Umeå universitets logga

Background: The association between socioeconomic status (SES) and environmental burden is context-dependent. Those with low SES may be more likely to live near major roads and industries where environmental exposures are high but may also be more likely to live further away from city centers. The aim of this study was to investigate the association between SES and environmental burden in Sweden.

Methods: The Swedish CArdioPulmonary bioImage Study (SCAPIS) cohort recruited participants from six Swedish cities during 2013–2018. Residential environmental exposures (particulate matter <2.5 µm [PM2.5], road traffic noise, and lack of greenspace) and neighborhood-level SES factors (proportions with low-income, low education, unemployment, rental units, foreign born) were assessed from participants’ addresses in 2018. Individual-level SES factors (financial buffer, education, occupation, type of living, foreign born) were obtained from a questionnaire and neighborhood data from Statistics Sweden. Linear regression models were used to analyze the link between environmental exposures and SES factors. Correlations between SES factors were analyzed using Cramér's V and Spearman rank correlations.

Results: The study included 23 320 SCAPIS participants in 1939 neighborhoods. The explanatory power of SES factors differed substantially between factors and cities. A model with all ten SES factors could account for 25–88 %, 36–60 %, and 49–81 % of the spatial variance in PM2.5, noise, and lack of greenspace respectively. Neighborhoods with more low-income earners, rental units and foreign born had more environmental burden. Neighborhoods with more inhabitants with low education and unemployment had less environmental burden. Associations were generally non-linear. Individual-level SES factors were not associated with environmental burdens after adjusting for neighborhood SES factors. Individual and neighborhood SES were weakly correlated.

Conclusion: Neighborhood SES factors accounted for a large proportion of the variance in environmental burdens, whereas individual SES factors did not. The relationship between SES and environmental burden differed greatly between indicators and cities.