Mortality and causes of death in schizophrenia in Stockholm County, Sweden
Introduction
An increased mortality in schizophrenia approximately twice that of the general population has been reported in a number of studies from different countries over a long period of time, and this increase in mortality is still observed (Allebeck and Wistedt, 1986, Alström, 1942, Black and Fisher, 1992, Black and Winokur, 1988, Brown, 1997, Buda et al., 1988, Lindelius and Kay, 1973, Mortensen and Juel, 1993, Ødegård, 1952a, Saugstad and Ødegård, 1979). Important risk modifiers are gender and age at admission (Simpson, 1988). The risk is increased during periods of hospitalization and after the first discharge from hospital (Babigian and Odoroff, 1969, Ødegård, 1952b, Rorsman, 1974). The excessive mortality risk has been primarily attributed to unnatural causes, mainly suicide (Drake et al., 1985, Roy, 1986), but many studies also have reported an increased mortality from natural (somatic) causes (Allebeck and Wistedt, 1986, Herrman et al., 1983, Mortensen and Juel, 1990, Saugstad and Ødegård, 1979). In a meta-analysis of studies on mortality in schizophrenia by Brown (1997), standardized mortality ratios (SMRs) were 1.5 for all causes of death, 1.3 for all natural causes and 4.3 for all unnatural causes of death. The SMR for suicide was 8.4, suicide being the largest single cause of excess mortality.
Although there have been many studies on mortality in schizophrenia, only two have used data from first diagnosed patients (Eastwood et al., 1982, Mortensen and Juel, 1993) which gives the most accurate estimation of mortality rates. Data from patients where follow-up starts later than the first diagnosis will tend to underestimate the mortality rates in suicide in the period close to the first diagnosis (Brown, 1997). The aim of the present study is to assess the mortality among patients with first schizophrenia diagnosis, and specifically to analyze different mortality outcomes such as cancer, cardiovascular disease and suicide; and to what extent gender, age at first diagnosis, duration of follow-up and diagnostic subgroup modified the mortality in the patient group relative to the general population.
Section snippets
Materials and methods
The Swedish psychiatric in-patient register covers all in-patient treatments since 1971. The Stockholm part consists of the psychiatric in-patient treatments from the Stockholm County medical register, which started in 1969 and is complete from 1971. For each hospitalization, the unique national registration number, date of admission and discharge, as well as diagnosis, are registered. No private in-patient facilities exist in Sweden, and the psychiatric in-patient register is therefore
Results
A total of 9162 individuals residing in Stockholm County were discharged at least once with a schizophrenia diagnosis during 1973–95. Of those, 868 had an admission with a schizophrenia diagnosis prior to 1973, and 510 were residing outside Stockholm County at the first schizophrenia admission and were excluded. Thus, 7784 individuals (3929 males and 3855 females) with a first schizophrenia diagnosis were included in the study. In males, first admission was most common at <30 years of age,
Discussion
Our main finding was that, despite higher SMRs for unnatural than for natural causes of death, the total number of excess deaths was larger for natural than for unnatural causes for females and of the same magnitude for males. The largest single cause of death was cardiovascular disease followed by suicide, in both males and females. Cardiovascular disease was the main cause of excess deaths in females, while in males this was the case with suicide. The number of excess deaths, rather than
Conclusion
This study confirmed a marked increase in standardized mortality ratios in schizophrenia both in males and females. Mortality was increased in natural as well as unnatural causes of death, with an increased mortality by natural causes as the main cause of excess deaths. Mortality ratios for suicide were very high, particularly in young age and in the first year after the first schizophrenia admission. These results are important for planning preventive measures and improvement of the
Acknowledgements
This study was supported by grants from Stockholm County Council. We want to thank our reviewers for a valuable comment on the non-differential effect of smoking on SMR for cardiovascular, cerebrovascular and respiratory disease. We are also indebted to professors Lars Terenius and Göran Sedvall, Department of Clinical Neuroscience, Karolinska Institutet, Stockholm for useful opinions and critical revisions of the manuscript.
References (27)
- et al.
Mortality in DSM-IIIR schizophrenia
Schizophr. Res.
(1992) - et al.
Age, mortality and chronic schizophrenia
Schizophr. Res.
(1988) - et al.
Suicide among schizophrenics: a review
Compr. Psychiatry
(1985) - et al.
Mental illness and mortality
Compr. Psychiatry
(1982) The occurrence of cancer in first admitted schizophrenic patients
Schizophr. Res.
(1994)- et al.
Suicide and violent death among patients with schizophrenia
Acta Psychiatr. Scand.
(1986) - et al.
Mortality in schizophrenia: a ten-year follow-up based on the Stockholm County in-patient register
Arch. Gen. Psychiatry
(1986) Schizophrenia: a life-shortening disease
Schizophr. Bull
(1989)Mortality in mental hospitals with especial regard to tuberculosis
Acta Psychiatrica et Neurologica Suppl.
(1942)- et al.
The mortality experience of a population with psychiatric illness
Am. J. Psychiatry
(1969)