ORIGINAL CONTRIBUTIONS Epidemiology

Blood PressureVariability and CardiovascularRisk in Treated Hypertensive Patients

Sante D. Pierdomenico, Domenico Lapenna,Roberta Di Tommaso, Silvio Di Carlo, Anna L. Esposito,Rocco Di Mascio, Enzo Ballone, Franco Cuccurullo, and Andrea Mezzetti

AJH 2006; 19:991–997

Background: The independent prognostic value ofblood pressure (BP) variability in treated hypertension isnot yet clear. We investigated the relationship between BPvariability, evaluated by noninvasive monitoring, and car-diovascular outcome in treated hypertensive patients.

Methods: The occurrence of fatal and nonfatal cardio-vascular events was evaluated in 1472 treated patients. Sub-jects with the standard deviation of daytime or night-timesystolic BP below or above the median of the populationwere classified as having low or high BP variability. Specif-ically, 738 and 734 patients had low and high daytime BPvariability, respectively, and 739 and 733 subjects had lowand high night-time BP variability, respectively.

Results: During follow-up (4.88 � 2.9 years, range0.2–11.6 years) there were 119 events. The event rates per100 patient-years in subjects with low and high BP vari-ability according to daytime BP were 1.18 and 2.01, re-spectively, and in those with low and high BP variability

according to night-time BP were 1.2 and 2.05, respec-

Università “Gabriele d’Annunzio,” and Centro di Ricerca Clinica, Fonda-

© 2006 by the American Journal of Hypertension, Ltd.Published by Elsevier Inc.

tively. Event-free survival was significantly different be-tween the low and high BP variability groups (P � .006for both daytime and night-time BP). However, after ad-justment for other covariates in a Cox multivariate analy-sis, the adverse prognostic relevance of high BP variabilitywas no longer detectable, whereas age, smoking habit,LDL cholesterol, diabetes, previous events, LV hypertro-phy, and daytime or night-time systolic BP resulted inde-pendent predictors of risk.

Conclusions: Increased BP variability is associated withhigher incidence of cardiovascular events, but also with otherrelevant prognostic factors. Indeed, in multivariate analysisthe possible adverse prognostic impact of BP variability is nolonger evident. Thus, in treated hypertension, BP variabil-ity evaluated by noninvasive monitoring is not an inde-pendent predictor of outcome. Am J Hypertens 2006;19:991–997 © 2006 American Journal of Hypertension, Ltd.

Key Words: Ambulatory blood pressure monitoring,

blood pressure variability, prognosis.

N oninvasive ambulatory blood pressure (BP) mon-itoring is progressively more used in clinicalpractice.1,2 It gives the opportunity to estimate

the average BP level, circadian BP variations, and BPvariability.1,2

There is growing evidence that average ambulatory BPand circadian BP variations are superior to clinic BP inpredicting cardiovascular risk in both untreated3–10 andtreated11–15 hypertensive patients.

The impact of BP variability on cardiovascular out-come in different populations has also been evaluated.There are few and contrasting data in the literature regardingthe relationship between BP variability and prognosis in

Received February 8, 2006. First decision March 11, 2006. AcceptedMarch 18, 2006.

From the Dipartimento di Medicina e Scienze dell’Invecchiamento,

untreated subjects at baseline.16–19 Two studies have eval-uated the influence of BP variability on cardiovascular riskin untreated and treated subjects analyzed together.20,21 Atpresent, to the best of our knowledge, there is a singlestudy in the literature assessing the prognostic impact ofBP variability in untreated and treated hypertensive pa-tients separately investigated.22

Thus other studies are needed to understand better theclinical relevance of BP variability, especially in treatedhypertension.

The aim of this study was to evaluate the prognosticsignificance of BP variability in treated hypertensive pa-tients.

zione Università “Gabriele d’Annunzio,” Chieti, Italy.Address correspondence and reprint requests to Dr. Sante D. Pierdo-

menico, Istituto di Patologia Medica, Policlinico “Santissima Annun-ziata,” Via dei Vestini, 66013, Chieti Scalo (Chieti), Italy; e-mail:

pierdomenico@unich.it or piersd@tiscalinet.it

0895-7061/06/$32.00doi:10.1016/j.amjhyper.2006.03.009

992 AJH–October 2006–VOL. 19, NO. 10BLOOD PRESSURE VARIABILITY AND PROGNOSIS

MethodsSubjects

We studied 1472 treated hypertensive patients. These sub-jects were selected from a larger treated hypertensivepopulation (1715 patients) who underwent to ambulatoryBP monitoring. Patients receiving single therapy with un-controlled clinic BP were excluded from this study. Sub-jects with secondary hypertension were also excluded. Allthe patients underwent clinical evaluation, electrocardiog-raphy, routine laboratory tests, echocardiographic exami-nation, and noninvasive ambulatory BP monitoring. Studypopulation came from the same geographical area (Chietiand Pescara, Abruzzo, Italy). The study was conducted inaccordance with the Second Declaration of Helsinki andwas approved by the institutional review committee. Sub-jects gave informed consent.

Office BP Measurements

Clinic systolic and diastolic BP recordings were per-formed by a physician with the subject in the supineposition after 10 min of rest, using a mercury sphygmo-manometer and appropriate-sized cuffs. Phase V was usedto determine diastolic BP. Measurements were performedin triplicate, 2 min apart, and the average value was usedas the BP for the visit.

Ambulatory BP Monitoring

Ambulatory BP monitoring was performed with a portablenoninvasive recorder (SpaceLabs 90207, Redmond, WA)on a day of typical activity. Technical aspects have beenpreviously reported.23 Ambulatory BP readings were ob-tained at 15-min intervals from 6 AM to midnight, and at30-min intervals from midnight to 6 AM. The followingambulatory BP parameters were evaluated: average andstandard deviation of daytime (awake period), night-time(asleep period), and 24-h systolic and diastolic BP. Awakeand asleep periods were calculated from diary times. Re-cordings were automatically edited if systolic BP was�260 or �70 mm Hg or if diastolic BP was �150 or �40mm Hg and pulse pressure was �150 or �20 mm Hg.23

Subjects had recordings of good technical quality (at least70% of valid readings). Patients with the standard deviation(SD) of daytime systolic, daytime diastolic, night-time sys-tolic and night-time diastolic BP below or above the groupmedian (12.6, 9, 10.5, and 8.3 mm Hg, respectively) wereclassified as having low or high BP variability. This approachhas been previously reported.17 The standard deviation of24-h BP, which is largely influenced by day–night BPchanges, was not used because it is not an appropriateindex of short-term BP variability.20 Because the eventrate was different between patients with low and high BP

variability of systolic BP, but not between patients with

low and high BP variability of diastolic BP, patients’classification was based on systolic BP variability only.

Echocardiography

End-systolic and end-diastolic measurements of interven-tricular septal thickness, left ventricular (LV) internal di-ameter, and posterior wall thickness were made accordingto the American Society of Echocardiography recommen-dations.24 The LV mass was calculated using the formulaintroduced by Devereux et al.25 Individual values for LVmass were indexed by height2.7 and LV hypertrophy wasdefined as LV mass/height2.7 �50 g/m2.7 in men and �47g/m2.7 in women.26

Follow-up

Subjects were followed-up in our hospital outpatient clinicor by their family doctors. Patients’ characteristics and theoccurrence of cardiovascular events were recorded duringfollow-up visits or by telephone interview of the patientfollowed by a clinical visit. Hospital record forms werecollegially reviewed by the authors of this study. Cardio-vascular events included fatal and nonfatal myocardialinfarction (at least two of three standard criteria: typicalchest pain, electrocardiographic changes, transient eleva-tion of conventional myocardial enzymes by more thantwofold the upper normal limits), coronary or peripheralrevascularization (bypass surgery or angioplasty), heartfailure requiring hospitalisation (at least two major or onemajor, plus two minor, Framingham criteria),27 fatal ornonfatal stroke (rapid onset of localizing neurologic deficitlasting �24 h as evidenced by computed tomography),and renal failure requiring dialysis.

Statistical Analysis

Data are expressed as mean � SD or percentage. Theunpaired t test, Mann-Whitney U test, and �2 test, wereused where appropriate.28 Event rates are expressed as thenumber of events per 100 patient-years based on the ratioof the observed number of events to the total number ofpatient-years of exposure up to the terminating event orcensor. Survival curves were estimated using the Kaplan-Meier product–limit method and compared by the Mantel(log-rank) test.28 The effect of various covariates on sur-vival was evaluated by using the backward stepwise Coxregression model (significance levels for inclusion andexclusion were 0.05 and 0.1, respectively).28 Covariatesincluded in the Cox model were as follows: age (years),gender (male v female), family history of premature car-diovascular disease (yes v no), smoking habit (yes v no),previous cardiovascular events (yes v no), body massindex (kg/m2), clinic BP (mm Hg), low density lipoprotein(LDL) cholesterol (mg/dL), creatinine (mg/dL), diabetes(yes v no), LV hypertrophy (yes v no), daytime or night-time BP (mm Hg), BP variability group (high v lowvariability) (forced into the models) and antihypertensive

drug class. Adjusted relative risk and 95% confidence

993AJH–October 2006–VOL. 19, NO. 10 BLOOD PRESSURE VARIABILITY AND PROGNOSIS

interval for the significant Cox model factors were calcu-lated. Statistical significance was defined at the level ofP � .05. Analyses were performed with the SPSS version12 software package (SPSS Inc., Chicago, IL).

Results

Characteristics of study groups according to daytime BPvariability are reported in Table 1. Age, LDL choles-terol, clinic BP, and daytime, night-time and 24-h sys-tolic BP were significantly higher in subjects with highBP variability. Percentage of men, family history ofpremature cardiovascular disease, and single-drug ther-apy were lower in patients with high BP variability.Prevalence of LV hypertrophy and double therapy werehigher in subjects with high BP variability. The otherparameters were not significantly different between thegroups.

Aspirin and statin therapy were not significantly differ-ent between the low and high BP variability groups (11 v10.8% and 6.1 v 5.3%, respectively).

Characteristics of study groups according to night-timeBP variability are reported in Table 2. Age, clinic, day-time, night-time, and 24-h systolic BP and night-time and24-h diastolic BP were significantly higher in patients withhigh BP variability. Percentage of subjects with familyhistory of premature cardiovascular disease was signifi-cantly lower in the high BP variability group. The otherparameters were not significantly different between thegroups.

Table 1. Characteristics of study groups according

ParameterLo

Male sex, n (%)Age, yearsBody mass index, kg/m2

Smoking, n (%)Family history of CVD, n (%)LDL cholesterol, mg/dLCreatinine, mg/dLLV hypertrophy, n (%)Diabetes, n (%)Previous events, n (%)Clinic systolic BP, mm HgClinic diastolic BP, mm HgDaytime systolic BP, mm HgDaytime diastolic BP, mm HgNighttime systolic BP, mm HgNighttime diastolic BP, mm Hg24-h systolic BP, mm Hg24-h diastolic BP, mm Hg1 Antihypertensive drug, n (%)2 Antihypertensive drugs, n (%)�3 Antihypertensive drugs, n (%)

CVD � cardiovascular disease; LDL � low-density lipoprotein; LV �

* P � .05 v low BP variability.

Aspirin and statin therapy were not significantly differ-ent between subjects with low and high BP variability(11.5 v 10.2% and 5.1 v 6.3%, respectively).

Antihypertensive drug classes of study groups are re-ported in Table 3. Diuretics and angiotensin-convertingenzyme inhibitors were more frequently used in subjectswith high daytime BP variability and angiotensin-convert-ing enzyme inhibitors were more recurrently used also inpatients with high night-time BP variability. The use of theother antihypertensive drug classes was not significantlydifferent between the groups.

When the study population was divided into those withnormal clinic and ambulatory BP (responder hyperten-sion), normal clinic and high ambulatory BP (maskedhypertension), high clinic and normal ambulatory BP(white-coat hypertension) and high clinic and ambulatoryBP (sustained hypertension), BP variability was signifi-cantly higher in subjects with masked and sustained hy-pertension, ie, in those with higher ambulatory BP. Inpatients with responder, masked, white-coat, and sustainedhypertension, the SD of daytime systolic BP was 12.2 � 3.8,13.7 � 4.4, 12.8 � 3.4, and 14.3 � 4.3 mm Hg, respectively,and the SD of night-time systolic BP was 9.9 � 3.3, 12 �4.0, 10.6 � 3.6, and 12 � 4.0 mm Hg, respectively.

When we evaluated the relationship between the white-coat effect (defined as the difference between clinic anddaytime systolic BP) and the SD of daytime systolic BP,we found only a weak association (r � 0.09).

During the follow-up (4.88 �2.9 years, range 0.2–11.6years), 119 cardiovascular events occurred (108 first and

aytime blood pressure (BP) variability

P variabilityn � 738)

High BP variability(n � 734)

79 (51.3) 311 (42.4)*57 � 11 61 � 11*

7.9 � 4.3 27.7 � 4.058 (21.4) 138 (19)77 (10) 29 (4)*

6.5 � 29 129.6 � 29*.87 � 0.21 0.88 � 0.2326 (17) 176 (24)*48 (6.5) 39 (5.3)20 (2.7) 29 (4)42 � 13 152 � 17*89 � 10 90 � 11*31 � 12 138 � 14*81 � 9 81 � 918 � 13 123 � 16*70 � 9 70 � 927 � 12 134 � 14*78 � 9 78 � 974 (24) 103 (14)*06 (55) 453 (62)*58 (21) 178 (24)

ventricular.

to d

w B(

3

21

1201

1

1

1

1

141

left

994 AJH–October 2006–VOL. 19, NO. 10BLOOD PRESSURE VARIABILITY AND PROGNOSIS

11 second events). Specifically, there were cases of 31myocardial infarctions (7 fatal), 19 coronary revasculariza-tions, 9 heart failure requiring hospitalization, 49 strokes (10fatal), 8 peripheral revascularizations, and 3 renal failurerequiring dialysis.

The event rates per 100 patient-years in subjects withlow and high BP variability according to daytime BP were,1.18 and 2.01, respectively, and in those with low and highBP variability according to night-time BP were 1.2 and2.05, respectively. Event-free survival curves of studygroups according to daytime and night-time systolic BPvariability are shown in Figure 1. As reported, event-freesurvival was significantly different between the low andhigh BP variability groups.

To identify independent predictors of outcome, a back-ward stepwise Cox regression analysis was then per-

Table 2. Characteristics of study groups according

ParameterLo

Male sex, n (%)Age, yearsBody mass index, kg/m2

Smoking, n (%)Family history of CVD, n (%)LDL cholesterol, mg/dLCreatinine, mg/dLLV hypertrophy, n (%)Diabetes, n (%)Previous events, n (%)Clinic systolic BP, mm HgClinic diastolic BP, mm HgDaytime systolic BP, mm HgDaytime diastolic BP, mm HgNighttime systolic BP, mm HgNighttime diastolic BP, mm Hg24-h Systolic BP, mm Hg24-h Diastolic BP, mm Hg1 Antihypertensive drug, n (%)2 Antihypertensive drugs, n (%)�3 Antihypertensive drugs, n (%)

CVD � cardiovascular disease; LDL � low-density lipoprotein; LV �* P � .05 v low BP variability.

Table 3. Antihypertensive drug classes of study gr

Daytime BP

Parameter Low variability High

Diuretic, n (%) 423 (57.3) 46�-Blocker, n (%) 233 (31.6) 22CC-blocker, n (%) 244 (33.1) 25ACE inhibitor, n (%) 345 (46.7) 41AR blocker, n (%) 144 (19.5) 12�-Blocker, n (%) 113 (15.3) 9

ACE � angiotensin converting enzyme; AR � angiotensin receptor; CC �* P � .05 v low variability for each classification.

formed that included various covariates. In multivariateanalysis, age, smoking habit, LDL cholesterol, diabetes,previous events, LV hypertrophy, and daytime (model 1)or night-time (model 2) systolic BP resulted independentpredictors of events, whereas BP variability did not. Themain results are reported in Table 4.

When the standard deviation of daytime or night-timesystolic BP was included in the Cox model as a continuousvariable, in place of BP variability group (high BP vari-ability v low BP variability), the results did not change.

Moreover, when the number of antihypertensive drugsand use of aspirin and statin were included in the models,the results did not change.

When cardiac and cerebrovascular events were ana-lyzed separately, the results showed the same trend: that is,there was no significant impact of BP variability.

ight-time blood pressure (BP) variability

P variabilityn � 739)

High BP variability(n � 733)

42 (46.3) 348 (47.5)57 � 11 61 � 11*.7 � 4.0 28 � 4.354 (20.8) 142 (19.4)69 (9.3) 37 (5)*28 � 29 128 � 2987 � 0.22 0.88 � 0.2142 (19.2) 160 (21.8)39 (5.3) 48 (6.5)24 (3.2) 25 (3.4)44 � 15 150 � 17*89 � 10 89 � 1131 � 13 138 � 14*81 � 9 81 � 1017 � 14 125 � 16*70 � 9 71 � 10*27 � 13 134 � 14*78 � 8 79 � 9*51 (20) 126 (17)27 (58) 432 (59)61 (22) 175 (24)

ventricular.

s

Nighttime BP

iability Low variability High variability

3.2)* 439 (59.4) 448 (61.1)0.7) 240 (32.5) 218 (29.7)5.1) 254 (34.4) 248 (33.8)5.9)* 352 (47.6) 403 (55.0)*6.3) 140 (18.9) 124 (16.9)2.9) 109 (14.7) 99 (13.5)

to n

w B(

3

271

10.1

1

1

1

1

141

left

oup

var

4 (65 (38 (30 (50 (15 (1

calcium channel.

DTtithanthose

with

lowB

Pvariability,the

adverseprognostic

Ta ents by Cox regression analysis

) P value Model 2 RR (95% CI) P value

Ag 3) �.001 Age (10 y) 1.51 (1.24–1.83) �.001Sm 6) �.05 Smoking habit (yes v no) 1.67 (1.04–2.67) �.05LD 5) �.05 LDL cholesterol (1 SD) 1.22 (1.03–1.45) �.05Di 8) �.01 Diabetes (yes v no) 2.66 (1.51–4.70) �.01Pr ) �.001 Previous events (yes v no) 4.24 (2.22–8.1) �.001LV 5) �.01 LVH (yes v no) 2.01 (1.35–2.98) �.01Da 2) �.001 Night-time systolic BP (10 mm Hg) 1.14 (1.03–1.26) �.01BP 2) .55 BP variability (high v low) 1.29 (0.87–1.92) .20

BP protein; LVH � left ventricular hypertrophy; RR � relative risk; 1 SD of LDL cholesterol is 29 mg/dl.ascular disease, body mass index, clinic BP, creatinine and antihypertensive drug class did not reach statistical significance.

Fb(b

99

5AJH

–Octo

ber

2006

–VO

L.19,N

O.10

BLO

OD

PR

ES

SU

RE

VA

RIA

BIL

ITY

AN

DP

RO

GN

OS

IS

iscussio

nhis

studyshow

sthat,

althoughtreated

hypertensivepa-

entsw

ithhigh

BP

variabilityhave

ahigher

eventrate

ble 4. Independent predictors of cardiovascular ev

Model 1 RR (95% CI

e (10 y) 1.58 (1.30–1.9oking habit (yes v no) 1.66 (1.03–2.6L cholesterol (1 SD) 1.22 (1.03–1.4abetes (yes v no) 2.53 (1.43–4.4evious events (yes v no) 4.06 (2.12–7.8H (yes v no) 1.76 (1.17–2.6ytime systolic BP (10 mm Hg) 1.26 (1.12–1.4variability (high v low) 1.13 (0.75–1.7

� blood pressure; CI � confidence interval; LDL � low density lipoBeyond BP variability, gender, family history of premature cardiov

IG.

1.

Even

t-freesu

rvivalcu

rvesin

subjects

with

low

and

hig

hlo

od

pressu

revariab

ilityacco

rdin

gto

daytim

e(to

p)an

dnig

ht-tim

eotto

m)

systolic

blo

od

pressu

re.BPV

�blo

od

pressu

revariab

ility.

996 AJH–October 2006–VOL. 19, NO. 10BLOOD PRESSURE VARIABILITY AND PROGNOSIS

relevance of BP variability is no longer present in multi-variate analysis because of the overwhelming influence ofother covariates (age, smoking habit, LDL cholesterol,diabetes, previous events, LV hypertrophy, and averageambulatory BP).

Kikuya et al,20 by using noninvasive monitoring withinterval recordings of 30 min, reported higher mortalityrate in subjects with high BP variability in a generalpopulation. Indeed, in the highest quintile of the SD ofdaytime and night-time systolic BP, the adjusted RR was2.51, 95% CI 1.1 to 5.7, P � .05, and 2.21, 95% CI 1.11to 4.43, P � .05, respectively. These data obtained in ageneral population, including untreated and treated sub-jects (�30%), cannot be extrapolated to a treated hyper-tensive population. Moreover, cardiovascular morbiditywas not reported in that study.

Bjorklund et al,21 studied a population of elderly men.They reported that the variability of daytime systolic BP,reflected by the SD, was an independent predictor ofcardiovascular morbidity (adjusted RR 1.24 for each SDincrease, 95% CI 1.07 to 1.42, P � .05). These resultswere obtained in elderly men aged 70 years, and untreatedand treated subjects (�30%) were analyzed together.21

Thus, also these data cannot be extrapolated to a treatedhypertensive population including men and women with awider age range.

Pringle et al,22 evaluated the impact of BP variability inthe Syst-Eur trial, including elderly patients with isolatedsystolic hypertension. The placebo and the active-treat-ment groups were analyzed separately.22 In the activetreatment group, neither daytime nor night-time BP vari-ability resulted an independent predictor of cardiovascularmortality and of cerebrovascular and cardiac events.22

Our findings are similar to those reported by Pringleet al22 in treated elderly hypertensive patients. The maindifference is that we studied a treated hypertensive popu-lation with systolic and diastolic hypertension and a widerage range. Thus the lack of association between BP vari-ability and prognosis in treated hypertension is extended inour study.

At present, it is sill uncertain whether the clinicalrelevance of BP variability evaluated by noninvasive mon-itoring is comparable to that of continuous monitoring.The potential prognostic value of BP variability evaluatedby invasive and noninvasive beat-to-beat monitoring re-mains to be established.

This study has some limitations. First, we studied onlysubjects of our ethnicity, and our results cannot be appliedto other ethnic groups. Second, ambulatory BP monitoringcould not be repeated in all the subjects during the study.Third, with regard to night-time BP variability, it wasevaluated with a sampling interval of 30 min which givesa less accurate estimation of BP variability29; however,this interval was also used in the Ohasama20 and Syst-Eur22 studies. Finally, it has been recently reported thataverage real variability could be a more reliable index of

BP variability30; however, other studies are needed to

confirm whether the results obtained with this more com-plex index can be extrapolated to a general untreated ortreated hypertensive population. Indeed the sample stud-ied30 showed a rate of events exceedingly high consideringits risk profile.

In conclusion, in treated hypertension BP variabilityevaluated by noninvasive monitoring is not an indepen-dent predictor of outcome.

References

1. Pickering TG: Recommendations for the use of home (self) and am-bulatory blood pressure monitoring. Am J Hypertens 1996;9:1–11.

2. Pickering TG: A review of national guidelines on the clinical use ofambulatory blood pressure monitoring. Blood Press Monit 1996;1:151–156.

3. Verdecchia P, Porcellati C, Schillaci G, Borgioni C, Ciucci A,Battistelli M, Guerrieri M, Gatteschi C, Zampi I, Santucci A, San-tucci C, Reboldi G: Ambulatory blood pressure. An independentpredictor of prognosis in essential hypertension. Hypertension 1994;24:793–801.

4. Khattar RS, Senior R, Lahiri A: Cardiovascular outcome in white-coat versus sustained mild hypertension: a 10-year follow-up study.Circulation 1998;98:1892–1897.

5. Staessen JA, Thijs L, Fagard RH, O’Brien ET, Clement D, deLeeuw PW, Mancia G, Nachev C, Palatini P, Parati G, TuomilehtoJ, Webster J: Predicting cardiovascular risk using conventionalv ambulatory blood pressure in older patients with systolic hyper-tension. J Am Med Assoc 1999;282:539–546.

6. Fagard RH, Staessen JA, Thijs L, Gasowski J, Bulpitt CJ, Clement D,de Leeuw PW, Dobovisek J, Jaaskivi M, Leonetti G, O’Brien E,Palatini P, Parati G, Rodicio JL, Vanhanen H, Webster J: Response toantihypertensive therapy in older patients with sustained and nonsus-tained systolic hypertension. Circulation 2000;102:1139–1144.

7. Kario K, Shimada K, Schwartz JE, Matsuo T, Hoshide S, PickeringTG: Silent and clinically overt stroke in older Japanese subjects withwhite-coat and sustained hypertension. J Am Coll Cardiol 2001;38:238–245.

8. Celis H, Staessen JA, Thijs L, Buntinx F, De Buyzere M, Den HondE, Fagard RH, O’Brien ET: Cardiovascular risk in white-coat andsustained hypertension. Blood Press 2002;11:352–356.

9. Bjorklund K, Lind L, Zethelius B, Andrén B, Lithell H: Isolatedambulatory hypertension predicts cardiovascular morbidity in el-derly men. Circulation 2003;107:1297–1302.

10. Pierdomenico SD, Lapenna D, Bucci A, Di Iorio A, Neri M,Cuccurullo F, Mezzetti A: Cardiovascular and renal events in un-complicated mild hypertensive patients with sustained and whitecoat hypertension. Am J Hypertens 2004;17:876–881.

11. Redon J, Campos C, Narciso ML, Rodicio JL, Pascual JM, RuilopeLM: Prognostic value of ambulatory blood pressure monitoring inrefractory hypertension: a prospective study. Hypertension 1998;31:712–718.

12. Verdecchia P, Reboldi G, Porcellati C, Schillaci G, Pede S, Ben-tivoglio M, Angeli F, Norgiolini S, Ambrosio G: Risk of cardio-vascular disease in relation to achieved office and ambulatory bloodpressure control in treated hypertensive subjects. J Am Coll Cardiol2002;39:878–885.

13. Clement DL, De Buyzere ML, De Bacquer DA, de Leeuw PW,Duprez DA, Fagard RH, Gheeraert PJ, Missault LH, Braun JJ, SixRO, Van Der Niepen P, O’Brien E: Prognostic value of ambulatoryblood-pressure recordings in patients with treated hypertension.

N Engl J Med 2003;348:2407–2415.

997AJH–October 2006–VOL. 19, NO. 10 BLOOD PRESSURE VARIABILITY AND PROGNOSIS

14. Ohkubo T, Kikuya M, Metoki H, Asayama K, Obara T, HashimotoJ, Totsune K, Hoshi H, Satoh H, Imai Y: Prognosis of maskedhypertension and white coat hypertension detected by 24-h ambu-latory blood pressure monitoring. J Am Coll Cardiol 2005;46:508–515.

15. Pierdomenico SD, Lapenna D, Bucci A, Di Tommaso R, Di MascioR, Manente BM, Caldarella MP, Neri M, Cuccurullo F, Mezzetti A:Cardiovascular outcome in treated hypertensive patients with re-sponder, masked, false resistant, and true resistant hypertension.Am J Hypertens 2005;18:1422–1428.

16. Pickering TG, James GD: Ambulatory blood pressure and progno-sis. J Hypertens 1994;12:S29–S33.

17. Verdecchia P, Borgioni C, Ciucci A, Gattobigio R, Schillaci G,Sacchi N, Santucci A, Santucci C, Reboldi G, Porcellati C: Prog-nostic significance of blood pressure variability in essential hyper-tension. Blood Press Monit 1996;1:3–11.

18. Sander D, Kukla C, Klingelhofer J, Winbeck K, Conrad B: Rela-tionship between circadian blood pressure patterns and progressionof early carotid atherosclerosis. A 3-year follow-up study. Circula-tion 2000;102:1536–1541.

19. Pierdomenico SD, Lapenna D, Bucci A, Manente BM, Mancini M,Cuccurullo F, Mezzetti A: Blood pressure variability and prognosisin uncomplicated mild hypertension. Am Heart J 2005;149:934–938.

20. Kikuya M, Hozawa A, Ohokubo T, Tsuji I, Michimata M, Matsub-ara M, Ota M, Nagai K, Araki T, Satoh H, Ito S, Hisamichi S, ImaiY: Prognostic significance of blood pressure and heart rate variabil-ities. The Ohasama study. Hypertension 2000;36:901–906.

21. Bjorklund K, Lind L, Zethelius B, Berglund L, Lithell H: Prognosticsignificance of 24-h ambulatory blood pressure characteristics forcardiovascular morbidity in a population of elderly men. J Hyper-tens 2004;22:1691–1697.

22. Pringle E, Phillips C, Thijs L, Davidson C, Staessen JA, de Leeuw

PW, Jaaskivi M, Nachev C, Parati G, O’Brien ET, Tuomilehto J,

Webster J, Bulpitt CJ, Fagard RH: Systolic blood pressurevariability as a risk factor for stroke and cardiovascular mortality inthe elderly hypertensive population. J Hypertens 2003;21:2251–2257.

23. Pierdomenico SD, Lapenna D, Guglielmi MD, Antidormi T, Schia-vone C, Cuccurullo F, Mezzetti A: Target organ status and serumlipids in patients with white coat hypertension. Hypertension 1995;26:801–807.

24. Sahn DJ, DeMaria A, Kisslo J, Weyman A, for the Committee onM-mode standardization of the American Society of Echocardiog-raphy: Recommendations regarding quantitation in M-mode echo-cardiography: results of a survey of echocardiographic mea-surements. Circulation 1978;58:1072–1083.

25. Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, SachsI, Reichek N: Echocardiographic assessment of left ventricularhypertrophy: comparison to necropsy findings. Am J Cardiol 1986;57:450–458.

26. de Simone G, Devereux RB, Daniels SR, Koren MJ, Meyer RA,Laragh JH: Effect of growth on variability of left ventricular mass:assessment of allometric signals in adults and children and theircapacity to predict cardiovascular risk. J Am Coll Cardiol 1995;25:1056–1062.

27. McKee PA, Castelli WP, McNamara PM, Kannel WB: The naturalhistory of congestive heart failure. N Engl J Med 1971;285:1441–1446.

28. Altman DG: Practical Statistics for Medical Research. Chapman &Hall, London, 1991.

29. Di Rienzo M, Grassi G, Pedotti A, Mancia G: Continuous v inter-mittent blood pressure measurements in estimating 24-hour averageblood pressure. Hypertension 1983;5:264–269.

30. Mena L, Pintos S, Queipo NV, Aizpùrua JA, Maestre G, Sulbaran T:A reliable index for the prognostic significance of blood pressure

variability. J Hypertens 2005;23:505–511.