Ventricular-arterial coupling derived from proximal aortic stiffness and aerobic capacity across the heart failure spectrum

Aims Ventricular–arterial coupling (VAC) can be evaluated as the ratio between arterial stiffness (pulse wave velocity, PWV) and myocardial deformation (global longitudinal strain, GLS). To evaluate VAC across the spectrum of heart failure (HF). Methods and results We introduced a Doppler-derived, single-beat technique to estimate aortic arch PWV (aa-PWV) in addition to tonometry-derived carotid-femoral PWV (cf-PWV). We measured PWVs and GLS in 155 healthy controls, 75 subjects at risk of developing HF (American College of Cardiology/American Heart Association Stage A-B) and 236 patients in HF Stage C with preserved (HFpEF, n = 104) or reduced ejection fraction (HFrEF, n = 132). We evaluated peak oxygen consumption (VO2) and peripheral extraction (AVO2diff) using combined cardiopulmonary-echocardiography exercise stress. aa-PWV was obtainable in all subjects and significantly lower than cf-PWV in all subgroups (P < 0.01). PWVs were directly related and increased with age (all P < 0.0001). cf-PWV/GLS was similarly compromised in HFrEF (1.08 ± 0.36) and HFpEF (1.05 ± 0.22), while aa-PWV/GLS was more impaired in HFpEF (0.69 ± 0.11) than HFrEF (0.60 ± 0.15; P < 0.01). Stages A and B had values of cf-PWV/GLS and aa-PWV/GLS (0.66 ± 0.25 and 0.47 ± 0.12) higher than controls (0.47 ± 0.10 and 0.40 ± 0.10) but lower than Stage C (all P < 0.01). Peak AVO2diff was inversely related with cf-PWV/GLS and aa-PWV/GLS (all P < 0.01). cf-PWV/GLS and aa-PWV/GLS independently predicted peak VO2 in the overall population (adjusted R2 = 0.32 and 0.35; all P < 0.0001) but only aa-PWV/GLS was independently associated with flow reserve during exercise (R2 = 0.51; P < 0.0001). Conclusions Abnormal VAC is directly correlated with greater severity of HF and worse functional capacity. HFpEF shows a worse VAC than HFrEF when expressed by aa-PWV/GLS.