Ambulatory Blood Pressure Is Superior to Clinic Blood Pressure in Predicting Treatment-Induced Regression of Left Ventricular Hypertrophy
Background In cross-sectional studies, ambulatory blood pressure (ABP) correlates more closely than clinic BP with the organ damage of hypertension. Whether ABP predicts development or regression of organ damage over time better than clinic BP, however, is unknown.
Methods and Results In 206 essential hypertensive subjects with left ventricular hypertrophy (LVH), we measured clinic supine BP, 24-hour ABP, and left ventricular mass index (LVMI, echocardiography) before and after 12 months of treatment with lisinopril (20 mg UID) without or with hydrochlorothiazide (12.5 or 25 mg UID). Measurements included random-zero, clinic orthostatic, and home BP. In all, 184 subjects completed the 12-month treatment period. Before treatment, clinic supine BP was 165±15/105±5 mm Hg (systolic/diastolic), 24-hour average BP was 149±16/95±11 mm Hg, and LVMI was 158±32 g/m2. At the end of treatment, they were 139±12/87±7 mm Hg, 131±12/83±10 mm Hg, and 133±26 g/m2, respectively (P<.01 for all). Before treatment, LVMI did not correlate with clinic BP, but it showed a correlation with systolic and diastolic 24-hour average BP (r=.34/.27, P<.01). The LVMI reduction was not related to the reduction in clinic BP, but it was related to the reduction in 24-hour average BP (r=.42/.38, P<.01). Treatment-induced changes in average daytime and nighttime BPs correlated with LVMI changes as strongly as 24-hour BP changes. No substantial advantage over clinic supine BP was shown by clinic orthostatic, random-zero, and home BP.
Conclusions In hypertensive subjects with LVH, regression of LVH was predicted much more closely by treatment-induced changes in ABP than in clinic BP. This provides the first longitudinally controlled evidence that ABP may be clinically superior to traditional BP measurements.
Frequent use of ambulatory blood pressure monitoring has become customary in many countries to improve the diagnosis of hypertension and the estimation of the efficacy of antihypertensive treatment. However, evidence that ambulatory blood pressure is clinically superior to the blood pressure obtained by a doctor or a nurse rests on cross-sectional studies showing that the target-organ damage of hypertension is more closely related to 24-hour average than to clinic blood pressure.1 It also rests on the results of a small number of follow-up studies that suggest a closer association of cardiovascular morbidity or target-organ deterioration with ambulatory than with clinic blood pressure but that lack a controlled design and fail to report key data, such as number of patients treated, type of treatment, and blood pressure values achieved during treatment.2 3 4 The evidence of the clinical superiority of ambulatory to clinic blood pressure is thus so incomplete as to prevent a definite conclusion on the advantage of routine use of ambulatory blood pressure in the clinical management of the hypertensive patient.
Our study has addressed this issue by prospectively examining whether in patients with echocardiographic evidence of LVH, a reduction of LVM induced by long-term antihypertensive treatment is more accurately predicted by the reduction in clinic or 24-hour average blood pressure. Because little information exists on the prognostic importance of home blood pressure,5 the study also examined the ability of home blood pressure measurements to predict regression of LVH. Echocardiographic LVH was selected because its prognostic importance is documented by epidemiological studies in hypertensive patients6 and in the general population.7
Our study was named SAMPLE, for Study on Ambulatory Monitoring of Blood Pressure and Lisinopril Evaluation. The study was performed in 11 hypertension clinics located in Italy. Each center was instructed to recruit 20 to 22 essential hypertensive patients of either sex according to the following criteria: (1) age ranging from 20 to 65 years; (2) a DBP between 95 and 115 mm Hg after a 4-week period without antihypertensive drugs (previously treated patients) or a 3-week observation period (previously untreated patients), and (3) an “echocardiographic” LVH8 (see below). Exclusion criteria were (1) secondary hypertension, (2) history and/or signs of cardiovascular complications (eg, congestive heart failure, myocardial infarction, stroke, angina pectoris) or major target-organ damage (eg, serum creatinine >1.5 mg/dL), (3) major cardiovascular or noncardiovascular diseases besides hypertension, (4) pregnancy or lactation, (5) contraindications to the antihypertensive drugs to be used during the treatment period (see below), and (6) conditions that would prevent collection of technically adequate echocardiograms (eg, obesity or pulmonary emphysema) or ambulatory blood pressure monitoring (eg, atrial fibrillation or other major arrhythmias). Patients were also excluded from the study if previous antihypertensive treatment consisted of more than two drugs to reduce the risk of subsequent withdrawal because of inadequate blood pressure control (see below). All patients consented to the study after being informed of its nature and purpose. The study protocol was approved by the Ethical Committees of the centers involved.
Clinic Blood Pressure
In all patients, blood pressure was measured in a hypertension clinic by a mercury sphygmomanometer, with the first and fifth Korotkoff sounds taken to identify systolic and diastolic values, respectively. Two measurements were collected with the patient in the supine position for 5 and 8 minutes, and the average of the two values was taken as the “clinic” supine blood pressure for inclusion in the study, determination of the efficacy of treatment, and subsequent statistical analysis. Blood pressure was also measured with the patient in the upright position for 1 minute. In the supine position, heart rate was assessed by palpation of the radial artery for 30 seconds. In each individual, all measurements were performed in the morning, in the same arm, by a single doctor.
Random-Zero Blood Pressure
Blood pressure was also measured by a random-zero sphygmomanometer (Hawksley and Sons). This measurement was obtained with the patient supine immediately before the first measurement obtained by the mercury sphygmomanometer.
Home Blood Pressure
At the time of the first medical visit (see below), each patient was given a commercial device for semiautomatic oscillometric measurement of blood pressure (model HP 5331, Philips), the accuracy of which had been established in a previous study.9 The patient was asked to obtain a morning and an evening measurement during the 24-hour period in which ambulatory blood pressure monitoring was performed. Instructions were given to (1) obtain the measurement after maintaining the sitting position for 5 minutes, (2) use the arm contralateral to the one used for ambulatory blood pressure, and (3) report the digital display of the blood pressure values on a diary. Morning and evening values were subsequently averaged.
Ambulatory Blood Pressure
Ambulatory blood pressure monitoring was performed with oscillometric Spacelabs 90202 or 90207 equipment. The monitoring equipment was applied in the morning at the end of the medical visit. The cuff was fixed to the nondominant arm, and the device was set to obtain automatic blood pressure readings at 15-minute intervals during the day (from 6 am to midnight) and at 20-minute intervals during the night (from midnight to 6 am). The patient was then sent home with instructions to perform his or her usual activities, hold the arm immobile at the time of the measurements, note on a diary the occurrence of unusual events or poor sleep quality, and return 24 hours later. The blood pressure monitoring was always performed over a working day (Monday through Friday). Before each monitoring session, a few blood pressure readings were taken simultaneously with readings provided by a mercury sphygmomanometer to ensure that on average, the two sets of values did not differ by more than ±5 mm Hg.
Ambulatory blood pressure monitorings were analyzed in a single center. The monitorings in which (1) blood pressure readings regarded as valid by the machine software were <70% of the expected number of readings and/or (2) ≥2 hours showed no valid readings were not considered for further analysis. This consisted of (1) the calculation of 24-hour average SBP, DBP, and heart rate and (2) separate calculations of daytime and nighttime average blood pressures and heart rate values and differences. In the patients whose ambulatory blood pressure data were accepted for the above analysis, the number of 24-hour ambulatory readings were never <95.9% of the expected number of readings. This was the case for the recordings performed before treatment, after 3 and 12 months of treatment, and after the final placebo period (see below).
The echocardiographic studies were performed in the morning, with the subject in a supine left lateral decubitus position, after 30 minutes of rest. Only one physician in each center was responsible for recording the echocardiograms. Echocardiographic tracings were recorded on light-sensitive paper at a speed of 50 mm/s. Two-dimensional imaging of the longitudinal parasternal view was checked to avoid angulation of the ultrasonic beam and consequent changes in the left ventricular shape. Left ventricular internal dimensions, posterior LVWT, and interventricular septum thickness were measured according to the recommendations of the American Society of Echocardiography.10 Left ventricular volumes were calculated with the cube formula. LVM was calculated according to the Penn Convention and indexed to body surface area, calculated by the formula of Dubois and Dubois.11 On the basis of the physician’s calculation, LVH was considered to be present (and the patient was then recruited) if LVMI exceeded 110 g/m2 in women or 131 g/m2 in men.8 All echocardiographic tracings, however, were examined by four expert readers from a previously established center to remove echocardiographic tracings of poor quality by uniform criteria and recalculate the data blindly. The LVMI provided by the control analysis (which was on average 2.9% less than that reported by peripheral centers) was used for calculations of mean data and correlations with blood pressure values. The intraobserver and interobserver coefficients of variation of the “central” measurements were, respectively, 0.5% and 0.8% for left ventricular end-diastolic diameter, 3.2% and 3.9% for septal wall thickness, and 3.4% and 3.9% for posterior LVWT.
The study was designed to prospectively examine whether in patients with hypertension and LVH, reduction of LVM is accounted for to a different degree by treatment-induced reduction of 24-hour average blood pressure compared with clinic blood pressure. On the basis of cross-sectional studies,1 the hypothesis was made that the reduction of LVMI induced by a 12-month antihypertensive treatment would correlate with an r value of .50 to the concomitant reduction in 24-hour average SBP but only with an r value of .30 to the concomitant reduction in clinic supine SBP. This gave a minimum of 158 patients to be studied to detect this difference at a (two-sided) value of α=.05 with 90% power. The number of patients to recruit was set between 220 and 242 to account for dropouts and technical inadequacies of data collection.
After an initial medical visit, previously treated hypertensive patients underwent a 4-week washout period from antihypertensive treatment, and untreated hypertensive patients underwent a 3-week observation period.
After a second medical visit, patients who satisfied recruitment criteria were given lisinopril at a morning dose of 20 mg. After a 1-month treatment, nonresponders to lisinopril (ie, patients in whom clinic DBP at trough was not reduced to <90 mm Hg or by ≥10 mm Hg) were given additional treatment with hydrochlorothiazide at a morning dose of 12.5 mg. After 1 more month, the dose of hydrochlorothiazide was increased to 25 mg daily if there was no satisfactory response. Effective treatment was continued to complete an overall treatment period of 12 months, after which antihypertensive drugs were substituted with placebo tablets, which were administered for 1 additional month. This was done to allow blood pressure to return to pretreatment values and thus confirm that the preceding blood pressure fall had been due to active treatment.
Clinic supine, orthostatic, and random-zero blood pressures were measured before treatment, after 1, 2, 3, 6, and 12 months of treatment, and at the end of the final placebo period. Home blood pressure, ambulatory blood pressure, and echocardiographic data were collected before treatment, after 3 and 12 months of treatment, and at the end of the final placebo period.
On the assumption of a normal data distribution (see below), correlation coefficients (and their 95% confidence limits) were calculated between the reduction in LVMI and LVWT induced by the 12-month treatment and the concomitant SBP reductions. Results were analyzed separately for various subgroups. Correlation coefficients were also calculated for changes in all other blood pressures versus LVMI or LVWT changes and for entry blood pressures versus entry LVMI or LVWT values.
Data from individual patients were summarized as mean±SD. The statistical significance of differences in correlation coefficients and mean values was examined by ANOVA and by Student’s t test for unpaired observations. The Bonferroni correction was applied when multiple comparisons were performed. Normality of data distribution was tested by the Shapiro-Wilks nonparametric test. A value of P<.05 was taken as the minimal level of statistical significance.
The hemodynamic and echocardiographic values before treatment, during treatment, and after the final placebo period are shown in the Table⇓. The numbers of patients analyzed were 206 at entry (age, 50± 9 years), 184 after 12 months of treatment, and 172 after placebo; 4 patients were withdrawn for treatment inadequacy, 12 for side effects, and 21 for lack of compliance to follow-up examinations. Of the 184 patients who completed the 12 months of treatment, 45.6% were treated only with lisinopril, whereas the others required lisinopril plus hydrochlorothiazide. Suitable echocardiographic data were obtained in 184, 160, and 152 patients before treatment, at the end of treatment, and after the final placebo period, respectively. The most important findings were that entry random-zero blood pressures were only slightly (and nonsignificantly) higher than clinic blood pressures, whereas entry 24-hour average blood pressures were markedly lower than clinic blood pressures and entry home blood pressures were lower than clinic blood pressure but significantly greater than daytime and 24-hour average blood pressures. Clinic and 24-hour average SBPs showed a marked reduction at month 3 of treatment with no further change at month 12 and a return toward pretreatment values after the final placebo period (Fig 1⇓, top). The correlation coefficients between the changes in these pressures after 12 months of treatment were r=.47 and r=.40 for SBP and DBP, respectively (P<.01 for both). Clinic and 24-hour heart rates were similar before treatment and showed no change during treatment. LVMI was much above the cutoff values at entry and showed a modest but significant reduction even after only 3 months of treatment. The reduction was more marked after 12 months of treatment, with a slight increase after the final placebo period. LVWT was also significantly reduced at month 3 and more so at month 12 of treatment, with no increase after the final placebo period (Fig 1⇓, bottom). Similar results were obtained when comparisons were limited to subjects who completed the whole study. Entry blood pressures and LVMI showed a normal distribution. This was also the case for the changes in blood pressure and LVMI induced by the 12-month treatment (data not shown).
Fig 2⇓ shows the relationship between SBP and LVMI. At entry, LVMI showed no correlation with clinic SBP but a significant correlation with 24-hour average SBP. The change of LVMI after the 12-month treatment was also not related to the 12th-month change in clinic SBP, whereas the correlation with the 12th-month change in 24-hour average SBP was statistically significant. The treatment-induced changes in LVMI correlated significantly with the changes in 24-hour average SBP but not with the change in clinic SBP in men (r=.50, P<.01 and r=.15, P=NS, respectively) and women (r=.29, P=NS and r=.07, P=NS, respectively); in patients previously treated (r=.39, P<.01 and r=.01, P=NS, respectively) or untreated (r=.49, P<.01 and r=.32, P<.05, respectively); and in patients treated with lisinopril (r=.32, P<.05 and r=.09, P=NS, respectively) or lisinopril plus hydrochlorothiazide (r=.48, P<.01 and r=.12, P=NS, respectively).
The relationship between entry and treatment-induced changes in 24-hour average SBP, clinic SBP, and LVMI are also shown in Fig 3⇓ (left panels). Fig 3⇓ further shows, however, that (1) entry values and changes in random-zero and orthostatic blood pressures did not correlate with corresponding LVMI values and (2) entry home blood pressures did not correlate with entry LVMI, but changes in LVMI induced by the 12-month treatment showed a correlation with the concomitant home blood pressure changes, albeit at a barely significant level. Similar results were obtained when LVWT rather than mass index was used, in which case, however, the correlation with home blood pressure changes was seen even less (Fig 4⇓). At variance from entry values, the reduced LVMI and LVWT values achieved after 12-month treatment showed no correlation with the achieved 24-hour average blood pressure values (r never >.14, P=NS in all instances).
The Table⇑ also shows that entry nighttime SBP and DBP were much less than the daytime values and that both were reduced to a similarly marked degree by the 3- and 12-month treatment. As shown in Fig 5⇓, entry and treatment-induced changes in daytime blood pressures were closely correlated with entry and treatment-induced changes in nighttime blood pressures. Entry daytime and nighttime blood pressures correlated significantly with LVMI, as did changes in daytime and nighttime blood pressures induced by treatment with treatment-induced changes in LVMI. LVMI values and changes showed no relationship with corresponding “clinic” and 24-hour average, daytime average, and nighttime average heart rate values (r always <.15). Similar results were obtained for LVWT.
In our hypertensive subjects, long-term administration of lisinopril without or with hydrochlorothiazide allowed blood pressure to be markedly and steadily reduced. This reduction was accompanied by a clear-cut reduction of the LVH or LVWT that existed before administration of these drugs. Pretreatment LVMI or LVWT showed no relationship with clinic supine SBP but a significant relationship with 24-hour average SBP. Furthermore, and more importantly, the reduction in LVMI or LVWT induced by a 12-month antihypertensive treatment showed no relationship with the treatment-induced reduction in clinic supine SBP but a significant relationship with the treatment-induced reduction in 24-hour average SBP. These differences were not due to a bias in clinic blood pressure readings because random-zero SBP values were very similar to clinic SBP values. They were also not due to the prevailing (and perhaps fortuitous) contribution of a specific subgroup of patients, because similar results were observed in men and women, patients previously treated and untreated, and patients given lisinopril as well as those given lisinopril plus hydrochlorothiazide. Finally, the correlation of LVMI with ambulatory but not with clinic blood pressure could not be explained by a wider range of 24-hour average compared with clinic SBP values, because (1) entry SD was similar for clinic and 24-hour average SBPs and (2) similar SDs characterized the clinic and 24-hour average SBP reductions induced by treatment. It can thus be concluded not only that the degree of LVH is more closely reflected by 24-hour average than by clinic SBP but also that a treatment-induced regression of LVH is accounted for more by a reduction in 24-hour average than by a reduction in clinic blood pressure. This provides the first longitudinal evidence from a study with a controlled experimental design of a greater clinical importance of ambulatory SBP than the SBP measured in the traditional fashion. This conclusion holds for ambulatory versus clinic DBP as well as for ambulatory blood pressures versus the clinic blood pressures obtained in the orthostatic position.
Home blood pressure is widely used in the medical practice to diagnose white-coat hypertension, increase patient compliance, and obtain more numerous blood pressure values on which to assess treatment efficacy.12 13 However, the clinical importance of home blood pressure data has not been assessed in epidemiological or interventional studies. In our patients, no relationship was found between entry LVMI and home blood pressure. However, the reduction of LVMI induced by treatment was related to the reduction of home blood pressure, albeit to a barely significant degree and without a significant relationship with changes in LVWT. Thus, as far as regression of LVH is concerned, treatment-induced changes in home blood pressure may be slightly more predictive than treatment-induced changes in clinic blood pressure but by no means as predictive as ambulatory blood pressure. It should be emphasized, however, that our study design included only two home blood pressure readings over a single day and that therefore the clinical importance of obtaining many more home blood pressure values over different days remains to be assessed.
The clinical importance of daytime and nighttime blood pressures is a controversial issue because some investigators give little value to ambulatory blood pressure monitoring at night,14 whereas others suggest that daytime and nighttime blood pressures are equally important.15 Nighttime blood pressure is also regarded by some investigators as being more important than daytime blood pressure, on the basis of the observation that in hypertensive patients showing little or no nighttime blood pressure fall, end-organ damage is greater16 and cardiovascular events are more frequent4 than in hypertensive patients showing a clear nocturnal blood pressure fall.
Our study offers several data that may increase understanding of this issue. (1) Entry LVMI correlated to a similar degree with daytime and nighttime blood pressure. (2) Reduction of LVMI induced by treatment also showed a similar correlation with reduction in daytime and nighttime blood pressures. (3) In line with previous findings,17 entry and treatment-induced reductions of daytime and nighttime blood pressure values were closely related to each other. Thus, daytime and nighttime blood pressures seem to be similarly important in reflecting LVH and predicting its regression by antihypertensive treatment. However, because they are always linked by a close relationship, these pressures are not totally independent variables to compare for their effect on LVM. This means that the information provided by one variable is contained in the other one, making collection of both sets of data redundant. This appeared to be the case in our patients, in whom reduction of nighttime blood pressure did not add to the prediction of LVH regression provided by daytime blood pressure reduction, suggesting that as far as improvement of end-organ damage by antihypertensive treatment is concerned, daytime blood pressure monitoring suffices.
Several other findings of our study deserve to be discussed. (1) Previous studies suggested that lack of nocturnal blood pressure fall is particularly frequent in severe hypertension, thereby representing a marker of this condition.18 However, the patients of the SAMPLE study had a marked LVH and a pronounced elevation of clinic, home, and ambulatory blood pressure values above the respective normal ranges.9 Yet, both before and during treatment, they showed a nocturnal reduction in blood pressure that was on average similar to the one described in a recent large-scale population study,9 indicating that under these circumstances, day-to-night blood pressure modulation is largely preserved. (2) In the previously mentioned population study,9 home average blood pressure was found to be much lower than clinic blood pressure, lower than daytime blood pressure, and similar to 24-hour average blood pressure. In the same study, however, the difference between home and 24-hour average blood pressure increased progressively as clinic blood pressure increased. Furthermore, in the patients of the present study, home blood pressure was significantly lower than clinic blood pressure but also significantly greater than daytime and 24-hour average blood pressures. Thus, home blood pressure reflects 24-hour average blood pressure in the general population, whereas in hypertensive patients, its value is closer to clinic blood pressure. The reasons for this phenomenon are unknown, but an obvious possibility is that awareness of their clinical problem makes hypertensive patients prone to an alerting reaction and a blood pressure rise19 during home blood pressure measurements as well. (3) Entry and treatment-induced changes in LVMI showed a slightly but consistently greater relationship with ambulatory SBP than with DBP. This might depend on the greater reading accuracy of the oscillometric device we used for SBP than for DBP.20 It may also reflect, however, the more direct relationship of SBP to an important determinant of LVH such as cardiac stress.14 In contrast, LVMI and its changes did not show any relationship with clinic or 24-hour average heart rate, despite the potential relevance of heart rate values to cardiac workload and oxygen consumption. (4) LVM showed some reduction even after 3 months of treatment and a tendency to increase after the final month of placebo, which was characterized by a blood pressure return to almost pretreatment values. Thus, LVM may change rather quickly in response to blood pressure changes. Even a 1-year satisfactory blood pressure control does not prevent blood pressure from rapidly regaining hypertensive values when treatment is withdrawn.
Three final considerations should be made. (1) It may appear surprising that contrary to several cross-sectional data,1 clinic blood pressure showed no correlation whatsoever with LVMI. One possible explanation of this finding is that by focusing on the presence of a marked LVH and a relatively pronounced blood pressure elevation, the patients selected had a limited range of LVMI and blood pressure values that negatively affected the correlation between these two variables. (2) It may also appear surprising that LVMI achieved after 12 months of treatment no longer correlated with the achieved ambulatory blood pressure values. Again, this might have occurred because after treatment, LVMI and ambulatory blood pressure were distributed over a narrower range. Another possibility (not exclusive of the previous one), however, is that 24-hour blood pressure contributes to only a portion of the increase in LVM and that what is left after ambulatory blood pressure lowering represents the influence of trophic factors that affect cardiac growth to variable degrees.21
Finally, LVH is an intermediate end point, and thus the prognostic value of ambulatory versus home and clinic blood pressures will have to be confirmed by a study based on cardiovascular morbidity and mortality. However, in hypertensive patients and in the general population, LVH is associated with a greater risk of myocardial infarction, stroke, sudden death, and death from any cause.6 7 Furthermore, regression of LVH has been shown to be associated with an improvement of cardiac functions, an increase in coronary reserve, and a reduction of cardiac arrhythmias.22 Finally, a regression of LVH has been associated with a reduction of cardiovascular morbid events.23 24 The prognostic value of this intermediate end point therefore appears to be sufficiently well established, making the conclusion of the present study clinically relevant.
Selected Abbreviations and Acronyms
|DBP||=||diastolic blood pressure|
|LVH||=||left ventricular hypertrophy|
|LVM||=||left ventricular mass|
|LVMI||=||left ventricular mass index|
|LVWT||=||left ventricular wall thickness|
|SBP||=||systolic blood pressure|
The following investigators participated in the SAMPLE Study Group.
Coordinators: G. Mancia and A. Zanchetti (University of Milan).
Investigators: E. Agabiti-Rosei, M.L. Muiesan (University of Brescia); G. Benemio, G. Schillaci (Città della Pieve Hospital); R. De Cesaris, G. Ranieri (University of Bari); R. Fogari (University of Pavia); G. Mancia, R. Sega (University of Milan, S. Gerardo Hospital, Monza); A. Pessina, P. Palatini (University of Padua); C. Porcellati, P. Verdecchia (Silvestrini Hospital, Perugia); A. Rappelli, P. Dessı̀-Fulgheri (University of Ancona); A. Salvetti, V. Di Legge (University of Pisa); B. Trimarco (University of Naples).
Ambulatory Blood Pressure Analysis Center: S. Omboni, G. Parati, A. Ravogli, A. Villani (Centro di Fisiologia Clinica e Ipertensione, Ospedale Maggiore, Milan).
Echocardiographic Reading Center: C. Cuspidi, L. Sampieri (Ospedale Maggiore, Milan); S. Calebich, M.L. Muiesan (University of Brescia).
Statistical analysis: A. Vanasia, G.M. Villa (Zeneca, Milan).
This study was sponsored by a research grant from Zeneca.
- Received June 5, 1996.
- Revision received September 23, 1996.
- Accepted October 7, 1996.
- Copyright © 1997 by American Heart Association
Mancia G, Di Rienzo M, Parati G. Ambulatory blood pressure monitoring in hypertension research and clinical practice. Hypertension. 1993;21:510-524.
Perloff D, Sokolow M, Cowan R. The prognostic value of ambulatory blood pressure monitoring in treated hypertensive patients. J Hypertens. 1991;9(suppl 1):S33-S40.
Verdecchia P, Porcellati C, Schillaci G, Borgioni C, Ciucci A, Battistelli M, Guerrieri M, Gatteschi C, Zampi I, Santucci A, Santucci C, Reboldi G. Ambulatory blood pressure: an independent predictor of prognosis in essential hypertension. Hypertension. 1994;24:793-801.
Guidelines Subcommittee of the WHO/ISH Mild Hypertension Liaison Committee. 1993 guidelines for the management of mild hypertension: memorandum from a WHO/ISH meeting. Hypertension. 1993;22:392-403.
Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass and geometry to morbidity and mortality in uncomplicated essential hypertension. Ann Intern Med. 1991;114:345-352.
Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in man. Circulation. 1977;55:613-618.
Sahn DY, De Maria A, Kinlo J, Weyman A, the Committee on M-mode Standardization of the Americal Society of Echocardiography. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation. 1978;58:1072-1083.
Dubois D, Dubois EF. Clinical calorimetry: a formula to estimate the approximate surface mass if height and weight are known. Arch Intern Med. 1916;17:863-871.
Julius S, Mejia A, Jones K, Krause L, Schork N, van de Ven C, Johnson E, Petrin J, Sekkarie MA, Kjeldsen SE, Schmouder R, Gupta R, Ferraro J, Nazzaro P, Weissfeld J. ‘White coat’ versus ‘sustained’ borderline hypertension in Tecumseh, Michigan. Hypertension. 1990;16:617-623.
Ménard J, Chatellier G, Day M, Vaur L. Self-measurement of blood pressure at home to evaluate drug effect by the trough:peak ratio. J Hypertens. 1994;12(suppl 8):S21-S25.
Devereux RB, Pickering TG, Harshfield GA, Kleinert HD, Denby L, Clark L, Pregibon D, Jason M, Kleiner B, Borer JS, Laragh JH. Left ventricular hypertrophy in patients with hypertension: importance of blood pressure response to regularly recurring stress. Circulation. 1983;68:470-476.
Mancia G, Parati G, Albini F, Villani A. Circadian blood pressure variations and their impact on disease. J Cardiovasc Pharmacol. 1988;12(suppl 7):S11-S17.
Groppelli A, Omboni S, Parati G, Mancia G. Evaluation of noninvasive blood pressure monitoring devices Spacelabs 90202 and 90207 versus resting and ambulatory 24-hour intra-arterial blood pressure. Hypertension. 1992;20:227-232.
Frohlich ED. Overview of hemodynamic and non-hemodynamic factors associated with LVH. J Mol Cell Cardiol. 1989;21:3-10.
Devereux RB. Hypertensive cardiac hypertrophy: pathophysiologic and clinical characteristics. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis and Management. New York, NY: Raven Press Ltd; 1990:359-377.
Yurenev AP, Dyakonova HG, Novikov ID, Vitols A, Pahl L, Haynemann G, Wallrabe D, Tsifkova R, Romanovska L, Niderle P, Tsiskarishvili DL, Davarashvili T, Gelovani K, Kochachidze T, Balash A. Management of essential hypertension in patients with different degrees of left ventricular hypertrophy: multicenter trial. Am J Hypertens. 1992;5:182S-189S.