PURPOSE: Knowledge of the epidemiology of suicide is a necessary prerequisite of suicide prevention. We aimed to conduct a nationwide study investigating suicide risk in relation to level of psychiatric treatment.

METHODS: Nationwide nested case-control study comparing individuals who died from suicide between 1996 and 2009 to age-, sex-, and year-matched controls. Psychiatric treatment in the previous year was graded as "no treatment," "medicated," "outpatient contact," "psychiatric emergency room contact," or "admitted to psychiatric hospital."

RESULTS: There were 2,429 cases and 50,323 controls. Compared with people who had not received any psychiatric treatment in the preceding year, the adjusted rate ratio (95 % confidence interval) for suicide was 5.8 (5.2-6.6) for people receiving only psychiatric medication, 8.2 (6.1-11.0) for people with at most psychiatric outpatient contact, 27.9 (19.5-40.0) for people with at most psychiatric emergency room contacts, and 44.3 (36.1-54.4) for people who had been admitted to a psychiatric hospital. The gradient was steeper for married or cohabiting people, those with higher socioeconomic position, and possibly those without a history of attempted suicide.

CONCLUSIONS: Psychiatric admission in the preceding year was highly associated with risk of dying from suicide. Furthermore, even individuals who have been in contact with psychiatric treatment but who have not been admitted are at highly increased risk of suicide.