Social Psychiatry and Psychiatric Epidemiology

, Volume 39, Issue 3, pp 165–170

Social fragmentation, severe mental illness and suicide




Geographic patterns of suicide are associated with area levels of social fragmentation. It is unknown whether this reflects higher levels of severe mental illness in socially fragmented areas.


Data on psychiatric inpatient admissions and suicides amongst people aged 15–64 living in the City of Bristol [1991–1992] were postcode matched to the city’s 34 electoral wards. Ecological associations of psychiatric admission (used as a ‘proxy’ measure of prevalence of severe mental illness) and suicide rates with levels of social fragmentation were investigated using negative binomial regression models.


Psychiatric hospital admission rates were higher in areas with high levels of socioeconomic deprivation than in areas with high levels of social fragmentation. In contrast, associations with suicide were stronger in relation to social fragmentation than socioeconomic deprivation. Association of suicide with social fragmentation was only moderately attenuated in models controlling for psychiatric admission rate and socio-economic deprivation, RR 1.23 (95 % C. I. 1.09–1.38) per quartile increase in social fragmentation, compared to 1.29 (95% C. I. 1.16–1.44) before adjustment.


The association between social fragmentation and suicide is not explained by socioeconomic deprivation or the prevalence of severe mental illness within socially fragmented areas as measured by psychiatric admission rate.

Key words

social fragmentation suicide hospital admission deprivation severe mental illness 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Steinkopff Verlag 2004

Authors and Affiliations

  • Jonathan Evans
    • 1
  • Nicos Middleton
    • 2
  • David Gunnell
    • 2
  1. 1.Division of PsychiatryUniversity of Bristol, Cotham HouseBristol BS6 6JLUK
  2. 2.Dept. of Social MedicineUniversity of BristolBristolUK

Personalised recommendations