Day 256 - 04 Jun 96 - Page 27
1 really changing their intake of animal based foods.
2
3 Under that particular scenario, the evidence would suggest
4 that, if anything, particularly for breast cancer, it
5 should not decrease. If anything, it would tend to
6 increase, so this particular analysis of these cohort
7 studies really did not take that into consideration.
8
9 There is no evidence in this paper, for example, to show
10 what is the nutrient intakes of the individuals on these
11 low fat diets, so-called low fat diets. If anything, as
12 I just mentioned, these individuals on these low fat diets
13 here, although few in number, it was to be expected on
14 theoretical grounds that they should not see a decrease in
15 breast cancer risk, so I think this study is to be
16 questioned on those grounds.
17
18 As I said, I have spoken to Professor Willett about whether
19 he had any such data on this particular point and he does
20 not, not yet.
21
22 Q. If you look at the chart on page 358, it is table 2?
23 A. Right.
24
25 Q. Can you just read the points underneath that chart and
26 underneath that table and comment as you go through on
27 anything that you feel is of significance. If you read it,
28 I do not know if you can read it?
29 A. Yes, I can. I am familiar with what is there; I really
30 do not need to read it.
31
32 What is being done here is fairly traditional in the
33 field. That is to say, when a hypothesis is focused on a
34 specific component of the food, in this particular case it
35 is dietary fat, and the investigator wishes to know what is
36 that specific effect, they tend to adjust for all the
37 so-called confounding variables which are listed in that
38 footnote.
39
40 They would adjust for fibre intake, for example. They
41 adjust for age at menarche, alcohol intake, and history of
42 benign breast disease, and so forth and so on, and this is
43 in reference to what I was asked before by the judge about
44 reduction of science: Namely, when we make that kind of
45 adjustment we are essentially taking away some of the
46 effect that would otherwise be expected and, at the end of
47 the day, what we are really doing is looking at the
48 specific effect of that alone and that has been my
49 complaint with these kind of studies is that when we tend
50 to do that kind of thing, we end up with a very, very
51 narrow slice of the information to analyse and that is a
52 mistake. It is simply wrong.
53
54 It is right in the sense that if a person, if the
55 investigator wishes to just focus on that one thing but, as
56 I say, that is where the problem lies. Basically the basic
57 hypothesis is formulated the wrong way.
58
59 Q. The points under that table about:
60
