The article I was reading seemed to parallel—in effect augment—your argument, and then there seemed to be this jump. Your argument, then, is that you have to look at it in a five-year block. In truth, the article doesn't look at a five-year block. The last year they looked at was 2007.
When you're tracking suicides, do you track those who have been discharged from the military? If I had been discharged two years ago, and for whatever reason I committed suicide this year, then it's an interesting argument as to whether you put that into the military suicide or you keep it out of the military suicide. How do you account for that, or not, as the case may be?