Mode
Text Size
Log in / Sign up

Granger causality analysis identifies structural determinants predicting beat-to-beat mitral regurgitation variability in functional mitral regurgitation.

Granger causality analysis identifies structural determinants predicting beat-to-beat mitral regurgi…
Photo by Bozhin Karaivanov / Unsplash
Key Takeaway
Note that Granger causality identifies structural determinants predicting beat-to-beat mitral regurgitation variability in functional mitral regurgitation subtypes.

This exploratory Granger causality study investigated structural determinants predicting beat-to-beat mitral regurgitation variability within a cohort of 41 patients with functional mitral regurgitation, comprising 21 with atrial and 20 with ventricular subtypes. The intervention involved beat-to-beat echocardiographic time series analysis to assess predictive relationships between structural metrics and regurgitation area. No comparator was reported, and the study setting was not specified. Funding or conflicts of interest were not reported.

Main results indicated that left ventricular volume Granger-predicting mitral regurgitation area was the strongest predictor at short lags, with atrial p=0.011 and ventricular p=0.006. Left atrial volume Granger-predicting mitral regurgitation area emerged at longer lags (lag 7), showing atrial p=0.043 and ventricular p=0.011. Systolic papillary muscle length was not predictive in pooled analysis, whereas papillary muscle length Granger-predicting regurgitation was predictive only in the ventricular subtype (p=0.001). Conversely, regurgitation predicting papillary muscle displacement was observed only in the atrial subtype (p<0.001). No structural determinant correlated with severity.

Safety and tolerability were not reported, as adverse events, serious adverse events, discontinuations, and tolerability data were not collected or reported. Key limitations include marked heterogeneity in individual models and subtype-specific dissociation in full-cycle analysis. Static analyses cannot capture temporal patterns, and relative contributions vary between atrial and ventricular subtypes. The practice relevance lies in a framework that may support patient-specific temporal phenotyping of functional mitral regurgitation, though this is an exploratory study with uncertain certainty.

Study Details

Sample sizen = 41
EvidenceLevel 5
PublishedApr 2026
View Original Abstract ↓
Background Functional mitral regurgitation results from interacting mechanisms whose relative contributions vary between atrial and ventricular subtypes and shift dynamically within each heartbeat, producing temporal patterns that static analyses cannot capture. Objectives To identify which structural determinants predict mitral regurgitation variability beat to beat using Granger causality within vector autoregression, focusing on papillary muscle dynamics across subtypes. Methods Frame-level echocardiographic time series from 41 patients (21 atrial, 20 ventricular; 1,959 frames) were z-score standardised within patient. Individual (lag 3) and pooled (lag 2) vector autoregression models tested whether left ventricular volume, left atrial volume, papillary muscle length, and annulus diameter Granger-predict mitral regurgitation area. Results Individual models revealed marked heterogeneity. In pooled analysis, left ventricular volume was the strongest Granger predictor at short lags (atrial p=0.011; ventricular p=0.006), while left atrial volume emerged at longer lags (lag 7: atrial p=0.043; ventricular p=0.011). Systolic papillary muscle length was not predictive. Full-cycle analysis revealed a subtype-specific dissociation: papillary muscle length Granger-predicted regurgitation only in the ventricular subtype (p=0.001), while regurgitation predicted papillary muscle displacement only in the atrial subtype (p<0.001). Left ventricular volume dominated within-beat prediction but lost cross-beat relevance in the ventricular subtype, while left atrial volume gained cross-beat predictive relevance in the atrial subtype. No structural determinant correlated with severity cross-sectionally. Conclusions Beat-to-beat vector autoregression and Granger modelling reveals heterogeneous, subtype-specific temporal patterns with distinct temporal windows of predictability for ventricular loading and papillary geometry. This framework may support patient-specific temporal phenotyping of functional mitral regurgitation.
Free Newsletter

Clinical research that matters. Delivered to your inbox.

Join thousands of clinicians and researchers. No spam, unsubscribe anytime.