Abstract
Rationale: Chronic lung diseases are associated with increased risk of mortality due to coronary heart disease (CHD). Nonetheless, the population attributable fraction (PAF) of lung function impairment relative to other established cardiovascular risk factors is unclear. Objectives: To evaluate the PAF of low lung function for CHD mortality Methods: We harmonized and pooled lung function and clinical data across eight U.S. general population cohorts. Impaired lung function was defined as forced expiratory volume in 1 second (FEV(1)) and/or forced vital capacity ≤ 95% predicted on baseline spirometry. The association between CHD mortality and risk factors was assessed using cause-specific proportional hazards and Fine-Gray proportional subdistribution hazard models, treating non-CHD mortality as a competing risk. Models were adjusted for lung function as well as age, sex, race/ethnicity, educational attainment, body mass index, smoking status, pack-years of smoking, diabetes mellitus, high-density lipoprotein, and high low-density lipoprotein (≥130 mg/dl). PAF was calculated as the relative change in the average absolute risk of 10-year CHD mortality by elimination of lung function lower than 95% predicted. Results: Among 35,143 participants, 1,844 of 13,174 (14.0%) deaths were due to CHD. Compared with percentage predicted FEV(1) (FEV(1)pp) > 95%, the subdistribution adjusted hazard ratio for low FEV(1)pp was 1.30 (95% confidence interval, 1.18-1.44). The PAF for FEV(1)pp ≤ 95% was 12%, ranking low FEV(1) third on the list of PAF for CHD mortality, after hypertension and diabetes. Low FEV(1)pp ranked second in the subgroup of active smokers (PAF 14%), after hypertension. Conclusions: Low lung function, even in the range considered clinically normal, ranks high on the list of attributable risk factors for CHD mortality and should be considered in cardiovascular risk stratification.