Day 036 - 13 Oct 94 - Page 13
1 5,485 women, including 99 breast cancer cases (those being
2 34 premenopausal women and 65 postmenopausal at NHANES 1
3 base-line), was examined for associations with dietary
4 intake of fat, per cent energy from fat, total energy,
5 saturated fat, polyunsaturated fat, monosaturated fat, and
6 cholesterol on the basis of a 24 hour recall administered
7 at the baseline NHANES 1 examination. No significant
8 differences in dietary fat intake between cases and
9 non-cases were evident when mean intakes for each group
10 were compared. For total fat in grammes and saturated fat
11 in grammes, a significant inverse association was
12 indicated in proportional hazards analyses. Adjustment of
13 fat for total energy intake resulted in a smaller effect
14 that was no longer statistically significant. Adjustment
15 for accepted breast cancer risk factors did not change
16 these findings".
17
18 Pausing there, Dr. Barnard, the accepted breast cancer
19 risk factors must be other than those considered in this
20 study, must they not?
21 A. One would presume so.
22
23 Q. "This prospective study of a sample from the US population
24 does not support the hypothesis that high dietary fat
25 intake increases breast cancer risk. Indeed, some lower
26 risk associated with high-fat intake may be indicated,
27 although this result may be influenced by methodologic
28 problems with the dietary assessment". If we turn
29 over ----?
30 A. Pardon me, let me amend my earlier answer. I am not
31 quite sure what they were referring to in that statement
32 regarding "other cancer risk factors". If that is
33 important, I would be glad to read the study in order to
34 see what they are referring to.
35
36 Q. I do not believe it is. It is a feature of epidemiology
37 and studies of this kind, is it not, that what you try to
38 do is to make adjustments so as to remove as many of the
39 confounding variables as you possibly can?
40 A. Also in recognition that there is more than one way to
41 get cancer.
42
43 Q. Yes, exactly. That would be a confounding variable, would
44 it not?
45 A. It could be, yes, potentially.
46
47 Q. Please turn to page 469. Here is the qualification, as
48 I take it, which was expressed in the Abstract. In the
49 right-hand column: "It should be noted that the
50 distribution of dietary fat intakes examined in this
51 analysis is heavily shifted toward relatively high
52 intakes. Thus the disparity between these findings and
53 cross-cultural studies may be because fat exerts an effect
54 only up to a certain threshold level (eg. 15 or 25 per
55 cent of energy), which could not be addressed in this
56 sample. It may also be that dietary fat intake exerts its
57 major influence early in life, such as during puberty, and
58 an analysis of diets in adult women is not relevant. In
59 addition, the influence of any dietary changes that may
60 have occurred after the baseline dietary assessment could
