the confounding factor is that the veteran cohort contains a large amount of 50+ males who are the group most likely to complete suicide. If the cohort was made up of group that more reflected the general population I am sure those stats would drop considerably
I’ll bet you’re right.