Biostatistician/Research Associate Harvard School of Dental Medicine Boston, United States
Background: Given the health threats posed by climate change, it is important to identify individual-level drivers of vulnerability to environmental and climate hazards. However, few studies have examined the influence of commonly-used drugs, such as anticoagulants, on susceptibility to the adverse health effects of air pollution
Objectives: We investigate whether the risk of bleeding-related hospitalization is exacerbated by exposure to PM2.5 and anticoagulants, both independently and synergistically.
Methods: Our study population consists of a 50% random sample of 2008-2016 Medicare Part D-eligible Fee-for-Service beneficiaries who are at high risk for cardiovascular thromboembolic events (CTEs). Individuals were considered to have increased risk for CTE if they had pre-existing cardiovascular diseases, prior venous thromboembolic events, total joint arthroplasty, or cancer. This high-risk Medicare cohort is representative of the population of older adults who often require anticoagulant therapies. The primary exposures of interest were PM2.5 and anticoagulant drugs (Apixaban, Dabigatran, Edoxaban, Rivaroxaban, or Warfarin). The outcomes of interest were first hospitalizations for gastrointestinal bleeding, intracranial bleeding, or epistaxis. To adjust for a large set of individual and neighborhood-level confounders, we constructed inverse probability weights for each of PM2.5 exposure, drug exposure, and censoring due to death. We estimated the main effects of PM2.5 exposure, anticoagulant use, and their interaction, by fitting marginal structural Cox proportional hazards models for each outcome. Finally, we estimated the relative excess risk due to interaction (RERI) to assess the sufficient cause interaction between the two exposures on the additive scale.
Results: In the absence of anticoagulant use, we observed a significant hazard ratio (HR) associated with a 5 microgram per cubic meter increase in PM2.5 for gastrointestinal bleeding (HR = 1.22 [1.20, 1.23]), intracranial bleeding (HR = 1.26 [1.22, 1.29]), and epistaxis (HR = 1.25 [1.17, 1.33]). We also observed a significant additive interaction between PM2.5 exposure and anticoagulant use for gastrointestinal bleeding (RERI = 12.40% [15.88%, 23.22%]), and intracranial bleeding (RERI = 25.89% [8.29%, 43.49%]).
Conclusions: Among older adults in the US at high risk for CTEs, we found novel evidence that PM2.5 exposure and anticoagulant use may act independently and synergistically to increase the risk of severe gastrointestinal bleeding and intracranial bleeding.
