PCDD and PCDF depletion in milk from dairy cows according to the herd metabolic scenario

Chemosphere 73 (2008) S216–S219

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

PCDD and PCDF depletion in milk from dairy cows according to the herd metabolic scenario Gianfranco Brambilla a,*, Igor Fochi a, Michele Falce b, Stefania Paola De Filippis a, Alessandro Ubaldi c, Alessandro di Domenico a a

Italian National Institute for Health, Toxicological Chemistry Unit, Viale Regina Elena, 299, I-00161 Rome, Italy Cirio Agricola S.r.L., Caserta, Italy c Istituto Zooprofilattico delle Regioni Lazio e Toscana, Via Appia Nuova, 1411 I – 00198 Rome, Italy b

a r t i c l e

i n f o

Article history: Accepted 8 November 2007 Available online 6 May 2008 Keywords: Dairy cows Transition period PCDD PCDF Carry over Depletion

a b s t r a c t High level of PCDD + PCDF contamination in bulk milk (9.7 pgWHO-TE g1 fat) from 1604 Holstein Fresian lactacting cows was observed just four weeks after the beginning of their exposure to a feed supplement contaminated at 10.4 ngWHO-TE kg1 dry matter. In-farm produced hay and silage showed levels not exceeding 0.2 ng WHO-TE kg1 dry matter. After the supplement withdrawal, it was possible to monitor the depletion phase for a following 75-day period in milk, until the levels dropped well below 3.0 pgWHO-TE g1 fat, the EU regulatory Maximum Residue Level for PCDD + PCDF. During this phase, the half-life was calculated as 17 ± 3 days, on WHO-TEQ basis. The full availability of farm data on both cow nutrition and milk production allowed the calculation of the carry-over rate (COR) (PCDD + PCDF milk excretion vs. feed), which was 46% at the end of the exposure. This COR value is justified from the main TE contribution of Penta-CDD and -CDF congeners (63%), and the half-life is among the shortest of all those described in the literature both for experimental and naturally-exposed dairy cows. A fugacity-based model predicts a bulk milk contamination of 5 pgWHO-TE g1 fat, compared to the 10 pgWHOTE g1 fat level observed. Such findings are discussed in light of the lactation and metabolic status of the herd for which the transition period, characterised by a negative metabolic energy balance and a consequent adipose tissue mobilization, could play a relevant role. Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction The European Union has recently established provisional maximum and alert levels for PCDDs and PCFDs in cow milk at 3.0 and 2.0 pgWHO-TE g–1 fat, respectively, to prevent unacceptable high human exposures (Council Regulation, 2001). During routine monitoring activity, when non compliant samples for PCDD/F contamination are found in large milk production farms, one of the more pressing questions is how much time is needed to restore the soundness of the milk after the identification and removal of the source of PCDD/Fs. Much effort has been made in the past for modelling the bioaccumulation and carry-over of persistent organic pollutants from contaminated feeds to cows’ milk. Nevertheless, some models are derived from experimental trials and possibly biased due to single-dose exposures (Olling et al., 1991; Slob et al., 1995) or to the reduced number of animals (McLachlan and Richter, 1998; Fries et al., 1999; Huwe and Smith, 2005). On the contrary, the data arising from bulk milk collected from longterm naturally-exposed herds, i.e. as in the case of contaminated * Corresponding author. E-mail address: [email protected] (G. Brambilla). 0045-6535/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2007.11.071

citrus pulp (Traag et al., 1999; Malisch, 2000), could be more informative as it is representative of field situations, for both animal welfare and food safety management purposes. It may be envisaged that lactating cows may not reach a steady state due to metabolic changes occurring during the milking period after labour (Sweetman et al., 1999). For instance, during the first three months of lactation, the metabolic difference between the energetic intake from feeds and the needs represented by body weight maintenance and the quantity of milk produced determines a status of ‘‘transition” cow, where up to 7% of the body weight can be lost mainly due to adipose tissue mobilization (Dechow et al., 2004) (Fig. 1). In this work, we report a case where the carry-over and the depletion of PCDDs and PCDFs were observed in a Holstein Fresian herd, fed on a contaminated feed supplement for 28 days, along with the description of its feed regimen, metabolic state and composition.

2. Materials and methods (case report) In late autumn 2005, a dairy farm, during periodical internal quality milk checking, found unacceptable levels of PCDD/F in the bulk milk (9.7 pgWHO-TE g–1 fat). The herd consisted of 1604

G. Brambilla et al. / Chemosphere 73 (2008) S216–S219

An empirical fugacity-based model to calculate the relationship between PCDD and PCDF intake with feeds and milk contamination was derived from Travis and Hattemer-Frey (1991) by the following equation:

Mcal NEI/day

Energy

+ –

28 26 24 22 20 18 16 14 12 4 2 0 -2 -4 -6 -8 -10 -12

Energy needs Energy intakes Risk of over-conditioning

Energy stored as fat

starvation Lipid mobilisation

S217

C M ¼ Q F C F K OW ð1:0  109 Þ where CM is the PCDD and PCDF contamination, expressed in pgWHO-TE g–1 whole milk, QF is the amount of feed ingested as dry matter, CF is the feed contamination, in pgWHO-TE g–1 dry matter; KOW was averaged at 107.

energy balance = energy needs – energy intake

3. Results and discussion

30

60

90 120 140 168 196 224 252 280 308 336 365

time (days) Fig. 1. Typical metabolic balance of an Holstein Fresian cow during lactation period. Time reported in days from parturition. Energy calculated as net energy intake (NEI).

lactating animals (32% at the first calving) for a total 37 598 l/day production (26.4 kg cow/day as an average) of milk with 3.8% fat. The detailed composition of the herd is reported in Table 1. In-farm produced hay and silage were given at 24 ± 2 kg dry matter per head per day and the feed supplement purchased from the market was administered only to lactacting cows at 1.5 kg per head per day. Both feeds were sampled and analysed for PCDD + PCDF content. The analyses revealed that the contaminated feed supplement was introduced just 28 days before the first non compliant finding. Bulk milk samples representative of the exposed group were then sub sequentially drawn at 14, 21, 28, 35, and 50 days from the removal of the exposure. A milk sample from 40 non-exposed dairy cows belonging to the same farm and receiving the same feed regimen was considered as an indicator of the natural environmental background levels. The local Public Health Veterinary unit followed the case and collected all the analytical reports both coming from inter calibrated private and officially-appointed public laboratories (extended uncertainty within 25%, upper bound approach), according to international standardised criteria (Commission Directive, 2002). Feed intake (in kg day–1 of dry matter) and the daily milk yield were recorded as basis to calculate the carry-over ratio (COR) and the Bio-concentration factors (BCF) (Huwe and Smith, 2005), along with the herd composition data. Contamination half-life was estimated by a logarithmic (ln) plot of concentrations (pgWHO-TE g–1 fat) vs. time (days). Table 1 Structure of the herd exposed to the contaminated feed supplement and relative milk yields Days from labour

Cows (N)

Milk (l) per head per day

Total amount (l) per day

% on the total daily production

16–30 31–60 61–90 91–120 121–150 151–180 181–210 211–240 241–270 271–300 301–330 331–360 361–390 >390

78 163 160 147 111 123 95 73 73 101 122 108 55 195

23 27 29 29 27 25 24 23 23 21 21 19 17 16

1.794 4.401 4.640 4.263 2.997 3.075 2.280 1.679 1.679 2.121 2.562 2.052 935 3.120

4.7 11.7 12.3 11.3 7.9 8.2 6.0 4.6 4.6 5.6 6.8 5.4 2.6 8.3

Feed supplement was found contaminated at 10.4 pgWHOTE g–1 dry matter, while in-farm produced feeds revealed 0.2 pgWHO-TE g–1 dry matter, respectively. The depletion curve for PCDDs and PCDFs in milk is reported in Fig. 2. The linear regression was characterised by high significance (R2 = 0.9712 and F = 136 with p > 99.9%). Plotted half-life on WHO-TEQ basis was estimated as 17 ± 3 day. The milk from non-exposed cows showed a 0.7 pgWHO-TE g–1 fat contamination, in line with the previous internal checks, thus representative of the overall environmental situation of the farm. The congener profile of milk samples from exposed and non-exposed cows and of the feed supplement are summarised in Fig. 3. The CORs calculated as ratio between total milk excretion of PCDDs and PCDFs and overall intake of PCDDs and PCDFs in mineral feed was 46% in exposed, and 13% in nonexposed cows, on WHO-TE basis. The COR and the BCF of most representative congeners are reported in Table 2, on analytical basis, along with their estimated half-lives. Extrapolated milk contamination, calculated according to the fugacity-based model of Travis and Hattemer-Frey (1991) gave 5.0 pgWHO-TE g–1 fat and 1.3 pgWHO-TE g–1 fat for the two groups, respectively. It is well recognised that, in dairy cows, the steady state for PCDD and PCDF in milk and body fat could be reached only after long-term exposures, in conditions of energetic homeostasis, and substantial equivalence of quality and quantity of feed intake. In this case, steady state was not reached due to the relatively short exposure (28 days). Moreover, the herd composition revealed a quite relevant number of transition cows (cows within the 90th day of lactation characterised by a negative energetic balance) (Fig. 1), accounting for the 25% of the herd and contributing 29% to the overall daily milk production (Table 1). According to the fugacity-based model at steady state, the predicted milk contamination should be around 5.0 pgWHO-TE g–1 fat, just half of that observed in field conditions (10 pgWHO-TE g–1 fat) (Fig. 2). Further evidence of possible body burden contribution from mobilised adipose tissues arises from the observed mass balance between the PCDD and PCDF milk excretion and the feed intake at the end of the exposure. The calculated COR value of 46% is in line with those already reviewed (Hoogenboom, 2004), according to the pattern of the milk contamination, where contribution of penta and hexa congeners accounted for 63% and 32% of the total WHO/TEQ value (Table 2) (Fig. 3), respectively. The body burdens of fresh first calving cows may have represented an unexpected contribution of higher chlorinated molecules to the observed milk contamination, as far as hexa and hepta congeners are more prone to bioaccumulate in adipose tissues in non-lactating animals (Tuinstra et al., 1992) as consequence of long-term exposures (two years before the first calving and lactation). On the contrary, the value of 13% found in the non-exposed group seems more in line with the scientific literature documenting the lower oral bioavailability (<11%) of those PCDD and PCDF congeners with Kow >6.5, as in the case of hexa and heptaCCD/CCF (McLachlan and Richter, 1998).

S218

G. Brambilla et al. / Chemosphere 73 (2008) S216–S219

of the production. Within this context, the contribution of the very short alpha-phase (5 days) seems negligible (Tuinstra et al., 1992). The above reported half-life and the half-lives calculated on analytical basis for the most relevant congeners (Table 2) are among the shortest every described for non-single exposures, especially considering the profile of the contamination (highly liphophilic congeners are not so efficiently metabolised, as in the case of TCDD and TCDF) (Olling et al., 1991). This finding can be reasonably addressed by the timing of contamination (lactation in progress), the amount of adipose tissue present in cows at the early stages of lactation, the amount of fat excreted with milk and the herd composition (Tuinstra et al., 1992). At the beginning of the exposure, most of cows were in a negative or neutral energetic balance, thus preventing the bioaccumulation of persistent contaminants in adipose tissue and allowing their mobilisation from such depot (Richter and McLachlan, 2001) (Fig. 1) (Table 1). With the progression of the lactation towards a situation of positive energetic balance, highly chlorinated congeners were basically subtracted to milk excretion, and driven to the reconstituted fat depots. Such explanation is supported by the finding of Huwe and Smith (2005), who reported about hexa and hepta congeners concentration higher in adipose tissue than in the milk in two cows 184 and 262 day into the milking circle, well over the transition period. This case could be representative of situations at farms, where feed supplements could constitute a risk in diary production (Commission Recommendation, 2001). Feed supplement administration is strongly recommended during the transition period, to support milk and fat yields, when the cow cannot ingest larger volumes of feed to cover the energy unbalance. Our findings highlight also that the maximum contamination level for PCDDs and PCDFs in compound feed in accordance with the European Union (0.75 pgWHOTE g–1 dry matter) (Commission Directive, 2002) determined on an inventory basis (European Commission, 2000), could not guarantee the soundness of milk: The sum of the contribution from inhouse hay and silage (24 kg dry matter, 0.2 pgWHO-TE g–1) and commercial supplement (1.5 kg dry matter, 10 pgWHO-TE g–1) results in a compound feed with 0.78 pgWHO-TE g–1 dry matter. This value is very close to the above reported regulatory level nevertheless it was able to cause a milk contamination of 9.7 pgWHO-TE g1

10

8

6

4

2

y = -0.0418x + 2.3232 2

R = 0.9713 0 0

5

10

15

20

25

30

35

40

45

50

55

time (days) Fig. 2. Linear (ln pg/g lb) and exponential (pg/g lb) depletion curves of PCDDs and PCDFs in milk, expressed as pg WHO-TEQ/g lb; time 0 (days) corresponds to the last day (28th) of exposure.

The recorded congeners BCF reflects the quite high variability of such parameter already considered by Huwe and Smith (2005) from several other studies with dairy cows, as possible consequence of the vehicle of the contamination and of the herd management (feed regimen, milk yield and the fat content. In this study, the observed BCF higher in penta - and hexa- chlorinated congeners, is in line with the toxicokinetics models that correlate such parameter positively with the hydrophobicity and inversely with the number of chlorine atoms (Van den Berg et al. 1994; Dimitrov et al., 2002). The short notice about the first report of non compliant results, did not allow a time set of the sampling appropriate to check for an alpha-phase during the first days of the depletion, as reported by Tuinstra et al. (1992) and Huwe and Smith (2005), respectively. However, the good fitting of the linear regression allowed us to extrapolate an overall half-life, of 17 ± 3 day (p = 95%), robust to calculate the withdrawal time needed to recover the soundness

100% 80% 60%

40% 20%

Background profile

Day 0 profile

DF

F

C

D pC H

98, 7,

1,

2,

3,

4,

4, 3, 2,

1,

O

D pC

87, 6,

6, 4, 3,

2,

H

H 87,

98, 7,

F

F D

F xC

D

F xC H

H 87,

3,

6,

2, 1,

3,

4,

2, 1,

3,

D

F xC

D

F

xC

87,

87, 4,

2, 1,

3, 2,

H

Pe

CD

CD

F

F D

Pe 8-

7, 3,

1,

2,

3,

7,

8-

TC

DD

D

2,

D

C O

D 6, 4,

3, 1,

2,

2, 1,

H 7,

8-

98, 7,

7,

3,

6, 3,

pC

D xC

D H

xC H 8-

H 87,

4,

2, 1,

3, 2, 1,

D

D D xC

CD Pe

87,

3, 2,

1,

2,

3,

7,

8-

TC

D

D

D

0%

Feed Supplement profile

Fig. 3. Comparison of the normalised PCDD/F congeners analytical profiles from milk of exposed (dark bars) and unexposed (blank bars) cows and of the contaminated feed supplement (spotted bars), respectively.

G. Brambilla et al. / Chemosphere 73 (2008) S216–S219

S219

Table 2 PCDD and PCDF carry -over rates (COR), bioaccumulation factors (BAF) and averaged half-lives (HL) of most relevant congeners at the 28th day of exposure

the kind co-operation and Mrs. Fabiola Ferri for the graphical assistance.

Congener

Contribution (%) to the total TEQ

COR (a)

COR (b)

BCF (a)

HL (days)

References

2,3,7,8-TCDD 1,2,3,7,8-PeCDD 1,2,3,4,7,8HxCDD 1,2,3,6,7,8HxCDD 1,2,3,7,8,9HxCDD 1,2,3,4,6,7,8HpCDD OCDD

3 24 2

29.72 54.20 40.87

56.22 59.69 41.55

0.44 0.81 0.61

NC 20 19

6

48.83

52.00

0.73

14

2

28.56

29.33

0.43

NC

1

10.30

11.26

0.15

15

0.44

1.36

0.01

NC

2,3,7,8-TCDF 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8HxCDF 1,2,3,6,7,8HxCDF 1,2,3,7,8,9HxCDF 2,3,4,6,7,8HxCDF 1,2,3,4,6,7,8HpCDF 1,2,3,4,7,8,9HpCDF OCDF

0.1 0.2 39 7

0.60 4.21 55.59 42.44

2.84 4.89 63.14 46.72

NC 0.06 0.83 0.63

NC NC 12 11

6

43.91

45.98

0.66

8

8.09

12.53

0.12

NC

9

38.88

40.69

0.58

8

1

10.15

10.69

0.15

9

0.1

10.95

13.47

0.16

NC

<0.1

0.37

1.99

0.01

NC

Commission Directive 2003/57/EC amending Directive 2002/32/EC of the European Parliament and of the Council on undesirable substances in animal feed Official Journal of the European Union L 151, June 19, 2003. Commission Directive 2002/69/EC laying down the sampling methods and the methods of analysis for the official control of dioxins and the determination of dioxin-like PCBs in foodstuffs Official Journal of the European Communities L 209, August 06, 2002. Commission Recommendation 2001/459/EC concerning the reduction of the presence of dioxins, furans and PCBs in feedingstuffs and foodstuffs. Official Journal of the European Communities L 161, June 16, 2001. Council Regulation 2375/2001/EC amending Commission Regulation (EC) No 466/ 2001 setting maximum levels for certain contaminants in foodstuffs. Official Journal of the European Communities L321, December 06, 2001. Dechow, C.D., Rogers, G.W., Klei, L., Lawlor, T.J., VanRaden, P.M., 2004. Body condition scores and dairy form evaluations as indicators of days open in US Holsteins. J. Dairy Sci. 10, 3534–3541. Dimitrov, S.D., Dimitrova, N.C., Walker, J.D., Veith, G.D., Mekenyan, O.G., 2002. Predicting bioconcentration factors of highly hydrophobic chemicals. Effects of molecular size. Pure Appl. Chem. 74, 1823–1830. European Commission, Health and Consumer Protection: Opinion of the scientific committee on animal nutrition on the dioxin contamination of feeding stuffs and their contribution to the contamination of food of animal origin adopted on 06 November 2000. Available at: . Fries, G.F., Paustenbach, D.J., Mather, D.B., Luksenburg, W.J., 1999. A congener specific evaluation of the transfer of polychlorinated dibenzo-p-dioxins and dibenzofurans to milk of cows following ingestion of pentachlorophenoltreated wood. Environ. Sci. Technol. 33, 1165–1170. Hoogenboom, L.A.P., 2004. Behaviour of polyhalogenated and polyciclic aromatic hydrocarbons in food producing animals. In: Preedy, V.R., Watson, R.R. (Eds.), Review in Food and Nutrition Toxicity, vol. 2. CRC Press, London (UK), pp. 269– 299. Huwe, J.K., Smith, D.J., 2005. Laboratory and on-farm studies on the bioaccumulation and elimination of dioxins from a contaminated mineral supplement fed to dairy cows. J. Agric. Food Chem. 53, 2362–2370. Malisch, R., 2000. Increase of the PCDD/PCDF-contamination of milk, butter and meat samples by use of contaminated citrus pulp. Chemosphere 40, 1041–1053. McLachlan, M.S., Richter, W., 1998. Uptake and transfer of PCDD/Fs by cattle fed naturally contaminated feedstuffs and feed contaminated as a result of sewage sludge application. 1. Lactating cows. J. Agric. Food. Chem. 41, 474–480. Olling, M., Derks, H.J.G.M., Berende, P.L.M., Liem, A.D.K., de Jong, A.P.J.M., 1991. Toxicokinetics of eight13 C-labelled polychlorinated dibenzo-p-dioxins and -furans in lactating cows. Chemosphere 23, 1377–1385. Richter, W., McLachlan, M.S., 2001. Uptake and transfer of PCDD/Fs by cattle fed naturally contaminated feedstuffs and feed contaminated as a result of sewage sludge application. 2. Non lactating cows. J. Agric. Food Chem. 49, 5857–5865. Slob, W., Olling, M., Derks, H.J.G.M., de Jong, A.P.J.M., 1995. Congener-specific bioavailability of PCDD/F and coplanar PCBs in cows: laboratory and field measurements. Chemosphere 31, 3827–3838. Sweetman, A.J., Gareth, O.T., Jones, K.C., 1999. Modelling the fate and behaviour of lipophilic organic contaminants in lactating diary cows. Environ. Pollut. 104, 261–270. Traag, W.A., Mengelers, M.J.B., Kann, C.A., Malisch, R., 1999. Studies on the uptake and carry-over of polychlorinated dibenzodioxins and dibenzofurans from contaminated citrus pellets to cow’s milk. Organohalogen Comp. 42, 201–204. Travis, C.C., Hattemer-Frey, H.A., 1991. Human exposure to dioxin. Sci. Total Environ. 104, 12–97. Tuinstra, L.G.M.Th., Roos, A.H., Berende, P.L.M., van Rhijn, J.A., Traag, W.A., Mengelers, M.J.B., 1992. Excretion of polychlorinated dibenzo-p-dioxins and furans in milk of cows fed on dioxins in the dry period. J. Agric. Food Chem. 40, 1172–1176. Van den Berg, M., De Jong, J., Poiger, H., Olson, J.R., 1994. The toxicokinetics and metabolism of polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) and their relevance for toxicity. Crit. Rev. Toxicol. 24, 1–74.

<0.1

0.2

Calculation accounted for a 26.4 kg (3.8% fat) milk daily production and for a 1.5 kg mineral supplement intake, respectively: (a) background contamination observed in milk from not exposed cows subtracted; (b) background not subtracted; NC = not calculated.

fat, a value 3.2-fold higher than the maximum residue level of 3.0 pgWHO-TE g1 fat (Council Regulation, 2001). The weak consistence between feed and milk regulatory levels are confirmed by the fugacity-based model derived in this work that predicted a value 5.0 pgWHO-TE g1 fat. To conclude, as risk management options, the following critical control points can be identified: (a) prevention of excessive PCDD + PCDF body burdens in non-lactating cows, (b) in dairy farms where labour is seasonally scheduled, adequate sampling procedures should be taken to monitor bulk milk contamination levels while reducing possible seasonal biases (Sweetman et al., 1999); (c) when high body burdens are suspected, it is advisable to avoid pooling milk from transition period cows, because of their potential body burden. The above considerations could also be of some utility for reducing possible sampling bias, which is at this time not completely framed in the specific EU legislation (Commission Directive, 2003). Acknowledgements This work granted by Ministry of Health, Project ARACNA 2002-2005. Authors wish to thank ASL CE 1 veterinary staff for