Abstract 4367306: Non-invasive Diastolic Stress Testing in a Large Observational Cohort Study: Insights from the Dallas Heart Study

Abstract

Introduction: Despite its increasing prevalence and significant morbidity/mortality in the aging population, heart failure with preserved ejection fraction (HFpEF) remains challenging to diagnose as its symptoms may occur only during exercise. In smaller studies, non-invasive diastolic stress testing has been validated in diagnosing exercise-induced increase in diastolic filling pressures. However, its feasibility in a large real-world population has not been tested. Aims: The Dallas Heart Study is a multiethnic, probability-based, population cohort study of Dallas County residents with deliberate oversampling of black individuals. The study sample consisted of 712 participants with LVEF≥50% and free of prevalent HF who attended the 3 rd phase of the Dallas Heart Study (2021-2024) and underwent diastolic stress echocardiography. Methods: Participants underwent comprehensive resting echo followed by a submaximal stress echo protocol which includes pedaling on a supine echocardiography bicycle at a fixed workload of 30W. Per 2016 stress echocardiography recommendations by the American Society of Echocardiography (ASE), a definitive abnormal test was defined when the following criteria are met: septal E/e′ ratio>15, average E/e′>14, peak TR velocity>2.8 m/s with exercise, and either septal e′<7 or lateral e′<10 cm/s at rest. Results: Mean age was 59±11 years, and 56% were women ( Table 1 ). 48% of the participants were Black, 32% White, and 18% Hispanic. Mean BMI was 31±7 kg/m 2 , 23% had diabetes, and 59% had hypertension. Between rest and stress ( Table 2 ), there were statistically significant increases in LVEF, LV GLS, e’, mean E/e’, peak TR velocity, LA volume index, and LA reservoir strain. At rest, 5% of the participants had abnormal E/e’ (defined as >14), while at stress 8% did. To identify predictors of abnormal E/e’, backwards (P <0.001) multivariable stepwise regression modeling was performed on baseline demographics and resting echo variables and found lateral E/e’ as the only independent predictor (OR 1.86, 95% CI 1.40-2.46). Notably, 2% of the participants were found to have a definite abnormal diastolic stress test per ASE recommendation. Conclusion: We have demonstrated the feasibility of performing non-invasive diastolic stress testing on a large contemporary real-world cohort. With rapidly emerging HFpEF therapies, more widespread adoption of an effective and safe diagnostic tool for HFpEF is needed to warrant adequate and individualized treatment.