Day 017 - 25 Jul 94 - Page 33
1 experiments by Kritchevsky and, indeed, I think somebody
called Alvarnes (?) as well noticed that caloric intake
2 seemed to be more effective than fat in tumour promotion
------
3 A. Yes, they actually gave rats different diets; one with
high calories and low fat and one with high fat and -----
4
Q. At the top of 595, one sees that. A Low fat -----
5 A. A low fat, high calorie diet and a high fat, low
calorie diet, and it was the caloric reduction which
6 inhibited the growth of mammary tumours in the rats.
7 Q. Despite the high fat?
A. Despite the high fat.
8
Q. Then there is a section about fat and other cancers:
9 Large bowel cancer. I will leave that since we are going
to come on to that in due course. I want to go to the top
10 of page 596, just above the Conclusion. "Despite
plausible theoretical mechanisms the positive geographical
11 correlations and the results of animal experiments and the
human evidence for fat as a cause of large bowel cancer
12 remains weak." Do you agree with that or not?
A. I would agree with that, yes.
13
Q. "The similarity in geographical correlations with fat
14 between colon and breast cancers may be coincidental,
though the lack of reliable international data makes it
15 difficult to evaluate one possible explanation, namely,
the postulated protective role in colon cancer of
16 non-starch polysaccharides (fibre).
17 Conclusion: The evidence for fat as a cause of breast
cancer has been exaggerated, with crude geographical
18 correlations holding sway over the findings of
individual-based studies, counter to the accepted
19 hierarchy for such types of evidence."
20 Dr. Arnott, is there a hierarchy? You started with
cohorts at the top and here we find Dr. Kinlen putting
21 individual case control studies next and populations
studies at the bottom; is that a hierarchy you agree with?
22 A. Yes, certainly cohort studies are the most reliable;
case control are intermediate and the population studies
23 are open to so many problems of interpretation that
I think they have the least value.
24
Q. Dr. Kinlen says: "The present case does not seem to
25 warrant such an exception. The work most often quoted in
support in fact stressed the weaknesses of international
26 correlations, and cautioned that they 'should be taken
only as suggestions for further research and not as
27 evidence of causation or as bases for preventative
action'." Can I take it that comes from Armstrong and
28 Doll?
A. That is the Armstrong and Doll paper.
29
Q. Can we skip the next part and start halfway down the
30 penultimate paragraph. Do you see the words: "We lack
any good evidence". Do you have that? After two words at
