What is the central question of this study? During exercise, there are fluctuations in conduit artery blood flow (BF) caused by both cardiac and muscle contraction-relaxation cycles. We identified an optimal method to process Doppler ultrasound-measured BF for the purpose of characterizing the dynamic response of BF during step-transitions in exercise. What is the main finding and its importance? Continuous BF data were processed in relation to either cardiac or muscle contraction-relaxation cycles and computed based on `binned` or `rolling` averages over 1, 2, or 5 consecutive cycles. Kinetics characterization revealed no data processing technique-specific differences in steady-state BF, but variability in the rapidity at which BF attained steady-state (i.e., mean response time) was observed. The overall rate of blood flow (BF) adjustment (i.e., kinetics) from the onset of an exercise transition can be quantified by the mean response time (MRT). However, the BF response profile can be distorted during rhythmic, dynamic exercise consequent to variations caused by the cardiac cycle (HR) and the muscle contraction-relaxation (CR) cycle. We examined the extent to which distortions imposed by HR and CR cycles affected BF kinetics. Eight healthy, young men (27 (4) yrs; mean (SD)) performed transitions of alternate-leg knee-extension exercise from 3 W to either a moderate- (MOD) or heavy-intensity (HVY) power output. Femoral artery BF was continuously measured by Doppler ultrasound and averaged over 1, 2, or 5 `binned` (e.g., HR2b, etc.) or `rolling` (e.g., CR5r, etc.) HR and CR cycles. Amongst analysis techniques, there were no differences for steady-state BF values at the 3 W baseline. In MOD, MRT using CR1 was smaller than most other analysis techniques. For both MOD and HVY, the confidence interval for MRT (CI95) was generally larger when using HR- compared to CR-related methods, and monoexponential fits based on `rolling` averages (HR2r, HR5r, CR2r, CR5r) had a poorer ability to estimate the true end-exercise BF in HVY than in MOD. When modeling BF kinetics, we conclude that the CR1 method is a good option because of its ability to accurately estimate the `data-determined` end-exercise BF value from the `model-derived` response, maintain a relatively high density of data points during the transition, and yield a relatively small CI95.
© Copyright 2022 Experimental Physiology. The Physiological Society. All rights reserved.