Day 014 - 20 Jul 94 - Page 15
1
Q. Where you take some people who are suffering from a
2 condition and you take a parallel group who are not?
A. Yes.
3
Q. You observe the effects of diet or supposed effects of
4 diet on those two different groups?
A. Yes.
5
Q. Have I got it right? Experimental studies, am I right,
6 are those largely performed on animals?
A. In general, yes.
7
Q. Are there also what you scientists call invitro studies?
8 A. Yes, there could be.
9 Q. Clinical studies, I suppose that is obvious; that is the
observation of morbidity in humans, largely speaking, is
10 it?
A. Yes.
11
Q. Tell us first, please -- use them as a broad category for
12 present purposes -- what sort of conclusions can safely be
drawn in general from epidemiological results?
13 A. Well, what they will usually provide is associations
between diet or a characteristic of the diet and the
14 incidence of the disease, but they do not, of course,
demonstrate that there is any cause and effect.
15
Q. Let me ask you a further question: As a responsible,
16 objective, dispassionate scientist -- hypothesize such a
person if you will -- what more is required before one can
17 sensibly or responsibly postulate cause and effect?
A. Well, ideally, one would need to try and understand
18 the actual mechanism which would explain in detail what
the link was between the diet and the changes that occur
19 within the body and the final disease state.
20 Q. To that end are experimental studies and clinical studies
of assistance?
21 A. Again they make a contribution.
22 Q. If the epidemiological evidence, the experimental and
clinical evidence, all pointed in the same direction over
23 a period of time, could one safely say it is probably so;
there is a cause and effect?
24 A. One probably could, but again we must not
under-estimate the very considerable difficulties there
25 are before we come to definite conclusions on the links
between diet and health.
26
Q. How would you describe, if I may put it like this, in
27 degrees of certainty the conclusion or proposition which
is often expressed that there is a causal relationship
28 between tobacco smoking and lung cancer?
A. On a scale of one to 10?
29
Q. If you like?
30 A. OK, probably nine.