Mental disorders in the general population are associated with increased mortality, but few studies have diagnosis made by interviewing psychiatrist.

To investigate long-term risk of death for specific diagnosis in a random population-based sample interview by psychiatrist.

To examine excess mortality associated with psychiatric diagnosis during 41 year follow-up

REBUS is a community sample collected in Stockholm County in 1970. a total of 1896 men and women 18-65 years of age participated in clinical examination, including interview by a psychiatrist (89% participation). Diagnoses were made according to the 8th version of International Classification of Disease. Information on mortality was obtained from tax registers. Excess mortality was assessed with Cox proportional hazard regression models.

There were 883 death during the 41 year long follow-up. Age, sex and comorbidity adjusted hazard ratios were for functional psychosis 2.22 (95% confidence Intervals (96%CI), 1.15-4.30), psycho-organic symptoms 1.94 (95%CI, 1.31-2.87), depressive neurosis 1.71 (95%CI, 1.22-2.39), psychopathy 2.88 (95%CI, 1.02-8.16), alcohol use disorder 1.91 (95%CI, 1.40-2.61) and substance use disorder 3.71 (95%CI, 1.80-7.65). There was no significant risk for those with neurosis other than depressive neurosis, for character neurosis, psychosomatic disorders or mental retardation.

This study confirms the increased risks of death associated with functional psychosis, psycho-organic symptoms, depression, alcohol and substance use disorder and antisocial personality disorder found in psychiatric community samples based on diagnosis made from structured interview.