Day 017 - 25 Jul 94 - Page 11
1 Q. Of the three kinds of epidemiological studies, population,
case control -- do you prefer cohort or prospective?
2 A. Cohort.
3 Q. Which to your mind is, as it were, the least fraught with
difficulties in interpretation?
4 A. I would say the cohort studies.
5 Q. Put another way, are they the most reliable?
A. They are probably the most reliable, yes.
6
Q. What can one derive from studies using animals?
7 A. One of the problems with animal studies is that you
are subjecting animals to specific stimulae in artificial
8 circumstances. The animals themselves are also specially
bred to make them more susceptible to develop, for
9 example, since we are talking about cancer, tumours. So
whilst they can actually sometimes act as pointers towards
10 further lines of research, or possible associations in
human beings, because of their very nature they cannot be
11 transposed directly from the laboratory situation to our
lives. So they have serious limitations, but they can
12 sometimes provide pointers.
13 Q. One reason for that might be the physiology of animals is
different from that of humans; is that right?
14 A. That is certainly the case. Often though what they
are trying to do is to measure a certain effect of a
15 certain substance, but, as I say, these are highly
artificial situations, often where the dosage of the
16 substance you are measuring is given in very, very large
doses, those which we would not be exposed to in normal
17 life, and whether there is, therefore, a linear
relationship between what may happen in the laboratory and
18 what may happen in life we just do not know the answer to
that.
19
Q. They are suggestive of hypotheses that need to be looked
20 at; is that right?
A. That is right.
21
Q. Is that also, in effect, what you told us about the
22 results of epidemiology?
A. Yes. For example, if you take population studies,
23 they can only suggest avenues of research; they do not in
themselves provide evidence of cause and effect. There
24 are many reasons why there may be differences between
people in different countries. It is not only that there
25 may be real differences. For example, we have very good
cancer registries in this country, but in the Third World
26 the making of a diagnosis and the registries are very
poor.
27
So you can imagine that there are obvious differences that
28 may arise simply from that. I am not suggesting that is
the only reason why there may be differences, but it is
29 one reason.
30 Q. Can I just pick that up? Again, I am afraid it is an
obvious question. If the cancer registration is poor,
