In a previous study, a side-effect was discovered in which the gender of the researchers handling the rodents had some impact on their tumor rates. They have a more acute sense of smell and male hormones induced a higher level of stress which caused a higher rate of tumors.
So, let's suppose that approximately 1 member of each group will develop a tumor for some reason unrelated to the effect being studied.
If you have 5 groups, one control and 4 experiments, then potentially you'll see a 4x effect increase if you lump the experiment groups together.
This would only hold of course if the effect isn't directly proportional to the population size -- if it's proportional to the number of groups instead. So, I dunno, maybe. But I'd be reluctant to draw any conclusions from numbers for "all experimental groups" vs "one control group".
And while the percentages and ratios involved can look conclusive (varying from 0% to 2.2% -- that's an infinite percentage increase!), the effects in absolute numbers are still quite small.
If anything this warrants replication.
Based on the evidence available right now, I don't think there's a link between cellular technology and cancer. If there was about one more study like this one -- from an overall reputable research outfit, with a good population size -- I'd start to reconsider.
And if another study is done and doesn't find the same effects, then that's still another interesting case of the decline effect. So, either way, it's a win.
4/817 = .004
19/1631 = .011
Malignant tumors in rats exposed from 5-50 RFR appear to be 2.75x more likely to have some kind of malignant brain cancer than the control group. That hardly seems inconsequential.