Fractionating E-wave deceleration time into its stiffness and relaxation components distinguishes pseudonormal from normal filling

Background: Pseudonormal Doppler E-wave filling patterns indicate diastolic dysfunction but are indistinguishable from the normal filling pattern. For accurate classification, maneuvers to alter load or to additionally measure peak E' are required. E-wave deceleration time (DT) has been fractionated into its stiffness (DT_s) and relaxation (DT_r) components (DT=DT_s+DT_r) by analyzing E-waves via the parametrized diastolic filling formalism. The method has been validated with DT_s and DT_r correlating with simultaneous catheterization-derived stiffness (dP/dV) and relaxation (τ) with r=0.82 and r=0.94, respectively. We hypothesize that DT fractionation can (1) distinguish between unblinded (E' known) normal versus pseudonormal age-matched groups with normal left ventricular ejection fraction, and (2) distinguish between blinded (E' unknown) normal versus pseudonormal groups, based solely on E-wave analysis. Methods and Results: Data (763 E-waves) from 15 age-matched, pseudonormal (elevated E/E') and 15 normal subjects were analyzed. Conventional echocardiographic and parametrized diastolic filling stiffness (k) and relaxation (c) parameters and DT_s and DT_r were compared. Conventional diastolic function parameters did not differentiate between unblinded groups, whereas k, c (P<0.001) and DT_s, DT_r (P<0.001) did. Independent, blinded (E' not provided) analysis of 42 subjects (30 subjects from unblinded training set and 12 additional subjects from validation set, 581 E-waves) showed that R (=DT_r/DT) had high sensitivity (0.90) and specificity (0.86) in differentiating pseudonormal from normal once E' revealed actual classification. Conclusions: Parametrized diastolic filling-based E-wave analysis (k, c or DT_s and DT_r) can differentiate normal versus pseudonormal filling patterns without requiring knowledge of E'.