Social fragmentation,severe mentalillness and suicide

Abstract.Background: Geographic patterns of suicide are associated with arealevels of social fragmentation. It is unknown whether thisreflects higher levels of severe mental illness in sociallyfragmented areas.Method: Data on psychiatric inpatient admissions and suicidesamongst people aged 15–64 living in the City of Bristol[1991–1992] were postcode matched to the citys 34 electoralwards. Ecological associations of psychiatric admission (used asa proxy measure of prevalence of severe mental illness) andsuicide rates with levels of social fragmentation wereinvestigated using negative binomial regression models.Results: Psychiatric hospital admission rates were higher in areaswith high levels of socioeconomic deprivation than in areas withhigh levels of social fragmentation. In contrast, associationswith suicide were stronger in relation to social fragmentationthan socioeconomic deprivation. Association of suicide withsocial fragmentation was only moderately attenuated in modelscontrolling for psychiatric admission rate and socio-economicdeprivation, RR 1.23 (95 % C. I. 1.09–1.38) per quartileincrease in social fragmentation, compared to 1.29 (95% C. I.1.16–1.44) before adjustment.Conclusion: The association between social fragmentation and suicideis not explained by socioeconomic deprivation or the prevalenceof severe mental illness within socially fragmented areas asmeasured by psychiatric admission rate.