I'm curious how significant this is. The article sounds like it might be abuse of subgroup analysis as in https://xkcd.com/882/ i.e. check every state, check a dozen demographics in each state, and one of those is going to have higher suicide rates than the rest. Doesn't mean it's some feature of that demographic instead of random chance.
You should be able to calculate the stddev and how many from the mean western Colorado is if you have a federal data set for all the subgroups. I'd be somewhat surprised if they hadn't, but you're right that I don't see any mention of it.
