Insomnia and urban neighbourhood contexts – are associations modified by individual social characteristics and change of residence? Results from a population-based study using residential histories
1 Faculty of Spatial Planning, Institute of Spatial Planning, TU Dortmund University, Dortmund, Germany
2 IUF- Leibniz Research Institute for Environmental Medicine, Düsseldorf, Germany, and Heinrich-Heine University of Düsseldorf, Medical Faculty, Düsseldorf, Germany
3 Institute of Medical Sociology, Heinrich Heine University of Düsseldorf, Düsseldorf, Germany
4 West German Heart Centre, University of Duisburg-Essen, University Hospital Essen, Essen, Germany
5 Institute of Medical Informatics, Biometry and Epidemiology, University of Duisburg-Essen, University Hospital Essen, Essen, Germany
6 Institute of Clinical Epidemiology, Martin-Luther-University of Halle-Wittenberg, Halle, Germany
7 Faculty of Spatial Planning, Department of Transport Planning, TU Dortmund University, Dortmund, Germany
BMC Public Health 2012, 12:810 doi:10.1186/1471-2458-12-810Published: 20 September 2012
Until now, insomnia has not been much of interest in epidemiological neighbourhood studies, although literature provides evidence enough for insomnia-related mechanisms being potentially dependent on neighbourhood contexts. Besides, studies have shown differences in sleep along individual social characteristics that might render residents more vulnerable to neighbourhood contextual exposures. Given the role of exposure duration and changes in the relationship between neighbourhoods and health, we studied associations of neighbourhood unemployment and months under residential turnover with insomnia by covering ten years of residential history of nearly 3,000 urban residents in the Ruhr Area, Germany.
Individual data were retrieved from the Heinz Nixdorf Recall Study, a population-based study of randomly chosen participants from adjacent cities, which contains self-rated insomnia symptoms and individual social characteristics. Participants’ residential addresses were retrospectively assessed using public registries. We built individually derived exposure measures informing about mean neighbourhood unemployment rates and months under high residential turnover. These measures were major predictors in multivariate logistic regressions modelling the association between social neighbourhood characteristics and insomnia in the whole sample and subgroups defined by low income, low education, social isolation, and change of residence. Traffic-related noise, age, gender, economic activity, and education were considered as covariates.
Nearly 12 per cent of the participants complained about insomnia. Associations of neighbourhood unemployment with insomnia were more consistent than those of residential turnover in the whole sample (adjusted OR 1.42, 95% CI 1.00-2.03 for neighbourhood unemployment and OR 1.33, 95% CI 0.78-2.25 for residential turnover in the highest exposure categories). In low-income and socially isolated participants, neighbourhood unemployment odds of reporting insomnia were particularly elevated (adjusted OR 2.90, 95% CI 1.39-6.02 and OR 3.32, 95% CI 1.11-9.96, respectively). Less educated participants displayed relatively high odds of reporting insomnia throughout all upper neighbourhood unemployment exposure categories. Change of residence weakened associations, whereas undisrupted exposure sharpened them by trend.
Our findings hint at multiple stressors being effective in both the neighbourhood context and individual resident, possibly reflecting precarious life situations undermining residents’ sleep and health chances. Moreover, our results suggest a temporal dependency in the association between neighbourhood and insomnia.