Predictors of congestive heart failure in the elderly: the Cardiovascular Health Study.

OBJECTIVES We sought to characterize the predictors of incident congestive heart failure (CHF), as determined by central adjudication, in a community-based elderly population. BACKGROUND The elderly constitute a growing proportion of patients admitted to the hospital with CHF, and CHF is a leading source of morbidity and mortality in this group. Elderly patients differ from younger individuals diagnosed with CHF in terms of biologic characteristics. METHODS We analyzed data from the Cardiovascular Health Study, a prospective population-based study of 5,888 elderly people >65 years old (average 73 +/- 5, range 65 to 100) at four locations. Multiple laboratory measures of cardiovascular structure and function, blood chemistries and functional assessments were obtained. RESULTS During an average follow-up of 5.5 years (median 6.3), 597 participants developed incident CHF (rate 19.3/1,000 person-years). The incidence of CHF increased progressively across age groups and was greater in men than in women. On multivariate analysis, other independent predictors included prevalent coronary heart disease, stroke or transient ischemic attack at baseline, diabetes, systolic blood pressure (BP), forced expiratory volume 1 s, creatinine >1.4 mg/dl, C-reactive protein, ankle-arm index <0.9, atrial fibrillation, electrocardiographic (ECG) left ventricular (LV) mass, ECG ST-T segment abnormality, internal carotid artery wall thickness and decreased LV systolic function. Population-attributable risk, determined from predictors of risk and prevalence, was relatively high for prevalent coronary heart disease (13.1%), systolic BP > or =140 mm Hg (12.8%) and a high level of C-reactive protein (9.7%), but was low for subnormal LV function (4.1%) and atrial fibrillation (2.2%). CONCLUSIONS The incidence of CHF is high in the elderly and is related mainly to age, gender, clinical and subclinical coronary heart disease, systolic BP and inflammation. Despite the high relative risk of subnormal systolic LV function and atrial fibrillation, the actual population risk of these for CHF is small because of their relatively low prevalence in community-dwelling elderly people.