Causal insights into the mechanisms underlying functional mitral regurgitation

Abstract Background Functional mitral regurgitation (FMR) arises from partially overlapping and interrelated atrial and ventricular mechanisms. However, the dominant mechanism and the interplay between etiological pathways remain incompletely defined. Distinguishing the mechanisms underlying atrial FMR (AFMR) from those of ventricular FMR (VFMR) is essential for guiding targeted therapeutic strategies. Purpose This study aimed to evaluate the directional relationships between known etiological determinants and the two distinct FMR subtypes through Granger causality testing. This statistical approach is based on time series analysis and provides insight into potential causal relationships. Methods We included 26 patients with at least mild FMR: 20 in the AFMR group, all in atrial fibrillation or flutter and with preserved left ventricular ejection fraction, and 6 in the VFMR group, all in sinus rhythm and with left ventricular dysfunction and/or dilatation. Each patient underwent transthoracic echocardiography, capturing up to ten consecutive cardiac cycles (25–102 frames per patient), totalling 969 time points. For each frame, we measured mitral regurgitant jet area (MR area), left atrial and ventricular volumes (LA, LV volumes), mitral valve annulus diameter (MV annulus), and tenting height to assess FMR variability based on its potential determinants. We applied multivariate autoregression and Granger causality tests within each group to evaluate the direction and significance of associations. Additionally, univariate models with group-specific standardized coefficients (β) were used to estimate effect sizes for significant relationships. Results Compared to the VFMR group, patients in the AFMR group were significantly older (mean age 78.7±7.2 vs. 67.5±9.7 years, p=0.005), had larger left atrial volumes (median 57.8 45.8–81.3 vs. 30.7 26.0–39.6 mL/m², p=0.007), and smaller left ventricular volumes (median 53.0 45.0–59.0 vs. 99.0 73.0–126.0 mL/m², p=0.001). The two groups were similar in terms of NYHA class, comorbidities (except for ischemic heart disease, which was more frequent in the VFMR group), or laboratory parameters. In the AFMR group, LA volume was the only variable that Granger-caused MR area (β=0.742, p0.001, Figure 1A), while in the VFMR group, MR area was Granger-caused by both LV volume (β=0.245, p0.001) and LA volume (β=0.175, p0.001, Figure 1B). Additional significant group-specific Granger causalities are summarized in Figure 1. Individual patient-level analyses (excluding three AFMR patients with insufficient frame data) enabled characterization of causal pattern distribution within each group (Figure 2). Conclusions AFMR is mainly driven by left atrial remodelling, while VFMR reflects a complex interplay of ventricular dysfunction and atrial remodelling. The patient-specific nature of these patterns highlights the need for tailored therapies targeting the underlying mechanisms of each FMR subtype.Fig. 1.Granger causality Fig. 2.Causality distribution per group