Abstract
Firearm violence is a public health crisis in the United States and negatively impacts individuals and communities. We used data from The Healthy Chicago Survey, a citywide survey of 8754 respondents from 2020 to 2021, and police-reported shootings to estimate the effect of residential exposure to shootings on sleep duration. Exposure was measured as fatal or non-fatal shootings occurring within 400 meters of the respondent's residence, within 7 days before taking the survey. Sleep was categorized into <6, 6 to <7, 7 to <9 and ≥9 h/day. We used log-binomial regression models to estimate risk ratios (RR). We used probabilistic quantitative bias analysis to estimate the magnitude and direction of the impact of differential outcome misclassification of self-reported sleep on results. Of 8584 survey respondents, 589 (6.9%) were exposed to at least one shooting. The estimated associations between shootings and very short and moderately short sleep duration were null, while the RR for long sleep duration was 1.10 (95% CI, 1.00-1.22); results of the quantitative bias analysis for long sleep duration were null (RR = 1.04 [95% simulation interval: 0.92, 1.15]). Our findings suggest that simply living close to a recent shooting may not affect sleep duration among adults living in Chicago, but further investigation is warranted.