Day 017 - 25 Jul 94 - Page 36
1 Just so that we can see what those methodological problems
are (or were) thought to be, please turn to page 469, the
2 right-hand column underneath the table, where the authors
write: "It should be noted that the distribution of
3 dietary fat intake examined in this analysis is heavily
shifted toward relative high intakes. Thus the disparity
4 between these findings and cross-cultural studies may be
because fat exerts an effect only up to a certain
5 threshold level (eg. 15 or 25% of energy)."
6 Do you know of any evidence which suggests whether or not
there is a threshold in this sense?
7 A. No. There is no knowledge whatsoever that there is a
threshold effect.
8
Q. "... which could not be addressed in this sample. It may
9 also be that dietary fat intake exerts its major influence
early in life, such as during puberty, and an analysis of
10 diets in adult women is not relevant. In addition, the
influence of any dietary changes that may have occurred
11 after the baseline dietary assessment could not be
determined in these analyses."
12
Is that a proper qualification of the findings of this
13 study, so far as you are concerned?
A. Well, I think it is reasonable. I mean, what they are
14 saying is that the method of assessment of dietary intake
is flawed; it is not perfect. They are just raising the
15 query that, you know, there are other effects which may
have taken place which may have influenced dietary
16 changes, for example.
17 I just think that in a paper like this it is very
reasonable people should put forward these
18 qualifications. This is what one would hope to see in
most publications, that not only are they producing a
19 result, but they are qualifying it by saying: "Look,
there may be some problems attached to this".
20
Q. May I draw your attention to one final part of this paper
21 which is the last paragraph on page, it must be 470. It
is the next page; it is the very last paragraph where the
22 authors say this: "This study is an important addition to
the examination of diet and breast cancer because of its
23 prospective nature in which careful dietary assessment was
carried out prior to ascertainment of breast cancer
24 status. The NHANES 1 survey that provided this cohort
sampled from a large cross-section of the US population,
25 and thus these data do not focus primarily on the
upper-socioeconomic-class women at higher risk of breast
26 cancer who are usually the focus of breast cancer
studies".
27
Dr. Arnott, was it the case in those days, 1987, perhaps
28 it still is, that the women upon whom these studies
focused were in the upper socioeconomic-classes?
29 A. There is a tendency for this because, of course,
people in the upper socioeconomic-classes tend to be more
30 interested in disease and in themselves and so on, often
better educated and, therefore, they are more readily
