Evaluating same-source bias in the association between neighbourhood characteristics and depression in a community sample from Toronto, Canada

Background and purpose It is common in prior studies of the influence of neighbourhood characteristics on mental health to use participant-assessed neighbourhood exposures, which can lead to same-source bias since an individual's mental health status may influence their judgement of their neighbourhood. To avoid this potential bias, we evaluated the use of individually assessed neighbourhood exposures to understand how they compare to collectively assessed measures (by aggregating multiple responses within the same neighbourhood). This would increase the validity of the measure by decoupling the neighbourhood measure from an individual's mental health status. Methods We conducted a stratified-randomised survey of 2411 adults across 87 census tracts in Toronto, Canada (mean of 28 per census tract) to investigate how self-reported (individually assessed) social environmental neighbourhood measures compared to aggregated, collectively assessed, measures for neighbourhood problems/disorder, safety, service quality, and linking, bonding and bridging social capital. The outcome, experience of major depression in the past 12 months, was measured using the Composite International Diagnostic Studies Depression Scale Short Form. Results (1) Individually assessed neighbourhood problems, (2) low (individually assessed) neighbourhood safety, (3) low (individually assessed) neighbourhood service quality, and (4) low (individually assessed) linking social capital were independently associated with depression (all at least p<0.05). However, when the individually assessed exposures were aggregated over residents in the same neighbourhood, none of them were significantly associated with depression. Conclusions Our study provides evidence for same-source bias in studies of social environmental determinants of depression that relies on individually assessed neighbourhood measures. We caution future studies from solely relying on individually assessed neighbourhood exposures especially in the study of social environmental influences on mental health outcomes.