This letter was already briefly discussed at http://forums.phoenixrising.me/inde...-muscarinic-receptor-gene-polymorphism.43027/None of the findings are statistically significant when correcting for multiple comparisons.
Basically, the way they calculated percentages to find statistical differences is completely meaningless. Instead of looking at allele prevalence in each group (patients and controls), they listed the prevalence for each group as a percentage of the prevalence in the combined group. That's why the two values in each row always add up to 100%.
This is a problem because the ME group has 100 patients and the control group only has 90. So of course the ME group is going to have a bigger share of the genotypes listed.
I'm amazed they were even able to publish this as a letter. And I can't see how anything useful can result from this.