Associations between community characteristics and psychiatric admissions in an urban area

Neighbourhood characteristics are known to be associated with higher rates of hospital admission for psychiatric disorders. An ecological study with aggregated data was carried out. All cases for schizophrenia and depression in the 42 city districts of Augsburg were identified over a 4-year-period (2006-2009) and neighbourhood variables were obtained. Negative binomial regression adjusted the effects for year of admission and accommodation in inpatient centres. There was significant association of high unemployment rate, low proportion of working population and high population density with higher rates of admission for schizophrenia. An increase of 1 % in unemployment rate [incidence rate ration (IRR) 1.0451, 95 % CI 1.0175-1.0734] was associated with 5 % raise of admission rates for schizophrenia and an increase of 1 % in working population (IRR 0.9793, 95 % CI 0.9605-0.9985) with a decrease of admission rates by 2 %. High proportion of single households and high percentage of persons eligible for social security increased admission rate for depression. Thus 1 % increase in the proportion of single households (IRR 1.0095, 95 % CI 1.0030-1.0162) and of the proportion of persons eligible for social security (IRR 1.0148, 95 % CI 1.0002-1.0297) both independently were associated with an increased rate of admission for depression of 1 %. Our analysis demonstrated that measures of social isolation in neighbourhoods and social contacts at work influenced admission for schizophrenia and depression: in neighbourhoods with less social contacts and with a higher proportion of persons not working the admission rates increased. The problem of confounding in ecological studies need to be considered.