Assessing the causal effect of tobacco retail density on cardiovascular and pulmonary disease hospitalizations in the United States

Objectives: High tobacco retail density is linked to greater tobacco-related disease, but most evidence is cross-sectional limiting causal inference. Here, we use longitudinal data (2002-2016) from 327 counties in five U.S. states to estimate the causal effect of reducing tobacco retail density on hospitalizations for cardiovascular and pulmonary diseases. Methods: County-level data for Kentucky, Maryland, New Jersey, New York, and North Carolina were included. Annual rates of hospitalizations for acute myocardial infarction, asthma, chronic obstructive pulmonary disease (COPD), ischemic heart disease (IHD), and stroke were analyzed; annual tobacco retail density was dichotomized at ≥1.26 retailers per 1000 population. Marginal structural models with stabilized inverse probability weights were used to adjust for time-varying confounders. Results: High tobacco retail density was associated with increased hospitalization incidence rate ratios (IRR) for asthma (IRR = 1.43 [1.21, 1.69]), COPD (IRR = 1.13 [95%CI: 1.01, 1.25]), IHD (IRR = 1.09 [95%CI: 1.00, 1.18]), and stroke (IRR = 1.08 [95%CI: 1.01, 1.15]). Incidence rate differences indicated that had all counties been below the cut point, thousands of hospitalizations would have been prevented over the 15-year study period. Conclusion: These findings provide further evidence supporting retail-focused tobacco control policies as effective strategies to reduce the burden of tobacco-related disease.