There are indications that psychiatric patients do not receive adequate treatment of their somatic diseases.

To evaluate the effect of mental disorder on the risk of mortality.

To study whether severity of psychiatric disease increases the risk of mortality.

A register-based cohort study comprising all individuals alive and registered in Sweden in 2004 and 2005 aged 20–64. We followed 5,181,743 individuals with respect to mortality 2006 and 2007, generating 10,261,263 person-years and 24,475 deaths. Psychiatric patients were defined as those treated with a main diagnosis of psychiatric disease within the last five years and were classified according to type of mental disorder: psychoses, depression/anxiety, substance abuse, and other psychiatric diagnoses. We studied all cause mortality, mortality from ischemic heart disease, suicide, violent deaths, smoking related cancer and policy-related and health care-related avoidable mortality. Estimates of risk of mortality were calculated as incidence rate ratio (IRR) with 95% confidence intervals (CI) using Poisson Regression Analysis.

Patients with schizophrenia or other psychosis in general had the highest mortality for most mortality outcomes. For both women and men, those treated for substance abuse had the highest all-cause mortality risk (Women: IRR = 12.56 (95% CI = 10.82–14.57); Men: IRR = 9.38 (95% C = 8.73–10.08). For all psychiatric diseases and for all mortality outcomes those on social benefits had a higher mortality.

Psychiatric patients have a higher overall mortality and cause-specific mortality. The more severe the psychiatric disease the higher is the risk of mortality even after disregarding violent deaths and suicide.