The relative importance of left atria reservoir strain compared with components of HFA-PEFF diagnostic algorithm

Abstract

Abstract Background/Introduction The relative importance of the left atrial reservoir strain (LARS) compared to the parameters constituting the HFA-PEFF score developed as a diagnostic tool for patients with heart failure with preserved ejection fraction (HFpEF) is not well known. Purpose We aimed to identify the relative importance of LARS compared with variables associated with HFpEF and HFA-PEFF score including echocardiographic parameters, demographic data, and NT-proBNP. Methods From August 2021 to March 2022, we obtained data retrospectively from the participants visiting our clinic in a single cardiovascular center. The 1,252 participants with sinus rhythm and ejection fraction more than 50% were enrolled in the present study (61.8±16.6 years and 55% female). Multivariable regression analysis using backward elimination and random forest analysis, supervised learning algorithm, were performed to identify the relative importance of LARS on the HFA-PEFF score. Results The median HFA-PEFF score was 3 points (interquartile range 2 to 4 points). Two hundred forty-one subjects (19.2%) had more than 5 points. LARS showed a moderate correlation with the HFA-PEFF score (r=−0.632, p<0.001). In multivariable regression analysis, LARS was independent variable affecting HFA-PEFF score with female sex, left atrial volume index (LAVI), septal e', E over e' (septal), the maximal tricuspid regurgitation velocity (TR-Vmax), left ventricular global longitudinal strain (LV-GLS), NT-proBNP, interventricular septal wall thickness (standardized β=−0.106; P<0.001). In random forest analysis, LARS was a more relatively important variable than TR-Vmax, NT-proBNP, septal E/e', septal e', the major echocardiographic components of HFA-PEFF score estimating more than 5 points of HFA-PEFF score. Conclusions LA reservoir strain is one of the important factors with components of the HFA-PEFF score which is a useful tool to assess the patients with HFpEF. Funding Acknowledgement Type of funding sources: None.