Day 022 - 12 Sep 94 - Page 08
1 this, there is no consistency in the findings of these
2 studies. Many studies have failed to demonstrate that
3 there is a relationship between the dietary factors which
4 have been implicated in these illnesses and the actual
5 illnesses themselves. There is no consistency.
6
7 Q. You mean show no cause and effect?
8 A. No. What I am saying is, there is not even
9 consistency in the finding of an apparent relationship
10 between that aspect of the diet and the development of the
11 cancer. We have not even got near the question of looking
12 at cause and effect. Some studies actually dispute there
13 is even a relationship between fat intake and, say, colon
14 cancer or breast cancer.
15
16 Q. Is it not true there have been some studies in relation to
17 smoking and cancer which have denied there is a link or
18 any association?
19 A. I think some of the studies which were carried out on
20 cigarette smoking in early days -- well, perhaps
21 criticised the possibility, but, in fact, there has been
22 such a consistency of evidence over the last 25 years,
23 since the original study on British doctors was carried
24 out, that I do not think there is any anybody in the world
25 who has any real doubt that there is a direct relationship
26 between cigarette smoking and lung cancer.
27
28 Q. But is it not true there are, as I was saying, very
29 respectable organisations that do consider that there is a
30 link between diet and cancer?
31 A. Yes, I have already said this, but what is -- the
32 problem with diet and cancer is that there is no
33 consistency. For something to be accepted as fact there
34 has to be consistency in the finding and the finding of
35 the study has to be reproducible, not only by that group
36 of people carrying out the study, but also by other
37 groups. There is no doubt about it in medical research,
38 if you have a group of enthusiasts, they may find a
39 particular finding because of their enthusiasm and when a
40 subsequent group looks at it dispassionately and tries to
41 reproduce the findings, sometimes they cannot be
42 reproduced. For anything to be accepted as medically
43 proven, you have to have not only consistency of the
44 findings but it has to be reproducible from study to study
45 to study.
46
47 Q. You are always likely to get some studies which show
48 something, though, are you not?
49 A. You are indeed. I accept that. But here with diet
50 and cancer we have such an overwhelming, almost 50/50 type
51 of situation that some very respected studies and well
52 carried out studies, the prospective studies, looking at
53 people before they develop the illness, so there is no
54 bias that can creep in. They have studied these groups of
55 people as time has gone on and they have failed to find
56 this association between diet and cancer. On the whole,
57 the studies that have shown a relationship between diet
58 and cancer are the studies which have looked
59 retrospectively at people with cancer, and then tried to
60 find a group of people that match that group and see if
