Arterial function can be obtained by noninvasive finger pressure waveform

Abstract

Arterial function, expressed as augmentation index (AIx) and pulse wave velocity (PWV), predicts cardiovascular disease [1–4] and can be assessed by specific validated devices that subsequently record peripheral (carotid, radial, femoral) arterial waveforms. It could have practical implications to obtain the same information from easily applicable finger blood pressure monitors, which are commonly used to test autonomic function.Methods to obtain PWV from finger photoplethysmogram (PPG, [5]), or AIx from finger blood pressure [6] have been proposed.We tested whether reliable AIx and PWV data could be obtained from 5-minute recordings of noninvasive finger blood pressure (Finapres® 2300, Ohmeda, USA), and compared them with data simultaneously obtained from applanation tonometry (SphygmoCor®, Atcor, Australia). The protocol was approved by Helsinki University Ethical Committee, and written consent was obtained from each subject. To compare the twomethods over awide range of values we studied 46 consecutive unselected subjects of different ages (31 male, age 44 ± 16 yr, range 24–90, SBP: 129 ± 21 mmHg, DBP: 74 ± 16 mmHg, BMI: 24.8 ± 4.5 kg/m), in supine position, under spontaneous breathing at rest. Two reliable measurements were obtained from 29 (AIx) and 31 (PWV) participants by applanation tonometry (AIx-at, PWV-at), and one measurement in the remaining subjects. When two values were obtained their average was used for comparison. Measurementof AIx by transfer functionoffingerbloodpressure (AIx-f). In theory, thebestmethod toderivenoninvasiveAIx is throughsimultaneous recording of peripheral noninvasive and central intra-arterial blood pressure, with subsequent evaluation of the mathematical relationship (transfer function, TF) linking these two signals together (TF1). The TF1 can then be used to reconstruct the central blood pressure profile and obtain the AIx [2–4,6,7]. This rather complex approach requires expensive high-fidelitymicro-tiparterial catheters [7], butnowthisprocedure canbe simplified since our tonometer provided both the radial artery and the reconstructed central pressure waveforms (hence also TF1). Since the recordings of the radial artery and finger blood pressures were simultaneous, it was possible to obtain the TF that links the finger pressure to the central pressure (TF2)with the samedegree of precision as for TF1. By using TF2, the AIx-f can then be calculated (on-line supplement). The obtained TF1 was almost identical in all subjects, confirming that a single TF1 was used by the Sphygmocor® to link the central and radial pressures [7] (Fig. 1). Measurement of AIx by direct analysis of finger blood pressure contour (AIx-fbpc). From the previous considerations it follows that the blood pressure profile alone provides all the necessary information to directly derive AIx. To verify this we normalised the averaged amplitude of the finger blood pressure pulses (first systolic peak= 100, lowest point = 0) and determined the location of the dicrotic notch. Then we calculated the normalised amplitude at 21 equally-spaced points from the first systolic peak to the dicrotic notch (on-line supplement). Each of the amplitudes was comparedwith the AIx-at by regression analysis. The regression equation linking the PWV-at with the point showing the best correlation was used to an AIx estimate: AIx-fbpc. Pulse wave velocity from the finger blood pressure. Millasseau [5] described and validated a simple technique to estimate the PWV from the PPG without the need of two separate recordings at two different International Journal of Cardiology 175 (2014) 169–217