Day 017 - 25 Jul 94 - Page 52
1 Dietary consumptions with vegetables and grains and
regular use of asprin were the only factors having an
2 independent and statistically significant association with
fatal colon cancer. Participants who consumed the least
3 vegetables and grains and no asprin have a higher risk
compared with those who consumed the most vegetables and
4 use asprin 16 or more times per month. For men in the
former category the RR was 2.4, for women it was 2.9.
5
Weaker associations were seen for physical inactivity,
6 obesity, total dietary fat and family history. No
associations were seen with consumption of red meat or
7 total or saturated fat. In either case this finding must
be interpreted cautiously. Conclusions -- these findings
8 support recommendations that increased consumption of
vegetables and grains may reduce the risk of fatal colon
9 cancer. Regular use of low doses of asprin may prove to
be an important supplemental dosage."
10
MR. JUSTICE BELL: Mr. Rampton, could we just take the P and
11 the relative risk and the confidence interval there and
ask Dr. Arnott to spell out just what it means, then
12 I will know I have understood what he said earlier -- or
I hope I will.
13
MR. RAMPTON: I hole I will too. Which ones, my Lord?
14
MR. JUSTICE BELL: If we start under Results, we get: Risk of
15 fatal colon cancer decreased with more frequent
consumption of vegetables and high fibre grains; and then
16 P for Trend. Can you spell out in ordinary English what
the words in the parentheses mean?
17 A. The P value of 0.031 means there is only a 3 per cent
chance that that could have arisen by chance. In women it
18 is a 1 per cent risk that it has arisen by chance. So
these are very highly statistically significant figures.
19
MR. MORRIS: How do you get the figures of 3 per cent and 1 per
20 cent from those figures?
A. Because this is the way that the statistics worked.
21 If you take P as 0.05 that is a 5 per cent risk that it
may have arisen by chance.
22
MR. JUSTICE BELL: Yes, but where do you get that from?
23 A. That is taken from the analysis of the data. As
I say, when you have large numbers of patients who fall
24 into one category or the other and then compare the
relative incidence of things happening, you can then put
25 in from your statistics a measure of whether that might
have arisen by chance or not. That is how they have come
26 up with these particular figures. As far as the
confidence intervals are concerned, this means again that
27 -----
28 MR. JUSTICE BELL: Can we do the relative risk then, although
you have explained it before?
29 A. What they are saying is that the relative risk of men
taking the highest versus the lowest amount of vegetables
30 was .76. In other words, if the norm is one, then these
people are only having a three-quarters chance of getting
