[Marxism] Study finds strong GM corn-cancer link

Jeff meisner at xs4all.nl
Sun Sep 23 12:15:31 MDT 2012

At 04:35 23-09-12 -0400, Greg McDonald wrote:
>Rule #1: YOU MUST clip all extraneous text when replying to a message.

I stand by everything I wrote before regarding the poorly designed aspects 
of the experiment and possibly invalid conclusions. This 2.5 minute video 
Greg has pointed to was produced by an advocacy group (the Sustainable food 
trust) which therefore strengthens my initial skepticism: I would have been 
expecting to hear a confirmation of the claim by other experts in the field 
of toxicity and animal testing, if there were any. Just reading something 
about their experimental design (in the original article) had already 
shocked me. The excerpt that was quoted states:

"By the beginning of the 24th month, 50-80 percent of female animals had
developed tumours in all treated groups, with up to three tumours per
animal, whereas only 30 percent of controls were affected," it said

But just before that the article innocently mentioned that the experimental 
design included a number of test groups of only 10 animals per group, but 
very shockingly the control group (not exposed to either the chemical or the 
GM corn) was itself only 10 animals! (That is why every percentage mentioned 
is a multiple of 10%!). So the above statement means that 3 of the 10 rats 
in the control group were affected but 5 to 8 of the females in the various 
test groups, with that same (tiny) control group amazingly being used as a 
baseline for all of the tests! Thus, for instance, suppose the actual rate 
of tumors for non-treated animals was 60% (why it would be that high I have 
no idea, but 30% of healthy animals developing cancer already sounds high 
unless there were some other problem in their lab!). With the control group 
of 10 thus expecting a mean of 6 cancers (under this assumption) there is 
already over a 5% a priori probability of no more than 3 of them having 
developed cancer. And if the actual rate of cancer really had been 60%, then 
the results for ALL of the test groups (with each having 5 to 8 out of 10 
getting cancer) would evaporate. In other words, there would have been a 5% 
chance a priori of obtaining such results just by chance (without any 
toxicity from the agents they were fed). Put another way, not only was the 
control group tiny, but because of that the results of all the various test 
groups became rather correlated, a result of their poor experimental design 
which should have placed a much larger proportion of the animals in the 
control group.

By way of disclaimer (should anyone be trying to make a political point) I 
absolutely hate Monsanto and don't necessarily advocate use of the Roundup 
herbicide along with this variety of corn. And I have absolutely no reason 
to expect that that chemical itself is safe; it is surprising/suspicious 
that they would have conducted an experiment testing both the GM corn and 
the chemical, which are very different issues! But as I said before, the 
fact that an artificial chemical (which kills plants by design) and a 
variety of corn both produced the same apparent health effects is already 
suspicious and points to dubious experimental conditions. In fact, an honest 
scientist faced with such results would have either held off on publishing 
them, or included some words of surprise in that regard. Unless..... unless 
the science was taking a back seat to a political agenda.

Again, I am not an expert in either chemical toxicity or food toxicity (two 
very different fields) but I can spot a poorly designed experiment and 
questionable statistical conclusions just from the few details mentioned in 
that news article. The Reuters article below mentions other scientists (who 
are actually in the relevant fields) who are also skeptical of this 
experiment, and none that endorsed the conclusions except for those involved 
in conducting it or writing the paper.

- Jeff


More information about the Marxism mailing list