Objectives: To evaluate the association of self-reported health with residential area type defined by the population density in individual's local environment using a population-based cohort in Finland.
Methods: Young adults of the Northern Finland Birth Cohort 1966 (4,201 women and 3,835 men), surveyed by a postal questionnaire in 1997, were linked to population density of their resident grid (1 km(2)) depicting different types of residential areas. Self-reported health was regressed on residential area type using ordinal logistic analysis, adjusting for psychosocial well-being, social relationships, health behaviour, education and residence time.
Results: Cumulative odds ratios (COR) for poor health were lowest in high-rise centres, highest in scattered settlement areas and second highest in transitional zones. Adjustments (especially for education and time of residence) reduced the CORs to insignificance except the persistently high COR for women in scattered settlement areas.
Conclusion: Poor self-reported health is associated with individual's residential area type, with the lowest occurrence in high-rise centres and higher elsewhere. The difference is likely explained, at least partly, by a complex of psychosocial factors, possibly different for women and men.