Article Text

Death rate in older people with schizophrenia elevated compared with the general population
  1. David Healy, MD
  1. Hergest Unit, Bangor, Wales LL57 2PW, UK;

Statistics from


Question: How do the death rate and cause of death in older people with schizophrenia compare with the general population?

Population: 9461  Adults aged over 65 on 1 January 1999 with schizophrenia or schizoaffective disorder (international classification of diseases (ICD)8, ICD9 and ICD10). Data were taken from the PERFormance, Effectiveness and Cost of Treatment episodes (PERFECT) project which links information from different registers using the unique personal identification number of each Finnish citizen.

Setting: General population, Finland; January 1999–2008.

Prognostic factors: Diagnosis of schizophrenia or schizoaffective disorder. People with these diagnoses were identified using the Finnish Hospital Discharge Register and the register of disability pensions forms the Social Insurance Institution and the Pension Security Centre.

Outcomes: Death rate and cause of death. Outcomes were identified using the National Causes-of-Death Register of Statistics Finland (ICD10). Causes of death were classified as unnatural (suicides, accidents and murder) and natural (all other diagnoses). Standard mortality ratios (SMRs) were calculated for people with schizophrenia or schizoaffective disorder, using the age and gender-matched general Finnish population as the reference population.


Design: Retrospective cohort study.

Follow-up period: Ten years (assessed retrospectively).

Main results

During follow-up, 59% of people with schizophrenia died (4.7% unnatural causes of death). Overall, death rates increased for people with schizophrenia compared with the general population (overall SMR 2.69, 95% CI 2.62 to 2.76; men: SMR 3.00, 95% CI 2.87 to 3.14; women: SMR 2.55, 95% CI 2.47 to 2.63). For unnatural causes of death, death rates greatly increased for people with schizophrenia compared with the general population (SMR 11.04, 95% CI 9.75 to 12.47; men: SMR 11.52, 95% CI 9.35 to 14.04; women: SMR 10.78, 95% CI 9.20 to 12.56). For natural causes of death, death rates also increased for people with schizophrenia compared with the general population (SMR 2.58, 95% CI 2.51 to 2.65; men: SMR 2.87, 95% CI 2.74 to 3.00; women: SMR 2.45, 95% CI 2.37 to 2.53). Death due to neoplasms, infectious, genitourinary, endocrine, digestive, respiratory and circulatory diseases were higher in people with schizophrenia compared with the general population (see webextra table S1). People hospitalised for schizophrenia 5 years before follow-up had increased risk of mortality compared with people with schizophrenia who were not hospitalised during this time (SMR 3.92, 95% CI 3.73 to 4.11 vs SMR 2.37, 95% CI 2.29 to 2.44).


Overall mortality among older people with schizophrenia is nearly three times higher than in the general population (matched for age and gender). Death from unnatural causes was greatly elevated in this population. People with relapsing schizophrenia are at increased risk of mortality compared with people in remission.

Abstracted from



Against a background of concerns about decreased life expectancy in schizophrenia and worries that antipsychotic drugs may aggravate this, this study by Talaslahti and colleagues is the first to specifically tackle the question of mortality in older patients with schizophrenia. Its findings are consistent with other studies in showing elevations of mortality in patients with psychoses.1 The causes of death in the elderly reflect those found in the general population and also those linked to antipsychotic use; this differs from the findings in the younger population where the primary cause of death is suicide.

There is an important ambiguity in the data that the authors note concerning the reliability of the diagnosis of schizophrenia. Another ambiguity lies in a lack of clarity as to whether some patients had an onset of their illness in their 60s or later. This study has features of both a cross-sectional and cohort design, but it is not possible to pick out those elderly patients with a first exposure during their older years. Thus, while it is assumed that antipsychotic drugs are contributing to the excess of mortality, this study does not distinguish between those exposed to antipsychotics for the first time when elderly and those exposed to them for decades, which are probably two different groups in terms of their risk profiles.

The findings will confirm the risks that this group of drugs poses to elderly patients and may support across the board restrictions on their use, when, in fact, there may be particular populations who are vulnerable. The findings paradoxically suggest strongly that some patients are at little risk from ongoing treatment, but they do not help us identify which patients. Given the black box warnings on administering these drugs to older patients in some countries, further research is called for to pinpoint the characteristics of those who seem relatively unaffected.


View Abstract
  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • Sources of funding: Not stated.


  • Competing interests None.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.