Abstract

Background:

Published normative data of noninvasive blood pressures (BPs) and autonomic modulations have been primarily derived from the finger arteriole using the Finapres (Ohmeda Co., Englewood, CO), a device that is no longer manufactured. Currently, beat-to-beat BP are obtained from the radial artery using the Colin tonometer.

Methods:

We compared BP and autonomic parameters in a crossover design between the two devices in 29 subjects during seated rest and a 0.1-Hz breathing protocol. In addition, we tested whether finger arteriolar BP differences were due to pressure changes exerted by the radial tonometer.

Results:

Uniformly, BP measured at the radial artery were significantly higher than those from the finger arteriole. Radial BP (106 ± 19.5 mm Hg) were higher than finger arteriolar BP (95.8 ± 13.7 mm Hg) (P < .005). Tonometric baroreflex sensitivity (BRS) (24.0 ± 18 msec/mm Hg) was higher compared to photoplethysmographic BRS (12.0 ± 7.7 msec/mm Hg; P < .0003). Systolic BP (radial artery) (115 ± 25 mm Hg) were higher compared to finger arteriolar BP (97.7 ± 19 mm Hg; P < .0025) during breathing, as was BRS (25.9 ± 11.6 msec/mm Hg v 21.5 ± 11.6 msec/mm Hg; P < .05). Differences in the low frequency systolic BP (LFSBP), representative of sympathetic vasomotor modulation, between the two methods, whether absolute, normalized, or log-transformed were not observed.

Conclusions:

There were no differences in arteriolar BP values in the presence or absence of radial artery tonometric pressure. These findings indicate that differences exist in systolic BP and BRS using the tonometer (radial artery) versus the Finapres (Ohmeda Co.) (finger arteriole). Furthermore, these differences are not due to pressure exerted by the radial artery tonometer that supplies blood to the finger arteriole. Am J Hypertens 2003; 16:371–374 @ 2003 American Journal of Hypertension, Ltd.

Noninvasive techniques in the measurement of baroreflex sensitivity (BRS) have gained popularity, due in part to the availability of beat-to-beat blood pressure (BP) monitors using photoplethysmographic and tonometric principles, both of which have been shown to be practical and safe. Two methods have predominantly been used to determine BRS: time–domain and frequency–domain analyses. The time–domain method calculates the slope of the relationship between spontaneous changes in systolic BP (SBP) and R–R intervals, whereas the frequency–domain method calculates the power of the heart rate variability (HRV) divided by the power of BP variability for both low-frequency (LF) and high-frequency (HF) modulations. Both methods are dependent on the mechanical properties of the vessel and these pressure changes vary from one site of the arterial tree to another.1 For example, the photoplethysmographic device (Finapres, Ohmeda Co.) derives SBP changes from the finger arteriole, whereas the tonometric device (Colin tonometer, Colin Medical Instruments Corp., San Antonio, TX) derives SBP changes from the radial artery. Thus, BP measured at these sites may differ significantly from one another as the systolic portion of the pulse wave dampens as the vessel narrows.1 Therefore, the aim of this investigation was threefold: 1) to compare BRS measurements using both time–and frequency–domain methods derived from the radial artery (wrist) and finger arteriole (middle phalanx of the middle finger); 2) to compare SBP and the LF component of SBP variability (LFSBP) measured at both sites; and 3) to investigate the effect, if any, of radial artery pressure on the downstream finger arteriolar pressure. All measurements were obtained simultaneously at rest and during deep breathing (0.1 Hz).

Methods

Subjects

Twenty-nine male and female subjects between the ages of 21 to 38 years (mean age 32.6 ± 3.6 years) were tested. Before enrollment into the study, all subjects were provided an explanation of the purpose and procedures of the protocol. In accordance with the Columbia University Institutional Review Board, all subjects provided written informed consent before testing. All subjects were normotensive and were not taking medications that would interfere with autonomic function. Upon arrival to the laboratory, subjects were introduced to the BP sensing devices and were instrumented. An applanation tonometer (Colin Medical Instruments) was attached to the subject's right radial artery and the photoplethysmographic device (Finapres 2300, Ohmeda Co.) was attached to the middle phalanx of the middle finger of the right hand and was supported by an adjustable table to the level of the atria. Electrocardiograms (ECG) (Max 2000, Marquette Instruments, Marquette, WI) were continuously monitored. A temperature sensor was placed under the right nostril of the subject and recorded respiratory excursions. Subjects remained seated and equilibrated to the environment for 15 min before data collection. After the equilibration period, data were collected and all signals (ECG, beat-to-beat radial, and finger BP and respiratory waves) were collected at 500 Hz and channeled into a computer (Dell computers, Austin, TX) through an analog-to-digital (ATMIO-16, National Instruments, Austin, TX) conversion board. Baroreflex and autonomic data acquisition and analyses were carried out off-line in accordance with the recommendations of the European and North American Task Force2 using customized and validated LabView programs (LabView 6.0, National Instruments). After 5 min of resting data were acquired, subjects were instructed to breathe at six breaths per minute (0.1 Hz) while observing a timer and auditory cues for an additional 5 min.

Blood pressures in the radial artery and finger arteriole were measured on the same side. To test for a pressure effect, a group of 12 subjects had 5 min of resting BP measured with the tonometer on the wrist with the pressure on, and once with the tonometer still on the wrist without the pressure. Moreover, to determine whether the radial tonometry pressures dampened LFSBP modulation in the finger arteriole, we also compared LFSBP modulation between the two methods. A counterbalanced sequence was used to avoid an order effect.

Baroreflex sensitivity assessment

Time–domain method: sequence method

The digitized R–R intervals and in-phase systolic peaks obtained after the spectral analyses were used to determine baroreflex by a modification of the sequence technique.3 Linear regressions were performed on any episode of three consecutive beats with R–R interval lags in the same direction (either up or down). The resultant correlation coefficients were averaged and yielded an index considered to be representative of BRS.4 These spontaneous beat-by-beat interactions of SBP and R–R interval have been shown to reflect true baroreflex events rather random interactions.5,6,7

Frequency–domain method: alpha index

The baroreflex alpha index assessment was measured on the same epochs as the sequence method. Values were determined by the ratio between the band-pass filtered R–R interval values and filtered SBP values in the frequency region wherein the phase difference was zero. This method is based on the calculation the transfer function or modulus between SBP and pulse interval powers in the LF band (0.04 to 0.14 Hz) where the coherence between the two signals is highest. The cross-spectrum between pulse interval and SBP (alpha index) was computed in the frequency regions and defined as LF (0.1 Hz) and HF (0.3 Hz) bandwidths. Validity and reliability of this method have been shown to be qualitatively and quantitatively similar to the sequence method using noninvasive and invasive techniques.5,6

Autonomic modulation assessment

Sympathetic modulation (LFSBP)

Beat-to-beat BP were derived from the continuous radial (Colin tonometer, Colin Medical Instruments) and finger arteriolar BP (Finapres, Ohmeda Co.). Systolic BP peaks were detected by an established peak detection algorithm. For both BP waveforms, the sequence of the systolic peaks (SPs) was interpolated to provide a continuous data stream and the resultant SP data were detrended using a robust locally weighted regression procedure. Data were split in sections; each section was tapered using a split cosine window. Finally, the data were transformed into frequency spectra using fast Fourier transform algorithms and smoothed across blocks of frequencies to produce a spectrum. The power of the SPs was calculated by measuring the area (integration) under the peak of the spectra. Power spectra within the 0.04- to 0.15-Hz range were defined as LF components and are considered to be representative of sympathetic vasomotor modulation.7,8,9,10,11 The validity and reliability of these noninvasive pressure-monitoring systems has been accurately demonstrated by linearly correlating pulsatile intra-arterial pressure monitoring systems (correlation coefficients ranged from 0.89 to 0.98).12,13

Statistical analyses

Dependent variables obtained from both systems were compared using paired t tests. All statistical procedures were performed using SPSS 10.5 software (SPSS Inc., Chicago, IL). A probability level of P < .05 was used for statistical significance.

Results

Physical characteristics of the subjects are listed in Table 1. Data are presented as mean (SD). A total of 29 subjects completed this study. The gender breakdown consisted of 14 female and 15 male subjects; the racial distribution was composed of whites (n = 11), African Americans (n = 10), and Asians (n = 8).

Table 1

Physical characteristics of subjects (n = 29)

Variable  
Age (y) 32.6 (3.6) 
Height (cm) 171.5 (6.7) 
Weight (kg) 67 (10.4) 
BMI [wt(kg)/ht(m2)] 23.1 (2.9) 
HR (beats/min) 68.8 (8.8) 
SBP (brachial) (mm Hg) 114.9 (12.6) 
Variable  
Age (y) 32.6 (3.6) 
Height (cm) 171.5 (6.7) 
Weight (kg) 67 (10.4) 
BMI [wt(kg)/ht(m2)] 23.1 (2.9) 
HR (beats/min) 68.8 (8.8) 
SBP (brachial) (mm Hg) 114.9 (12.6) 

BMI = body mass index; HR = heart rate; SBP = systolic blood pressure.

Data are mean and standard deviation (SD).

Rest

Radial systolic BP (106 ± 19 mm Hg) were significantly higher than finger arteriolar BP (95.8 ± 13.7 mm Hg; P < .005) during rest, as were radial diastolic BP (76.2 ± 9.6 mm Hg) versus finger arteriolar diastolic BP (73.8 ± 8.7 mm Hg; P < .00). No significant differences were observed in either method for BRS using the sequence method. Alpha indices (frequency–domain method) were significantly higher for the LF modulation bandwidth associated with radial tonometry (24.0 ± 18.2 msec/mm Hg) compared to finger photoplethysmography (12.0 ± 7.7 msec/mm Hg) (P < .0003). No significant differences were observed for any of the LFSBP modulations between the two methods, whether absolute, log-transformed, or normalized (Table 2).

Table 2

Mean, standard deviations, and probability levels for blood pressures and autonomic indices for Colin tonometry and Finapres photoplethysmography while at rest and during deep breathing (0.1 Hz)

Variable Tonometry Photoplethysmography P 
Rest    
SBP (mm Hg) 106 (19.5) 95.8 (13.7) .005 
DBP (mm Hg) 76.2 (9.6) 73.8 (8.7) .0001 
Alpha-index (msec/mm Hg) 24 (18) 12 (7.7) .0003 
LFSBP (a) 444 (299) 457 (283) .73 
LFSBP (ln) 5.7 (.84) 5.6 (0.98) .28 
LFSBP (nu) 81.9 (6.2) 80.7 (7.4) .09 
Deep breathing    
SBP (mm Hg) 115 (25) 97.7 (19) .0025 
Alpha Index (msec/mm Hg) 25.9 (11.6) 21.5 (11.6) .05 
Variable Tonometry Photoplethysmography P 
Rest    
SBP (mm Hg) 106 (19.5) 95.8 (13.7) .005 
DBP (mm Hg) 76.2 (9.6) 73.8 (8.7) .0001 
Alpha-index (msec/mm Hg) 24 (18) 12 (7.7) .0003 
LFSBP (a) 444 (299) 457 (283) .73 
LFSBP (ln) 5.7 (.84) 5.6 (0.98) .28 
LFSBP (nu) 81.9 (6.2) 80.7 (7.4) .09 
Deep breathing    
SBP (mm Hg) 115 (25) 97.7 (19) .0025 
Alpha Index (msec/mm Hg) 25.9 (11.6) 21.5 (11.6) .05 

DBP = diastolic blood pressure; LF = low-frequency; a = absolute; In = log transformed; nu = normalized; other abbreviations as in Table 1.

Deep breathing

Breathing at 0.1 Hz (0 breaths per minute), resulted in significant differences in SBP with radial tonometry presenting higher values (115 ± 25 mm Hg) compared to finger BP photoplethysmography (97.7 ± 19 mm Hg) (P < .0025). There were no differences in BRS using the sequence method between the two methods. The alpha index was significantly higher measured by radial tonometry compared to photoplethysmography (25.9 ± 11.6 msec/mm Hg v 21.5 ± 11.6 msec/mm Hg; P < .05). No differences were observed in the determination of LFSBP autonomic modulation between the two methods, whether absolute or normalized (Table 2).

Carry-over effects

There were no carry-over effects in BP changes between radial tonometric pressures and finger arteriolar pressures, whether in the presence of pressure or in the absence of radial tonometric pressure.

Discussion

Our findings confirm that differences exist in SBP and BRS values when comparing BP measurements of the radial artery to the finger arteriole. Moreover, these differences are not due to pressure exerted by the radial tonometer that could affect the downstream pressure in the finger arteriole. The use of these noninvasive BP analyses techniques has become popular because of their noninvasive nature and ease of application. The information derived from these analyses has prognostic value in determining autonomic and baroreflex sensitivity status, both of which are predictors of future cardiovascular morbidity and mortality.14,15 The differences in SBP values found in this investigation may be explained by the mechanical properties of the arteries that vary (radial artery versus finger arteriole) along the arterial tree.16 Specifically, Fourier analysis indicates that the BP curve may be divided into two components: a steady component and a pulsatile component. The steady component is related to the caliber of the small arteries, whereas the pulsatile component to the mechanical properties of the large arteries.16 Therefore, the arterial pressure contour changes as the wave is transmitted down the arterial tree. This, in turn, dampens both the oscillatory components of the pulse wave and systolic portion of the pressure wave.17 The changes in the above properties may be largely attributed to the viscoelastic properties of the arterial walls.17 These properties help explain the differences observed in the results of this investigation for SBP and the alpha indices for different branches of the arterial tree. It is these differences in viscoelastic properties between radial arteries and finger arterioles that result in dissimilar SBP. These BP differences elicit different magnitudes of compensatory responses as seen in the current baroreflex modulations (alpha index). These differences ought to be considered in light of several recent published reports on normative data of baroreflex modulations in healthy men and women.18,19,20,21 The normative data reported in all these studies were derived from finger arterioles exclusively. Currently the finger arteriole BP monitor (Finapres, Ohmeda Co.) is no longer manufactured and available. Therefore, scientists and clinicians interested in the acquisition of noninvasive beat-to-beat BP and associated autonomic and baroreflex sensitivity analyses will most likely use radial tonometry and should take these methodologic differences into consideration. In summary, findings in the current investigation revealed significant differences in SBP and BRS derived from the radial artery versus the finger arteriole. However, no differences in LFSBP modulations between the two methods were found. This is encouraging as LFSBP modulations derived from either system are frequently used as markers of sympathetic vasomotor activity.

References

1.
Darne’
B
,
Girerd
X
,
Safar
ME
,
Cambien
F
,
Guize
L
:
Pulsatile versus steady component of blood pressure: a cross-sectional and a prospective analysis on cardiovascular mortality
.
Hypertension
 
1989
;
13
:
392
400
.
2.
Task Force of the European Society of Cardiology and North American Society of Pacing and Electrophysiology
Heart rate variability, standards of measurement, physiological interpretation, and clinical use
.
Circulation
 
1996
.
1043
1065
.
3.
Bertinieri
G
,
Dirienzo
M
,
Cavallazzi
A
:
Evaluation of baroreceptor reflex by blood pressure monitoring in unanesthetized cats
.
Am J Physiol
 
1988
;
254
(
Heart Circ Physiol 23
):
H377
H383
.
4.
Blaber
AP
,
Yamamoto
Y
,
Hughson
RL
:
Methodology of spontaneous baroreflex relationship assessed by surrogate data analysis
.
Am J Physiol
 
1995
;
268
(
Heart Circ Physiol 37
):
H1682
H1687
.
5.
Akselrod
S
,
Gordon
D
,
Madwed
J
,
Snidman
NC
,
Shannon
DC
,
Cohen
RJ
:
Hemodynamic regulation: Investigation by spectral analysis
.
Am J Physiol
 
1985
;
249
:
H867
H875
.
6.
Saul
JP
,
Berger
RD
,
Chen
MH
,
Cohen
RJ
:
Transfer function analysis of autonomic regulation. II Respiratory sinus arrhythmia
.
Am J Physiol
 
1989
;
256
:
H153
H161
.
7.
Saul
JP
,
Berger
RD
,
Albrecht
P
,
Stein
SP
,
Chen
MH
,
Cohen
RJ
:
Transfer function analysis of the circulation: unique insights into cardiovascular regulation
.
J Physiol
 
1991
;
30
:
H1231
H1245
.
8.
Pagani
M
,
Pizzinelli
P
,
Furlan
R
,
Guzetti
S
,
Rimoldi
O
,
Malliani
A
:
A sympathetic hypertensive reflex from the heart of conscious dog
.
Clin Sci
 
1989
;
61
:
181
183
.
9.
Pagani
M
,
Furlan
R
,
Dell’Orto
S
,
Pizzinelli
P
,
Lanzi
G
,
Baselli
G
,
Santoli
C
,
Cerutti
S
,
Lombardi
F
,
Malliani
A
:
Continuous recording of direct high fidelity arterial pressure and electrocardiogram in ambulant patients
.
Cardiovascular Res
 
1986
;
20
:
384
388
.
10.
Goldstein
D
,
Horwitz
H
,
Keiser
R
:
Comparison of techniques for measuring baroreflex sensitivity in man
.
Circulation
 
1982
;
66
:
432
439
.
11.
Smyth
HS
,
Sleight
P
,
Pickering
GW
:
Reflex regulation of arterial pressure during sleep in man: a quantitative method of assessing baroreflex sensitivity
.
Circ Res
 
1969
;
24
:
109
121
.
12.
Sato
T
,
Nishinaga
M
,
Kawamoto
A
,
Ozawa
T
,
Takasutji
H
:
Accuracy of a continuous blood pressure monitor based on arterial tonometry
.
Hypertension
 
1993
;
21
:
866
874
.
13.
Smith
NT
,
Wesseling
KH
,
de Witt
B
:
Evaluation of two prototype devices producing noninvasive, pulsatile calibrated blood pressure measurement of a finger
.
J Clin Monitoring
 
1985
;
1
:
17
29
.
14.
Ross
R
:
Atherosclerosis: A problem of the biology of arterial wall cells and their interactions with blood components
.
Arteriosclerosis
 
1981
.
293
311
.
15.
Dzau
VJ
:
The role of mechanical and humoral factors in growth regulation of vascular smooth muscle and cardiac myocytes
.
Curr Opin Nephrol Hypertens
 
1983
;
2
:
27
32
.
16.
Safar
ME
,
Benetos
A
:
The shape of the blood pressure curve and genetic hypertension
.
Trends Cardiovasc Med
 
1995
;
5
:
6
102
.
17.
Berne
R
,
Levy
MN
Cardiovascular Physiology
 .
C.V. Mosby Company
:
St. Louis
,
1981
.
18.
Fukusaki
C
,
Kawakubo
K
,
Yamamoto
Y
:
Assessment of the primary effect of aging on heart rate variability in humans
.
Clin Aut Res
 
2000
;
10
:
123
130
.
19.
Piccirillo
G
,
Fimognari
F
,
Munizzi
MR
,
Bucca
C
,
Cacciafesta
M
,
Marigliano
V
:
Age-dependent influence on heart rate variability in salt-sensitive hypertensive subjects
.
J Am Geriatr Soc
 
1996
;
44
:
530
538
.
20.
Tank
J
,
Baevski
R
,
Fender
A
,
Baevski
A
,
Graves
KF
,
Ploewka
K
,
Weck
M
:
Reference values of indices of spontaneous baroreceptor reflex sensitivity
.
Am J Hypertens
 
2000
;
13
:
268
275
.
21.
Kardos
A
,
Watterich
G
,
de Menezes
R
,
Csanady
M
,
Casadei
B
,
Rudas
L
:
Determinants of spontaneous baroreflex sensitivity in a healthy working population
.
Hypertension
 
2001
;
37
:
911
916
.

Author notes

*
This study was supported by the VIDDA Foundation (REDM), grant HL61287 (RPS, REDM) and Department of Veterans Affairs, Veterans Health Administration, Rehabilitation Research and Development Service (JMW, REDM).