Does social cohesion modify the association between area income deprivation and mental health? A multilevel analysis

Int J Epidemiol. 2007 Apr;36(2):338-45. doi: 10.1093/ije/dym004. Epub 2007 Feb 28.


Background: Despite the increasing belief that the places where people live influence their health, there is surprisingly little consistent evidence for their associations with mental health. We investigated the joint effect of community and individual-level socio-economic deprivation and social cohesion on individual mental health status.

Methods: Multilevel analysis of population survey data on 10,653 adults aged 18-74 years nested within the 325 census enumeration districts in Caerphilly county borough, Wales, UK. The outcome measure was the Mental Health Inventory (MHI-5) subscale of the SF-36 instrument. A social cohesion subscale was derived from a factor analysis of responses to the Neighbourhood Cohesion scale and was modelled at individual and area level. Area income deprivation was measured by the percentage of low income households.

Results: Poor mental health was significantly associated with area-level income deprivation and low social cohesion after adjusting for individual risk factors. High social cohesion significantly modified the association between income deprivation and mental health: the difference between the predicted mean area mental health scores at the 10th and 90th centiles of the low income distribution was 3.7 in the low cohesion group and 0.9 in the high cohesion group (difference of the difference in means = 2.8, 95% CI: 0.2, 5.4).

Conclusions: Income deprivation and social cohesion measured at community level are potentially important joint determinants of mental health. Further research on the impact of the social environment on mental health should investigate causal pathways in a longitudinal study.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Female
  • Health Status
  • Humans
  • Income*
  • Male
  • Mental Health / statistics & numerical data*
  • Middle Aged
  • Poverty Areas*
  • Risk Factors
  • Small-Area Analysis
  • Social Welfare / economics*
  • Social Welfare / statistics & numerical data
  • Socioeconomic Factors
  • Surveys and Questionnaires
  • Wales / epidemiology