hypertension Circadian blood pressure changes and left ventricular hypertrophy in essential

P Verdecchia, G Schillaci, M Guerrieri, C Gatteschi, G Benemio, F Boldrini and C Porcellatihypertension.

Circadian blood pressure changes and left ventricular hypertrophy in essential

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1990 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.81.2.528

1990;81:528-536Circulation. 

http://circ.ahajournals.org/content/81/2/528the World Wide Web at:

The online version of this article, along with updated information and services, is located on

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer information about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. FurtherEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from

528

Circadian Blood Pressure Changes and LeftVentricular Hypertrophy in

Essential HypertensionPaolo Verdecchia, MD, Giuseppe Schillaci, MD, Massimo Guerrieri, MD,

Camillo Gatteschi, MD, Guglielmo Benemio, MD,

Francesca Boldrini, MD, and Carlo Porcellati, MD

The effects of circadian blood pressure (BP) changes on the echocardiographic parameters ofleft ventricular (LV) hypertrophy were investigated in 235 consecutive subjects (137 unselecteduntreated patients with essential hypertension and 98 healthy normotensive subjects) whounderwent 24-hour noninvasive ambulatory blood pressure monitoring (ABPM) and cross-sectional and M-mode echocardiography. In the hypertensive group, LV mass index correlatedwith nighttime (8:00 PM to 6:00 Am) systolic (r=0.51) and diastolic (r=0.35) blood pressuremore closely than with daytime (6:00 AM to 8:00 PM) systolic (r=0.38) and diastolic (r-0.20) BP,or with casual systolic (r=0.33) and diastolic (r=0.27) BP. Hypertensive patients were dividedinto two groups by presence (group 1) and absence (group 2) of a reduction of both systolic anddiastolic BP during the night by an average of more than 10% of the daytime pressure. CasualBP, ambulatory daytime systolic and diastolic BP, sex, body surface area, duration ofhypertension, prevalence of diabetes, quantity of sleep during monitoring, funduscopic changes,and serum creatinine did not differ between the two groups. LV mass index, after adjustmentfor the age, the sex, the height, and the daytime BP differences between the two groups (analysisof covariance) was 82.4 g/m2 in the normotensive patient group, 83.5 g/m2 in hypertensivepatients of group 1 and 98.3 g/m2 in hypertensive patients of group 2 (normotensive patients vs.

group pNS roup 1 vs. group 2, p- 0.002). The other echocardiographic parameters ofLVanatomy (i.e., interventricular septum and posterior wall thickness, relative wall thickness,cross-sectional area) differed between the groups as did LV mass index. A statisticallysignificant inverse correlation was found between LV mass index and percentage of nocturnalreduction of daytime ambulatory systolic (r=-0.34; p<0.001) and diastolic (r=-0.30;p<0.001) BP. These findings suggest that in unselected hypertensive patients, an ambulatoryBP decline from day to night is associated with a lo*ei LV muscle mass. In these patients, anocturnal reduction of systolic and diastolic BP by more than 104 of daytime values could delayor prevent the development of cardiac LV hypertrophy. (Circulation 1990;gl:528-536)

L eft ventricular hypertrophy (LVH) detected byelectrocardiography (ECG) is an importantpredictor of cardiovascular morbidity and

mortality.In the Framingham study about 45% of all cardio-

vascular deaths were preceded by left ventricularhypertrophy detected by electrocardiography (ECG-LVH), and the 5-year mortality rate in men with

From the Division of Medicine, Civic Hospital "Beato G. Villa,"'Citta della Pieve (Perugia), Italy.Address for correspondence: Dr. Paolo Verdecchia, Civic Hos-

pital "Beato G. Villa," Division of Medicine, 06062 - Citta dellaPieve PG, Italy.

Received December 29, 1988; revision accepted September 27,1989.

ECG-LVH was about 35%, as compared with 10-15% in the absence of ECG-LVH.1-3 LVH is a usualconsequence of arterial hypertension, and it has beenshown to appear within 12 years in about 50% of

See p 700

patients with systolic blood pressure (BP) above 180mm Hg.2Compared with ECG, echocardiography shows a

higher sensitivity and an equally high specificity forthe diagnosis of LVH.4 Echocardiography allowsquantitation of LV muscle mass and provides valuesreasonably near to those found at necropsy.4-7 ECGallowed detection of LVH in no more than 38% ofhypertensive patients with echocardiographic LVH.8

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from

Verdecchia et al Nighttime Blood Pressure and Cardiac Hypertrophy 529

Some findings suggest that echocardiographic LVHcould be an independent predictor of cardiovascularmorbidity9 and mortality.10The weak relation of casual BP to echocardio-

graphic LVH1" becomes closer by using 24-hour BPaverages obtained with noninvasive12'13 or intra-arterial14 ambulatory blood pressure monitoring(ABPM).Daytime BP, particularly that recorded during

recurring stressful situations,13 seems to be more

closely related to LVH than nighttime BP,1213 but inother studies, both awake and sleep BP correlatedequally well with LV muscle mass,14 or nighttime BPshowed a better relation to LV muscle mass than diddaytime BP.15ABPM shows that, in hypertensive patients, day-

time BP usually decreases during night.16"17 Thedegree and duration of nocturnal BP and wall stressreduction could, thus, influence development of car-diac structural changes in hypertensive patients butsuch hypothesis has never been specifically investi-gated.

Consequently, the present study was designed toexamine the relation of BP decline from day to nightto the echocardiographic parameters of LV anatomyin a large population of unselected hypertensivepatients and healthy normotensive subjects undergo-ing 24-hour noninvasive ABPM.

MethodsPatient Population

One-hundred thirty-seven patients with essentialhypertension (47% female) and 98 healthy nor-motensive subjects (49% female) were studied. Theywere consecutively chosen from patients with docu-mented essential hypertension and healthy nor-motensive subjects examined in our laboratory andmeeting all the following criteria: 1) no antihyperten-sive drugs from at least 4 weeks, 2) good quality ofechocardiographic tracings, 3) agreement within 5mm Hg between ambulatory BP recording unit andmercury sphygmomanometer in at least three consec-utive measurements taken simultaneously on thesame arm, 4) absence of clinical, ECG, or echocar-diographic evidence of coronary artery disease, val-vular disease (Doppler echocardiography), or renaldisease. Seven patients with mitral valve prolapseand mild (grade 118) mitral regurgitation wereincluded.

Hypertensive patients were identified in a ratherrestricted rural area of 6,500 inhabitants by a groupof general practitioners who referred all the subjectswith supine diastolic BP of 90 mm Hg or higher toour hypertension clinic.To be admitted to the study, supine diastolic BP

had to be 90 mm Hg or higher in at least three visitsat 1-week intervals, and all the four aforementionedcriteria had to be fulfilled.

Healthy normotensive patients were voluntary sub-jects chosen among medical students, hospital staff,

and other subjects examined for clinical check-upand found healthy.

Experimental ProceduresBlood pressure measurements. Clinic BP was mea-

sured after 10 minutes of supine rest by using aHawksley Random Zero manometer (phase 1 and 5)(Hawksley and Sons, Ltd., Lancing, West Sussex,England). Ambulatory BP was recorded by using thefully automatic unit ICR 5200 system (Spacelabs,Redmond, Washington) that, in our laboratory,showed a correlation of 0.97 with both systolic anddiastolic pressures measured simultaneously on thesame arm (by a Y connector) with a mercury sphyg-momanometer.19 Others have reported a similar cor-relation using the same unit.20The reading, editing, and analysis of data provided

by the unit was done by the ABP5600 interface(Spacelabs, Redmond, Washington) installed on anIBM/XT personal computer. Systolic readingsgreater than 260 or less than 70 mm Hg, diastolicreadings greater than 150 or less than 40 mm Hg, andpulse pressure readings greater than 150 or less than20 mm Hg were automatically discarded. The unitwas set to take readings automatically every 15minutes throughout the 24 hours. Overall, there were107.4 (SD, 12) BP readings per patient, with 89.5(SD, 11) readings per patient fulfilling the editingcriteria. Error percentage was 15.9% (SD, 12).

Echocardiographic MethodsM-Mode echocardiograms were performed under

cross-sectional control, with the patient in partial leftdecubitus position, using an ATL Ultramark 8 system(Advanced Technology Laboratories, Bellevue,Washington) with an ATL 3.00 MHz mechanicaltransducer. LV measurements were made at end-diastole and end-systole according to the recommen-dations of the American Society of Echocardiogra-phy (ASE).21LV mass was calculated using the following equa-

tion, based on necropsy validation studies7: LVmass=0.80(ASE-cube LV mass)+0.6 g, where ASE-cube LV mass= 1.04* 1 {(IVSd+LVIDd+PWd)3 -LVIDd3 D }; IVSd is interventricular septal thicknessat end-diastole, LVIDd is left ventricular internaldimension at end diastole, and PWd is posterior wallthickness at end diastole.

All echocardiographic examinations were per-formed by the same experienced sonographer. Read-ing of echocardiographic tracings was made in ran-dom order by two investigators who had noknowledge of patients' BPs and other clinical data.Both investigators marked locations on stop frames

on the screen of the ATL Ultrasound System, and themean values from at least five measurements for eachparameter for observer were computed. Only frameswith optimal visualization ofLV interfaces and showingsimultaneous visualization of interventricular septalthickness (IVS), posterior wall thickness (PW), and leftventricular internal dimension (LVID) throughout thewhole cardiacccycle were considered for reading.

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from

530 Circulation Vol 81, No 2, February 1990

TABLE 1. Main Findings in the Study Population

Normotensive patients Hypertensive patientsParameter (n =98) (n = 137) F pAge (yr) 51.9 (14) 52.5 (11) 0.10 0.74Height (cm) 165.6 (8) 166.6 (9) 0.80 0.36Weight (kg) 73.0 (13) 75.2 (14) 1.57 0.21BSA (m2) 1.80 (0.19) 1.83 (0.20) 1.89 0.169Casual BP (mm Hg)

Systolic 142.6 (18) 159.5 (19) 47.36 <0.0001Diastolic 79.6 (7) 97.8 (7) 417.52 <0.0001

Average 24-hr BP (mm Hg)Systolic 128.7 (18) 141.8 (16) 36.35 <0.0001Diastolic 79.6 (8) 92.0 (8) 127.40 <0.0001

Daytime BP (mm Hg)Systolic 133.3 (18) 147.6 (16) 43.38 <0.0001Diastolic 83.6 (9) 96.3 (8) 138.34 <0.0001

Nighttime BP (mm Hg)Systolic 121.9 (18) 134.5 (18) 27.93 <0.0001Diastolic 74.4 (9) 86.1 (10) 86.58 <0.0001

Heart rate (beats/min)Casual 72.2 (10) 76.8 (13) 8.36 0.0042Average 24-hr 72.7 (7) 76.8 (8) 15.46 0.0001Daytime 76.7 (9) 80.3 (9) 10.12 0.0017Nighttime 68.8 (8) 72.9 (9) 13.32 0.0003

Data are expressed as mean (SD). Daytime is 6:00 AM to 8:00 PM; nighttime is 8:00 PM to 6:00 AM.BSA, body surface area; BP, blood pressure.

Forty-three subjects, most of whom were affectedby long-term obstructive lung disease or obesity, wereexcluded from the study because of unsatisfactoryechocardiographic tracings. Forty-two treatedpatients with complicated or severe hypertensionwere not investigated because it would have beenunethical to discontinue therapy.

Relative wall thickness (RWT)22 and cross-sectional area (CSA)23 were calculated as previouslydescribed. Fractional shortening (FS) was calculatedaccording to the formula: FS(%)= {(LVIDd-LVIDs)/LVIDd }*100.

Nocturnal Blood Pressure BehaviorStudy subjects were divided into two groups by the

presence (group 1) or absence (group 2) of significantnocturnal BP reduction, defined as daytime (6:00 AM to8:00 PM) systolic and diastolic BP averages decreasingby more than 10% during night (8:00 PM to 6:00 AM).

Statistical AnalysisAll data were handled and stored by using the

DBASE III software, and analyzed by using the BMDPpackage24 (1985 version) installed on an IBM AT/3personal computer. One-way analysis of variance(ANOVA) and nonparametric tests were used tocompare the groups in their demographic, bloodpressure, and heart rate findings. Analysis of covari-ance (ANCOVA) (age, sex, height, and daytimeambulatory systolic and diastolic BP acting as simul-taneous covariates) was used to compare the nor-

motensive group, the hypertensive subgroup with,and the hypertensive subgroup without nocturnal BPfall in the echocardiographic parameters (Plv rou-tine of BMDP). Multiple comparison between thegroups were performed on the adjusted data by usingthe t statistics.24p values less than 0.05 were consid-ered statistically significant.

ResultsThe main characteristics of the study population,

including the statistical significance of differencesbetween hypertensives and normotensives, arereported in Table 1. Hypertensive patients and nor-motensive controls did not differ for age, height,weight, sex prevalence, and body surface area. Theheart rate, both casual and ambulatory, was slightlyhigher in the hypertensive group. Prevalence ofLVH, defined as LV mass index above 120 g/m2 25,26was 14.6% in the hypertensive groups.

Reproducibility ofAmbulatory Blood PressureFigure 1 shows the hourly ambulatory BP profile on

the two recording sessions in a subgroup of 32 hyper-tensive patients who repeated ABPM within 3-5 daysin the absence of therapy. Average 24-hour systolic/diastolic BP values were 164.4/97.9 mm Hg (SD, 19/7)on the first session, and 158.2/95.8 mm Hg (SD, 19/8)on the second session (all, p=NS). A statisticallysignificant difference (p<0.05) between the two ses-sions was found only from hours 1:30 to 2:30 PM in the

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from

Verdecchia et al Nighttime Blood Pressure and Cardiac Hypertrophy 531

170

150

110

90J

18 20 22 0 2 4 6 8 10 12 14 16

Time of day (hour)

FIGURE 1. Graph showing 24-hour blood pressure profile in32 off-therapy hypertensive patients who repeated noninvasiveambulatory blood pressure monitoring within 3-5 days. Over-all, blood pressure was slightly lower on second session,particularly in first hours after application of system. Statisti-cally significant differences were found only between 1:30 and2:30 PM on hourly averages of systolic blood pressure.

hourly averages of systolic BP, the values recorded on

the second occasion being slightly lower.

Relation of Blood Pressure to Left VentricularHypertrophyAs shown in Table 2, nighttime systolic BP showed

the best correlations with LV mass (r=0.54), LVmass index (r=0.51), RWT (r=0.38), CSA (r=0.57),and CSAI (r=0.51) in the hypertensive group. Over-all, in the hypertensive population, the relationsbetween the echocardiographic indexes of LVH andBP values were slightly closer with the nighttimevalues, and progressively weaker with the daytimeand casual values.

TABLE 2. Relation of Casual and Ambulatory Blood Pressure toThickness, Cross-sectional Area, and Cross-sectional Area Index

160

140 X_

120

100

80

6018 20 22 0 2 4 6 8 10 12 14 16

Time of day (hour)

FIGURE 2. Graph showing 24-hour blood pressure profile inhypertensive patients with noctumal reduction of systolic anddiastolic blood pressure by 10% or less of daytime values(n=55) (-), and in patients with greater noctumal reduc-tion (n=82) (----).

Effects of Noctumal Blood Pressure ReductionFigure 2 shows the hourly ambulatory BP profile in

hypertensive patients with and without significantnocturnal BP reduction. The main demographic,blood pressure, and heart rate values in either of thetwo subgroups are reported in Table 3.The two subgroups differed slightly for age

(p=0.03), but did not differ for height, weight, bodysurface area, duration of hypertension (5.72 vs. 5.84years,p=NS), prevalence of diabetes mellitus (10.1%vs. 14.8%,p=NS) and serum creatinine (0.78 vs. 0.79mg/dl, p=NS). There were slightly more women ingroup 1 (53.7%) than in group 2 (36.9%), a differ-ence bordering statistical significance (x2=3.9,p =0.05).

Casual and daytime BPs were slightly higher in thegroup of hypertensive patients with a smaller noctur-

Left Ventricular Mass, Left Ventricular Mass Index, Relative Wall

All subjects (n =235) Normotensive patients (n=98) Hypertensive patients (n=137)

LVM LVMI RWT CSA CSAI LVM LVMI RWT CSA CSAI LVM LVMI RWT CSA CSAI

Casual SBP 0.34* 0.38* 0.26* 0.36* 0.41* 0.26* 0.36* 0.26* 0.28* 0.37* 0.29* 0.33* 0.16t 0.30* 0.35*

Casual DBP 0.32* 0.29* 0.24* 0.34* 0.30* 0.02 0.02 0.10 0.04 0.01 0.29* 0.27* 0.17t 0.32* 0.28*

Average 24-hrSBP 0.51* 0.48* 0.38* 0.53* 0.47* 0.36* 0.33* 0.34* 0.39* 0.33* 0.54* 0.51* 0.34* 0.56* 0.50*

Average 24-hrDBP 0.41* 0.36* 0.30* 0.43* 0.35* 0.23t 0.15 0.26* 0.25* 0.15t 0.41* 0.34* 0.22* 0.41* 0.33*

Daytime SBP 0.43* 0.40* 0.31* 0.45* 0.40* 0.35* 0.31* 0.31* 0.37* 0.30* 0.41* 0.38* 0.23* 0.43* 0.38*

Daytime DBP 0.33* 0.28* 0.23* 0.35* 0.28* 0.26* 0.16 0.24* 0.28* 0.15 0.24* 0.20* 0.09 0.25* 0.19tNighttime SBP 0.50* 0.47* 0.38* 0.52* 0.46* 0.31* 0.29* 0.28* 0.32* 0.27* 0.54* 0.51* 0.38* 0.57* 0.51*

Nighttime DBP 0.41* 0.37* 0.30* 0.42* 0.36* 0.22t 0.17 0.24* 0.22t 0.17 0.40* 0.35* 0.23* 0.41* 0.34*

Daytime is 6:00 AM to 8:00 PM; night time is 8:00 PM to 6:00 AM.LVM, left ventricular mass; LVMI, left ventricular mass index; RWT, relative wall thickness; CSA, cross sectional area; CSAT, cross

sectional area index; SBP, systolic blood pressure; DBP, diastolic blood pressure.*p<0.01; tp<0.05.

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from

532 Circulation Vol 81, No 2, February 1990

TABLE 3. Main Findings in Hypertensive Patients With (Group 1) and Without (Group 2) Significant Nocturnal BloodPressure Fall*

Hypertensive patients (n=137)

Group 1 Group 2Parameter (n=82) (n=55) F pAge (yr) 50.7 (11) 55.2 (12) 4.92 0.03Height (cm) 165.9 (9) 167.5 (7) 1.13 0.28Weight (kg) 74.4 (15) 76.5 (13) 0.71 0.39BSA (m2) 1.81 (0.21) 1.85 (0.18) 0.87 0.35Casual BP (mm Hg)

Systolic 160.4 (18) 163.8 (20) 1.12 0.29Diastolic 97.8 (8) 98.8 (7) 1.06 0.32

Average 24-hr BP (mm Hg)Systolic 137.3 (14) 148.5 (16) 18.05 <0.0001Diastolic 88.9 (8) 96.6 (6) 35.35 <0.0001

Daytime BP (mm Hg)Systolic 146.7 (16) 148.9 (16) 0.64 0.426Diastolic 96.7 (9) 97.4 (7) 1.67 0.199

Nighttime BP (mm Hg)Systolic 127.2 (15) 145.3 (17) 43.72 <0.0001Diastolic 80.8 (8) 94.1 (7) 92.51 <0.0001

Heart rate (beats/min)Casual 77.5 (13) 75.9 (14) 0.45 0.506Average 24-hr 77.3 (8) 75.9 (8) 0.85 0.357Daytime 81.6 (9) 79.4 (9) 0.78 0.402Nighttime 72.3 (9) 73.9 (10) 0.94 0.332

Data are expressed as mean (SD). Daytime is 6:00 AM to 8:00 PM; nighttime is 8:00 PM to 6:00 AM.BSA, body surface area; BP, blood pressure.*Nighttime systolic and diastolic blood pressure more than 10% lower as compared with daytime values.

nal BP decline but the differences from the othergroup were not statistically significant.No funduscopic changes, focal or generalized nar-

rowing of the retinal arteries, or retinal haemorragesand exudates were found, respectively, in 48, 26, andeight hypertensive patients with, and in 34, 15, and

six patients without significant nocturnal BP reduc-tion (2=O.32, p=NS).

In comparison with the usual sleeping habits, therewere no changes in the duration of sleep, or a

reduction up to 2 hours, or of 2-4 hours, or of morethan 4 hours in 36, 13, 26, and seven patients with,

TABLE 4. Echocardiographic Findings in the Normotensive and Hypertensive Groups*

Multiple comparisons (p)

Hypertensive patients Analysis ofNormotensive covariance Normotensives vs. Group 1patients Group 1 Group 2 vs.

Parameter (n=98) (n=82) (n=55) F p Group 1 Group 2 Group 2

End-diastolic IVS thickness (cm) 0.82 (0.27) 0.83 (0.24) 0.93 (0.25) 3.63 0.028 NS 0.021 0.0143End-diastolic PW thickness (cm) 0.74 (0.18) 0.73 (0.15) 0.83 (0.16) 7.44 0.01 NS 0.002 0.0003LV end-diastolic diameter (cm) 5.38 (0.66) 5.33 (0.60) 5.37 (0.62) 0.15 0.86 NS NS NSShortening fraction (%) 40.60 (7.6) 37.9 (6.9) 38.3 (7.2) 2.70 0.07 NS NS NSLV mass (Penn) (g) 151.1 (59) 151.8 (53) 179.7 (55) 5.41 0.005 NS 0.007 0.002LV mass index (Penn) (g/m2) 82.4 (31) 83.5 (28) 98.3 (29) 5.72 0.004 NS 0.004 0.002Relative wall thickness 0.28 (0.08) 0.28 (0.07) 0.32 (0.07) 5.05 0.007 NS 0.008 0.004Cross-sectional area (cm2) 15.2 (4.6) 15.2 (4.1) 17.5 (4.4) 5.67 0.004 NS 0.006 0.002Cross-sectional area index (cm2/m2) 8.29 (2.4) 8.36 (2.2) 9.63 (2.3) 6.55 0.002 NS 0.002 0.001

Data are expressed as mean (SD).IVS, interventricular septum; PW, posterior wall; LV, left ventricular.*Hypertensive group divided by presence (group 1) and absence (group 2) of significant nocturnal blood pressure reduction (nighttime

systolic and diastolic blood pressure more than 10% lower as compared with daytime values). For either of three groups, data are adjustedfor age, sex, height, and daytime systolic and diastolic blood pressure differences between groups (see text for explanation).

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from

Verdecchia et al Nighttime Blood Pressure and Cardiac Hypertrophy 533

nGrCoil) A

Croup B

Group C

70 so 90 100Left ventricular mass index (g/m2)

FIGURE 3. Bar graph showing adjusted values (age-, sex-,

height-, and daytime systolic and diastolic BP differencesbetween groups acting as simultaneous covariates) of leftventricular mass index in the normotensive group (Group A),in hypertensive subgroup with noctumal blood pressure reduc-tion by more than 10% of daytime values (Group B), and insubgroup with a smaller noctumalpressure decline (Group C).There is significant overall difference between three groups

(one-way ANCOVA, F=5. 72, p=0.0004) as well as betweenGroup B and Group C (p=0.002), and between GroupA andGroup C (p=0.004). Group A and Group B did not differsignificantly.

and in 27, nine, 14, and five patients without signifi-cant nocturnal BP reduction (X2=0.65, p=NS).Table 4 shows the age-, sex-, height-, and daytime

systolic and diastolic BP-adjusted echocardiographicfindings in the normotensive group and both hyper-tensive groups. All the echocardiographic indexes ofLV hypertrophy differed in the overall comparison

between the three groups. None of the differencesbetween normotensive and hypertensive patientswith nocturnal BP reduction by more than 10%yielded statistical significance. In contrast, all differ-ences between the two groups of hypertensivepatients were statistically significant. FS and LVend-diastolic diameter did not differ among the threegroups. Adjusted values of LV mass index in thethree groups are also depicted in Figure 3. A score offive or more by Romhilt-Estes criteria27 was found in46% of patients with echocardiographic LVH.An inverse relation was found (all, p<0.01) in the

hypertensive group between percentage of nocturnaldecline of systolic BP and LV mass (r= -0.35), LVmass index (r=-0.34), RWT (r=-0.29), CSA(r= -0.37), CSA index (r= -0.34), IVSd (r= -0.30),PWd (r= -0.38).

In the hypertensive group, there was also an

inverse relation (all, p<0.01) between percentage ofnocturnal decline of diastolic BP and LV mass(r=-0.32), LV mass index (r=-0.30), RWT(r=-0.22), CSA (r=-0.31), CSA index (r=-0.27),IVSd (r= -0.22), PWd (r= -0.32).None of the relations between percentage of noc-

turnal decline of systolic or diastolic BP and echo-cardiographic measures of LV anatomy was statisti-cally significant in the normotensive group.

DiscussionIn agreement with previous investigations,12-14 our

findings confirm that ambulatory BP is more closelyrelated to LV muscle mass than is casual BP. The mostimportant finding, however, of this study was theinverse relation between nocturnal BP decline and LVmuscle mass in a large population of unselected anduntreated patients with essential hypertension.

TABLE 5. Correlations Between Casual or Ambulatory Blood Pressure (24-Hour, Daytime, and Nighttime) and LeftVentricular Mass Index in Hypertensive Patients

Reference

Rowlands et al'4 Devereux et al13 Drayer et al12Blood pressure (n=50) (n=100) (n=12)

Casual blood pressureSystolic 0.45* 0.24t 0.55Diastolic 0.46* 0.20t 0.10

Ambulatory blood pressureAverage 24-hr systolic 0.60* 0.38* 0.81*Average 24-hr systolic 0.35* 0.31* 0.56Daytime systolic 0.57* ... 0.82*Daytime diastolic 0.41* ... 0.49Nighttime systolic 0.56* 0.10 0.70tNighttime diastolic 0.49* 0.24t 0.60

Devereux et al13 (n=100) is mixed population of 81 hypertensive and 19 normotensive patients. Rowlands et al'4(n =50) gives daytime systolic and diastolic ambulatory blood pressure reported as awake blood pressure. Rowlands etal'4 (n =50) and Devereux et al13 (n= 100) gives nighttime systolic and diastolic ambulatory blood pressure reported assleep blood pressure. Devereux et al'3 (n=100) gives daytime systolic and diastolic ambulatory blood pressureseparately reported for clinic, work, and home blood pressure. Drayer et al'2 (n = 12) gives daytime systolic and diastolicambulatory blood pressure reported as blood pressure averages from 6 AM to 8 PM. Drayer et al'2 (n= 12) gives nighttimesystolic and diastolic ambulatory blood pressure reported as blood pressure averages from 8 PM to 6 AM.

*p<0.01; tp<0.05.

1

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from

534 Circulation Vol 81, No 2, February 1990

Relation of Blood Pressure to Left Ventricular MassTable 5 shows the correlation coefficients between

LV mass index and casual or ambulatory BP in thethree main investigations comparable with our own.

In the study by Rowlands et a114 who performedintra-arterial BP monitoring in 50 patients with mildto moderate hypertension, the relations of casual andambulatory BP to LV mass index were similar as inthis study (see Table 2).

In the study by Devereux et al,13 the correlationcoefficients between casual or ambulatory BP and LVmass index were generally lower, and in that by Drayeret al,12 generally higher than our values. As we couldnot differentiate between work, clinic, and homeambulatory BP measurements, our findings about theinverse relation of nocturnal BP decline to LV musclemass do not rule out the possibility that daytime BPduring stress can be an independent determinant ofLVH development in hypertensive patients.

Prevalence of Left Ventricular Hypertrophy inEssential HypertensionOur prevalence figure (14.6%) is comparable with

that reported by Hammond et al,28 who studied an

unselected group of hypertensive patients drawnfrom an employed population (prevalence of LVHapproximating 12% in borderline hypertension and20% in sustained hypertension). In another studydone in 93 patients with mild-to-moderate hyperten-sion, prevalence of LVH was 16%.29

Prevalence of echocardiographic LVH was higherin other studies1,"'330 presumably done on more

selected hypertensive populations. In the study byHammond et a128 as well as in our own, severaltreated patients (about 30% of the study population)with complicated or severe hypertension and, pre-

sumably, a higher prevalence of LVH,2 3could not beincluded in the study, it being unethical to discon-tinue antihypertensive therapy.These findings suggest that the prevalence of LVH

in essential hypertension is strongly influenced by therecruiting criteria of the study population, and impli-cate that prevalence figures in different studies mightnot always be comparable.

Relation of Daytime and Nighttime Blood Pressure toLeft Ventricular Hypertrophy

Both in the study by Devereux et al13 and in that byDrayer et al,'2 daytime BP showed a generally closerrelation to LV mass than did nighttime BP. In thestudy by Rowlands et al,14 awake BP and sleep BPcorrelated equally well with LV mass, whereas inanother study performed on a mixed population ofhypertensive and normotensive patients, sleep BPshowed a closer relation to LV mass than did awakeBP.1' Therefore, the available data do not allow todraw definite conclusions about the possible predom-inance of daytime or nighttime BP on LVH develop-ment in hypertensive patients.

A constant finding in this and other studies12-15 isthe generally closer relation of systolic over diastolicBP to the degree of hypertrophy, another evidencethat wall stress, which is mostly related to systolic BP,is a key factor influencing LVH development.

In our normotensive population, the relation ofambulatory BP to the echocardiographic parametersof LV anatomy was slightly closer using daytimevalues, whereas in the hypertensive population, it wascloser with nighttime values. We referred to night-time BP, and not to sleep BP, because in ourexperience, sleep can be interrupted by cuff inflationand accompanying noise. Thus, one cannot be confi-dent that all indirect BP measurements over a givennocturnal period have been really obtained duringsleep.As shown with intra-arterial16 and noninvasive31

monitoring, ambulatory BP decreases during night byan average of 20% of the waking values. Individualprofiles, however, can diverge widely with somepatients showing greater nocturnal changes associ-ated with normal BP values even for several hoursand some others showing negligible changes.16 Toour knowledge, the possibility that nocturnal BPreduction might tend to outweigh the effect of day-time hypertension on LV hypertrophic response toincreased afterload has never been investigated.

In one study,32 intra-arterial BP variation from dayto night in 23 untreated hypertensive patients wasapparently similar in the presence versus absence ofLVH.32 An inspection to individual data of suchstudy, however, shows that BP decreased by less than10% from day to night in one of six patients withLVH, against one of 17 patients without LVH.Another study33 with intra-arterial BP monitoringreported no difference in the diurnal rhythm of BP inpatients with and without LVH.

In another study13 with noninvasive monitoring,BP reduction from day to night was similar in thepresence versus absence of LVH. In a recent prelim-inary report, Lang et a134 have shown increasedvalues of LV mass in 20 patients diagnosed as havingboth diurnal and nocturnal hypertension, in compar-ison with values in 10 patients with only daytimehypertension. In our study, patients with and withoutnighttime BP reduction by more than 10% of daytimevalues slightly differed for age, but not for othervariables that might have independently influencedLV muscle mass, such as sex, height, weight, bodysurface area, duration of hypertension, renal func-tion, prevalence of diabetes mellitus, and fundu-scopic changes. Also, casual BP and ambulatory BPduring the day did not differ significantly between thegroups, however, both being slightly higher in thegroup with a smaller nocturnal BP decline.

Therefore, only nighttime ambulatory BP discloseda very evident differentiation between the two hyper-tensive groups. All the echocardiographic indexes ofLVH, after adjustment for age differences, sex dif-ferences, height differences, and daytime systolic anddiastolic BP differences, were increased in the group

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from

Verdecchia et al

of hypertensive patients with a smaller nocturnal BPdecline, and the prevalence of LVH was increased inthis latter group.

Thus, the differences in LV muscle mass betweenthe groups appeared to be mostly explained by thedifferent nocturnal BP decline. The inverse relationbetween percentage of nocturnal reduction of sys-tolic and diastolic BP and degree of LVH givesfurther support to this point.

ConclusionsThe prognostic and therapeutic implications of the

changes in blood pressure from day to night are stillunknown. It has been suggested that hypertensivepatients losing nocturnal BP reduction as detectedwith noninvasive monitoring might be at higher riskof stroke.35The findings of this study support the concept that

the duration of exposure to increased levels of BPand wall stress over the 24 hours can play an impor-tant role in the pathogenesis of LVH in unselectedpatients with arterial hypertension. In these patients,a nocturnal reduction of systolic and diastolic BP bymore than 10% of daytime values can exert thebeneficial effect of delaying or preventing develop-ment of cardiac LVH.

References1. Kannel WB, Gordon T, Castelli WP, Margolis JR: Electrocar-

diographic left ventricular hypertrophy and risk of coronaryheart disease: The Framingham Study. Ann Intem Med 1970;72:813-822

2. Kannel WB, Gordon T, Offut D: Left ventricular hypertrophyby electrocardiogram. Prevalence, incidence and mortality inthe Framingham Study. Ann Intem Med 1969;71:89-105

3. Kannel WB: Prevalence and natural history of electrocardio-graphic left ventricular hypertrophy. Am J Med 1983;75(suppl3A):4- 11

4. Reichek N, Devereux RB: Left ventricular hypertrophy. Rela-tion of anatomic, echocardiographic and electrocardiographicfindings. Circulation 1981;63:1391-1398

5. Devereux RB, Reichek N: Echocardiographic determinationof left ventricular mass in man. Anatomic validation of themethod. Circulation 1977;55:613-618

6. Woythaler JN, Singer SL, Kwan OL, Melzer RS, Reulner B,Bommer W, DeMaria A: Accuracy of echocardiography versuselectrocardiography in detecting left ventricular hypertrophy:Comparison with postmortem mass measurements. JAm CollCardiol 1983;2:305-311

7. Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E,Sachs I, Reichek N: Echocardiographic assessment of leftventricular hypertrophy: Comparison to necropsy findings. AmJ Cardiol 1986;57:450-458

8. Carr AA, Prisant M, Watkins LO: Detection of hypertensiveleft ventricular hypertrophy. Hypertension 1985;7:948-954

9. Casale PN, Devereux RB, Milner M, Zullo G, Harshfield GA,Pickering TG, Laragh JH: Value of echocardiographic mea-surement of left ventricular mass in predicting cardiovascularmorbid events in hypertensive men. Ann Intem Med 1986;105:173-178

10. Savage DD, Garrison DJ, Castelli WP, Kannel WB, AndersonSJ, Feinleib M: Echocardiographic left ventricular hypertro-phy in the general population is associated with increased2-year mortality, independently of standard coronary riskfactors-the Framingham Study (abstract). AIA Council Car-diovasc Epidemiol Newslett 1985;37:33

11. Savage DD, Drayer JIM, Henry WL, Methews EC, Ware JH,Gardin JM, Cohen ER, Epstein SE, Laragh JH: Echocardio-

Nighttime Blood Pressure and Cardiac Hypertrophy 535

graphic assessment of cardiac anatomy and function in hyper-tensive subjects. Circulation 1979;59:623-632

12. Drayer JIM, Weber MA, DeYoung JL: BP as a determinant ofcardiac left ventricular muscle mass. Arch Intern Med 1983;143:90-92

13. Devereux RB, Pickering TG, Harshfield GA, Kleinert HD,Denby L, Clark L, Pregibon D, Jason M, Kleinert B, Borer JS,Laragh JH: Left ventricular hypertrophy in patients withhypertension: Importance of blood pressure response to reg-ularly recurring stress. Circulation 1983;68:470-476

14. Rowlands DB, Glover DR, Ireland MA, McLeay RAB, Stal-lard TJ, Watson RDS, Littler WA: Assessment of left ventric-ular mass and its response to antihypertensive treatment.Lancet 1982;1:467-470

15. Smith V-E, White WB, Karimeddini MK, McCabe EJ, KatzAM: Lowest not highest blood pressure may determine leftventricular filling (abstract). Circulation 1986;74(suppl II):II-290A

16. Littler WA, West MJ, Honour AJ, Sleight P: The variability ofarterial pressure. Am Heart J 1978;95:180-186

17. Littler WA: Sleep and blood pressure: Further observations.Am Heart J 1979;97:35-37

18. Abbasi AS: The role of Doppler echocardiography in valvularinsufficiency, in Berger M (ed): Doppler Echocardiography inHeart Disease. New YorklBasel, Marcel Dekker, Inc, 1987,pp 133-177

19. Verdecchia P, Gatteschi C, Benemio G, Boldrini F, GuerrieriM, Porcellati C: Home ambulatory blood pressure readings donot differ from clinic readings taken at the same time of day.J Hum Hypertens 1988;2:235-240

20. Harshfield GA, Pickering TG, Bkank S, Lindahl C, Stroud L,Laragh JH: Ambulatory blood pressure monitoring: Record-ers, applications and analyses, in: Weber MA, Drayer JIM(eds): Ambulatory Blood Pressure Monitoring. Darmstadt,Steinkopff Verlag, 1984, pp 1-7

21. Sahn DJ, DeMaria A, Kisslo J, Weyman A, The Committee onM-Mode Standardization of The American Society of Echo-cardiography. Recommendations regarding quantitation inM-Mode echocardiography: Results of a survey of echocardio-graphic measurements. Circulation 1978;58:1072-1083

22. Reichek N, Devereux RB: Reliable estimation of peak leftventricular systolic pressure by M-Mode echographic deter-mined end-diastolic relative wall thickness: Identification ofsevere valvular aortic stenosis in adult patients. Am Heart J1982;103:202-209

23. Ditchey RV, Schuler G, Peterson KL: Reliability of echocar-diographic and electrocardiographic parameters in assessingserial changes in left ventricular mass. Am J Med 1981;70:1042-1050

24. Dixon WJ (ed): BMDP Statistical Software. 1985 Printing. Berkeley,Calif, University of California Press, LTD, 1983, pp 345-358

25. Devereux RB, Pickering TG, Alderman MH, Chien S, BorerJS, Laragh JH: Left ventricular hypertrophy in hypertension.Prevalence and relationship to pathophysiologic variables.Hypertension 1987;9(suppl II):II-53-II-60

26. Devereux RB, Lutas EM, Casale PN, Kligfield P, EisenbergRR, Hammond IW, Miller DH, Reis G, Alderman MH,Laragh JH: Standardization of M-Mode echocardiographicleft ventricular anatomic measurements. J Am Coll Cardiol1984;4:1222-1230

27. Romhilt DW, Estes EH Jr: A point score system for theelectrocardiographic diagnosis of left ventricular hypertrophy.Am Heart J 1968;75:752-758

28. Hammond IW, Devereux RB, Alderman MH, Lutas EM,Spitzer MC, Crowley JS, Laragh JH: The prevalence andcorrelates of echocardiographic left ventricular hypertrophyamong employed patients with uncomplicated hypertension. JAm Coll Cardiol 1986;7:639-650

29. Kleinert HD, Harshfield GA, Pickering TG, Devereux RB,Sullivan PA, Marion RM, Mallory WK, Laragh JH: What isthe value of home blood pressure measurement in patientswith mild hypertension? Hypertension 1984;6:574-578

30. Devereux RB, Savage DD, Drayer JIM, Laragh JH: Leftventricular hypertrophy and function in high, normal, and

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from

536 Circulation Vol 81, No 2, February 1990

low-renin forms of essential hypertension. Hypertension 1982;4:524-531

31. Pickering TG, Harshfield GA, Kleinert HD, Blank S, LaraghJH: Blood pressure during normal daily activities, sleep andexercise. JAMA 1982;247:992-996

32. Heber ME, Brigden GS, Prince H, Lahiri A, Raftery EB: Isthere a relationship between ambulatory intra-arterial bloodpressure and left ventricular function? Hypertension 1988;11:464-469

33. Raftery EB: Understanding hypertension. The contribution ofdirect ambulatory blood pressure monitoring, in Weber MA,

Drayer JIM (eds): Ambulatory Blood Pressure Monitoring.Darmstadt, Steinkopff Verlag, 1984, pp 105-116

34. Lang RM, Borow KM, Weinert L, Bednarz J, Neumann A,Nelson K, David D, Murphy MB: Circadian blood pressurepatterns in hypertensive patients: Importance of nocturnalblood pressure (abstract). Circulation 1988;78(suppl II): 517

35. O'Brien E, Sheridan J, O'Malley K: Dippers and non-dippers.Lancet 1988;2:397

KEY WORDS * echocardiography * ambulatory blood pressuremonitoring * hypertension * left ventricular hypertrophy

by guest on July 7, 2014http://circ.ahajournals.org/Downloaded from