Isolated Elevation of Diastolic Blood Pressure
Real or Artifactual?
Abstract Not infrequently, blood pressure measurement by the standard auscultatory technique yields a normal systolic pressure with an elevated diastolic pressure. The relatively narrow pulse pressure of such a measurement raises concern about the accuracy of the blood pressure measurement. The purpose of this study was to assess the accuracy of auscultatory blood pressure measurements in patients with an uncommonly narrow pulse pressure, particularly patients with an elevated diastolic but normal systolic pressure. Auscultatory blood pressure measurements were compared with an objective noninvasive standard, called K2 analysis, which has been shown to be more accurate than the auscultatory technique. Blood pressure was measured simultaneously by auscultatory and K2 techniques in 175 subjects. Comparisons were performed (1) in the group as a whole, (2) in four clinical subgroups (normotensive [<140/<90 mm Hg, n=69], hypertensive [≥140/≥90 mm Hg, n=53], isolated systolic hypertensive [≥140/<90 mm Hg, n=38], and isolated diastolic hypertensive [<140/≥90 mm Hg, n=15]), and (3) in two subgroups whose ratio of pulse pressure to diastolic pressure was greater than or equal to 0.45 (n=151) or less than 0.45 (n=24). Subjects in the isolated diastolic hypertensive group and in the group with a pulse pressure ratio less than 0.45 were considered to have a narrow pulse pressure. In the group as a whole, consistent with previous auscultatory-K2 comparisons, systolic pressure was slightly higher and diastolic pressure slightly lower when measured by K2 versus the auscultatory technique (auscultatory, 145/85 mm Hg; K2, 147/83 mm Hg). For diastolic pressure auscultatory measurements averaged 7 mm Hg greater than K2 in the isolated diastolic hypertensive group (94±4 versus 87±5 mm Hg) but were less than 3 mm Hg (greater) in the other three groups (P<.0004). For systolic pressure, differences were less than 3 mm Hg in all four clinical groups. Auscultatory-K2 differences of diastolic pressure exceeding 5 mm Hg (and 10 mm Hg) were seen in 73.3% (and 40.0%) of isolated diastolic hypertensive subjects versus only 14.5% (2.9%) of normotensive subjects, 22.6% (1.9%) of hypertensive subjects, and 7.9% (2.6%) of isolated systolic hypertensive subjects (P<.0001). Similarly, the auscultated diastolic pressure exceeded the K2 measurement by at least 5 mm Hg (and 10 mm Hg) in 62.5% (29.2%) of subjects with a pulse pressure ratio less than 0.45 versus 13.9% (2.0%) in subjects with a ratio greater than or equal to 0.45 (P<.0001). Auscultatory-K2 differences of diastolic pressure were strongly and inversely related to the pulse pressure ratio (r=−.68, P<.0001) independent of sex, race, or body weight. In conclusion, when the pulse pressure is particularly narrow, auscultation frequently overestimates the true diastolic pressure by 5 mm Hg or more. The treatment implications of this finding, particularly in patients with elevated diastolic but normal systolic pressure, merit further study.
Elevated blood pressure (BP) has been widely documented as an important risk factor for stroke and a major risk factor for coronary heart disease.1 Physicians have tended to regard elevation of diastolic blood pressure (DBP) as more important than elevation of systolic blood pressure (SBP) on the grounds that it is more closely related to end-organ damage.2 However, epidemiological studies have shown that SBP may be at least as important a risk factor as DBP, with many studies showing it to be more important with regard to cardiovascular morbidity and mortality.3 4 5
Patients with elevation of both SBP and DBP6 7 and patients with isolated systolic hypertension8 9 have been shown to benefit from drug treatment. Thus, when SBP is significantly elevated, the argument for treatment can be made independent of a small degree of inaccuracy or variation in DBP.
Not infrequently, patients present with elevation of DBP only. This could be regarded as “isolated diastolic hypertension.” In the National Health Examination Survey (NHES) and National Health and Nutrition Examination Survey (NHANES) studies, 6% to 9% of subjects with elevated BP readings had (isolated) elevation of DBP with a normal SBP.10 11 According to the Joint National Committee on the Detection, Evaluation, and Treatment of High Blood Pressure, as many as 58 million people in the United States may have high BP.1 Thus, a considerable number of individuals have isolated elevation of DBP.
Based on current widely accepted guidelines, a person with such a BP, for example, 125/100 mm Hg, may be treated pharmacologically, perhaps for life. However, in this example the pulse pressure (the numerical difference between SBP and DBP) is smaller than expected for this level of mean arterial pressure,12 raising concern about the accuracy of the BP measurement. Either underestimation of SBP or overestimation of DBP could contribute to such a reading. The existence of an auscultatory gap is widely known to affect SBP measurements. However, the accuracy of DBP measurements in such patients has not been adequately studied. Because the decision to treat is often based on DBP rather than SBP, overestimation of DBP by as little as 5 mm Hg may frequently affect the treatment decision.
Confirmation of the accuracy of auscultated measurements has traditionally required invasive intra-arterial measurement, which is not practical in the clinical setting. Recently, we have described an objective noninvasive method of BP measurement that uses analysis of the brachial wideband external pulse recorded during BP cuff deflation.13 14 15 Wideband external pulse recording is based on the ability of a pressure sensor to record inaudible frequencies (down to 0.1 Hz) during cuff deflation. Three distinct components of the wideband external pulse signal can be detected (called K1, K2, and K3), one of which, K2, appears and disappears at SBP and DBP, respectively. Fig 1⇓ demonstrates the disappearance of K2 as cuff pressure is reduced from above intra-arterial DBP to below. The visible appearance and disappearance of the K2 signal are not subject to the vagaries of auscultation. It should be noted that K1, K2, and K3 are not related to the five phases of the Korotkoff sound. We have previously shown that BP measured by K2 analysis is closer to intra-arterial pressure than auscultatory measurements13 and that no statistically significant differences of BP determination occurred between K2 and intra-arterial measurements with the use of a solid-state Millar catheter. Thus, K2 analysis offers an alternative to invasive intra-arterial measurement in assessing the accuracy of indirect cuff measurements.
The purpose of this study was to compare BP measurement by auscultation with K2 analysis in a wide range of subjects, including those with isolated elevation of DBP.
BP was measured simultaneously by the auscultatory method and by K2 analysis in 175 nonobese subjects (109 men, 66 women); 132 were from a larger waveform analysis study, and 43 were specifically recruited because their seated clinic BP, auscultated by a physician, was characterized by a narrow pulse pressure (defined below).
Twenty-eight of the 175 subjects also underwent auscultation simultaneously by two observers using a dual stethoscope (S.G.B. and one of the trained Cardiovascular Center staff, ie, nurse, technician, or physician). The second observer was unaware of the hypothesis of the study. Auscultation by two rather than one examiner was based on the availability of personnel but was emphasized when the referred subject was believed to have a narrow pulse pressure.
The wideband external pulse was recorded with a specially designed Foil Electret Sensor that is similar in principle to conventional electret microphones used for airborne sound reception. The Foil Electret Sensor amplifier system has a flat frequency response from below 0.1 Hz to above 2000 Hz. Detailed characteristics of this sensor16 17 and sensor-amplifier system18 have been described.
Recordings were obtained in a quiet room, with the subject in the supine position. The wideband external pulse was recorded during cuff deflation with the sensor positioned over the brachial artery and under the distal portion of a BP cuff. The cuff bladder (width) size (12- to 14-cm width×40-cm length) was selected to match the circumference of the subject’s arm as recommended by the American Heart Association.19 An electrocardiogram was also recorded.
The cuff was rapidly inflated to a pressure that was visually confirmed to be above the SBP by examination of the wideband external pulse recording, as previously described.18 The cuff pressure was controlled by a cuff pressure inflator/regulator (E-10, DE Hokanson) and was read by both a mercury column and pressure sensor coupled to a VR6 physiological recording system (Electronics for Medicine). Cuff pressure was manually deflated, with the use of the E-10 rotary pressure regulator handle, at a rate of 2 to 3 mm Hg/s.
The cuff inflation/deflation procedure enabled BP measurement simultaneously by the auscultatory and K2 methods. Inflation/deflation was performed three times on each subject. Auscultation was performed with the use of a switch to mark the onset and disappearance of sound (Korotkoff phase V). The stethoscope was positioned distal to the cuff and held firmly in position without excessive pressure with the use of an elastic strap. This eliminated extraneous noise caused by human touch and ensured that the surface of the stethoscope was maintained flush against the skin.
Each channel of the VR6 physiological recorder was sampled at 500 Hz by a 12-bit analog-to-digital convertor for storage into an IBM PC/AT computer using the codas (Dataq) data acquisition software.
Analysis of Signals
SBP was identified as the cuff pressure at the electrocardiographic QRS complex immediately preceding the listener’s mark for the onset of Korotkoff sounds. DBP was identified as the cuff pressure two QRS complexes before the mark identifying the disappearance of sound. This procedure, which allows for the expected lag time before the appearance or disappearance of sound can be marked, is exactly analogous to extrapolation backward to the last auscultated sound during standard sphygmomanometry.
Identification of SBP and DBP from the K2 analysis has been previously described.13 Briefly, SBP was identified by the cuff pressure at the cardiac cycle in which the K2 signal initially appears and DBP by the cuff pressure at the last cardiac cycle before the K2 signal disappears.
Comparison by Clinical Subgroup
Based on the auscultated BP, subjects were categorized into four clinical subgroups characterized by normal BP (NT: systolic <140 mm Hg, diastolic <90), hypertension (HT: systolic >140, diastolic ≥90), isolated systolic hypertension (ISH: systolic ≥140, diastolic <90), and isolated diastolic hypertension (IDH: systolic <140, diastolic ≥90). Subjects in the IDH group were considered to have a narrow pulse pressure. Selected characteristics of each group are shown in Table 1⇓.
Analysis Based on Pulse Pressure Ratio
Because of the nonlinear elastic properties of the arterial system, a higher mean arterial pressure is normally associated with a wider pulse pressure than is a lower mean pressure.12 Thus, a pulse pressure as narrow as 30 mm Hg is encountered commonly with a BP of 90/60 but not 140/110 mm Hg. Therefore, the ratio of pulse pressure to DBP more meaningfully defines a pulse pressure as “narrow.” Thus, in the above examples although the pulse pressure was the same, the pulse pressure ratios were 0.50 and 0.27, respectively. Therefore, the ratio of pulse pressure to DBP (pulse pressure ratio) was used for analysis rather than the pulse pressure by itself. Subjects with a supine auscultated pulse pressure ratio less than 0.45 were considered to have a narrow pulse pressure (n=24). The 0.45 value was selected because this value corresponds roughly to the ratio obtained with a DBP of 90 mm Hg (hypertensive cutoff) with a normal SBP (130 mm Hg).
All BP values used for analysis were an average of three measurements. Auscultated and K2 values for SBP and DBP and in the pulse pressure ratio were compared with Student’s t test. The correlation between the two methods was determined with regression analysis. The auscultated pressures of the two observers (S.G.B. and the other) were also compared by a paired t test. For the defined subgroups the likelihood of BP differences greater than 5 or 10 mm Hg between measurement techniques was examined with a contingency analysis.
Overall, the BP values as measured by the two methods were closely associated (systolic, r=.99; diastolic, r=.95; P<.001) (Table 2⇓). SBP by K2 analysis averaged 2 mm Hg higher and the K2 DBP averaged 2 mm Hg lower than the corresponding auscultatory values. The two methods were in agreement in all clinical subgroups except in the IDH group (Table 2⇓). In this group the auscultated DBP was 7 mm Hg higher than the K2 DBP (94±4 versus 87±5 mm Hg), which was significantly greater than the auscultatory-K2 difference in the three other groups (P<.0004).
Auscultated DBP exceeded the K2 measurement by at least 5 mm Hg in 73.3% (11 of 15) of IDH subjects, compared with only 14.5% (10 of 69), 22.6% (12 of 53), and 7.9% (3 of 38) of NT, HT, and ISH subjects, respectively (P<.0001) (Table 3⇓). Similarly, auscultatory readings were at least 10 mm Hg higher in 40% (6 of 15) of IDH subjects versus 2.9% (2 of 69), 1.9% (1 of 53), and 2.6% (1 of 38) of NT, HT, and ISH subjects, respectively (P<.0001).
For SBP, significant discrepancies were uncommon and did not differ by clinical subgroup. The auscultated SBP was at least 10 mm Hg lower than the K2 value in only 2 of the 175 subjects. Both subjects were in the IDH group.
Pulse Pressure Ratio
For DBP, the magnitude of the difference between auscultatory and K2 measurements was strongly and inversely related to the pulse pressure ratio (r=−.68, P<.0001) (Fig 2⇓). This relationship was independent of age, sex, race, or body weight.
The auscultated DBP was at least 5 mm Hg higher than the K2 value in 62.5% (15 of 24) of subjects with a pulse pressure ratio less than 0.45, compared with only 13.9% (21 of 151) of subjects with a pulse pressure ratio of 0.45 or greater (P<.0001). Similarly, differences of 10 mm Hg or greater were seen in 29.2% (7 of 24) versus 2.0% (3 of 151), respectively (P<.0001).
For SBP, auscultatory-K2 differences as large as 5 or 10 mm Hg were less common, occurring in only 17 and 2 subjects, respectively. Auscultated SBP tended to underestimate the K2 value by at least 5 mm Hg slightly more frequently in subjects with a pulse pressure ratio less than 0.45 (P=.07). The auscultated value was at least 10 mm Hg lower in 2 of 24 subjects with a ratio less than 0.45 versus 0 of 151 with a ratio of 0.45 or greater (P<.02).
Subjects with a narrow pulse pressure ratio (<0.45) were younger (44±13 versus 52±17 years; P<.02). The mean weight of the two groups was similar (170±28 versus 162±30 lb, respectively).
Auscultated values determined by the two observers were the same (136/93±16/9 versus 136/94±16/10 mm Hg). For each observer the discrepancy between auscultated and K2 DBP correlated highly (P<.0001) with the corresponding pulse pressure ratio (observer 1 (S.G.B.), r=−.76; observer 2 (trained staff), r=−.67). Thus, regardless of the observer the likelihood of a discrepancy (5 mm Hg or greater) between auscultated and K2 DBP was similar. Interestingly, when one observer found a narrow ratio and the other did not, the K2 value for DBP consistently coincided with the lower of the two auscultated measurements.
The major finding of this study is that when the auscultatory measurement of the pulse pressure is narrow (arbitrarily defined either as isolated diastolic hypertension or pulse pressure ratio less than 0.45, regardless of the DBP), a clinically significant overestimation of the DBP by the auscultatory technique often occurs. Inaccuracy in the estimation of SBP occurs far less frequently. In individuals with a normal SBP, elevation of DBP is usually the sole reason for initiating pharmacological treatment. Thus, an overestimation of 10 or even 5 mm Hg in the measurement of DBP can result in unnecessary treatment, perhaps lifelong, for a substantial number of individuals.
Diastolic pseudohypertension has previously been reported, particularly in the elderly20 21 22 23 and obese,24 25 with elevated levels of SBP. The present report extends these observations to a nonelderly and nonobese population with more normal levels of SBP.
Cardiovascular Risk and Isolated Diastolic Hypertension
Several factors point toward a benign outcome in patients with isolated diastolic hypertension (by definition, a group with a narrow pulse pressure): (1) SBP is more strongly correlated than DBP with adverse cardiovascular outcome3 4 5 ; (2) an increased pulse pressure (pulsatile component of the BP), independent of mean pressure, is associated with cardiovascular morbidity,26 mortality,27 and target-organ damage28 and has been found to be the strongest pressure-related stimulus for vascular hypertrophy and remodeling in animal models29 30 ; and (3) isolated systolic hypertension, a condition characterized by a wide pulse pressure, is associated with a high cardiovascular event rate, even though the DBP is normal.5 31
In an accompanying report, Fang et al32 present data that demonstrate a more favorable prognosis for young and middle-aged hypertensive subjects with isolated elevation of DBP compared with individuals with both systolic and diastolic hypertension. In fact, not one IDH subject experienced a myocardial infarction during the years of follow-up. We suggest that a clinically significant overestimation of diastolic BP occurred in many IDH subjects and that some were erroneously classified as hypertensive.
Thus, the entity of isolated diastolic hypertension may constitute a low-risk condition. Our finding that the DBP is frequently overestimated in this group is consistent with the reported benign outcome in such individuals, further raising questions concerning the diagnosis of hypertension and the need for lifelong therapy.
Auscultated and K2 BP Measurements
Auscultation of Korotkoff sounds through a stethoscope has been the most widely used noninvasive method for BP measurement.33 Most of the data about human arterial pressure have been obtained with its use. However, indirect BP measurement by auscultation does not always accurately represent the actual intra-arterial pressure.
Intra-arterial measurement, the traditional gold standard in assessing the accuracy of indirect, auscultated recordings, is invasive and not easily amenable to the study of large numbers of normotensive or mildly hypertensive individuals. K2 analysis, the standard used in the present study, has been previously shown to come closer to intra-arterial BP measurements than the auscultatory technique.13 This new technique measures BP objectively and provides the opportunity for noninvasively investigating the accuracy of auscultated BP in large numbers of individuals with relatively mild hypertension.
We believe that the K2 component of the wideband external pulse reflects the initial dynamic events related to the opening of the compressed artery segment and reestablishment of fluid contact between the proximal and distal portions of the occluded artery. Therefore, the presence of the K2 signal may be a more specific indicator of transmural pressure gradient reversals (ie, intra-arterial pressure increasing above the cuff pressure) than audible sound and may differ from auscultated measurement of DBP in situations in which Korotkoff sounds disappear early or late.
The origin of Korotkoff sounds during cuff deflation is still poorly understood. The sounds have been attributed to both pressure and flow phenomena.34 One may think of the sounds occurring under the BP cuff as sounds generated by the physical events created by the initial opening of the compressed artery segment (ie, when the intra-arterial pressure exceeds the cuff pressure) modulated by the resulting blood flow (eg, blood turbulence).35 When pulsatile blood flow is reduced, as in the state of peripheral vasoconstriction (eg, shock), sound disappears prematurely, resulting in overestimation of DBP.36 37 In contrast, in patients with aortic regurgitation,38 which is associated with a wide pulse pressure and turbulent blood flow, and in pregnant women,39 40 who have peripheral vasodilation, sound often persists below the true DBP. Thus, physiological and/or physical mechanisms that modify peripheral blood flow may modify the accuracy of the Korotkoff sound technique. In the present study in subjects with a narrow pulse pressure, a reduced pulsatile peripheral blood flow may similarly explain the premature disappearance of sound, with resultant overestimation of the DBP.
A commonly offered explanation for an abnormally narrow pulse pressure is an auscultatory underestimation of the SBP, as in the “auscultatory gap” (ie, appearance followed by temporary disappearance of Korotkoff sounds as the cuff is deflated).18 However, in the present study the wideband external pulse K1 pattern was observed on initial cuff inflation, assuring that the cuff pressure was above the true SBP. We found that underestimation of SBP was far less common than overestimation of DBP.
In performing the study we were careful to avoid errors in technique that commonly cause overestimation of DBP. We used BP cuffs whose width (12 to 14 cm) corresponded to American Heart Association recommendations.19 We performed cuff deflation slowly (2 to 3 mm Hg/s). Furthermore, any overestimation of DBP related to the rate of cuff deflation would have been independent of the clinical subgroup or the pulse pressure ratio.
Of 43 subjects referred to the study because of a narrow pulse pressure, only 24 were found to have a narrow pulse pressure ratio (ie, <0.45) during the study. This may have resulted from the use of the supine position, when pulse pressure would be expected to be wider than during seated measurements obtained clinically. In addition, many of the recruited subjects had “soft,” low-intensity Korotkoff sounds. Thus, careful BP measurement technique made during the study, including the use of an elastic strap to eliminate extraneous noise, may have contributed to these soft sounds being heard below what might be heard during a standard clinical measurement. Other factors that might have contributed to this phenomenon are subject and observer variabilities, which were not assessed. The frequency of overestimation of DBP in the seated position was not assessed. Further studies, with measurement in the seated, standing, and supine positions, are indicated.
Finally, when two observers performed auscultation simultaneously, they were generally in agreement in their measurement of DBP. In instances in which one observer found a narrow pulse pressure and the other did not, the K2 measurement consistently agreed with the observer who obtained the lower diastolic reading and “normal” pulse pressure ratio. This suggests that when the auscultated DBP significantly varies between observers, a reading with an uncommonly narrow pulse pressure is likely to be overestimating the DBP.
In summary, we have found that in subjects with an uncommonly narrow pulse pressure, the auscultated BP measurement frequently overestimates DBP by 5 mm Hg or more. This overestimation, particularly in view of the lower cardiovascular risk of isolated diastolic hypertension, suggests the need for reconsideration of the widespread and automatic practice of treating patients with isolated diastolic hypertension. Intra-arterial measurement could be helpful in assessing such patients but is not clinically practical. Wideband external pulse recording would be useful in assessing the accuracy of the diastolic reading, but an instrument based on K2 analysis is not at present commercially available. Nevertheless, our findings suggest that when an auscultated BP is characterized by a narrow pulse pressure, the possibility of an overestimation of DBP should be considered.
This work was supported in part by grant HL-18323-19 from the National Heart, Lung, and Blood Institute.
- Received February 4, 1995.
- Revision received March 21, 1995.
- Accepted May 30, 1995.
Vokonas PS, Kannel WB, Cupples LA. Epidemiology and risk of hypertension in the elderly: The Framingham Study. J Hypertens. 1988;6(suppl 1):s3-s9.
Stamler J, Neaton JD, Wentworth DN. Blood pressure (systolic and diastolic) and risk of fatal coronary heart disease. Hypertension. 1989;13(suppl I):I-2-I-12.
Wikstrand J, Westergren G, Berglund G, Bracchetti D, Van Couter A, Feldstein CA, Ming KS, Kuramoto K, Landahl S, Meaney E, Pederson EB, Rahn KH, Shaw J, Smith A, Waal-Manning H. Antihypertensive treatment with metoprolol or hydrochlorothiazide in patients 60 to 75 years: report from a double-blind international multicenter study. JAMA. 1986;255:1304-1310.
MRC Working Party. Medical Research Council trial of treatment of hypertension in older adults: principal results. Br Med J. 1992;304:405-416.
Gordon G.. Blood Pressure of Adults by Age and Sex, United States, 1960-62. National Center for Health Statistics, Series II, No. 4. Washington, DC; 1964.
Drizd T, Dannenberg AL, Engel A. Blood pressure levels in persons 18-74 years of age in 1976-1980 and trends in blood pressure from 1960 to 1980 in the United States. In: United States Vital and Health Statistics, Series II, No. 234, DHHS publication No. (PHS) 86-1684, National Center for Health Statistics, July 1986 (NHANES II study).
Berne RM, Levy MN. Cardiovascular Physiology. 6th ed. St Louis, Mo: Mosby Year Book; 1992:135-151.
Blank SG, West JE, Mueller FB, Cody RJ, Harshfield GA, Pecker MS, Laragh JH, Pickering TG. Wideband external pulse recording during cuff deflation: a new technique for evaluation of the arterial pressure pulse and measurement of blood pressure. Circulation. 1988;77:1297-1305.
O’Brien E, Fitzgerald D. The history of indirect blood pressure measurement. In: O’Brien E, O’Malley K, eds. Handbook of Hypertension, Volume 14: Blood Pressure Measurement. New York, NY: Elsevier Science; 1991:1-54.
Pickering TG, Blank SG. Blood pressure measurement and ambulatory blood pressure monitoring. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis, and Management. New York, NY: Raven Press Publishers; 1990:1429-1441.
Blank SG, West JE, Mueller FB, Pecker MS, Laragh JH, Pickering TG. Characterization of auscultatory gaps with wideband external pulse recording. Hypertension. 1991;17:225-233.
Frohlich ED, Grim CV, Labarthe DR, Maxwell MH, Perloff D, Weidman WH. Recommendations for human blood pressure determination by sphygmomanometers: report of a special task force appointed by the Steering Committee, American Heart Association. Hypertension. 1988;11:210A-222A. (Published simultaneously in Circulation. 1988;77:502A-514A.)
Zweifler AJ, Shahab ST. Pseudohypertension: a new assessment. J Hypertens. 1983;11:1-6.
Spence JD, Sibbald WJ, Cape RD. Pseudohypertension in the elderly. Clin Sci Mol Med. 1978;55:399s-402s.
Nielson PE, Larsen B, Holstein P, Poulsen HL. Accuracy of auscultatory blood pressure measurements in hypertensive and obese subjects. Hypertension. 1983;5:122-127.
Mejia AD, Egan BM, Schork NJ, Zweifler AJ. Artifacts in measurement of blood pressure and lack of target organ involvement in the assessment of patients with treatment-resistant hypertension. Ann Intern Med. 1990;112:270-277.
Madhavan S, Ooi WL, Cohen H, Alderman MH. Relation of pulse pressure and blood pressure reduction to the incidence of myocardial infarction. Hypertension. 1994;23:395-401.
Darne B, Girerd X, Safar M, Cambien F, Guize L. Pulsatile versus steady component of blood pressure: a cross-sectional analysis and a prospective analysis on cardiovascular mortality. Hypertension. 1989;13:392-400.
Christensen KL. Reducing pulse pressure in hypertension may normalize small artery structure. Hypertension. 1992;18:722-727.
Kannel WB, Dawber TR, McGee DL. Perspectives on systolic hypertension: the Framingham Study. Circulation. 1980;61:1179-1182.
Fang J, Madhavan S, Cohen H, Alderman MH. Isolated diastolic hypertension: a favorable finding among young and middle-aged hypertensive subjects. Hypertension. 1995;26:377-382.
Korotkoff NS. On the subject of methods of measuring blood pressure. Bulletin of the Imperial Military Academy (St Petersburg). 1905;11:365-367.
McCutcheon EP, Rushmer RF. Korotkoff sounds: an experimental critique. Circ Res. 1967;20:149-161.
Cohn JN. Blood pressure measurement in shock. JAMA. 1967;199:972-976.
Goldstein S, Killip T. Comparison of direct and indirect arterial pressures in aortic regurgitation. N Engl J Med. 1962;267:1121-1124.
Blank SG, Helseth G, Pickering TG, West JE, August P. How should diastolic blood pressure be defined during pregnancy? Hypertension. 1994;24:234-240.