Pulse transit time and the pulse wave contour as measured by photoplethysmography: the effect of drugs and exercise

Abstract

Photoplethysmography (PPG) is a simple means of measuring the pulse wave in humans, exploitable for the purposes of timing the arrival of the pulse at a particular point in the arterial tree, and for pulse contour analysis. This thesis describes a methodology for measuring arterial pulse transit time (PTT) from cardiac ejection to pulse arrival at the finger. It describes the effect on PTT of drug and exercise induced changes in BP. The nature of the relationship between the PPG and arterial pressure is also examined, and the PTT technique extended to assessment of conduit vessel pulse wave velocity (PWV) during exercise.

PTT measured from ECG R-wave to PPG finger wave (rPTT) had a negative correlation (R2=0.39) with systolic BP (SBP), unaffected by vasoactive drugs in some but not all persons. rPTT showed similar beat-to-beat variability to SBP, unaffected by drugs. rPTT correlated weakly with diastolic (DBP) and mean (MAP) pressure. Cardiac pre-ejection period (PEP) formed a substantial and variable part of rPTT (12% to 35%). Transit time adjusted for PEP (pPTT) correlated better with DBP (R2=0.41) and MAP (R2=0.45), than with SBP. The PPG wave tracked changes in the peripheral pressure wave. Drugs had little effect on the generalised transfer function (GTF) describing the association between arterial and PPG waves.

Strenuous exercise induced a large decrease in rPTT, mainly accounted for by decreases in PEP (53% of the total change in rPTT) and in transit time from aorta to distal brachial artery (33%). In contrast, minimal change in transit time from wrist to finger tip occurred with exercise.

Simultaneous ear-finger PPG signals were used to measure conduit artery PWV during exercise. Ear-finger PWV (PWVef) overestimated carotid-radial PWV throughout exertion (overall bias 0.81±1.05ms-1, p<0.001), but the degree of difference remained constant. The increase in PWVef with exercise, was greater (1.18±0.54ms-1, p=0.035) in healthy subjects with a positive cardiovascular family history compared to those without.

PPG enables analysis of the pulse contour during exercise, but estimation of the radial pressure wave from finger PPG by use of a GTF derived at rest, resulted in inaccuracy following exertion. These effects were variable and relatively short-lived.

Furthermore, a resting GTF used to determine central pressure from the peripheral wave, resulted in underestimation of SBP (-5.9±2.1mmHg) and central pressure augmentation index (-8.3±2.9%), which persisted for 10 minutes post-exercise.

rPTT had a negative linear association with SBP (R2=0.94) during strenuous exercise, slightly stronger than during recovery (R2=0.85). Differences existed in area-undercurve of the rPTT/SBP relationship between exercise and recovery, due to discrepancies in rate and degree of recovery of SBP and PEP. The linear relationship between the rPTT/SBP during exercise was affected by aerobic capacity, and the regression slope was less in the anaerobic compared to aerobic phase of exercise due to minimal change in PEP during anaerobic exertion. The correlation between rPTT/SBP did not change with prolonged aerobic exercise. Finally, measures of baroreflex sensitivity during exercise, were not significantly different between actual beat-to-beat SBP and SBP estimated using rPTT.

In conclusion, absolute BP cannot be reliably estimated by measurement of rPTT following administration of drugs and during exercise. However, rPTT may have a role in measuring BP variability and in the assessing exercise capacity.

PPG may also be useful in determining the effects of exercise on arterial stiffness, and for estimating the pressure wave contour, although its use during exercise for the latter purpose must be treated with caution.