Ethnicity and excess mortality in severe mental illness: a cohort study

Abstract

Background: Excess mortality in severe mental illness (defined here as schizophrenia, schizoaffective disorders, and bipolar affective disorders) is well described, but little is known about this inequality in ethnic minorities. We aimed to estimate excess mortality for people with severe mental illness for five ethnic groups (white British, black Caribbean, black African, south Asian, and Irish) and to assess the association of ethnicity with mortality risk.

Methods: We conducted a longitudinal cohort study of individuals with a valid diagnosis of severe mental illness between Jan 1, 2007, and Dec 31, 2014, from the case registry of the South London and Maudsley Trust (London, UK). We linked mortality data from the UK Office for National Statistics for the general population in England and Wales to our cohort, and determined all-cause and cause-specific mortality by ethnicity, standardised by age and sex to this population in 2011. We used Cox proportional hazards regression to estimate hazard ratios and a modified Cox regression, taking into account competing risks to derive sub-hazard ratios, for the association of ethnicity with all-cause and cause-specific mortality.

Findings: We identified 18 201 individuals with a valid diagnosis of severe mental illness (median follow-up 6·36 years, IQR 3·26-9·92), of whom 1767 died. Compared with the general population, age-and-sex-standardised mortality ratios (SMRs) in people with severe mental illness were increased for a range of causes, including suicides (7·65, 95% CI 6·43-9·04), non-suicide unnatural causes (4·01, 3·34-4·78), respiratory disease (3·38, 3·04-3·74), cardiovascular disease (2·65, 2·45-2·86), and cancers (1·45, 1·32-1·60). SMRs were broadly similar in different ethnic groups with severe mental illness, although the south Asian group had a reduced SMR for cancer mortality (0·49, 0·21-0·96). Within the cohort with severe mental illness, hazard ratios for all-cause mortality and sub-hazard ratios for natural-cause and unnatural-cause mortality were lower in most ethnic minority groups relative to the white British group.

Interpretation: People with severe mental illness have excess mortality relative to the general population irrespective of ethnicity. Among those with severe mental illness, some ethnic minorities have lower mortality than the white British group, for which the reasons deserve further investigation.

Funding: UK Health Foundation and UK Academy of Medical Sciences.

Figure

Study flow for regression models *Some individuals had missing information in more than one category.