Heart Rate and V˙O2 Concordance in Continuous-Flow Left Ventricular Assist Devices

Abstract

The American College of Sports Medicine currently recommends the HR reserve (HRR) method to guide exercise in individuals who have heart failure with reduced ejection fraction. This recommendation is based on the known association between %HRR and percentage of V˙O2 reserve (%V˙O2R) in this population. However, to our knowledge, no studies exist regarding this relation in individuals with a left ventricular assist device (LVAD).

Purpose: This article aimed to describe the relation between V˙O2 and surrogate markers of exercise intensity among patients with LVAD.

Methods: Patients with continuous-flow LVAD (n = 24, seven females) completed a symptom-limited graded exercise test on a treadmill. HR and V˙O2 were measured continuously and averaged every 20 s. Regression equations were determined using a generalized estimating equation to predict %V˙O2R from %HRR, Borg RPE, and LVAD flow, overall and stratified by presence of pacing.

Results: Although the association between %HRR and %V˙O2R was good (R = 0.75), the slope and y-intercept for %HRR versus %V˙O2R was different from the line of identity (P = 0.002). However, when paced subjects were excluded (n = 8) from the analysis, there was no significant difference between the slope and y-intercept (= 0.036 + 0.937 × %HRR; SEE, 2%; P = 0.052). RPE showed a strong association with %V˙O2R (R = 0.84), whereas LVAD flow showed a weak (albeit statistically significant) association (R = 0.05). Both had slopes and y-intercepts that were different from the line of identity (P < 0.05).

Conclusions: In patients with LVAD who are not paced during exercise, the use of %HRR is a good predictor of %V˙O2R. However, for patients in this population who are also paced during exercise, RPE is a suitable surrogate measure of exercise intensity.