Day 036 - 13 Oct 94 - Page 23
1 part of Dr. Kushi's study, the same page, the right-hand
2 column halfway down there is a paragraph which begins:
3 "The above considerations notwithstanding", do you have
4 that?
5 A. Yes.
6
7 Q. "The above considerations notwithstanding, the association
8 of dietary fat with breast cancer did not approach
9 statistical significance in our study, and in this respect
10 our results are entirely consistent with the other four
11 prospective cohort studies that have investigated this
12 association. If there is a causal relationship between
13 fat intake and risk of breast cancer, it is likely to be
14 of small magnitude across the range of fat intakes
15 observed in studies such as this. Since the magnitude of
16 such an effect appears to be similar to the differences
17 that arise from differing analytic approaches, appropriate
18 consideration of the proper methods for analysis of
19 dietary data and adjustment for energy intake in
20 epidemiologic studies is warranted.
21
22 Several approaches that may provide additional insight
23 into the issue of dietary fat and breast cancer have been
24 proposed or are underway". We remind ourselves that we
25 are in July 1992 at the time this was published.
26
27 "First, efforts can be made in ecologic studies to
28 minimise errors associated with collection of data on
29 dietary habits and potential confounding variables in
30 geographic areas with contrasting breast cancer
31 incidences. With these improvements, it may be possible
32 to determine whether the strong ecologic associations are
33 causal. Second, as with the pooling project of
34 case-control studies, an effort is underway to pool data
35 from several prospective studies, including this one, to
36 relate a quantitative estimate of fat intake to risk of
37 breast cancer (Hunter, D: personal communication)" --
38 presumably through Dr. Kushi.
39
40 "This project will provide increased power to detect an
41 association within the ranges of fat intake under study,
42 but may be unable to address limitations that result from
43 relative homogeneity of fat intakes. Prospective studies
44 in populations with sufficiently wide variation in fat
45 intake and including a sizeable proportion of the
46 population with fat intakes well below 30% of energy may
47 be required to overcome these limitations. Finally,
48 randomized trials such as the Women's Health Trial or the
49 low-fat intervention component of the Women's Health
50 Initiative may also provide evidence as to whether low fat
51 intakes decrease risk of breast cancer."
52
53 Two things about that passage which I have just read,
54 Dr. Barnard. The first is this. Dr. Kushi, at any rate
55 in July 1992, does not appear to know of any prospective
56 studies which studied fat intakes much below 30 per cent
57 of energy, is that right? We see that from the last bit
58 of the page.
59 A. What he is saying is that such studies would be
60 helpful.
