Environmental Health Perspectives Polychiorinated Biphenyl Residues in Milk of Environmentally and Experimentally Contaminated Cows Nature and Occurrence of Residues Comparison of Dde and Pcb

Polychlorinated biphenyl (PCB) t residues have been found in the milk of cows. In some instances the residue levels exceeded the FDA guideline 0.2 ppm in milk (equivalent to 5.0 ppm in milk, fat), and the milk was removed from the market. The major source of PCB residues in milk is Aroclor 1254 that has been used in coatings for concrete silos. Aroclor 1254 is unaltered in silos, and most of the contamination is adjacent to the walls (1, 2). We have observed a number of farms with PCB-treated silos and have fed Aroclor 1254 to cows under controlled conditions. This paper summarizes our major findings. The PCB recovered from silage is identical to standard Aroclor 1254 (1). However, many of the components with short gas chromatographic retention times do not occur in milk residues (Fig. 1). The components with short retention times are apparently less chlorinated than those with long retention times. Our observations in cows are similar to observations in other species (3). A complication in analyzing environmental samples is a possibility of interference by DDT and some of its related compounds. We dehydro-t In this paper the term PCB is interchangeable with Aroclor 1254 or, in the case of milk and body fat, residues derived from Aroclor 1254. chlorinated the samples to remove possible interferences of DDD and DDT. DDE did not interfere, because those components of Aroclor 1254 with the same retention time as DDE did not occur in significant amounts in milk (Fig. 1). We obtained milk tank samples from six farms with PCB-contaminated silos. The PCB levels in milk fat are summarized in Table 1. In all cases PCB concentration in milk fat exceeded the FDA guideline, at least part of the time, when silage was fed. When the silage was not fed, levels of PCB in milk fat were always lower than the guidelines. The farms represent a variety of feeding and management conditions. The results suggest that any dairyman with a contaminated silo will probably exceed the guideline at least some of the time during the year. While a few higher levels have been reported for herds removed from the market, most were within the range of our observations. The number of farms with PCB-contaminated silos is not known. However, PCB contamination of milk is probably more prevalent than the regulatory record would suggest. We have observed a herd that was simultaneously …

chlorinated the samples to remove possible interferences of DDD and DDT. DDE did not interfere, because those components of Aroclor 1254 with the same retention time as DDE did not occur in significant amounts in milk (Fig. 1).
We obtained milk tank samples from six farms with PCB-contaminated silos. The PCB levels in milk fat are summarized in Table 1. In all cases PCB concentration in milk fat exceeded the FDA guideline, at least part of the time, when silage was fed. When the silage was not fed, levels of PCB in milk fat were always lower than the guidelines.
The farms represent a variety of feeding and management conditions. The results suggest that any dairyman with a contaminated silo will probably exceed the guideline at least some of the time during the year. While a few higher levels have been reported for herds removed from the market, most were within the range of our observations. The number of farms with PCB-contaminated silos is not known. However, PCB contamination of milk is probably more prevalent than the regulatory record would suggest.

Comparison of DDE and PCB
We have observed a herd that was simultaneously contaminated with high levels of DDE and Aroclor 1254 from a silo. By a fortuitous set of circumstances, the cows had been removed from the sources of both contaminations at about the same time. A detailed description of this work has been prepared (4). Silage was not being fed when italicized samples were obtained. b Not Sampled.    Relationship of the rate constants for DDE and PCBs within individual cows is presented in Fig. 2. The correlation (r = 0.82) between the rate constants of the two compounds was significant (P<0.01). The usual linear regression was calculated, and the intercept did not differ significantly from zero (6). Therefore, the regression in Fig. 2 was recalculated, forcing the intercept through zero. The regression coefficient, 0.974, was essentially unity.
In some cases there was considerable variation between the rate constants for DDE and for  PCBs within a given cow. The rate constants for the two compounds within a given cow were tested statistically (6), but the differences were not significant. The usefulness of phenobarbital alone, or in combination with activated carbon, in accelerating the reduction in milk concentration of DDE and PCBs was tested using three groups of cows. Neither phenobarbital nor phenobarbital in combination with activated carbon had an effect on the relative reduction in concentration of either compound (Table 3). We have previously reported a small effect of phenobarbital on DDE, but we did not consider it of practical significance (7). The absence of an effect here is consistent with that interpretation. The failure of activated carbon, even in combination with phenobarbital, to affect the reduction of DDE and PCB concentrations is consistent with our previous interpretation of laboratory experiments (8).

Aroclor 1254 Feeding Study
We have carried out a controlled feeding study using nine Holstein cows. The cows were fed 200 mg per day Aroclor 1254 for 60 days. Milk samples were obtained periodically during the 60-day feeding period and the subsequent 60-day period.
Body fat samples were obtained by biopsy at 30-day intervals.
The cows were selected to have a range in both the stage of lactation and level of milk production.
However, preliminary examination of the results indicates that there was little difference in the residue levels due to these factors. Therefore, only average data for all cows are presented.
The average level of PCBs in the milk fat during the feeding period is presented in Fig. 3. The standard deviation of any given point did not exceed :1 10% of the mean. The concentration of PCBs increased rapidly and approached a "steady state" at about 40 to 60 days. As expected from our field studies, the shape of the curve is similar to that for other chlorinated hydrocarbon pesticides, particularly DDE (5).
The decline in PCB concentration in the milk for the 60 days after feeding stopped is presented in Fig. 4. This curve is quite typical of previous findings with DDE and was resolved into a twocomponent first-order system (5). The equation for the curve normalized to an initial concentration of 1.0 ,ug/g is: C = 0.52e0°*25I+0.48e-o0 oos Where C is concentration, e is the base of the natural logarithms, and t is days.
The constants are similar to our previous observations for DDE (5). While the rate for the second component in this study was less than the values found in the field, it was within the range of values that one might expect.
The level of PCB in body fat at various times is presented in Table 4. For reference purposes, milk fat levels at these times are also presented.

Environmental Health Perspectives
While the PCBs were being fed, the level in milk fat was higher than that in body fat. When feeding stopped, milk fat levels dropped below and appear to reflect body fat levels.
We have conducted a number of studies with chlorinated hydrocarbon pesticides under conditions similar to those of this PCBs study. A summary of levels in milk fat at 20 and 60 days and body fat at 60 days with normalized intakes is presented in Table 5. The levels of PCBs are similar to levels of DDE and dieldrin in corresponding samples. In contrast DDD, p,p'-DDT, and o,p'-DDT are transferred to the body fat and milk fat at much lower rates than DDE and PCBs.

Conclusions
Farms having silos that have been treated with PCB-containing paints will frequently have residues of PCBs in milk which exceed the FDA guidelines. The behavior of this residue in the cows is similar to DDE and other chlorinated hydrocarbon pesticides resistant to metabolic degradation. The only practical countermeasures appear to be to decontaminate the silos sufficiently to remove most of the PCBs or to discontinue the use of these silos completely. Limited field experience and the similarity of PCBs to DDE suggest that there are no practical procedures to minimize the transfer of PCBs from the diet to the milk or to speed up the elimination of PCBs from cows.