CHANGES OF DIASTOLIC FUNCTION IN LATE LIFE AND NON-CARDIAC RISK FACTORS FOR HEART FAILURE IN THE BLACK- PERSPECTIVES ON HEART FAILURE IN HIGH-RISK POPULATIONS

Abstract

Paper 1 Introduction: Diastolic dysfunction is a potent risk factor for heart failure (HF). However, there is limited data regarding longitudinal changes of diastolic function in the very old, who are at the highest risk for HF. Methods: We studied 2,499 older adult participants in the prospective community-based Atherosclerosis Risk in Communities (ARIC) study who underwent protocol echocardiography at study Visits 5 (2011-2013) and 7 (2018-2019), were in sinus rhythm at both exams, and were free of interval myocardial infarction. The primary diastolic measures were tissue Doppler e’, E/e’ ratio, and left atrial volume index (LAVi). The associations of changes of diastolic function and changes of participant-reported dyspnea and functional capacity by 2-minute walk distance were assessed. Results: Mean age at Visit 5 was 74±4 years, 59% were women, and 24% black. At Visit 5, mean e’septal was 5.8±1.4 cm/s, E/e’septal 11.7±3.5, and LAVi 24.3±6.7 ml/m2. Over a mean of 6.6±0.8 years, e’septal decreased by 0.6±1.4 cm/s, E/e’septal increased by 3.1±4.4, and LAVi increased by 2.3±6.4 ml/ m2. Increases were also observed for the proportion of participants with abnormal e’ (24% at Visit 5 to 43 % at Visit 7), E/e’ (25% vs 54% respectively), and LAVi (17% vs 25% respectively; all p.001), and the proportion with two or more abnormal diastolic measures (16 vs 42%, p.001). Compared to participants free of cardiovascular (CV) risk factors or diseases at Visit 5 (n=234), those with prevalent CV risk factors or diseases (n=2,150) demonstrated greater increases in E/e’septal (2.2±3.9 vs 3.1±4.4 respectively; p=0.004) and LAVi (1.5±5.2 vs 2.3±6.4 ml/m2; p=0.047) while those who developed HF between Visits 5 and 7 (n=63) demonstrated the greatest increase in E/e’septal (4.6±5.1; p.001) and LAVi (3.6±6.8 ml/m2; p=0.011). Increases of E/e’septal and LAVi were both associated with progression of self-reported dyspnea between Visits. Conclusion: Diastolic function worsens over 6.6 years in late life, particularly among persons with CV risk factors, and is associated with development of self-reported dyspnea. Further studies are necessary to determine if risk factor prevention or control will mitigate these changes. Paper 2 Introduction: Heart failure (HF) disproportionately burdens Black Americans. However, sparce data exist regarding the contributions of subclinical impairments in cardiovascular and non-cardiovascular function to incident HF with reduced (HFrEF) and preserved (HFpEF) ejection fraction in this group, or the extent to which they account for the impact of adverse social determinants on HF risk. Methods: Among Black American participants in the Jackson Heart Study (JHS) who were free of prevalent HF at the first study visit (2000-2004), we assessed the associations of measures of cardiovascular and non-cardiovascular organ function with incident HF overall, HFpEF, and HFrEF over a median follow-up of 12 years using multivariable Cox proportional hazard models. Systems considered included: Cardiovascular: LV structure, LV systolic function, LV diastolic function, systemic arterial function; Non-cardiovascular: pulmonary function, renal function, body composition, and dysglycemia. A cardiovascular score was generated using LV mass index, LVEF, left atrial width, and pulse pressure, and a non-cardiovascular score was generated using percent predicted FEV1, eGFR, HbA1c, and waist circumference. Results: Among the 4,356 participants in this analysis, mean age was 54 years and 64% were women. Incident HF occurred in 315, including 152 HFpEF, 134 HFrEF and 29 unclassified HF. In multivariable models incorporating measures reflecting each organ system (LV mass index, LVEF, LA diameter, ppFEV1, eGFR, pulse pressure, HbA1c, waist circumference), significant predictors of incident HFpEF included greater LA diameter, higher pulse pressure, lower ppFEV1, lower eGFR, and higher HbA1c. Predictors of incident HFrEF included greater LVMI, lower LVEF, and lower eGFR. eGFR was the only independent predictor of both HFpEF and HFrEF. In separate models containing both the cardiovascular and non-cardiovascular risk scores, the magnitude of association of the cardiovascular risk score was greater with incident HFrEF while the magnitude of association of the non-cardiovascular risk score was greater with HFpEF. CV and non-CV risk scores partially accounted for the association of worse economic status with risk of HF, but did not appreciably attenuate associations of lower educational attainment and greater neighborhood problems with incident HF and HFpEF respectively. Conclusion: Subclinical impairments in both cardiovascular and non-cardiovascular organ function differentially associated with risk of incident HFpEF and HFrEF. These findings support partially distinct mechanisms underlying HFrEF and HFpEF, and a greater contribution of diverse non-cardiac impairments to HFpEF in particular.