Day 038 - 19 Oct 94 - Page 15
1 other?
2 A. Yes, indeed. Populations of laboratory animals are
3 typically genetically quite homogeneous. They will get a
4 group of very, very similar animals in order that they can,
5 hopefully, factor out, as it were, genetic variations with
6 the animals insofar as different effects arise. They want
7 to be able to confidently ascribe those to the compound
8 being tested and not to the variation within the
9 population.
10
11 So, they will use a genetically homogeneous group of
12 animals; whereas, of course, the human population is far
13 from genetically homogeneous. Moreover, all the animals in
14 the laboratory tests receive the same diet, though the only
15 variation is that systematic variation in dose. They are
16 kept in scrupulously hygenic and (insofar as it is
17 possible) pathogene free environments and they are,
18 therefore, free from extraneous illnesses which might
19 otherwise complicate the reaction of ordinary human
20 beings.
21
22 So, you are using small homogeneous groups of laboratory
23 animals in a constant environment to model the possible
24 effects in human populations which are genetically diverse,
25 occupationally diverse, nutritionally diverse and diverse
26 in so many other ways.
27
28 Q. And may have ailments?
29 A. May indeed have ailments.
30
31 Q. Complications?
32 A. Yes; they are not all equally healthy in the way one
33 would expect a group of laboratory animals to be.
34
35 Q. So does that have an implication for the ten and a 100 risk
36 factor multiplications? If you want to go into that in a
37 bit more detail?
38 A. Yes. Insofar as the regulatory authorities set these
39 figures called an acceptable daily intake, or ADI, by
40 reference to the results of animal studies, they will
41 typically seek to identify in the animal studies what is --
42 it used to be called "no effect level"; then they called it
43 "no adverse effect level"; now sometimes they will qualify
44 that even further and say "no observable adverse effect
45 level". But let us suppose you identify a level in the
46 feeding studies at which no conspicuous adverse effect
47 emerges. That level of exposure of the test compound in
48 the animal is estimated in terms of -- well, the typical
49 units are milligrams of chemical per kilogram body weight
50 of the rodent.
51
52 In then identifying what is referred to as an "acceptable
53 daily intake", nominally they are supposed to use a safety
54 factor which is nominally supposed to be 100. I should,
55 however, say that in my scrutiny of the practices,
56 particularly the Joint Expert Committee on Food Additives,
57 more commonly other safety factors are used than a 100;
58 sometimes they are as high as 5,000, sometimes they are as
59 low as 20. But let us stick with the nominally standard
60 safety factor of 100.
