Figure:
A) Clinical characterization of three HFpEF subgroups: hypertension/hypertrophy (Ht/Hp), obesity/diabetes (Ob/Dm), and Mixed. Violin plots show median and 25/75 percentiles for each variable. (Ht/Hp)/(Ob/Dm) index ratio, body mass index (BMI), sex-adjusted left ventricular mass index (LVMi), systolic blood pressure (SBP), pulmonary arterial systolic pressure (PASP), pulmonary arterial wedge pressure (PAWP), log(NTproBNP), and ejection fraction (EF). Kruskal Wallis test followed by Dunn’s multiple comparisons, p-values displayed in each panel. B) Passive myocyte tension-sarcomere length (SL) dependence for control and 3 HFpEF groups. Relations are fit to a mono-exponential. Group effect on Tension-SL relation p=7x10−6 (2-way repeated measures analysis of variance, 2W-RMANOVA). Symbols show Tukey multiple comparisons test at each SL: * p<0.01 CON vs Ht/Hp; p<0.05 CON vs Mixed; # p<0.001 CON vs Ht/Hp and Mixed, p<0.02 vs Ob/Dm; † p<0.005 CON vs Ht/Hp and Mixed; ‡ p<0.001 CON vs Ht/Hp and Mixed, p=0.007 CON vs Ob/Dm. C) Tension-calcium relations for control and HFpEF. Curves are fit to the Hill equation, analyzed by 2W-RMANOVA (overall group effect p=10−6), symbols for Tukey multiple comparisons test: * p<0.001, † p<0.0001, ‡ p<0.002 versus both Ob/Dm and Mixed; § p<0.03 Ht/Hp versus Ob/Dm and Mixed; ¶ p<0.05 Ht/Hp versus CON and Ob/Dm; # p<0.01 CON vs Ht/Hp and Mixed; p=0.025 Ob/Dm vs Ht/Hp and p=0.05 vs Mixed. D) Violin plots for Tmax, E) myocyte cross sectional area (CSA) F) EC50 G) Hill coefficient. 1-way ANOVA, Tukey multiple comparisons: **** p<10−6; *** p<3x10−4; ** <0.01; * p<0.05. H) Negative correlation between BMI and Tmax. Data are color coded to show HFpEF patients with or without diabetes mellitus (DM), and Controls. Linear regression and 95% confidence bands (employing all the data) are shown. Regression equation: Tmax = −0.462 x BMI + 31. I) Linear regression analysis for the relation between EC50 and BMI using same approach as in Panel H. Regression equation EC50 = −0.01315*X + 2.102.