Abstract
Wall shear rate (WSR) is an important stimulus for the brachial artery flow-mediated dilation (FMD) response. However, WSR estimation near the arterial wall by conventional Doppler is inherently difficult. To overcome this limitation, we utilized multigate Doppler to accurately determine the WSR stimulus near the vessel wall simultaneously with the FMD response using an integrated FMD system [Ultrasound Advanced Open Platform (ULA-OP)]. Using the system, we aimed to perform a detailed analysis of WSR-FMD response and establish novel WSR parameters in a healthy young population. Data from 33 young healthy individuals (27.5 ± 4.9 yr, 19 females) were analyzed. FMD was assessed with reactive hyperemia using ULA-OP. All acquired raw data were postprocessed using custom-designed software to obtain WSR and diameter parameters. The acquired velocity data revealed that nonparabolic flow profiles within the cardiac cycle and under different flow states, with heterogeneity between participants. We also identified seven WSR magnitude and four WSR time-course parameters. Among them, WSR area under the curve until its return to baseline was the strongest predictor of the absolute (R2 = 0.25) and percent (R2 = 0.31) diameter changes in response to reactive hyperemia. For the first time, we identified mono- and biphasic WSR stimulus patterns within our cohort that produced different magnitudes of FMD response [absolute diameter change: 0.24 ± 0.10 mm (monophasic) vs. 0.17 ± 0.09 mm (biphasic), P < 0.05]. We concluded that accurate and detailed measurement of the WSR stimulus is important to comprehensively understand the FMD response and that this advance in current FMD technology could be important to better understand vascular physiology and pathology.
NEW & NOTEWORTHY An estimation of wall shear rate (WSR) near the arterial wall by conventional Doppler ultrasound is inherently difficult. Using a recently developed integrated flow-mediated dilation ultrasound system, we were able to accurately estimate WSR near the wall and identified a number of novel WSR variables that may prove to be useful in the measurement of endothelial function, an important biomarker of vascular physiology and disease.
Keywords: endothelial function, reactive hyperemia, ultrasound, vasodilation, wall shear stress
INTRODUCTION
Brachial artery flow-mediated dilation (FMD) has been used extensively to assess the function and health of the vascular endothelium since its first use by Celermajer et al. (5). However, a persistent problem with this method is the inability to accurately measure the wall shear stress stimulus that produces the measured FMD response. This is a perplexing problem that has resulted in the extensive use of Doppler ultrasound peak velocity to estimate wall shear stress, or its surrogate, wall shear rate (WSR). This is currently the only practical solution available when measuring FMD, but this method underestimates WSR (22) and is a blunt instrument to dissect the complex and dynamic WSR events occurring during FMD.
In the absence of WSR data, the accurate interpretation of the FMD response may be confounded and this has likely limited its full potential as a scientific and clinical tool. For example, the mechanisms of diminished FMD could involve impaired brachial artery endothelial function or alternatively could result from altered microvascular function blunting the hyperemic response and, in turn, WSR stimulus. Knowledge of the WSR stimulus may also help to resolve and exploit issues relating to differences in baseline diameter and “low-flow” vasoconstriction occurring during cuff occlusion (2, 11, 26), providing a more comprehensive characterization of the underlying vascular physiology.
For physiological studies, the optimal tool for FMD needs to combine continuous and simultaneous measurement of WSR and vessel diameter to comprehensively characterize the WSR-FMD stimulus-response relationship. We have previously reported a method to simultaneously measure WSR and vessel diameter during FMD (28). Unlike the estimation of WSR from Doppler ultrasound, this system uses multigate spectral Doppler to acquire velocity data at different sites across the vessel diameter and close to the vessel wall, producing a continuous velocity profile from near to far wall (27). These data can be used to generate a detailed spectral Doppler profile across the vessel diameter and close to the vessel wall to accurately determine WSR (27). This overcomes the limitation of a single pulsed-wave Doppler sample gate and the need to assume a perfect parabolic velocity profile (12).
Using this method we were able to show that estimation of WSR by conventional Doppler ultrasound is inaccurate (28), at least partly because the presumed parabolic velocity profile (15) is often asymmetric during hyperemia and varies considerably through the cardiac cycle (12, 23, 28). Because of this, the extant literature presents a well-characterized arterial diameter response to reactive hyperemia but lacks a well-characterized WSR stimulus.
We have recently enhanced and refined this system in a collaborative effort among engineers, clinicians, and physiologists with extensive experience of using ultrasound, to provide a sophisticated tool that can advance current FMD technology. This novel tool has an additional acquisition system capable of storing the large amount of raw data generated during the FMD procedure (18, 19). This has vastly expanded data collection capacity so that it can continuously measure WSR and vessel diameter over an extended period of time to provide accurate, detailed, and simultaneous WSR-FMD stimulus-response data.
Here we report, for the first time, the use of this method. Our overall aim was to establish a new benchmark in WSR-FMD measurement. First, we wanted to establish the relevant variables derived from the measurement system. Second, because the WSR-FMD response in humans is currently unknown, we wanted to establish a “normal reference” by conducting a detailed analysis of the WSR-FMD response in a cohort of healthy young adults. Third, we wanted to determine which WSR variables were the best predictors of a normal FMD response. Fourth, we wanted to investigate the patterns of WSR during hyperemia and determine whether these influenced the FMD response. We had noticed two distinct WSR patterns during our pilot work: a “monophasic” pattern, where WSR increases sharply reaching its peak in one go during hyperemia; and a “biphasic” pattern, where WSR increases sharply followed by a slow increase before reaching its peak during hyperemia. These two distinct WSR patterns cannot readily be seen with conventional Doppler, and we wanted to know if the biphasic WSR pattern resulted in a greater magnitude of hyperemic WSR than the monophasic WSR pattern and, if so, whether this greater WSR produced a greater brachial artery vasodilatory response than the monophasic WSR pattern.
METHODS
Participants.
Thirty-three young individuals (27.5 ± 4.9 yr, 19 females) participated in this study. Of these, 15 participants were studied in Exeter, UK and 18 in Pisa, Italy. All were healthy, without hypertension, type 2 diabetes, dyslipidemia, or overt cardiovascular disease. The UK National Research Ethics Service South West Committee and the institutional ethics committee at University of Pisa approved all study procedures, and written informed consent was obtained from all participants.
Experimental procedures.
Participants arrived in our temperature-controlled laboratories after an overnight fast, had blood samples drawn for biochemical analysis, consumed a standardized meal, and rested for 20 min before initiation of the study protocol.
Brachial artery FMD was assessed noninvasively following established guidelines (6, 10, 24) and as previously described by us elsewhere (2, 8, 9). Briefly, participants lay supine on an examination bed with the right arm fixed in position and immobilized using a positioning pillow on a metal table. A small blood pressure cuff was placed around the proximal part of the forearm. A complete and open research system for ultrasound imaging and acquisition, the Ultrasound Advanced Open Platform (ULA-OP; Microelectronics Systems Design Laboratory, University of Florence, Florence, Italy), was connected to a high-frequency linear array transducer LA523 (Esaote, Florence, Italy) and used to obtain both B-mode and multigate velocity data from the brachial artery (4). Once the optimal ultrasound image was obtained, the transducer was carefully clamped to prevent movement during the procedure using a custom-designed transducer holder. If necessary, a microadjuster was used to obtain a precisely aligned image.
Baseline brachial artery image and blood velocity were recorded for 60 s. Reactive hyperemia was then induced by rapidly inflating the forearm cuff (AI6; Hokanson, Bellevue, WA) to 250 mmHg to occlude forearm blood flow for 5 min. At 5 min, the cuff was rapidly deflated. Recording was restarted 30 s before deflation and continued until 5 min following deflation.
In a subset of participants (n = 13), baseline brachial artery diameter measurements were repeated after a 15-min rest period, and endothelium-independent dilation was assessed using sublingual nitroglycerin spray (0.4 mg). A 60-s recording was started 9 min after administering the spray. We have previously found that measurement between 9 and 10 min after nitroglycerin administration captures maximal endothelial-independent vasodilation (unpublished observation).
Integrated FMD evaluation system.
An integrated system capable of estimating both stimulus (WSR) and response (diameter) during the FMD assessment was developed by suitably modifying the existing hardware and software of the ULA-OP as well as by developing a postprocessing software platform (19, 28). In particular, the ULA-OP offers much wider experimental possibilities than commercially available Doppler ultrasound systems as it can be reconfigured at run time and enables complete access to imaging data in every stage of the processing chain. The ULA-OP system also provides real-time imaging visualization functions in connection to a host personal computer. For the FMD data acquisition, an additional dedicated acquisition board was developed to store all raw (quadrature demodulated) echo data over a long time interval (up to 15 min). Furthermore, the ULA-OP system was recently upgraded to include the real-time measurement of arterial diameter (19).
In addition, a postprocessing software platform was developed to evaluate endothelial function (19). It is based on Matlab (The Mathwork, Natick, MA) and consists of several processing blocks that, starting from the baseband acquired data, compute B-mode images and multigate spectral Doppler profiles. B-mode processing organizes the data in lines and frames, interpolates them, applies two-dimensional spatial filters, and finally applies a logarithmic scale compression. Multigate spectral Doppler processing reconstructs the information about the distribution of blood velocities at different depths (256 up to 512 gates) through 256-point fast Fourier transforms; hence, the profiles are low-pass filtered to remove the low-frequency spectral components through a frequency domain mask. In cascade to the latter blocks, other specific processing blocks extract wall positions of the artery and its diameter through a first order absolute central moment algorithm (7), as well as WSR by a method that employs the direct measurement of the whole velocity profile (20). Next, diameter and WSR time trends are saved in output files that are finally loaded by a further Matlab interface that extracts the main parameters needed to investigate endothelial function.
Measurements of WSR and diameter.
The signal elaboration system extracts detailed WSR parameters for describing both magnitude and time course of WSR changes. Seven WSR magnitude parameters were analyzed: 1) WSR at baseline, 2) WSR during low flow, 3) WSR at peak hyperemia, 4) absolute WSR increase from baseline, 5) percent WSR increase from baseline, 6) area under the WSR curve until time to peak dilation (WSR AUCTTP), and 7) area under the WSR curve (WSR AUC), measured between cuff release and the point at which WSR returned to the baseline value. Four WSR time-course parameters were analyzed: 1) first slope of WSR increase during hyperemia (WSR SL1, an initial steep increase), 2) the second slope of WSR increase during hyperemia (WSR SL2, a gradual increase after the initial steep increase; the biphasic pattern only), 3) time to peak WSR (WSR TP), and 4) time to return to baseline WSR (WSR TB). The monophasic and biphasic patterns of WSR increase were defined as follows: 1) monophasic: the peak WSR value was reached with a single, continuous steep increase only; and 2) biphasic: the peak WSR value was reached with an initial steep increase followed by a gradual second increase.
The signal elaboration system also extracts detailed diameter parameters. Three diameter magnitude parameters were analyzed: 1) baseline diameter, 2) absolute diameter increase from baseline, and 3) percent diameter increase from baseline. Two diameter time-course parameters were analyzed: 1) time to peak diameter, and 2) time to return to baseline diameter, taken as the point at which diameter returns to its baseline values or plateaus.
Statistical analysis.
Data are presented as means ± SD for variables with a normal distribution and median (interquartile range) for variables with a skewed distribution. A partial correlation analysis was performed between WSR parameters and diameter changes (absolute and percentage) controling for the study center in the whole cohort. A stepwise multivariate regression analysis was also performed in the whole cohort to determine the strongest predictor(s) of WSR parameters that were significantly associated with both absolute and percent diameter changes in the partial correlation analysis (WSR at peak hyperemia, WSR AUC, and absolute WSR increase from baseline). The study center was also included as an independent variable in the model. Analysis of covariance (study center as a covariate) was used to examine the differences in variables between monophasic and biphasic groups. A log transformation was used for variables with skewed distribution before statistical analysis. All statistical analysis was conducted using IBM SPSS Statistics 22 (IBM, Armonk, NY). Significance was set at P < 0.05.
RESULTS
Selected baseline characteristics of the study participants are shown in Table 1. Body mass index (24.1 ± 3.0 vs. 21.7 ± 2.7 kg/m2) and systolic blood pressure (121.5 ± 7.5 vs. 108.6 ± 8.8 mmHg) were higher in male participants than female participants (both P < 0.05). Other participants’ characteristics were similar between males and females.
Table 1.
Selected characteristics of the study participants
| Values | |
|---|---|
| Participants, n | 33 |
| Age, yr | 27.5 ± 4.9 |
| Sex, male/female | 14/19 |
| BMI, kg/m2 | 22.7 ± 3.1 |
| Systolic BP, mmHg | 114.1 ± 10.4 |
| Diastolic BP, mmHg | 69.2 ± 6.6 |
| Heart rate, beats/min | 64.2 ± 9.5 |
Data are means ± SD or numbers. BMI, body mass index; BP, blood pressure.
Establishment of relevant WSR variables and their normal reference values.
Figure 1 exhibits examples of multigate Doppler spectral profiles obtained at different time points in the cardiac cycle at baseline and peak hyperemia, to illustrate the variability and complexity of the WSR events occurring during FMD. A schematic description of detailed WSR parameters extracted from the integrated FMD system is also presented in Fig. 2. In addition, parameters of WSR and diameter during brachial artery FMD and nitroglycerin-mediated dilation in a healthy young cohort are shown in Table 2.
Fig. 1.
Examples of multigate spectral Doppler profiles obtained from the Ultrasound Advanced Open Platform (ULA-OP) system, in red-to-white color scale, and the related mean frequency, which is overlaid in blue. A–C: multigate spectral Doppler profiles at baseline at different time points in the cardiac cycle from the same individual (A: early systole; B: peak systole; C: early diastole). D–F: multigate spectral Doppler profiles during peak hyperemia at the same time points in the cardiac cycle as A–C from the same participant (D: early systole; E: peak systole; F: early diastole). G–J: multigate spectral Doppler profiles during peak hyperemia at the same time points as D–F in the cardiac cycle but from a different participant (G: early systole; H: peak systole; J: early diastole). Note the asymmetry in velocity in some profiles (e.g., E, H, and J), blunt profile (F), and M-shaped profile (G). Also note the differences in spectral profile during cardiac cycle as well as between participants. Sub 1, subject 1; Sub 2, subject 2.
Fig. 2.
A schematic description of wall shear rate (WSR) parameters obtained from brachial artery flow-mediated dilation (FMD) assessment using continuous multigate Doppler and simultaneous diameter. Top: traces in light blue and red represent the peak and mean values of WSR, respectively. Bottom: traces in light blue and red represent the variations in diameter (due to cardiac cycle) and mean diameter of the brachial artery, respectively. SL1, first slope of wall shear rate increase; SL2, second slope of wall shear rate increase; AUCTTP, area under the curve until time to peak dilation (area shaded with turquoise); AUC, area under the curve until its return to baseline level (area shaded both with turquoise and grey); TP, time to peak value; TB, time to return to baseline value; Δ, changes.
Table 2.
Parameters of wall shear rate and diameter during brachial artery flow-mediated dilation and nitroglycerin-mediated dilation assessments in a healthy young cohort
| Values | Ranges | |
|---|---|---|
| Flow-mediated dilation (n = 33) | ||
| WSR magnitude parameters | ||
| WSR baseline, 1/s | 103.9 ± 54.9 | 17.5–222.6 |
| WSR low flow, 1/s | 32.6 ± 23.4 | −1.8–94.9 |
| WSR peak, 1/s | 594.9 ± 158.2 | 269.1–924.6 |
| WSR Δ, 1/s | 491.0 ± 160.7 | 188.4–845.0 |
| WSR %Δ, % | 564 (267–994) | 155.2–3,270.5 |
| WSR AUCTTP, AU | 13,414 ± 5,629 | 3,515–30,540 |
| WSR AUC, AU | 17,337 ± 6,724 | 4,064–39,473 |
| WSR time-course parameters | ||
| WSR SL1, 1/s2 | 79.9 ± 27.3 | 24.3–126.3 |
| WSR SL2, 1/s2* | 16.4 ± 8.8 | 0.33–35.4 |
| WSR TP, s | 12.2 ± 2.6 | 7.3–20.6 |
| WSR TB, s | 104.1 ± 36.6 | 57.5–182.5 |
| Diameter magnitude parameters | ||
| Diameter baseline, mm | 3.29 ± 0.45 | 2.57–4.24 |
| Diameter Δ, mm | 0.21 ± 0.10 | 0.03–0.47 |
| Diameter %Δ, % | 6.5 ± 3.5 | 0.95–14.8 |
| Diameter time-course parameters | ||
| Diameter TP, s | 55.3 ± 31.2 | 26.8–197.9 |
| Diameter TB, s | 113.6 ± 59.8 | 10.5–249.5 |
| Nitroglycerin-mediated dilation (n = 13) | ||
| WSR magnitude parameters | ||
| WSR baseline, 1/s | 58.6 ± 20.9 | 33.1–101.2 |
| WSR peak dilation, 1/s | 44.7 ± 24.6 | 17.9–87.8 |
| Diameter magnitude parameters | ||
| Baseline diameter, mm | 3.51 ± 0.56 | 2.48–4.30 |
| Diameter Δ, mm | 0.77 ± 0.15 | 0.56–1.10 |
| Diameter %Δ, % | 22.5 ± 4.8 | 13.8–28.4 |
Data are means ± SD for variables with normal distribution, median (interquartile range) for variables with skewed distribution, and range of each variable (minimum to maximum). WSR, wall shear rate; SL1, first slope of wall shear rate increase; SL2, second slope of wall shear rate increase; AUCTTP, area under the curve until time to peak dilation; AUC, area under the curve until its return to baseline value; TP, time to peak value; TB, time to return to baseline value; AU, arbitrary units; Δ, changes.
Obtained from 18 participants who showed the biphasic WSR increase response.
WSR variables that best predict a normal FMD response.
The results of partial correlation analysis between WSR parameters and diameter changes during brachial artery FMD in the whole cohort are shown in Table 3. WSR at peak hyperemia, WSR AUC, and absolute WSR increase were significantly associated with both absolute and percent diameter changes (all P < 0.05). WSR SL1 and WSR AUCTTP were significantly associated with percent diameter changes (both P < 0.05). We then performed a stepwise multivariate regression analysis to determine the strongest WSR predictor(s) of diameter changes during brachial artery FMD in the whole cohort. When WSR at peak hyperemia, WSR AUC, absolute WSR increase, and the study center were all included in the same model at the same time, WSR AUC was the best predictor of absolute brachial artery diameter change (β = 0.503, R2 = 0.25) and percent diameter change (β = 0.560, R2 = 0.31). Collinearity statistics (tolerance and variance inflation factor) did not indicate a collinearity problem in this model. These associations remained significant when including WSR at baseline or baseline brachial artery diameter in the model above.
Table 3.
Partial correlation analysis between parameters of WSR and diameter changes in the whole cohort
| Diameter Δ | Diameter %Δ | |
|---|---|---|
| WSR baseline | r = 0.18, P = 0.327 | r = 0.13, P = 0.470 |
| WSR low flow | r = 0.32, P = 0.072 | r = 0.32, P = 0.077 |
| WSR SL1 | r = 0.35, P = 0.053 | r = 0.36, P = 0.041 |
| WSR SL2* | r = 0.17, P = 0.524 | r = 0.24, P = 0.353 |
| WSR peak | r = 0.41, P = 0.020 | r = 0.47, P = 0.007 |
| WSR Δ | r = 0.41, P = 0.021 | r = 0.49, P = 0.004 |
| WSR %Δ | r = 0.04, P = 0.843 | r = 0.11, P = 0.558 |
| WSR AUCTTP | r = 0.34, P = 0.061 | r = 0.42, P = 0.017 |
| WSR AUC | r = 0.46, P = 0.008 | r = 0.56, P = 0.001 |
| WSR TP | r = 0.32, P = 0.072 | r = 0.32, P = 0.079 |
| WSR TB | r = 0.04, P = 0.109 | r = 0.35, P = 0.052 |
WSR, wall shear rate; SL1, first slope of wall shear rate increase; SL2, second slope of wall shear rate increase; AUCTTP, area under the curve until time to peak dilation; AUC, area under the curve until its return to baseline value; TP, time to peak value; TB, time to return to baseline value; Δ, changes.
Obtained from 18 participants who showed the biphasic WSR increase response.
Influence of the WSR patterns during hyperemia on the FMD response.
A schematic description of the monophasic and biphasic patterns of WSR increase is presented in Fig. 3. Table 4 shows the parameters of WSR and diameter during brachial artery FMD and nitroglycerin-mediated dilation stratified by monophasic and biphasic WSR increase patterns. During reactive hyperemia, we observed the monophasic pattern of WSR increase in 15 participants (9 females) and the biphasic pattern of WSR increase in 18 participants (10 females). Individuals with the biphasic pattern showed a significantly greater WSR SL1 than those with the monophasic pattern (P < 0.05). The parameters that were associated with diameter changes (WSR at peak hyperemia, WSR AUC, and absolute WSR increase from baseline) were significantly greater in the biphasic pattern than the monophasic pattern (Fig. 4, A–C). Similarly, individuals with the biphasic pattern showed a significantly greater WSR baseline and WSR AUCTTP than those with the monophasic pattern (all P < 0.05). WSR TP took longer in individuals with the biphasic pattern than in those with the monophasic pattern (P < 0.05). The absolute diameter increase following reactive hyperemia was significantly greater in individuals with the biphasic pattern than in individuals with the monophasic pattern (0.24 ± 0.10 vs. 0.17 ± 0.09 mm, P < 0.05; Fig. 5A). Percent diameter increase tended to be greater in individuals with the biphasic pattern than individuals with the monophasic pattern (7.6 ± 3.3 vs. 5.3 ± 3.5%, P = 0.08; Fig. 5B). However, to determine if WSR AUC stimulus (the strongest predictor of FMD) explained the difference in FMD response between the two groups, we used an analysis of covariance model that included WSR AUC. This analysis showed that there were no differences in absolute or percent brachial artery diameter change during FMD between the monophasic and biphasic groups when the WSR AUC stimulus was taken into account (data not shown). Following nitroglycerin spray, there were no differences in WSR or change in brachial artery diameter between the mono- and biphasic groups (Table 4).
Fig. 3.
Representative examples of biphasic (top) and monophasic (bottom) patterns of WSR increase and diameter changes during brachial artery FMD assessment. Traces in red represent a mean value of WSR or brachial artery diameter.
Table 4.
Parameters of wall shear rate and diameter during brachial artery flow-mediated dilation and nitroglycerin-mediated dilation assessments stratified by monophasic and biphasic patterns of wall shear rate increase
| Monophasic (n = 15) | Biphasic (n = 18) | |
|---|---|---|
| Flow-mediated dilation | ||
| WSR magnitude parameters | ||
| WSR baseline, 1/s | 89.6 ± 54.6 | 115.9 ± 53.8* |
| WSR low-flow, 1/s | 32.7 ± 23.8 | 32.5 ± 23.7 |
| WSR %Δ, % | 596 (181–1,374) | 552 (286–735) |
| WSR AUCTTP, au | 10,784 ± 4,961 | 15,607 ± 5,309* |
| WSR time-course parameters | ||
| WSR SL1, 1/s2 | 68.8 ± 30.0 | 89.1 ± 21.5* |
| WSR SL2, 1/s2 | — | 16.4 ± 8.8 |
| WSR TP, s | 11.1 ± 3.2 | 13.2 ± 1.6* |
| WSR TB, s | 92.1 ± 31.9 | 114.1 ± 38.1 |
| Diameter magnitude parameters | ||
| Diameter baseline, mm | 3.29 ± 0.44 | 3.28 ± 0.47 |
| Diameter time-course parameters | ||
| Diameter TP, s | 56.6 ± 43.8 | 54.2 ± 15.8 |
| Diameter TB, s | 110.8 ± 62.3 | 116.0 ± 59.3 |
| Nitroglycerin-mediated dilation (5 monophasic and 8 biphasic) | ||
| WSR magnitude parameters | ||
| WSR baseline, 1/s | 43.0 ± 10.2 | 68.4 ± 20.1* |
| WSR peak dilation, 1/s | 38.9 ± 21.5 | 48.3 ± 27.0 |
| Diameter magnitude parameters | ||
| Baseline diameter, mm | 3.38 ± 0.60 | 3.59 ± 0.56 |
| Diameter Δ, mm | 0.83 ± 0.10 | 0.74 ± 0.17 |
| Diameter %Δ, % | 25.0 ± 3.9 | 21.0 ± 4.8 |
Data are means ± SD for variables with normal distribution, and median (interquartile range) for variables with skewed distribution. WSR, wall shear rate; SL1, first slope of wall shear rate increase; SL2, second slope of wall shear rate increase; AUCTTP, area under the curve until time to peak dilation; AUC, area under the curve until its return to baseline value; TP, time to peak value; TB, time to return to baseline value; Δ, changes.
Significantly different from the monophasic group (P < 0.05).
Fig. 4.
WSR at peak hyperemia (A), WSR AUC (B), and absolute WSR increase from baseline (C) between monophasic (n = 15) and biphasic (n = 18) WSR increase patterns. Data are shown as means ± SE. *Significantly different from the monophasic group (P < 0.05).
Fig. 5.
Absolute diameter changes (A) and percent diameter changes (B) between monophasic (n = 15) and biphasic (n = 18) WSR increase patterns. Data are shown as means ± SE. *Significantly different from the monophasic group (P < 0.05).
DISCUSSION
Our main findings are that, using multigate spectral Doppler, we were able to acquire and extract seven WSR magnitude and four WSR time-course parameters over a long time period to more comprehensively characterize the FMD response. We were also able to derive the first WSR-FMD stimulus-response data in humans using this method to provide a first point of reference in healthy, young adults. Furthermore, we were able to show that in this cohort, the WSR AUC until its return to baseline was the strongest predictor of the brachial artery diameter change in response to hyperemia. For the first time, we were able to identify mono- and biphasic WSR stimulus patterns within our cohort that produced different magnitude of FMD response. However, these responses were not different when the strongest WSR predictor (WSR AUC) was statistically taken into account, illustrating the importance of knowing the WSR stimulus to correctly interpret the response.
Novel WSR variables, observations, and comments on their values.
We identified seven WSR magnitude variables and four WSR time-course variables that add novel measurements to the WSR-FMD response. These measurements took into account key phases of the FMD procedure and in combination with more traditional measurements provide a comprehensive characterization of FMD-WSR response. Our data show that WSR is reduced during cuff occlusion to about one-third that of baseline and that during reactive hyperemia, peak WSR is almost six-times greater compared with baseline and is over 18 times greater compared with WSR during cuff occlusion.
Our data also show that peak WSR was reached quickly (~12s) and preceded the time of peak dilation by an average of ~43 s. We also note that the time taken to reach peak WSR was relatively homogenous between subjects, whereas there was considerable interindividual variability in the time for diameter to reach peak diameter. The time taken for WSR to return to baseline (~104 s) preceded the time for diameter to return to baseline in 15 participants but in the remaining 18, arterial diameter returned to baseline before WSR. The physiological mechanisms underlying this variability are unknown and were beyond the scope of this study but putatively provide an opportunity to better understand the arterial response to the dynamic WSR changes during reactive hyperemia.
Multigate Doppler WSR AUC is a good predictor of FMD.
A key aim of the current study was to determine which WSR stimulus variable was the best predictor of the FMD response in healthy young adults. In a regression model that included the three WSR variables that were significantly associated with both absolute and relative diameter change (WSR peak, WSR AUC, and WSR absolute increase), we found that WSR AUC was the only predictor of both absolute (R2 = 0.25) and relative (R2 = 0.31) diameter changes, and this was the case irrespective of whether baseline diameter was included in the model.
Given that WSR AUC explains 25–31% of the variance in the regression model, it seems reasonable to conclude that this is an important predictor of the FMD response, while acknowledging the limitations of the R2 statistics. Other factors influencing diameter change likely include the stiffness of the brachial artery, mechanotransduction of WSR to the endothelium, cell signaling in response to the transduced stimulus, and the regulation of smooth muscle cell tone.
The association between WSR and the brachial artery FMD response has previously been reported in a cohort of young adults (14, 16, 17, 25). Using a similar cross-sectional study design as ours, Thijssen et al. found that WSR AUCTTP explained 14% of the variance in the FMD response in young healthy adults (25) and we found that WSR AUC using our method explained 25–31% of the variance in the FMD response. This suggests that the WSR stimulus is an important contributor to the FMD response, but whether FMD should be normalized to WSR is a contentious issue. Whereas some have suggested correcting (or normalizing) data by dividing FMD by WSR (17), recent guidelines (24) recommend reporting WSR and FMD together without corrections. Consistent with this, where WSR has been measured with multigate Doppler, we suggest that the normal stimulus-response characteristics should be reported as WSR AUC and FMD together and that the association between WSR AUC and FMD are determined using statistical models. It was not the purpose of this study to determine this relationship in characteristically different populations or diseased populations; instead, the data presented here, together with the novel method of data acquisition, provide a platform for future studies of this nature.
Mono- and biphasic WSR responses in healthy young adults.
An advantage of being able to continuously measure WSR was that we were able to observe, for the first time to our knowledge, two distinct patterns of WSR increase during reactive hyperemia: a monophasic and a biphasic WSR pattern that occurred between cuff release and peak WSR. Compared with the monophasic pattern, the biphasic pattern was associated with a greater magnitude as well as faster kinetics (steeper increase) of WSR increase during hyperemia. The biphasic pattern was also associated with a greater brachial artery vasodilatory response. However, when WSR AUC stimulus was statistically taken into account, the vasodilatory response was no longer different between mono- and biphasic groups; that is, the dilation was matched to the WSR stimulus. This finding shows that within an otherwise characteristically similar cohort, there are two distinct vasodilatory responses to reactive hyperemia but that vasodilation is, ultimately, matched to the WSR stimulus.
The two different WSR responses may allude to underlying physiological differences between the cohorts, although their significance is unknown. It is possible that different mechanisms regulating blood flow during hyperemia produce the two different stimulus-response relationships. There are two candidate sites for the regulation of FMD: first, the arterial endothelium above the cuff that is experiencing an increase in WSR during hyperemia; second, the downstream microvasculature below the cuff that has dilated during cuff occlusion and experiences a sudden increase in blood flow. During hyperemia, the endothelium of the brachial artery stimulates dilation to normalize the increased WSR brought about by hyperemic flow. Differences in mechanotransduction of WSR or differences in the paracrine response to the change in WSR could cause differences in the temporal pattern of WSR and magnitude of dilation. Microvascular function downstream of the cuff-occlusion site could also explain different WSR patterns (3, 13) because flow during reperfusion is influenced by downstream microvascular dilation (6). Microvessels also respond to reperfusion, including a vasoconstrictor response that likely influences upstream blood flow and, therefore, WSR. Consistent with this, we have previously shown distinct differences in the autoregulatory response to reperfusion by microvessels that temporally altered perfusion and oxygenation of tissue (1). Structural alterations in the microcirculation (21) can also influence its ability to respond to ischemia-reperfusion, which might influence the upstream WSR stimulus and explain differences between individuals. This is a first observation of two distinct WSR increase patterns, and we did not explore control mechanisms in the first instance. We also do not know if these WSR increase patterns are observed in characteristically different populations or diseased populations. Future studies will shed light on these issues and will ultimately determine whether there is value in determining two different WSR responses during the FMD assessment. However, the ability to measure different hyperemic responses itself may be useful for future physiological studies.
The flow velocity profile is not always parabolic and symmetrical.
One advantage of using multigate Doppler is that the flow velocity profile (normally assumed to be parabolic and symmetric, but which cannot be seen with conventional Doppler) can be seen in real time during data acquisition. One revelation from the current study was the heterogeneity in this flow profile within the cardiac cycle, under different flow states, and between subjects. During hyperemia, the shape of the parabola was seen to be blunt, M shaped, asymmetric, and symmetric (see examples in Fig. 1), and the shape of the parabola typically varied within the same cardiac cycle. Heterogeneity was also observed between subjects and was apparent at baseline, low flow, and during reactive hyperemia. These observations point to the importance of being able to detect flow velocity at different spatial points in the vessel to accurately measure WSR by extracting these local velocities.
Low-flow WSR and diameter during cuff occlusion.
We also found that all participants reduced WSR during cuff occlusion but only 17 exhibited reduced brachial artery diameter (1 remained similar to baseline and 15 exhibited increased brachial artery diameter). This finding suggests that the brachial artery response to low flow might be independent of WSR, at least in young, healthy adults.
Reduced vessel diameter immediately after cuff release and its influence on WSR increase and diameter response.
We noticed that immediately after cuff release, 24 participants showed a reduced brachial artery diameter and nine did not. To determine whether this reduction of diameter immediately after cuff release influenced the initial increase in WSR as well as the subsequent FMD response, we performed a sub-roup analysis of participants stratified by the presence or absence of reduced brachial diameter. We found that absolute WSR increase (496.1 ± 128.4 vs. 477.6 ± 134.4 1/s, P = 0.733) and WSR SL1 (80.3 ± 24.5 vs. 78.7 ± 25.5 1/s2, P = 0.874) were both similar between those with reduced brachial diameter and those without. Similarly, absolute diameter change (0.20 ± 0.10 vs. 0.22 ± 0.12 mm, P = 0.707) and relative diameter change (6.5 ± 3.9 vs. 6.8 ± 3.9%, P = 0.852) were not different between those with reduced diameter and those without. These observations indicate that the reduced brachial artery diameter immediately after cuff release does not play a major role in the initial increase in WSR as well as the subsequent FMD response, at least in our population of young, healthy adults. The reduction in diameter may be due to an acute pressure drop, a brief period of turbulent flow, and alteration of smooth muscle cell tone or some combination.
WSR response during nitroglycerin-mediated vasodilation.
There is a paucity of WSR data during the assessment of nitroglycerin-mediated vasodilation. It has been used as an endothelium-independent control test to ensure the validity of the FMD assessment. The application of nitroglycerin, thought to be an exogenous nitric oxide donor, reduces smooth muscle cell tone and induces vasodilation. As such, this has been considered WSR independent and thus that there is no clear benefit for acquiring WSR data during the assessment. Our observations support the position that nitroglycerin-mediated vasodilation is WSR independent, because WSR was slightly reduced (Table 2) at the time that peak diameter occurred following administration of nitroglycerin. Reduced WSR is likely a result of increased brachial artery diameter at the time of measurement and suggests that nitroglycerin administration overrides any effect of altered WSR.
Implications from this study.
Our study highlights the importance of being able to measure the WSR stimulus as well as the vasodilatory response to hyperemia. An accurate estimation of WSR close to the arterial wall by conventional Doppler ultrasound, especially during reactive hyperemia, is inherently difficult, further exacerbated by the uncertainties associated with the assumptions used to estimate WSR (especially that blood flow maintains a parabolic profile). We have shown that multigate Doppler overcomes these limitations by measuring blood flow velocity from near-to-far wall and by directly estimating WSR close to the arterial wall. By integrating these measurements with continuous and simultaneous measurement of arterial diameter, we were able to generate detailed information about the WSR stimulus and FMD response that has not been seen previously. For example, the system enabled us to reveal the monophasic and biphasic WSR patterns and allowed us to determine the most important WSR predictor of FMD. We are also able to characterize the normal WSR-FMD relationship in our cohort of healthy young adults, establishing a reference for these measurements. As such, we have demonstrated the usefulness of multigate Doppler as a modality for measuring arterial WSR, in this instance integrated into the ULA-OP system.
Overall, our study represents a technical advance that enables comprehensive WSR-FMD stimulus-response measurement within an integrated ultrasound system. The need to measure WSR continues to cause researchers to rely heavily on imprecise WSR data derived from vessel center-line peak velocity, creating uncertainties for the accurate interpretation of the FMD response. In a clinical setting, measurement of WSR has found little utility. Our broader aim is to provide a better tool for researchers and clinicians to augment the accuracy and usefulness of WSR-FMD measurement. This has the potential to expand current understanding of vascular physiology and pathophysiology, vascular aging, and the vascular response to interventions. In a clinical setting, this has the potential to improve clinical evaluation and management of patients with many diseases that involve blood vessels.
Limitations.
We did not assess brachial artery stiffness, which has been reported to affect the magnitude of FMD response (29) and may have contributed to the differences seen here. In addition, due to the cross-sectional nature of this study, we cannot infer any causation from our results. Finally, because multigate spectral Doppler is an extension of pulsed-wave Doppler, it has the same limitations of pulsed-wave Doppler; for example, analysis limited to the axial velocity component (28) or possible velocity detection difficulties in the presence of high-level clutter.
Conclusion.
Overall, our results demonstrate the importance of being able to accurately determine a simultaneous WSR-FMD measurement, provide a reference for the normal WSR-FMD response, and present a number of novel variables that may enable better understanding of vascular physiology and pathology.
GRANTS
This study was supported by the European Union’s Seventh Framework Programme (FP7/2007-2013) for the Innovative Medicine Initiative under Grant Agreement No. IMI/115006 (the SUMMIT consortium), in part by the National Institute of Health Research (NIHR) Exeter Clinical Research Facility, and by the Italian Ministry of University and Research (MIUR, Project PRIN 2010–2011). The views expressed are those of the authors and not necessarily those of the National Health Service, the NIHR, the Department of Health, or the MIUR.
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the authors.
AUTHOR CONTRIBUTIONS
K.A., P.T., A.C.S., P.E.G., and C.P. conceived and designed research; K.A., S.S., F.C., C.M., C.E.T., and P.E.G. performed experiments; K.A., A.R., and P.E.G. analyzed data; K.A., P.T., A.C.S., P.E.G., and C.P. interpreted results of experiments; K.A. and A.R. prepared figures; K.A., A.R., and P.E.G. drafted manuscript; K.A., S.S., A.R., P.T., F.C., C.M., C.E.T., A.C.S., P.E.G., and C.P. edited and revised manuscript; K.A., S.S., A.R., P.T., F.C., C.M., C.E.T., A.C.S., P.E.G., and C.P. approved final version of manuscript.
ACKNOWLEDGMENTS
We thank the staff of the Diabetes and Vascular Medicine Research Centre, University of Exeter Medical School, for valuable assistance to carry out this study.
REFERENCES
- 1.Adingupu DD, Thorn CE, Casanova F, Elyas S, Gooding K, Gilchrist M, Aizawa K, Gates PE, Shore AC, Strain DW. Blood oxygen saturation after ischemia is altered with abnormal microvascular reperfusion. Microcirculation 22: 294–305, 2015. doi: 10.1111/micc.12198. [DOI] [PubMed] [Google Scholar]
- 2.Aizawa K, Elyas S, Adingupu DD, Casanova F, Gooding KM, Strain WD, Shore AC, Gates PE. Reactivity to low-flow as a potential determinant for brachial artery flow-mediated vasodilatation. Physiol Rep 4: e12808, 2016. doi: 10.14814/phy2.12808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Anderson TJ, Charbonneau F, Title LM, Buithieu J, Rose MS, Conradson H, Hildebrand K, Fung M, Verma S, Lonn EM. Microvascular function predicts cardiovascular events in primary prevention: long-term results from the Firefighters and Their Endothelium (FATE) study. Circulation 123: 163–169, 2011. doi: 10.1161/CIRCULATIONAHA.110.953653. [DOI] [PubMed] [Google Scholar]
- 4.Boni E, Bassi L, Dallai A, Guidi F, Ramalli A, Ricci S, Housden J, Tortoli P. A reconfigurable and programmable FPGA-based system for nonstandard ultrasound methods. IEEE Trans Ultrason Ferroelectr Freq Control 59: 1378–1385, 2012. doi: 10.1109/TUFFC.2012.2338. [DOI] [PubMed] [Google Scholar]
- 5.Celermajer DS, Sorensen KE, Gooch VM, Spiegelhalter DJ, Miller OI, Sullivan ID, Lloyd JK, Deanfield JE. Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis. Lancet 340: 1111–1115, 1992. doi: 10.1016/0140-6736(92)93147-F. [DOI] [PubMed] [Google Scholar]
- 6.Corretti MC, Anderson TJ, Benjamin EJ, Celermajer D, Charbonneau F, Creager MA, Deanfield J, Drexler H, Gerhard-Herman M, Herrington D, Vallance P, Vita J, Vogel R; International Brachial Artery Reactivity Task Force . Guidelines for the ultrasound assessment of endothelial-dependent flow-mediated vasodilation of the brachial artery: a report of the International Brachial Artery Reactivity Task Force. J Am Coll Cardiol 39: 257–265, 2002. doi: 10.1016/S0735-1097(01)01746-6. [DOI] [PubMed] [Google Scholar]
- 7.Demi M, Paterni M, Benassi A. The first absolute central moment in low-level image processing. Comput Vis Image Underst 80: 57–87, 2000. doi: 10.1006/cviu.2000.0861. [DOI] [Google Scholar]
- 8.Gates PE, Boucher ML, Silver AE, Monahan KD, Seals DR. Impaired flow-mediated dilation with age is not explained by L-arginine bioavailability or endothelial asymmetric dimethylarginine protein expression. J Appl Physiol (1985) 102: 63–71, 2007. doi: 10.1152/japplphysiol.00660.2006. [DOI] [PubMed] [Google Scholar]
- 9.Gilchrist M, Winyard PG, Aizawa K, Anning C, Shore A, Benjamin N. Effect of dietary nitrate on blood pressure, endothelial function, and insulin sensitivity in type 2 diabetes. Free Radic Biol Med 60: 89–97, 2013. doi: 10.1016/j.freeradbiomed.2013.01.024. [DOI] [PubMed] [Google Scholar]
- 10.Harris RA, Nishiyama SK, Wray DW, Richardson RS. Ultrasound assessment of flow-mediated dilation. Hypertension 55: 1075–1085, 2010. doi: 10.1161/HYPERTENSIONAHA.110.150821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Irace C, Tripolino C, Scavelli FB, Carallo C, Gnasso A. Brachial low-flow-mediated constriction is associated with delayed brachial flow-mediated dilation. J Atheroscler Thromb 23: 355–363, 2016. doi: 10.5551/jat.32060. [DOI] [PubMed] [Google Scholar]
- 12.Krams R, Bambi G, Guidi F, Helderman F, van der Steen AF, Tortoli P. Effect of vessel curvature on Doppler derived velocity profiles and fluid flow. Ultrasound Med Biol 31: 663–671, 2005. doi: 10.1016/j.ultrasmedbio.2005.01.011. [DOI] [PubMed] [Google Scholar]
- 13.Mitchell GF, Parise H, Vita JA, Larson MG, Warner E, Keaney JF Jr, Keyes MJ, Levy D, Vasan RS, Benjamin EJ. Local shear stress and brachial artery flow-mediated dilation: the Framingham Heart Study. Hypertension 44: 134–139, 2004. doi: 10.1161/01.HYP.0000137305.77635.68. [DOI] [PubMed] [Google Scholar]
- 14.Padilla J, Johnson BD, Newcomer SC, Wilhite DP, Mickleborough TD, Fly AD, Mather KJ, Wallace JP. Adjusting flow-mediated dilation for shear stress stimulus allows demonstration of endothelial dysfunction in a population with moderate cardiovascular risk. J Vasc Res 46: 592–600, 2009. doi: 10.1159/000226227. [DOI] [PubMed] [Google Scholar]
- 15.Papaioannou TG, Karatzis EN, Vavuranakis M, Lekakis JP, Stefanadis C. Assessment of vascular wall shear stress and implications for atherosclerotic disease. Int J Cardiol 113: 12–18, 2006. doi: 10.1016/j.ijcard.2006.03.035. [DOI] [PubMed] [Google Scholar]
- 16.Pyke KE, Dwyer EM, Tschakovsky ME. Impact of controlling shear rate on flow-mediated dilation responses in the brachial artery of humans. J Appl Physiol (1985) 97: 499–508, 2004. doi: 10.1152/japplphysiol.01245.2003. [DOI] [PubMed] [Google Scholar]
- 17.Pyke KE, Tschakovsky ME. Peak vs. total reactive hyperemia: which determines the magnitude of flow-mediated dilation? J Appl Physiol (1985) 102: 1510–1519, 2007. doi: 10.1152/japplphysiol.01024.2006. [DOI] [PubMed] [Google Scholar]
- 18.Ramalli A, Bassi L, Lenge M, Palombo C, Aizawa K, Tortoli P. An integrated system for the evaluation of flow mediated dilation. IEEE Int Conf Acoust Speech Signal Proc, 2014, p. 5145–5148. doi: 10.1109/ICASSP.2014.6854583. [DOI] [Google Scholar]
- 19.Ramalli A, Byra M, Dallai A, Palombo C, Aizawa K, Sbragi S, Shore A, Tortoli P. A multiparametric approach integrating vessel diameter, wall shear rate and physiologic signals for optimized flow mediated dilation studies. IEEE Int Ultra Sym, 2015,. p. 1–4. [Google Scholar]
- 20.Ricci S, Swillens A, Ramalli A, Segers P, Tortoli P. Wall shear rate measurement: Validation of a new method through multi-physics simulations. IEEE Trans Ultrason Ferroelectr Freq Control 64: 66–77, 2017. doi: 10.1109/TUFFC.2016.2608442. [DOI] [PubMed] [Google Scholar]
- 21.Schiffrin EL. Remodeling of resistance arteries in essential hypertension and effects of antihypertensive treatment. Am J Hypertens 17: 1192–1200, 2004. doi: 10.1016/j.amjhyper.2004.05.023. [DOI] [PubMed] [Google Scholar]
- 22.Silber HA, Ouyang P, Bluemke DA, Gupta SN, Foo TK, Lima JA. Why is flow-mediated dilation dependent on arterial size? Assessment of the shear stimulus using phase-contrast magnetic resonance imaging. Am J Physiol Heart Circ Physiol 288: H822–H828, 2005. doi: 10.1152/ajpheart.00612.2004. [DOI] [PubMed] [Google Scholar]
- 23.Stoner L, Sabatier ML, Young JM. Examination of possible flow turbulence during flow-mediated dilation testing. Open J Med Imag 1: 1–8, 2011. doi: 10.4236/ojmi.2011.11001. [DOI] [Google Scholar]
- 24.Thijssen DH, Black MA, Pyke KE, Padilla J, Atkinson G, Harris RA, Parker B, Widlansky ME, Tschakovsky ME, Green DJ. Assessment of flow-mediated dilation in humans: a methodological and physiological guideline. Am J Physiol Heart Circ Physiol 300: H2–H12, 2011. doi: 10.1152/ajpheart.00471.2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Thijssen DH, Bullens LM, van Bemmel MM, Dawson EA, Hopkins N, Tinken TM, Black MA, Hopman MT, Cable NT, Green DJ. Does arterial shear explain the magnitude of flow-mediated dilation?: a comparison between young and older humans. Am J Physiol Heart Circ Physiol 296: H57–H64, 2009. doi: 10.1152/ajpheart.00980.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Thijssen DH, van Bemmel MM, Bullens LM, Dawson EA, Hopkins ND, Tinken TM, Black MA, Hopman MT, Cable NT, Green DJ. The impact of baseline diameter on flow-mediated dilation differs in young and older humans. Am J Physiol Heart Circ Physiol 295: H1594–H1598, 2008. doi: 10.1152/ajpheart.00669.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Tortoli P, Morganti T, Bambi G, Palombo C, Ramnarine KV. Noninvasive simultaneous assessment of wall shear rate and wall distension in carotid arteries. Ultrasound Med Biol 32: 1661–1670, 2006. doi: 10.1016/j.ultrasmedbio.2006.07.023. [DOI] [PubMed] [Google Scholar]
- 28.Tortoli P, Palombo C, Ghiadoni L, Bini G, Francalanci L. Simultaneous ultrasound assessment of brachial artery shear stimulus and flow-mediated dilation during reactive hyperemia. Ultrasound Med Biol 37: 1561–1570, 2011. doi: 10.1016/j.ultrasmedbio.2011.06.001. [DOI] [PubMed] [Google Scholar]
- 29.Witte DR, van der Graaf Y, Grobbee DE, Bots ML; SMART Study Group . Measurement of flow-mediated dilatation of the brachial artery is affected by local elastic vessel wall properties in high-risk patients. Atherosclerosis 182: 323–330, 2005. doi: 10.1016/j.atherosclerosis.2005.02.012. [DOI] [PubMed] [Google Scholar]