Day 017 - 25 Jul 94 - Page 15
1 are selected by writing to them. People who are
enthusiastic about diet are more likely to collaborate,
2 and in a number of studies one finds the actual takeup
rate of the controls is very low indeed. So, inevitably,
3 one is introducing some bias into an analysis of the
data.
4
I mean, this sort of study to me would say that -- well,
5 it is very interesting; we must look at it further. They
have actually compared population-based cancer incidence
6 rates and how relevant that is to these particular groups
of people because of their ethnic variation and is
7 difficult to ascertain.
8 Q. It is fair to say in relation to Mr. Kolonel or
Dr. Kolonel and his colleagues that they appear to
9 recognise some of these difficulties themselves. Can you
turn over the page to page 2400S. I do not know why it
10 has an S on it, "supplement" I suppose. See the passage
marked "Discussion". I will not read it out. I would
11 like you to read the first couple of paragraphs. Then
I will start reading after you have read those.
12 A. Yes.
13 Q. Before I read the particular passage which I want to draw
your attention to, so we make sure we get our terminology
14 right, a positive association between intake of whatever
it may be and incidence of the disease is when the two
15 march in step?
A. That is correct.
16
Q. More of one than more of the other?
17 A. Yes.
18 Q. An inverse association or relationship is when they go in
opposite directions?
19 A. Yes, taking more of one, less of the other.
20 Q. I will not say "cause", but less of the other is observed?
A. Yes.
21
Q. I would like to start with the paragraph which begins:
22 "One of the major findings", the third paragraph in the
right-hand column on page 2400S?
23 A. Yes.
24 Q. "One of the major findings of our research thus far
pertains to dietary fat. Strong associations of total fat
25 and certain of its components, particularly the saturated
fats, with breast, prostrate, and endometrial cancers are
26 notable in our correlational data and have been
reported". One can see, if one looks at the references,
27 those were reported in 1981. "The case-control studies of
breast and prostrate cancer reported here offer tentative
28 support for the finding, especially for saturated fat.
The latter, of course, is highly correlated with animal
29 protein intake. These findings are consistent with
several other reports in the literature, both
30 correlational analyses ... and case-control studies ...
The association with dietary fat appears to be much weaker
