract was then centrifuged, filtered and injected in the UPLC–ESI MS/MS system without any further purification step. The method was in-house validated on GF pasta and GF bread matrices. The method performances, including precision and accuracy, were evaluated by analyzing 6 replicates for two spiked levels. The in-house validation results were compliant with the criteria reported in the European Commission Regulation 401/2006 (EC, 2006a).
2.3. Exposure assessment The exposure was calculated by a deterministic approach combining the mycotoxin contamination value with the food consumption divided by the body weight (EFSA, 2011a; WHO, 2009). The exposure assessment calculations were performed according to four age categories: children (3–9.9), teenagers (10–17.9), adults (18– 64.5) and elderly (P65); with the exception of children age group, also the gender was evaluated. For each category lower and upper bound exposures were assessed. Exposure calculations at the 95th percentile (P95) of consumption were also performed with the aim of evaluating a worst-case scenario. 2.3.1. Consumption data In literature, few consumption data expressly assessed for celiac patient exist (Dall’Asta et al., 2012; Gibert et al., 2006), but they do not really fit with the purpose of this work; therefore the Authors preferred to refer to the food consumption survey published by the Italian Institute for Nutrition (INRAN) (Leclercq et al., 2009), which is highly representative of the Italian population in terms of geographical area considered and of the number of households involved in the survey (3323 respondents involved). The basic assumption of this choice is that the celiac patient has the same diet habits of the population involved in the survey, with the substitution of the gluten-based foods with GF products specifically formulated for celiac patients (pasta, bread, etc.). This assumption is also supported by the consumption data reported, for Italian celiac patients, by Gibert et al. (2006), which are in line with the assumed substitution. 2.3.2. Concentration data The concentration data used for exposure calculation were derived from the analyses conducted within this study. As regards the management of left-censored data, a substitution method was applied (EFSA, 2010; GEMS/Food-EURO, 1995) generating lower and upper bound contamination values by substituting with zero or LOQ (Limit Of Quantification) value the results reported to be below the LOQ. As a consequence lower bound and upper bound exposures were assessed.
3. Results and discussion 3.1. Mycotoxin occurrence The results obtained from the analysis of the collected samples, are reported in Table 1. The mean contamination reported in Table 1 is calculated attributing LOQ/2 value to all the samples reported to be lower than LOQ for all considered mycotoxins. As reported in Table 1, none of the analyzed samples was contaminated by AFB1, despite it often occurs in maize and maize-based foods. Only few samples were found positive for OTA, DON and T-2+HT-2 content. As regards the OTA contamination, some of the positive samples exceeded the maximum level set by the European legislation (EC, 2006b). However, the positive samples ranged between the 2% and 6% of the analyzed matrices and the mean value reported was well below the legal limit set for OTA. Also for DON contamination, one snack sample was found to contain about the double of the reported legal limit, but similarly to the OTA situation, the mean value for the snack samples was well below the legal limit. As regards the FBs content, the 29% of the analyzed samples resulted contaminated, the presence of FBs was detected in all the selected food categories. However, the number of positive samples was below the occurrence reported in other studies (Dall’Asta et al., 2009, 2012; Lo Magro et al., 2011) where the percentage of contaminated samples ranged between 54% and 88%. Zearalenone contamination was found in all the investigated food categories with the exception of GF pasta. The positive samples represented the 11% of the overall collected samples. 3.2. Exposure assessment Due to the poor amount of positive samples for AFB1, OTA, DON and T-2 and HT-2 contamination, the exposure assessment was performed only for FBs and ZON. The FBs exposure calculation was made for all food categories (i.e. pasta, bread, flour and snack), while for ZON, pasta was not included since none of the analyzed samples resulted contaminated. The exposure assessment for FBs and ZON is reported in Tables 2 and 3, respectively. To stress the
15
C. Brera et al. / Food and Chemical Toxicology 69 (2014) 13–17
Table 1 Results obtained from the analysis of the collected samples. The reported mean value is obtained attributing LOQ/2 value to all the samples reported to be lower than LOQ.
LOQ (lg/kg) GF pasta (N = 133)
GF bread (N = 87)
GF flour (N = 47)
GF snack (N = 109)
Positive samples Contamination range Mean (lg/kg) Positive samples Contamination range Mean (lg/kg) Positive samples Contamination range Mean (lg/kg) Positive samples Contamination range Mean (lg/kg)
AFB1 0.88
OTA 1.25
FBs 38 (FB1); 12 (FB2)
DON 173
ZON 19
T2 + HT2 46 (T2); 47 (HT2)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
4 (5%) 1.45–4.37 0.75 3 (6%) 2.48–3.42 0.78 2 (2%) 1.79–5.32 0.68
79 (59%) 55–421 113 4 (5%) 73–205 30 16 (34%) 52–544 116 10 (9%) 42–761 48
1 (1%) 296 89 0 (0%)
15 (17%) 19–31 13 8 (17%) 19–61 14 20 (18%) 19–82 17
1 (1%) 115 48 0 (0%)
(lg/kg)
(lg/kg) 0 (0%) (lg/kg) 0 (0%) (lg/kg)
2 (2%) 290–1120 183
0 (0%)
Table 2 Exposure assessment to FBs for total population. The results are reported per age/sex category; single matrix and overall exposures are reported.
*
Matrix
Age/sex (years)
Body weight (kg)*
Mean consumption (g)*
Consumption P95 (g)*
Contamination LB–UB (lg/kg)
Exposure LB–UB (ng/kg bw/day)
Exposure P95 LB–UB (ng/kg bw/day)
Pasta
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
58.2 63.6 56.6 60.3 47.4 61.1 50.7
104.9 128.0 105.3 118.4 100.0 109.6 100.6
101–125
225–279 112–139 116–144 78–96 77–95 79–98 79–98
406–502 226–280 217–268 153–189 162–201 142–175 156–193
Bread
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
73.6 121.4 85.1 127.2 86.1 137.6 102.1
180.0 280.0 206.5 285.0 210.0 298.9 240.0
6–54
17–152 13–115 10–94 10–88 8–75 11–95 9–85
41–372 29–265 25–227 22–196 20–182 23–207 22–199
Snack
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
5.9 7.4 10.6 5.7 9.2 10.3 11.7
30.0 40.0 45.7 32.0 38.3 40.0 43.3
25–71
6–16 3–9 5–15 2–5 4–11 3–9 5–13
29–82 18–50 23–66 10–29 15–44 13–36 17–47
Flour
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
26.5 50.6 43.9 44.1 36.5 27.2 18.9
84.1 149.3 124.3 120.1 97.5 79.3 61.6
98–133
100–135 87–118 88–119 55–75 58–78 34–46 28–39
316–429 256–348 248–337 150–204 154–208 100–135 93–126
Total
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
348–582 215–381 219–372 145–264 147–259 127–248 121–235
792–1385 529–943 513–898 335–618 351–635 278–553 288–565
Data reported by Leclercq et al. (2009).
contribution of the single food category, the calculations are reported for single matrix in each age/sex group considered. Also the total exposure, combining each single matrix contribution together, was assessed. As a worst case scenario, exposure calculation at the 95th percentile of consumption was evaluated. The higher exposure values, for all the matrices and for both the selected mycotoxins, were for children age group. Also focusing on the worst case, represented by the exposure calculated with the consumption value at the 95th percentile, the exposure values reported for children were the highest, namely 1385 ng/kg bw/ day for FBs and 237 ng/kg bw/day for ZON. Matching the results
with the TDI values, which are set at 2000 ng/kg bw for FBs (SCF, 2003) and at 250 ng/kg bw for ZON (EFSA, 2011b), it comes out that the resulting exposures represent the 69% and the 95% of the TDI, respectively. However, it has to be highlighted that in none of the case studied, even the 95th percentile exposures, the TDI value has been exceeded. Moreover, looking at the lower and upper bound exposure range, contrarily to the worst case considered above, the reported exposure values ranged between 348– 582 ng/kg bw/day for FBs and 22–83 ng/kg bw/day for ZON (Tables 2 and 3), that are well below the TDI for both the selected mycotoxins, representing the 17–29% and 9–33% of the TDI set for FBs
16
C. Brera et al. / Food and Chemical Toxicology 69 (2014) 13–17
Table 3 Exposure assessment to ZON for total population. The results are reported per age/sex category; single matrix and overall exposures are reported.
*
Matrix
Age/sex (years)
Body weight (kg)*
Mean consumption (g)*
Consumption P95 (g)*
Contamination LB–UB (lg/kg)
Exposure LB–UB (ng/kg bw/day)
Exposure P95 LB–UB (ng/kg bw/day)
Bread
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
73.6 121.4 85.1 127.2 86.1 137.6 102.1
180.0 280.0 206.5 285.0 210.0 298.9 240.0
5–20
14–56 11–43 9–35 8–32 7–28 9–35 8–31
34–138 25–98 21–84 18–73 17–68 19–77 18–74
Snack
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
5.9 7.4 10.6 5.7 9.2 10.3 11.7
30.0 40.0 45.7 32.0 38.3 40.0 43.3
9–24
2–5 1–3 2–5 1–2 1–4 1–3 2–4
10–28 6–17 8–22 4–10 6–15 5–12 6–16
Flour
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
26.1 57.1 49.1 78.4 62.2 78.1 65.0
26.5 50.6 43.9 44.1 36.5 27.2 18.9
84.1 149.3 124.3 120.1 97.5 79.3 61.6
6–22
6–22 5–19 5–20 3–12 4–13 2–8 2–6
19–71 16–58 15–56 9–34 9–34 6–22 6–21
Total
3–9.9 M/F 10–17.9 M 10–17.9 F 18–64.9 M 18–64.9 F P65 M P65 F
22–83 17–65 16–60 12–46 12–45 12–46 12–41
63–237 47–173 44–162 31–117 32–117 30–111 30–111
Data reported by Leclercq et al. (2009).
and ZON, respectively. The different width of the exposure ranges, always reported for FBs and ZON is due to the difference in positive samples percentage, the higher are the positive samples, the lower is the influence of substituting with zero or LOQ value the samples reported to be lower than LOQ. The higher value of exposure assessed for children age group can be explained considering that the consumption value influences the exposure assessment less than the body weight variation. In fact, looking more deeply in the data reported in Tables 2 and 3, it can be noted that the children body weight is 2 or 3 times lower than the body weight of other categories, while the children consumption are only in some case, higher than the ones reported for the other investigated age/sex groups. To better evaluate the children scenario, their intake was compared with the exposure calculated for the total population (all ages); this group covers all the ages/ sex groups, including children. The results for FBs and ZON are shown in Figs. 1 and 2, respectively. The histograms show that children exposures are always about the double of the exposure of the
total population (all ages), confirming the influence of the body weight in the calculation. In conclusion, the exposure assessed for celiac patients to FBs and ZON results below the TDI value. In
ZON 237
Exposure P95 UB
122 63
Exposure P95 LB
34 83
Exposure UB
46 Children (3-9.9) Total populaon (all ages)
22
Exposure LB
12 0
50
100
150
1385 661 792
Exposure P95 LB
360 582
Exposure UB
Children (3-9.9)
271
Total populaon (all ages) 348
Exposure LB
149
0
200
400
600
250
Fig. 2. Comparison between the exposures (ng/kg bw/day) to zearalenone assessed for children and total population (all ages).
FBs Exposure P95 UB
200
ng/kg bw/day
800
1000
1200
1400
1600
1800
2000
ng/kg bw/day Fig. 1. Comparison between the exposures (ng/kg bw/day) to fumonisins assessed for children and total population (all ages).
C. Brera et al. / Food and Chemical Toxicology 69 (2014) 13–17
particular, the FBs exposure assessment scenario outlined in this work, is in agreement with the evaluation reported by Dall’Asta et al. (2012) although they accounted for higher FBs contamination level and occurrence. In fact, their higher FBs contamination is counterbalanced with our higher consumption values since we assumed that celiac patients have the same diet habits of the population of the consumption survey used for the exposure calculation. The possible overestimation, which could result under this assumption, will be clarified when the consumption data of the celiac population, scheduled for the end of the project, will be assessed. Anyway, also the variation in mycotoxins content, caused by high influence of climatic conditions, should be taken into consideration when comparing different set of results for exposure assessment purpose. As regards the ZON exposure assessment this study represents the first evaluation of celiac patient exposure. The absence of previous study suggested to the Authors to perform the exposure calculations even if the number of positive samples was quite low, 43 out of 376, representing only the 11% of the overall collected samples. The results obtained from this study were compared with the exposure assessment reported in the EFSA Scientific Opinion (2011b). Even if the EFSA evaluations were not focused on celiac patient, the exposure assessment obtained in this study is in agreement with lower and upper bound exposure range reported by EFSA (2011b). Conflict of Interest The authors declare that there are no conflicts of interest. Transparency Document The Transparency document associated with this article can be found in the online version.
Acknowledgement Authors wish to thank Mrs. Maria Cristina Barea Toscan for the technical assistance for editing the manuscript. References Berry, C.L., 1998. The pathology of mycotoxin. J. Pathol. 154, 301–311. Cano-Sancho, G., Ramos, A.J., Marín, S., Sanchis, V., 2012. Occurrence of fumonisins in Catalonia (Spain) and exposure assessment of specific population groups. Food Addit. Contam. A 29, 799–808. Dall’Asta, C., Galaverna, G., Mangia, M., Sforza, S., Dossena, A., Marchelli, R., 2009. Free and bound fumonisins in gluten-free food products. Mol. Nutr. Res. 53, 492–499.
17
Dall’Asta, C., Scarlato, A.P., Galaverna, G., Brighenti, F., Pellegrini, N., 2012. Dietary exposure to fumonisins and evaluation of nutrient intake in a group of adult celiac patients on a gluten-free diet. Mol. Nutr. Res. 56 (4), 632–640. EC, 2006a. European Commission Regulation (EC) No 401/2006 of 19 December 2006. Off. J. Eur. Union L70/12-34. EC, 2006b. European Commission Regulation (EC) No 1881/2006 of 23 February 2006. Off. J. Eur. Union L364/5-24. European Food Safety Authority (EFSA), 2010. Management of left-censored data in dietary exposure assessment of chemical substances. EFSA J. 8, 1557. European Food Safety Authority (EFSA), 2011a. Overview of the procedures currently used at EFSA for the assessment of dietary exposure to different chemical substances. EFSA J. 9, 2490. European Food Safety Authority (EFSA), 2011b. Scientific opinion on the risks for public health related to the presence of zearelenone in food. EFSA J. 9 (6), 2197. GEMS/Food-EURO, 1995. Reliable evaluation of low-level contamination of food. Report on a workshop in the frame of GEMS/Food-EURO 26-27 May 2005, Kulmbach, Germany. World Health Organization Regional Office for Europe, Geneva, 8 pp.