Objective: The aim of this follow-up study is to examine whether the individual's social and physical environment predicts incidence rates of psychiatric illness, after adjusting for demographic and socio-economic characteristics.
Methods: A random sample of the Swedish population (9,170 women and men, aged 25-74) were interviewed between 1990 and 1991 and followed through December 1998 for incidence rates of psychiatric illness. A social participation index and a physical environment index were constructed, based on the interview data. Cox's regression was used to estimate relative risks expressed as hazard ratios (HR). We also stratified for self-reported long-term psychiatric illness at the time of the interview.
Results: Respondents with low scores in the social participation index exhibited an increased age- and sex-adjusted risk of psychiatric illness with an HR = 2.00 (95 % CI 1.32-3.03) compared with individuals with high scores in the social participation index. After inclusion of education, housing tenure and the physical environment index, the HR of psychiatric illness decreased only slightly to 1.69 (95% CI 1.07-2.66) for respondents with low scores in the social participation index. There was no relationship between the physical environment index and psychiatric illness.
Conclusions: In order to prevent and treat psychiatric illness, it is of importance also to consider the individual's social environment.