Threshold values for diastolic dysfunction parameters indicative of increased mortality risk in a real-world patient cohort
European Heart Journal

Abstract
Current diagnostic thresholds for echocardiographic diastolic dysfunction parameters are derived from healthy adult populations rather than outcomes-based metrics (1).
To identify diagnostic thresholds of diastolic dysfunction parameters associated with increased mortality risk.
Retrospective analysis of consecutive patients who underwent echocardiography between 2011 and 2021. Patients with reduced ejection fractions, significant mitral annular calcification or stenosis, those under 18, and critically ill individuals were excluded. Cox regression models with spline curve analyses (2) were used to determine thresholds associated with increased mortality risk. Multivariable analyses were adjusted for age, gender, hypertension, atrial fibrillation, and valvular heart disease.
Following exclusions, 58,090 cases (mean age 60.1 years, 54.5% male) were included. A maximal tricuspid regurgitation jet velocity <2.57 m/s, a systolic pulmonary artery pressure >31.2 mmHg, and a deceleration time <179 msec and left atrial volume index (<23.2 and >43 mL/m²) were identified as thresholds associated with increased mortality risk (Figure 1). Mitral inflow velocities and mitral annular early diastolic velocity (e′), were associated elevated mortality risk at both high and low thresholds (Figure 2: E wave <0.51 and >0.88 m/s; A wave <0.41 and >0.84 m/s; E/A ratios <0.81 and <1.86; E/Average e’ ratios >6 and <12.7; Average e’ <5.6 and >9.8 cm/s; Lateral e’ <6.2 and >11.4 cm/s; Septal e’ <5.1 and >9.0 cm/s).
Revised thresholds for maximal tricuspid regurgitation jet velocity and systolic pulmonary artery pressure were lower, and deceleration time thresholds were higher, than cutoffs employed in current recommendations. U-shaped spline curves were observed for left atrial volume index [as previously established (3)], mitral inflow velocities, and e’, indicating increased mortality risk at low and high thresholds for these parameters.
Contributors

L Zornitzki
Author

O Freund
Author

N Schamroth-Pravda
Author

N Flint
Author

Y Granot
Author

Z Rozenbaum
Author

Y Topilsky
Author



