AJH 1997;10:1302–1307 BRIEF COMMUNICATIONS

White-Coat Resistant Andrea Mezzetti, Sante D. Pierdomenico, Fabrizio Costantini, Ferdinando Romano, Anna Bucci, Mario Di Gioacchino, and Franco Cuccurullo Downloaded from https://academic.oup.com/ajh/article/10/11/1302/149451 by guest on 04 October 2021

The aim of this study was to evaluate whether hypertensives and 17 and 14, respectively, were sustained hypertensives with high clinic blood white-coat resistant hypertensives. Interestingly, in pressure, despite multiple drug treatment, show a white-coat resistant hypertensives the large true resistant hypertension or a ‘‘white-coat effect,’’ differences between clinic and ambulatory daytime and whether the pretreatment white-coat effect is (white-coat effect), recorded before maintained despite pharmacological therapy. The treatment assignment, were not affected by drugs occurrence of resistant hypertension was and remained constant over time. Left ventricular determined in 250 consecutive essential mass index in white-coat resistant hypertensives hypertensives who had had an ambulatory blood was significantly lower than in truly resistant pressure monitoring before treatment assignment. hypertensives, suggesting that prognosis could Twenty-seven of 250 hypertensives with differ between these groups. In this study, using persistently high clinic blood pressure despite 3 either our internal standards or some other cutoffs months of adequate pharmacological therapy reported in the literature, the white-coat underwent further ambulatory blood pressure phenomenon was an important cause of resistant monitoring. Using our internal standards, seven hypertension. The use of ambulatory blood patients had a true resistant hypertension whereas pressure monitoring in these patients may avoid 20 subjects showed a large white-coat effect (white- misdiagnosis of resistant hypertension, coat resistant hypertension), ie, high clinic blood unnecessary overtreatment, and expensive pressure (>140/90) but ‘‘normal’’ ambulatory procedures to look for possible secondary daytime (<139/90 mm Hg) and 24 h (135/85 mm hypertension. Am J Hypertens 1997;10:1302–1307 Hg) blood pressure. Using other cutoff points for © 1997 American Journal of Hypertension, Ltd. ambulatory blood pressure, 134/90 and 135/85 mm Hg for daytime blood pressure, 10 and 13 patients, KEY WORDS: White-coat effect, ambulatory blood respectively, were reclassified as true resistant pressure monitoring, resistant hypertension.

esistant hypertension has been defined as medical therapy, , and non- uncontrolled blood pressure despite stan- compliance.5 It has also been reported that some pa- dard triple therapy,1,2 and its prevalence tients receiving antihypertensive therapy show a has been estimated to be approximately be- ‘‘white-coat effect’’6–8 that could cause an overestima- tweenR 2.9% and 13%.3,4 Drug resistant hypertension tion of their real blood pressure. The amount of this may be due to different causes, such as suboptimal phenomenon may be variable and its impact on the

Received October 31, 1996. Accepted June 24, 1997. Address correspondence and reprint requests to Andrea Mezzetti, From the ‘‘Centro per lo Studio dell’ Ipertensione Arteriosa, delle MD, Istituto di Fisiopatologia Medica, Policlinico ‘‘S.S. Annun- Dislipidemie e dell’ Arteriosclerosi,’’ Istituto di Fisiopatologia Med- ziata,’’ Via dei Vestini, 66013—Chieti Scalo (Chieti), Italy; e-mail: ica (AM, SDP, FC, AB, MDG, FC) and the Chair of Hygiene (FR), [email protected] University ‘‘Gabriele D’Annunzio,’’ Chieti, Italy.

© 1997 by the American Journal of Hypertension, Ltd. 0895-7061/97/$17.00 Published by Elsevier Science, Inc. PII S0895-7061(97)00318-X AJH–NOVEMBER 1997–VOL. 10, NO. 11, PART 1 RESISTANT HYPERTENSION 1303

occurrence of resistant hypertension (white-coat resis- an excessive stiffness of the large brachial tant hypertension) has not yet been clearly evaluated. (pseudohypertension).13 Moreover, in previous studies,5–8 patients with white- coat hypertension,9–11 ie, high clinic blood pressure Ambulatory Blood Pressure Monitoring Ambulatory monitoring was performed with a portable noninva- but ‘‘normal’’ ambulatory blood pressure in the ab- 14 sence of drug treatment, were not clearly differenti- sive recorder (SpaceLabs 90207, Redmond, WA) on a ated from true hypertensives with a white-coat effect. day of typical activity after a clinic visit. Ambulatory Ambulatory blood pressure monitoring, performed blood pressure readings were obtained automatically before and after drug treatment, is the only method to at 15 min intervals from 6 am to midnight, and at 30 differentiate white-coat resistant hypertensives from min intervals from midnight to 6 am. Each time a patients with a true drug-resistant hypertension or reading was taken, subjects were instructed to remain motionless and to record their activity on a diary from subjects with white-coat hypertension. Downloaded from https://academic.oup.com/ajh/article/10/11/1302/149451 by guest on 04 October 2021 The aim of this study was to evaluate whether sus- sheet. On completion of ambulatory blood pressure tained hypertensives with high clinic blood pressure monitoring the data were analyzed by computer, us- despite multiple drug treatment show a resistant hy- ing a program designed to perform editing and inter- pertension or a white-coat effect, and whether the val statistics. All patients included in the study had pretreatment white-coat effect is maintained despite recordings of good technical quality. The following treatment. ambulatory blood pressure monitoring parameters were evaluated: average daytime (awake period) sys- METHODS tolic and diastolic blood pressure; average nighttime (sleep period) systolic and diastolic blood pressure; Patients and Study Design We studied 27 subjects and average 24 h systolic and diastolic blood pressure. (14 men and 13 women, age 56 Ϯ 11 years) with a White-coat resistant hypertension was defined as high previous diagnosis of true hypertension confirmed clinic blood pressure, despite triple treatment for at by ambulatory blood pressure monitoring, who had least 3 months, but ‘‘normal’’ ambulatory daytime and persistently elevated clinic blood pressure (Ͼ140/90 24 h blood pressure. There is no general agreement mm Hg) despite being on a rational triple drug concerning the normal limits of ambulatory blood regimen for at least 3 months. These patients under- pressure values, and different upper limits of nor- went a further ambulatory blood pressure monitor- malcy have been used in previous studies. Thus, ing to confirm the clinical diagnosis of resistant rather than using a previously reported cut-off point hypertension. They represented the 11% of 250 sus- and in light of possible influence of geographical lo- tained hypertensives who were consecutively sub- cation on blood pressure, we have chosen the blood mitted to ambulatory blood pressure monitoring pressure values representing the upper limits (mean ϩ and were regularly followed by our secondary hy- 2 SD) of a clinically normotensive population coming pertension referral center. Patients referred to our from our geographical area (Chieti, Abruzzo, Italy). center for suspected resistant hypertension who had 9 The resulting upper limits for average 24 h and av- not been submitted to ambulatory blood pressure erage daytime ambulatory blood pressure were monitoring before treatment assignment, and those 135/85 and 139/90 mm Hg, respectively. Patients found to have secondary hypertension and white were also reclassified according to different upper coat hypertension, were excluded from the study. limits of ambulatory blood pressure.9,15 Only antihypertensive medications were being taken at the time of the study. All participants read Echocardiographic Study Echocardiographic exam- and signed an informed consent. ination was performed using a Hewlett-Packard 77030A (Andover, MA) ultrasound imaging system Measurements Office Blood Pressure Clinic systolic equipped with a 2.5 or 3.5 MHz transducer. Interven- and diastolic blood pressure recordings were per- tricular septum, posterior wall, and left ventricular formed on the same arm, with the patient sitting after 12 dimension at the end of diastole were measured ac- 10 min quiet rest, according to a standard technique. cording to the American Society of Echocardiography Phase V was used to determine diastolic blood pres- recommendations.16 Left ventricular mass was calcu- sure. Measurements were performed in triplicate and lated by the formula introduced by Devereux et al17 on the average value was used as the blood pressure for the basis of necropsy validation studies. The left ven- the visit. Resistant hypertension was defined as clinic tricular mass index (LVMI) was calculated by dividing blood pressure Ͼ 140/90 mm Hg, despite a triple drug the left ventricular mass by body surface area. regimen, in at least 3 visits (1 week apart). In patients with resistant hypertension the ‘‘Osler maneuver’’ Medication Adherence Assessment All subjects was used to differentiate true hypertensives from were interviewed to collect information about nutri- those with falsely elevated blood pressure because of tional history, smoking habit, and other relevant life- 1304 MEZZETTI ET AL AJH–NOVEMBER 1997–VOL. 10, NO. 11, PART 1

TABLE 1. CLINIC CHARACTERISTICS AND DRUG noncompliance, the exact prevalence of true resistant THERAPY OF PATIENTS WITH RESISTANT hypertension is reduced from 2.8% to 1.2%. In all HYPERTENSION (RH) AND WHITE-COAT patients with resistant hypertension the ‘‘Osler ma- RESISTANT HYPERTENSION (WC-RH) neuver’’ proved negative. There was no significant Variable RH WC-RH difference in sex distribution, body mass index, so- dium intake (evaluated by urinary sodium excretion), Patients (n) 7 20 Men/women (n/n) 4/3 10/10 alcohol intake, smoking habits, prevalence of diabetes Age (years) 49 Ϯ 12 59 Ϯ 9* mellitus, and medication regimens (Table 1) between Body mass index (kg/m2)27Ϯ326Ϯ2 the two groups. Patients with a white-coat resistant Urinary sodium excretion (mmol/day) 90 Ϯ 15 95 Ϯ 11 hypertension were older (P Ͻ .03) than those with true Alcohol overconsumption (n) 1 1 resistant hypertension. There was no difference in

Smokers (n) 2 5 clinic blood pressure between the two groups (Table Downloaded from https://academic.oup.com/ajh/article/10/11/1302/149451 by guest on 04 October 2021 Diabetes mellitus (n) 1 3 2). Ambulatory blood pressures were significantly H ϩ A ϩ N (n) 3 9 higher in patients with resistant hypertension than in H ϩ E ϩ NorV(n) 2 8 those with white-coat effect (Table 2). Clinic blood H ϩ N ϩ C (n) 2 3 pressure was significantly higher than daytime ambu- Values are expressed as mean Ϯ SD. latory blood pressure in patients with white-coat re- H, hydrochlorothiazide (25 mg); A, atenolol (100 mg); N, nifedipine GITS sistant hypertension (149 Ϯ 4/97 Ϯ 2 v 127 Ϯ 7/78 Ϯ (90 mg); E, enalapril (20 to 40 mg); V, long-acting verapamil (240 or 360 5mmHg,PϽ.0001), whereas no significant differ- mg); C, clonidine patch (0.2 mg). ence was present between clinic and ambulatory day- *P Ͻ .03. time blood pressure in the group with true resistant hypertension. The difference between clinic and day- time systolic and diastolic blood pressure, recorded before treatment assignment, was significantly higher style factors. A comprehensive questionnaire and in- in the white-coat resistant group when compared with terview were used to evaluate the patient’s adherence that in a group of 20 age-, sex-, and weight-matched to the drug regimen as well as the adequacy and responders to antihypertensive therapy (20 Ϯ 3/18 Ϯ tolerability of the antihypertensive therapy. Items in- 2 v 7 Ϯ 2/5 Ϯ 2mmHg,PϽ.0001) and to truly cluded side effects experienced, reasons for not full- resistant hypertensives (20 Ϯ 3/18 Ϯ 2 v 5 Ϯ 2/3 Ϯ 2 filling prescriptions, difficulty in swallowing pills or mm Hg, P Ͻ .0001). Noteworthy, in white-coat resis- opening medication containers, and medication taking tant hypertensives, despite a significant reduction in behavior, such as taking drugs on the day of clinic systolic and diastolic blood pressure, the value of this visit or discontinuing drugs without physician con- difference was not modified by the drug treatment sent. (20 Ϯ 2/18 Ϯ 2 v 21 Ϯ 3/18 Ϯ 3 mm Hg, before versus Statistical Analysis Standard descriptive statistics, after treatment, respectively; P ϭ NS), showing very unpaired and paired Student’s t tests, and ␹2 tests were performed, where appropriate.18 Values are re- ported as means Ϯ SD. The P Ͻ .05 level of signifi- cance was adopted for all tests. TABLE 2. CLINIC AND AMBULATORY BLOOD PRESSURE VALUES IN SEVEN PATIENTS WITH RESULTS RESISTANT HYPERTENSION (RH) AND 20 PATIENTS WITH WHITE-COAT RESISTANT The ambulatory blood pressure monitoring showed HYPERTENSION (WC-RH) TREATED WITH that only seven out of 27 patients (26%) had a truly MULTIPLE THERAPY resistant hypertension, whereas 20 patients (74%) pre- sented a white-coat resistant hypertension. Subjects Variable RH WC-RH with resistant hypertension represented 2.8% of the Clinic systolic BP (mm Hg) 151 Ϯ 7 149 Ϯ 4 original population of 250 consecutive untreated es- Clinic diastolic BP (mm Hg) 98 Ϯ 397Ϯ2 sential hypertensives. The cause of resistant hyperten- 24 h systolic BP (mm Hg) 142 Ϯ 6 121 Ϯ 10* sion was suboptimal therapy in one patient (patient 24 h diastolic BP (mm Hg) 93 Ϯ 275Ϯ5* did not increase drug dosage as prescribed by the Daytime systolic BP (mm Hg) 148 Ϯ 9 127 Ϯ 7* physician), noncompliance in three patients (patients Daytime diastolic BP (mm Hg) 97 Ϯ 478Ϯ5* reported to not always take all the drugs), and unde- Nighttime systolic BP (mm Hg) 131 Ϯ 7 111 Ϯ 11* Nighttime diastolic BP (mm Hg) 83 Ϯ 370Ϯ7* termined in three patients. Recently, Alderman et al restricted the diagnosis of resistance only to patients Values are expressed as mean Ϯ SD. adherent to medication.3 In our study, if we eliminate BP, blood pressure. one patient with suboptimal treatment and three with *P Ͻ .0001. AJH–NOVEMBER 1997–VOL. 10, NO. 11, PART 1 RESISTANT HYPERTENSION 1305

low coefficients of variation estimated on an individ- ance were the most common causes of apparently ual basis (5.2% Ϯ 1.5%, mean Ϯ SD). resistant hypertension in their patients.5 However, Left ventricular mass index was significantly higher they used different selection criteria and did not in patients with resistant hypertension than in those systematically perform ambulatory blood pressure with white-coat resistant hypertension (150 Ϯ 9 v monitoring to rule out the presence of a white-coat 126 Ϯ 10 g/m2, P Ͻ .0001). Moreover, left ventricular effect in all patients suspected to have resistant mass index was not significantly different between hypertension.5 white-coat resistant hypertensives and 20 age-, sex-, In a recent study Mejia et al,7 who performed an and weight-matched responder hypertensives (126 Ϯ intraarterial monitoring in 15 subjects with suspected 10 v 125 Ϯ 10 g/m2, P ϭ NS). resistant hypertension, reported that only eight pa- In the white-coat resistant hypertensive group, drug tients had a true resistance, whereas seven showed therapy was reduced in 6 patients because of low high clinic blood pressure but normal mean intra- Downloaded from https://academic.oup.com/ajh/article/10/11/1302/149451 by guest on 04 October 2021 ambulatory blood pressure levels and symptoms sug- arterial blood pressure.7 Four of these seven subjects gestive of . (27%) had ‘‘office resistance,’’ whereas in the other When our patients were reclassified according to cases the cause of the resistance was represented by the cutoff points at first suggested by Pickering et al pseudohypertension or ‘‘cuff-inflation’’ hypertension. (134/90 mm Hg for daytime blood pressure),9 three This lower prevalence of white-coat resistant hyper- white-coat resistant hypertensives were reclassified as tension may depend on the fact that their study7 was true resistant hypertensives for systolic blood pressure performed in a small number of highly selected pa- (10 true resistant and 17 white-coat resistant hyperten- tients, whereas our data come from a population of sive patients). When more restrictive criteria were 250 consecutive untreated hypertensives. used (135/85 mm Hg for daytime blood pressure), as In our study, because we have used non-invasive suggested by Pickering,15 six white-coat resistant hy- ambulatory monitoring, it is not possible to exclude pertensives were reclassified as true resistant hyper- completely the possibility that pseudohypertension or tensives (13 true resistant and 14 white-coat resistant ‘‘cuff-inflation’’ hypertension may have contributed in hypertensive patients). determining resistant hypertension in our patients. Pseudohypertension is generally associated with a DISCUSSION positive response to the Osler maneuver.13 The rela- Resistant hypertension does not represent an un- tively young age of our patients with resistant hyper- common clinical problem and it may depend on tension (Table 1) and their negative response to the different factors.5 It has recently been reported that Osler maneuver seem to exclude the presence of some treated hypertensives show a white-coat effect pseudohypertension. Mejia et al7 have recently re- that causes an overestimation of their blood pres- ported that ‘‘cuff inflation hypertension,’’ ie, a marked sure and underrates the response to pharmacologi- rise of intraarterial blood pressure occurring during cal treatment.6–8 This phenomenon is distinct from cuff inflation, may represent a possible cause of pseu- white-coat hypertension, that is, high clinic blood doresistance. Ambulatory blood pressure monitoring, pressure but normal pressure at other times in un- as previously mentioned, cannot exclude the possibil- treated subjects.9–11 In our study, only seven of 27 ity that some of our patients resistant to treatment patients (26%) with high clinic blood pressure de- (that is, the three patients in which the cause of resis- spite triple therapy showed resistant hypertension, tance was undetermined) had a ‘‘cuff-inflation hyper- whereas the remaining (74%) presented a large tension’’; nonetheless this phenomenon does not in- white-coat effect in clinic and normal ambulatory fluence the high prevalence of white-coat resistant blood pressure, that is, white-coat resistant hyper- hypertension found in our study. tension. Thus, our data seem to demonstrate that, It is important to outline that ambulatory blood when secondary and white-coat hypertension have pressure decreased significantly in patients with a been excluded and volume overload is not present, large white-coat effect (Table 2), and in some cases a large white-coat phenomenon represents an im- we were forced to reduce antihypertensive treat- portant cause of ‘‘apparently’’ resistant hyperten- ment because of an excessive fall in blood pressure. sion. Moreover, its presence in untreated hyperten- Thus, in white-coat resistant hypertensives, ambu- sives may represent a predicting factor of a future latory blood pressure monitoring represents an im- false resistance to pharmacological treatment. In portant tool to check treatment efficacy to avoid contrast with our findings, Yakovlevitch and Black overtreatment. reported a very low prevalence of ‘‘office resis- The differences between clinic and daytime blood tance.’’5 In fact, suboptimal therapy, previously un- pressure recorded before treatment in white-coat re- diagnosed secondary hypertension, and noncompli- sistant hypertensives were significantly higher than 1306 MEZZETTI ET AL AJH–NOVEMBER 1997–VOL. 10, NO. 11, PART 1

those found in truly resistant and in a well matched Evaluation and Treatment of High Blood Pressure. group of sustained hypertensives who responded to Arch Intern Med 1993;153:154–183. the therapy. Interestingly, in white-coat resistant hy- 3. Alderman MH, Schoenbaum EE: Detection and treat- pertensives the treatment did not significantly influ- ment of hypertension at the work site. N Engl J Med ence this difference, which remained constant over 1975;293:65–68. time in each patient group, both for systolic and dia- 4. Swales JD, Bing RF, Pohl JE, et al: Treatment of refrac- stolic blood pressures. tory hypertension. Lancet 1982;i:894–896. It is known that clinic blood pressure is associated 5. Yakovlevitch M, Black HR: Resistant hypertension in with cardiovascular risk; however, it has also been a tertiary care clinic. Arch Intern Med 1991;151:1786– 1792. shown that ambulatory blood pressure is an inde- pendent predictor of prognosis.19–21 The prognostic 6. Waeber JB, Scherrer U, Petrillo A, et al: Are some hypertensive patients overtreated? A prospective significance of a large white-coat effect in true hy- study of ambulatory blood pressure monitoring. Lan- Downloaded from https://academic.oup.com/ajh/article/10/11/1302/149451 by guest on 04 October 2021 pertensive patients is at present unknown. Prospec- cet 1987;ii:732–734. tive studies are needed to evaluate whether or not a 7. Mejia AD, Egan BM, Schork NJ, Zweifler AJ: Artefacts large white-coat effect adds risk above that pro- in measurement of blood pressure and lack of target vided by ambulatory blood pressure. Verdecchia et organ involvement in the assessment of patients with al22 have recently reported that white-coat effect treatment resistant hypertension. Ann Intern Med 1990;112:270–277. does not predict cardiovascular morbidity and mor- tality in hypertensive patients. Moreover, we have 8. Myers MG, Reeves RA: White-coat phenomenon in found23 that in the white-coat resistant hyperten- patients receiving antihypertensive therapy. Am J Hy- pertens 1991;4:844–849. sives left ventricular mass index was significantly lower than in the truly resistant hypertensives and 9. Pickering TG, James GD, Boddie C, et al: How com- mon is white-coat hypertension? JAMA 1988;259:225– similar to that in the responders, potentially sug- 228. gesting that these groups are characterized by dif- 10. Pierdomenico SD, Mezzetti A, Lapenna D, et al: ferent prognostic implications. ‘‘White-coat’’ hypertension in patients with newly di- The present study has a limitation. We have used agnosed hypertension: evaluation of prevalence by the cut-off points obtained by a normotensive popu- ambulatory monitoring and impact on cost of health lation studied in our geographical area; however, we care. Eur Heart J 1995;16:692–697. cannot exclude the possibility that these upper limits 11. Pierdomenico SD, Lapenna D, Guglielmi MD, et al: of normalcy could somehow overestimate the ‘‘true’’ Target organ status and serum lipids in patients with white coat hypertension. Hypertension 1995;26:801– upper limit of ambulatoty blood pressure, thus lead- 807. ing to an overestimation of the prevalence of white- coat resistant hypertension. In any case, by our using 12. Kirkendall WM, Feinleib M, Freis ED, Mark AL: Rec- 15 ommendations for human blood pressure determina- more restrictive criteria as well, half of our clinically tions by sphygmomanometers. Circulation 1980;62: resistant hypertensives showed a white-coat resistant 1146A–1155A. hypertension, suggesting that this phenomenon is 13. Messerli FH, Ventura HO, Amodeo C: Osler’s maneu- clinically relevant. ver and pseudohypertension. N Engl J Med 1985;312: In conclusion, this study shows that many patients 1548–1551. with apparently resistant hypertension show a large 14. Groppelli A, Omboni S, Ravogli A, et al: Validation of white-coat effect. Ambulatory blood pressure moni- the SpaceLabs 90202 and 90207 devices for ambula- toring can reliably identify this subset of patients, thus tory blood pressure monitoring by comparison with intra-arterial resting and ambulatory measurements. avoiding unnecessary and possibly dangerous over- J Hypertens 1991;9:334–335. treatment or expensive and invasive procedures to 15. Pickering TG: Recommendations for the use of home look for possible secondary hypertension. Further- (self) and ambulatory blood pressure monitoring. more, we have shown that, in these patients, the large Am J Hypertens 1996;9:1–11. white-coat effect is not influenced by active pharma- 16. Sahn DJ, De Maria A, Kisslo J, Weyman A, for cological treatment. the Committee on M-mode Standardization of the American Society of Echocardiography: Recom- REFERENCES mendations regarding quantitation in M-mode echocardiography: results of a survey of echocar- 1. Gifford RW, Jr.: Resistant hypertension: introduction diographic measurements. Circulation 1978;58: and definitions. Hypertension 1988;11(3 Pt 2):II65– 1072–1083. II66. 17. Devereux RB, Alonso DR, Lutas EM, et al: Echocar- 2. Joint National Committee on Detection, Evaluation diographic assessment of left ventricular hypertrophy: and Treatment of High Blood Pressure: The fifth re- comparison to necropsy findings. Am J Cardiol 1986; port of the Joint National Committee on Detection, 57:450–458. AJH–NOVEMBER 1997–VOL. 10, NO. 11, PART 1 RESISTANT HYPERTENSION 1307

18. Altman DG: Practical Statistics for Medical Research. tory blood pressure: an independent predictor of Chapman & Hall, London, 1992, pp 179–276. prognosis in . Hypertension 19. Perloff D, Sokolow M, Cowan R: The prognostic value 1994;24:793–801. of ambulatory blood pressure. JAMA 1983;249:2792– 22. Verdecchia P, Schillaci G, Borgioni C, Ciucci A, Por- 2798. cellati C: Prognostic significance of the white-coat ef- 20. Perloff D, Sokolow M, Cowan RM, Justes RP: Prog- fect. Hypertension 1997;29:1218–1224. nostic value of ambulatory blood pressure measure- 23. Koren MJ, Devereux RB, Casale PN, et al: Relation of ments: further analyses. J Hypertens 1989;7(suppl 3): left ventricular mass and geometry to morbidity and S3–S10. mortality in uncomplicated essential hypertension. 21. Verdecchia P, Porcellati C, Schillaci G, et al: Ambula- Ann Intern Med 1991;114:345–352. Downloaded from https://academic.oup.com/ajh/article/10/11/1302/149451 by guest on 04 October 2021