Our goals were to 1) determine the prevalence of the auscultatory gap in a large cohort of asymptomatic patients with hypertension who were not receiving medication and 2) evaluate the relation between the auscultatory gap and cardiac and vascular structure and function, with special attention to findings that have been related to prognosis in previous studies [3].

Methods

Top Methods Results Discussion Author & Article Info References

Study Sample

One hundred sixty-eight patients with hypertension who were referred from the Hypertension Center of The New York Hospital were studied between December 1990 and November 1993. All patients were studied after they had not received medications for at least 3 weeks; 65 patients (39%) had never received antihypertensive medications. Essential hypertension was diagnosed by the presence of a sustained increase in blood pressure ( 140 mm Hg systolic pressure or more than equals 90 mm Hg diastolic pressure, or both, on three determinations done at least 2 weeks apart). Patients with diagnoses of secondary hypertension or diabetes mellitus and patients with clinical evidence of coronary artery disease (typical angina or echocardiographic evidence of myocardial infarction) or cerebrovascular disease (transient ischemic attack or stroke) were excluded. Isolated systolic hypertension (systolic pressure more than equals 160 mm Hg and diastolic pressure less than 90 mm Hg) was present in 25 patients (15%). Valvular heart disease was excluded by Doppler echocardiography. Standard blood laboratory analyses were done for all patients. The study was done in accordance with protocols approved by the Committee on Human Rights in Research of Cornell University Medical College.

Wideband External Pulse Recording

The presence of auscultatory gaps was evaluated using wideband external pulse recording in addition to standard auscultation. Wideband external pulses represent vibrations on the body surface that are caused by pulsatile blood flow, and they are detected using a wideband pressure sensor that includes frequencies below the audible range. We recorded wideband external pulses from the left arm while the cuff pressure was deflated from greater than systolic pressure to less than diastolic pressure while the patient was in the supine position. This was done using a specially designed foil electret sensor (AT&T Bell Laboratories, Murray Hill, New Jersey) positioned over the brachial artery under the distal portion of the blood pressure cuff. This equipment has been described in detail [4-7]. Simultaneous traditional auscultation was done by a trained observer who used a switch to mark the onset and disappearance of Korotkoff sounds. In the right arm, a Finapres noninvasive finger pulse recording device (model 2300, Ohmeda, Englewood, Colorado), which has been shown to accurately reflect changes in intra-arterial pressure [8], provided a beat-to-beat representation of arterial pressure.

The wideband external pulse recorded during cuff deflation was previously reported [7] to be composed of 1) a low-frequency signal [K1] that appears when the cuff pressure is greater than systolic pressure and disappears when the cuff pressure is less than diastolic pressure; 2) a high-frequency signal [K2] that is related to the Korotkoff sound and appears when the cuff pressure is at systolic pressure and disappears when the cuff pressure is at diastolic pressure; and 3) a large-amplitude, low-frequency signal (K3) that appears when the cuff pressure is between systolic and diastolic pressure and that is still present when the cuff pressure decreases to less than diastolic pressure.

Auscultatory gaps were characterized by comparing the auscultatory markers or stethoscope signals with the corresponding wideband external pulse. Three types of auscultatory gaps, which have been described previously [6], were identified. The nature and relative frequency of each type of gap were determined on the basis of each patient's normal breathing pattern and a cuff pressure deflation rate of 2 to 4 mm Hg per second. A gap characterized by the intermittent disappearance of Korotkoff sounds when the cuff pressure is slightly less than systolic pressure (G1 gap) is associated with the intermittent disappearance of the K2 signal when the sound disappears and with the presence of the K1 and K2 signals when the sound is heard. The G1 gap is caused by a phasic decrease in systolic pressure to less than the cuff pressure during inspiration. A gap that occurs when the cuff pressure is slightly greater than diastolic pressure (G2 gap) is characterized by the intermittent disappearance of the K2 signal when the sound disappears and the presence of the K1, K2, and K3 signals when the sound is heard. The G2 gap results from a phasic increase of arterial diastolic pressure during expiration. The classic auscultatory gap (G3 gap; Figure 1) occurs when the cuff pressure is between the systolic and diastolic pressures, independent of respiratory variations in arterial pressure; it is characterized by a loss of audible sound but the presence of an under-developed K2 signal [6]. Although wideband external pulse recording is the definitive method for evaluating and identifying auscultatory gaps [6], the three types of gaps may be easily detected and classified at the bedside using standard sphygmomanometry to detect the loss and reappearance of Korotkoff sounds in relation to the respiratory cycle and systolic and diastolic blood pressures.

View larger version (42K): [in this window] [in a new window]   Figure 1. Recording of classic auscultatory gap (G3). This auscultatory gap (auscultatory marker) occurs during deflation of the cuff pressure to between the systolic and diastolic pressures (Finapres recording). The G3 gap is characterized by the disappearance of Korotkoff sounds (K sound), persistence of activity on the wideband external pulse recording (FES), and the absence of phasic fluctuation in arterial pressure ECG equals electrocardiogram; FES equals foil electret sensor. Reproduced with permission from the American Heart Association [6].

Echocardiography

Standard M-mode and two-dimensional echocardiography were done in all patients by a highly skilled research technician using commercially available equipment. Left ventricular dimensions were measured according to the recommendations of the American Society of Echocardiography [9, 10]. A blinded investigator did M-mode measurements of tracings on several cycles using a digitizing tablet, and the measurements were averaged. Left ventricular mass was calculated using the Penn convention and was adjusted for body surface area [11]. Left ventricular volumes were obtained using the correction of Teichholz and colleagues [12]. Fractional shortening, ejection fraction, relative wall thickness, stroke volume, cardiac output, and total peripheral resistance were calculated using standard formulas.

Carotid Ultrasonography

We did carotid ultrasonography as previously described [13]. Extracranial carotid arteries on both sides were scanned with a 7.5-MHz imaging transducer using multiple projections to detect the presence of discrete carotid atherosclerosis [14]. A two-dimensionally guided M-mode tracing of the distal common carotid artery, about 1 cm proximal to the carotid bulb, was obtained with simultaneous electrocardiography and carotid pressure waveform (see below) and recorded on a 0.5-inch Super VHS videotape. Suitable frames for measurements were digitized using a frame-grabber (Imaging Technology, Inc., Woburn, Massachusetts) interfaced with a high-resolution (640 x 480 pixel) video monitor and stored on diskettes. The axial resolution of the M-mode system is 0.2 mm. A blinded investigator measured carotid tracings on the stored images using a mouse-driven computer program after calibration for depth and time. The simultaneous carotid pressure tracing was used to ensure that measurement of the carotid artery was done at end-diastolic and peak-systolic pressures. Measurements included that of the intima-media thickness of the far wall of the distal common carotid artery at end diastole (as has been validated in anatomic correlation studies [15, 16]) and the internal dimensions of the carotid artery at end diastole and peak systole, which were determined by continuous tracing of the intimalumen interface of the near and far walls. The relative wall thickness of the artery was also calculated [13]. All measurements were obtained on several cycles and were averaged. Standard measurements of wall thickness were never obtained at the level of a discrete plaque.

Arterial Pressure Waveform and Arterial Stiffness

Carotid pressure waveforms were obtained using a high-fidelity external pressure transducer functioning as an applanation tonometer (model SPT-301, Millar Instruments, Inc., Houston, Texas) that was applied to the cutaneous surface corresponding to a common carotid artery. Waveforms and modulus and phase of harmonic components obtained with this transducer closely resemble those derived from intra-arterial recordings [17]. The correlation between the tonometer and the simultaneous intra-arterial recordings has been validated [17-20]. The transducer registers absolute changes in blood pressure over a range of 300 mm Hg, but it requires external calibration to obtain actual carotid blood pressure values. Because mean arterial blood pressure does not vary substantially in the capacitance vessels [21, 22], we measured brachial arterial blood pressure using cuff and mercury sphygmomanometry while each patient was supine, and we assigned mean blood pressure (diastolic pressure + [1/3 x pulse pressure]) to the computer-derived mean blood pressure of the carotid waveforms. After calibration, carotid peak-systolic and end-diastolic pressures were calculated by computer.

Carotid arterial stiffness was evaluated by the following three methods [23].

1. The Peterson elastic modulus (Ep) according to the formula Ep = ([P_s – P_d]/[D_s – D_d]) · D_d, in which P_s is systolic pressure, P_d is diastolic pressure, D_s is systolic carotid diameter, and D_d is diastolic carotid diameter [24].

2. The Young elastic modulus (E), according to the formula E = ([P_s – P_d]/[D_s – D_d]) · (D/h), in which P_s, P_d, D_s, and D_d are as defined above; D is mean carotid diameter; and h is wall thickness [25].

3. Arterial stiffness index (ß), according to the formula ß equals (ln[P_s/P_d])/([D_s minus D_d]/D_d), which takes into account the curvilinear relation between arterial pressure and diameter [26, 27]. ß was also determined at the standardized pressures of 100 mm Hg (ß₁₀₀), 120 mm Hg (ß₁₂₀), and 130 mm Hg (ß₁₃₀) in all patients in whom the carotid waveform included these pressures, according to the formula ß₀ equals (ln[P_s/P_d])/([D_s minus D_d]/D₀), in which D₀ equals D₁₀₀ or D₁₂₀ or D₁₃₀ (carotid diameters at 100 mm Hg, 120 mm Hg, and 130 mm Hg, respectively) [23, 27]. This approach allows the influence of distending pressure on vascular stiffness to be standardized.

Pulse wave velocity was measured using pulse transducers (HP21050A, Hewlett Packard, Andover, Massachusetts) positioned over both the brachial and radial arteries and was defined as the brachial to radial distance divided by the pulse transit time. The foot was taken at the point of the initial rapid upstroke of the wave for both the brachial and radial pulse waves.

Statistical Analysis

Data were stored and analyzed using the Crunch4 Statistical Package (Crunch Software Corp., Oakland, California). Continuous variables, expressed as mean ±SD, were compared between the two groups using the Student t-test. Differences in prevalences were compared by the chi-square statistic with continuity correction. Measures of arterial stiffness were compared in the two groups using analyses of covariance to control for group differences at baseline. Independence of association with the presence of an auscultatory gap or carotid atherosclerosis was assessed using logistic regression analysis. A P value less than 0.05 was considered statistically significant.

Results

Top Methods Results Discussion Author & Article Info References

Auscultatory Gaps and Clinical Characteristics

Among the 168 patients analyzed, 36 (21%) (group 1) were found to have a G3 gap alone (n = 30) or in association with another type of auscultatory gap (G1 gap in 5 patients and G2 gap in 1 patient), whereas 132 patients (79%) (group 2) had no auscultatory gap (n = 111) or an isolated G1 (n = 19) or G2 (n = 2) gap (Table 1). Patients in group 1 were older (64 ± 11 years compared with 55 ± 13 years; P < 0.001) and were more likely to be women (67% compared with 44%; P < 0.05) than patients in group 2. Consequently, patients in group 1 had smaller body surface area, although body mass index was similar in the two groups. Brachial blood pressure values, prevalence of isolated systolic hypertension, duration of hypertension, and previous use of antihypertensive medications did not differ between the two groups. No differences were noted between the groups in resting heart rate, electrocardiographic or funduscopic abnormalities, family history of hypertension or cardiovascular disease, race, smoking history, serum electrolyte concentrations, lipid levels, and renal function.

Auscultatory Gaps and Left Ventricular Structure and Function

Absolute left ventricular mass, left ventricular mass indexed by body surface area, left ventricular internal dimensions, wall thicknesses, and prevalence of left ventricular hypertrophy did not differ between the two groups. Although cardiac index and the other hemodynamic variables were similar in the two groups, total peripheral resistance was significantly higher in group 1 (2030 ± 481 dynes/sec · cm^–5 compared with 1797 ± 485 dynes/sec · cm^–5; P < 0.05).

Auscultatory Gaps and Carotid Arterial Structure and Stiffness

The prevalences of both carotid atherosclerotic plaques and carotid atherosclerosis (defined as the presence of atherosclerotic plaque, diffuse carotid intima-media thickening, or both [14]) were increased more than twofold in patients with auscultatory gaps. Absolute and relative carotid wall thicknesses, dimensions of the carotid artery, and the prevalence of intima-media thickening alone were similar in the two groups (Table 2).

Regional arterial stiffness, the Peterson elastic modulus, and ß were increased in patients with auscultatory gaps Table 3, even when distending pressure was considered (ß₁₀₀, ß₁₂₀, ß₁₃₀). Differences in the Young elastic modulus between the groups were not statistically significant (P = 0.1).

After adjustment for baseline differences, significant differences persisted between the groups in arterial stiffness (ß, ß₁₀₀, ß₁₂₀, ß₁₃₀) (Table 4). Pulse wave velocity tended to be higher in group 1 than in group 2 (adjusted mean velocity, 12.5 m/sec compared with 11.5 m/sec; P = 0.09).

Stepwise backward logistic regression analysis with age, sex, body surface area, clinic systolic and diastolic blood pressures, serum cholesterol levels, smoking history, ß, and carotid plaque included in the model indicated that only female sex (P < 0.02), ß (P < 0.002), and discrete atherosclerotic plaque (P < 0.02) were independently associated with the presence of an auscultatory gap (Table 5). Logistic regression analysis was also done to determine independent correlates of the presence of carotid atherosclerosis, adding the presence of a G3 auscultatory gap to the list of variables given above. Only age (P < 0.001) and the presence of an auscultatory gap (P < 0.05) were independently related to carotid atherosclerosis.

Discussion

Top Methods Results Discussion Author & Article Info References

Our study provides detailed, quantitative information on the pathophysiologic correlates of auscultatory gaps by assessing the relation of this long-recognized physical finding with cardiac and vascular structure and function in otherwise healthy persons with hypertension. Our findings indicate that auscultatory gaps are associated with female sex, increased arterial stiffness, and carotid atherosclerotic plaque, independent of age, arterial blood pressure, and other cardiovascular risk factors. The known prognostic significance of carotid atherosclerosis [28] and of measures of arterial stiffness [29] and the loss of cardioprotection in women by the time they reach the age of our patients with auscultatory gaps (64 ± 11 years) suggest that this finding may have prognostic relevance.

Many theories have been postulated to explain the classic G3 auscultatory gap, which is related to properties of the arterial wall and not to phasic respiratory changes in arterial pressure [6]. In 1928, Mudd and White [30] hypothesized that a change in arterial wall tone or physical characteristics could prevent the vascular wall from resonating on opening, thereby causing the auscultatory gap [30]. In 1941, Ragan and Bordley [31] attributed the occurrence of gaps to a decreased pressure gradient between the pressure above and below the artery occluded during the cuff measurement due to venous congestion, whereas Rodbard and Margolis in 1957 [32] hypothesized that a reduction in peripheral blood flow was associated with the presence of an auscultatory gap. Tavel and coworkers [33] detected a direct relation between the intensity of the Korotkoff sound and the rate of increase in the intra-arterial pressure wave recorded during cuff deflation. More recently, the occurrence of the auscultatory gap was found to be related to the presence of an underdeveloped K2 signal [6]. During cuff deflation, audible Korotkoff sounds and the K2 signal occur when intra-arterial pressure exceeds the pressure imposed by the arm cuff and opens the compressed segment of the artery, allowing the proximal and distal portions of the artery to communicate. Thus, the auscultatory gap in this phase and the presence of a blunted K2 signal suggest the occurrence of changes in the physical properties of the arterial wall [6].

About one sixth of our hypertensive patients had either auscultatory gaps near the level of systolic pressure that were related to a phasic decrease in systolic pressure during inspiration (G1 gaps) or gaps near the level of diastolic pressure that were related to an increase in diastolic pressure during expiration (G2 gaps). The presence of these gaps was determined by the magnitude of blood pressure fluctuation caused by respiration interacting with the rate of cuff deflation. The high occurrence of these types of gaps at a fixed cuff deflation rate in our patients may have been caused by the enhanced blood pressure variability that has been reported in hypertensive patients [34] or by increased expiratory airway resistance, which also may be associated with greater fluctuation in respiratory blood pressure.

Our study provides the first objective demonstration of an association between auscultatory gaps and vascular structural and functional changes (increased sclerosis and atherosis), independent of age [35], blood pressure levels [36], and well-recognized risk factors for atherosclerosis, such as smoking history and serum lipid levels. In contrast to previous reports [37], we did not find age to be an independent predictor of the presence of auscultatory gaps in our patients, whereas increased arterial stiffness and carotid atherosclerosis (the occurrence of which has been extensively reported to be associated with age [14, 35]) were strong determinants of the presence of a G3 gap. However, because blood pressure variability has been reported to be increased in elderly persons (possibly because of impaired baroreflex sensitivity), the previously reported association of age with auscultatory gaps may be caused by this variability. This in turn may cause an increase in G1 and G2 gaps.

Whereas carotid atherosclerosis is a major cause of ischemic stroke, carotid arterial intima-media thickening and discrete atherosclerotic plaque have also been reported to be related to concurrent coronary artery disease [38, 39]) and to subsequent myocardial infarction [28], possibly as a result of generalized atherosclerosis [40], arterial remodeling due to a pressure load [13], and an increased likelihood of associated left ventricular hypertrophy [41]. These observations support the prognostic relevance of changes in carotid arterial structure.

In our study, left ventricular function and structure were not related to the presence of auscultatory gaps. This is consistent with the hypothesis that the modification of conduit arteries is a more sensitive indication of target organ damage in patients with hypertension [13], possibly stimulating subsequent alterations of cardiac structure [42].

In conclusion, our study supports the hypothesis that hypertensive patients with classic auscultatory gaps are at increased risk for cardiovascular disease, independent of age, blood pressure, and other cardiovascular risk factors. However, prospective studies are required to conclusively define the prognostic relevance of this auscultatory finding.

Drs. Roman and Devereux: Division of Cardiology, The New York Hospital-Cornell Medical Center, 525 East 68th Street, New York, NY 10021.

Dr. Blank: Department of Electromechanical Engineering Technology, New York City Technical College/City University of New York, 186 J Street, Brooklyn, NY 11201.

Dr. Pickering: Hypertension Center, The New York Hospital-Cornell Medical Center, 525 East 68th Street, New York, NY 10021.