I would expect that this should be rather easy to confirm (or disprove) by breaking down the suicides across the age/profession/social status.