Combined Effect of Air Pollution and Genetic Risk on Incident Cardiovascular Diseases

Abstract

Background: Whether genetic susceptibility to cardiovascular diseases (CVDs) enhances the vulnerability to adverse cardiovascular outcomes by air pollution is unknown. We assessed the combined effect of air pollution and genetic predispositions on CVD risk.

Methods and results: From the UK Biobank cohort, we selected genetically unrelated White British participants without CVD. Levels of ambient particulate matter with a diameter of <2.5 μm (PM_2.5) and <10 μm were estimated using land use regression models. An individual's genetic predisposition to CVDs was determined by polygenic risk scores for coronary artery disease, myocardial infarction, stroke, ischemic stroke, heart failure, and atrial fibrillation. We stratified mortality and CVD risk by PM_2.5 exposure across high and low genetic risk groups. A total of 249 082 participants (aged 56.9±8.0 years, 46.8% men) were followed for a median of 10.8 years. The combined effect of PM_2.5 exposure and the genetic predisposition of CVD demonstrated the highest risk of cardiovascular death in the high genetic risk group with the greatest PM_2.5 exposure (adjusted hazard ratios ranging from 1.73 to 2.12 across the polygenic risk score of each CVD). The combination of higher exposure to ambient PM_2.5 and high genetic risk was associated with higher incidence of all CVDs, although no significant interactions were observed between genetic risk and PM_2.5 exposure on cardiovascular death or CVD events.

Conclusions: A combination of greater PM_2.5 exposure and higher genetic predisposition to particular CVDs was modestly associated with elevated risks of cardiovascular death and CVDs. Not only alleviating PM_2.5 exposure in the general population but also implementing individualized preventive approach for those at high genetic risk might be beneficial.

Keywords: air pollution; cardiovascular disease; gene–environment interaction; particulate matter.

Figure 1. Study flowchart.

Flowchart of the study population is depicted. A total of 249 082 participants with adequate genetic data and without missing data on air pollution and covariates were analyzed. The analysis was performed in 5 separate data sets for 6 CVD outcomes, excluding those with any previous history of each CVD for each data set. AF indicates atrial fibrillation; CAD, coronary artery disease; CVD, cardiovascular disease; HF, heart failure; MI, myocardial infarction; and QC, quality control.

Figure 2. Relationship between genetic risk for each cardiovascular disease and the incidence of each disease.

The restricted cubic spline curves present the adjusted risks for (A) coronary artery disease, (B) myocardial infarction, (C) stroke, (D) ischemic stroke, (E) heart failure, and (F) atrial fibrillation by PRScs for each disease of interest. *Adjusted HRs and 95% CIs were calculated in multivariable Cox hazard models using age, sex, enrollment center, body mass index, current smoking, time spent outdoors at summer, time spent outdoors at winter, noise pollution, diabetes, hypertension on medication, dyslipidemia on medication, income score, housing score, physical activity, traffic on nearest road, traffic on nearest major road, distance from nearest major road, genotype array, and first to 10th genetic principal component. HR indicates hazard ratio; and PRScs, polygenic risk score under continuous shrinkage.

Figure 3. Combined effect of genetic risk for cardiovascular diseases and PM_2.5 exposure on the risk of cardiovascular death.

The forest plots show the adjusted HRs with 95% CIs for the risk of cardiovascular death according to the increase of annual average exposure to PM_2.5, stratified by the genetic risk group defined by the median value of PRScs for (A) CAD, (B) MI, (C) stroke, (D) ischemic stroke, (E) HF, and (F) AF. *Adjusted HRs and CIs were calculated in multivariable Cox hazard models using age, sex, enrollment center, body mass index, current smoking, time spent outdoors at summer, time spent outdoors at winter, noise pollution, diabetes, hypertension on medication, dyslipidemia on medication, income score, housing score, physical activity, traffic on nearest road, traffic on nearest major road, distance from nearest major road, genotype array, and first to 10th genetic principal component. AF indicates atrial fibrillation; CAD, coronary artery disease; HF, heart failure; HR, hazard ratio; i‐Stroke, ischemic stroke; MI, myocardial infarction; PM_2.5, particulate matter with a diameter of <2.5 μm; PRS, polygenic risk score; and PRScs, polygenic risk score under continuous shrinkage.

Figure 4. Combined effect of genetic risk for cardiovascular diseases and PM_2.5 exposure on the risk of each incident cardiovascular diseases.

The forest plots show the adjusted HRs with 95% CIs for the risk of each cardiovascular outcome according to the increase of annual average exposure to PM_2.5, stratified by the genetic risk group defined by the median value of PRScs for (A) CAD, (B) MI, (C) stroke, (D) ischemic stroke, (E) HF, and (F) AF. *Adjusted HRs and CIs were calculated in multivariable Cox hazard models using age, sex, enrollment center, body mass index, current smoking, time spent outdoors at summer, time spent outdoors at winter, noise pollution, diabetes, hypertension on medication, dyslipidemia on medication, income score, housing score, physical activity, traffic on nearest road, traffic on nearest major road, distance from nearest major road, genotype array, and first to 10th genetic principal component. AF indicates atrial fibrillation; CAD, coronary artery disease; HF, heart failure; HR, hazard ratio; i‐Stroke, ischemic stroke; MI, myocardial infarction; PM_2.5, particulate matter with a diameter of <2.5 μm; PRS, polygenic risk score; and PRScs, polygenic risk score under continuous shrinkage.