Abstract
RATIONALE: Although chronic obstructive pulmonary disease has been related to heart failure, the relationship between the restrictive spirometry pattern (forced vital capacity [FVC] < 80% predicted with preserved forced expiratory volume in 1 second [FEV(1)]/FVC ratio) and heart failure is poorly understood. OBJECTIVES: To determine whether having a restrictive spirometry pattern is associated with incident heart failure hospitalization. METHODS: Community-dwelling African Americans from the Jackson Heart Study (total n = 5,306; analyzed n = 4,210 with spirometry and heart failure outcome data) were grouped by restrictive spirometry (FEV(1)/FVC ≥ 0.70, FVC < 80%; n = 840), airflow obstruction (FEV(1)/FVC < 0.70; n = 341), and normal spirometry (FEV(1)/FVC ≥ 0.70, FVC ≥ 80%; n = 3,029) at the time of baseline examination in 2000-2004. We assessed relationships of echocardiographic parameters and biomarkers with spirometry patterns using regression models. Incident heart failure was defined as an adjudicated hospitalization for heart failure after January 1, 2005 in subjects with no self-reported heart failure history. We used multivariable-adjusted Poisson regression models and Cox proportional hazards models, with death treated as a competing risk in the Cox models, to test associations between spirometry patterns and incident heart failure. We also modeled the association of FVC% predicted with heart failure hospitalization risk using a restricted cubic spline after excluding subjects with airflow obstruction. RESULTS: At the time of baseline spirometry, participants with restrictive spirometry had a median age of 57.2 years (interquartile range, 47.8-64.1); 38.1% were male. Compared with normal spirometry, restrictive spirometry was associated with a higher transmitral early (E) wave velocity to atrial (A) wave velocity ratio, higher pulmonary artery systolic pressure, and higher endothelin levels. After a median follow-up time of 8.0 years, 8.0% of subjects with restrictive spirometry (n = 67) had developed incident heart failure, compared with 3.8% of those with normal spirometry (n = 115) and 10.6% of those with airflow obstruction (n = 36). After risk adjustment, both a restrictive pattern (hazard ratio [HR], 1.5; 95% confidence interval [CI], 1.1-2.0) and airflow obstruction (HR, 1.7; 95% CI, 1.1-2.5) were associated with increased rates of incident heart failure hospitalization compared with normal spirometry. Using flexible modeling, the lowest hazards of heart failure hospitalization were observed around FVC 90-100%, with lower FVC% values associated with an increased incidence of heart failure. CONCLUSIONS: Both a restrictive pattern on spirometry and airflow obstruction identify African Americans with impaired lung health at risk for heart failure.