Review

White-coat hypertension, as defined by ambulatory

blood pressure monitoring, and subclinical cardiac
organ damage: a meta-analysis
Cesare Cuspidi a,b, Marta Rescaldani c, Marijana Tadic d, Carla Sala c, Guido Grassi a,e, and
Giuseppe Mancia a,b

Aim: The clinical and prognostic relevance of white-coat

INTRODUCTION

W
hypertension (WCH) has not been fully elucidated; in hite-coat or isolated clinic hypertension currently
particular, the association of this blood pressure phenotype defines individuals whose blood pressure (BP) is
with suclinical organ damage remains unclear. We elevated in the medical setting, but normal when
performed a systematic meta-analysis in order to provide a assessed away from the medical environment, such as by
comprehensive information on cardiac structural and 24-h ambulatory BP recording and/or home BP measure-
functional changes in WCH, as defined by ambulatory ment [1–3]. Since the pioneering publication by Pickering
blood pressure monitoring. et al. [1] in which ‘white-coat hypertension’ (WCH) was
Design: Studies were identified by the following search used for the first time to define untreated hypertensive
terms: ‘white-coat hypertension’, ‘isolated clinic patients, the vast majority of studies agree that this con-
hypertension’, ‘cardiac organ damage’, ‘target organ dition accounts for a noticeable fraction of the hypertensive
damage’, ‘left ventricle’, ‘left ventricular hypertrophy’, population [4,5]. No agreement, however, exists on the
‘cardiac hypertrophy’, ‘ventricular dysfunction’, and prognostic significance of WCH: whether it is an innocent
‘echocardiography’. clinical entity or is associated with an adverse/increased
cardiovascular risk is still unsettled [6]. Despite numerous
Results: A total of 7382 untreated adult patients (2493
investigations, the presence and extent of increased car-
normotensive, 1705 WCH, and 3184 hypertensive
diovascular risk in WCH patients as compared to their true
individuals) included in 25 studies were considered. Left
normotensive counterparts remain controversial. This is
ventricular mass index was higher in WCH than in
because the extent of subclinical organ damage [that is left
normotensive patients [standardized difference in mean
ventricular hypertrophy (LVH), carotid atherosclerosis,
(SDM) 0.50, P < 0.01]; mitral E/A ratio was lower (SDM
microalbuminuria, and retinopathy] in WCH patients has

0.27, P < 0.01) and left atrium larger (SDM 0.29,
been reported to be similar as in normotensive patients by
P < 0.05) in WCH than in the normotensive counterparts.
some investigators, but as severe as in sustained hyper-
Hypertensive patients showed a greater left ventricular
tensive patients by others [7–10]. Likewise, the incidence of
mass index (SDM 0.42, P < 0.01), reduced E/A (SDM
cardiovascular morbid or fatal events in patients with WCH

0.15, P < 0.01), and larger left atrium diameter (SDM
has been reported to be similar either as in normotensive
0.27, P < 0.01) than WCH patients.
patients, as in hypertensive patients, or intermediate
Conclusions: Our meta-analysis shows that alterations in between these groups [11–13]. The relationship between
cardiac structure and function in WCH patients, as defined cardiovascular outcomes and WCH has been also inves-
by ambulatory blood pressure monitoring, are intermediate tigated by some recent meta-analyses.
between sustained hypertensive patients and normotensive
controls. The study supports the view that WCH should
not be further considered a fully benign entity.
Keywords: ambulatory blood pressure monitoring, cardiac Journal of Hypertension 2015, 33:24–32
damage, white-coat hypertension a
Department of Health Science, University of Milano-Bicocca, bIstituto Auxologico
Italiano, cDepartment of Clinical Sciences and Community Health, University of Milan,
Abbreviations: ABPM, ambulatory blood pressure Milan, Italy, dFondazione, Policlinico di Milano University Clinical Hspital Centre
monitoring; BP, blood pressure; BSA, body surface area; ‘Dragisa Misovic’, Belgrade, Serbia and eIstituto di Ricerche a Carattere Scientifico
Multimedica, Sesto San Giovanni, Milan, Italy
LVH, left ventricular hypertrophy; SDM, standardized
Correspondence to Professor Cesare Cuspidi, Istituto Auxologico Italiano, Clinical
difference in means; E/A ratio, ratio of early (E) to late (A) Research Unit, Viale della Resistenza 23, 20036 Meda, Italy. Tel: +39 0362/772433;
peak of mitral inflow velocity; WCH, white-coat fax: +39 0362/772416; e-mail: [email protected]
hypertension Received 4 May 2014 Revised 9 September 2014 Accepted 9 September 2014
J Hypertens 33:24– 32 ß 2014 Wolters Kluwer Health | Lippincott Williams &
Wilkins.
DOI:10.1097/HJH.0000000000000416

24 www.jhypertension.com Volume 33  Number 1  January 2015

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 White-coat hypertension and cardiac damage

In a pooled population of 7961 untreated patients (16% hypertrophy’, ‘cardiac hypertrophy’, ‘ventricular dysfunc-
with WCH), Pierdomenico and Cuccurullo [14] showed that tion’ and ‘echocardiography’. Checks of the reference lists
cardiovascular risk was not different in WCH compared to of selected papers and pertinent reviews complemented
true normotensive patients. The International Database on the electronic search. Data have been extracted by three
Ambulatory Blood Pressure Monitoring in Relation to Car- independent investigators (C.C., M.R., and C.S.); additional
diovascular Outcomes (IDACO) study [15], assessing the data have been obtained by personal contact with authors
significance of WCH in older persons with isolated systolic of the selected papers.
hypertension, free of cardiovascular disease at baseline and Inclusion criteria were: full articles published in English
stratified according to the presence or absence of antihy- in peer-reviewed journals; studies reporting quantitative
pertensive treatment, reported that untreated WCH and data on left ventricular structure, as defined by left ven-
normotensive patients were at a similar risk. This was tricular mass indexed to body size measures in at least 15
not true for treated WCH patients, who had a higher untreated adult patients with WCH; normal out-of-office BP
cardiovascular risk as compared with the untreated defined by ambulatory BP monitoring (ABPM) (that is mean
normotensive patients. 24-h or mean daytime levels).
The association between subclinical organ damage and Only updated or largest reports were considered when
WCH remains controversial; in particular, meta-analysis- multiple publications by the same research group were
based findings on this important topic are lacking. There- found in order to avoid double counting patients. The first
fore, the primary aim of this systematic review and meta- literature search identified 392 papers. After the initial
analysis was to provide a comprehensive and updated screening of titles and abstracts, 298 studies were excluded
information on the presence and extent of subclinical and 94 were reviewed; of these, 25 studies fulfilled the
structural and functional cardiac damage, as assessed by inclusion criteria and contained sufficient details to be
echocardiography in untreated patients with WCH. included in the final review [16–40] (Fig. 1).

METHODS Statistical analysis

The primary aim of the meta-analysis was to compare sub-
Search strategy and study selection clinical alterations in left ventricular structure, expressed as a
Medical literature was reviewed in order to identify all continuous variable (i.e. left ventricular mass index) and/or
articles evaluating the impact of WCH on cardiac structure as a categorical variable (i.e. prevalence of LVH), as assessed
and function as assessed by echocardiography. by echocardiography, in WCH compared to sustained hyper-
A computerized search was performed using PubMed, tensive and/or normotensive controls. To this purpose,
OVID, EMBASE, and Cochrane library databases from 1 pooled analysis of echocardiographic parameters was per-
December 1990 up to 31 January 2014. The studies were formed using fixed or random-effects meta-analysis by Com-
identified by the following search terms: ‘white-coat hyper- prehensive Meta-Analysis Version 2 (Biostat, Englewood,
tension’, ‘isolated clinic hypertension’, ‘cardiac organ dam- New Jersey, USA). In order to calculate the average preva-
age’, ‘target organ damage’, ‘left ventricle’, ‘left ventricular lence of LVH in the pooled population, we considered the

Studies identified and

screened for retrieval
(n = 392)

Studies excluded based on

title and abstract review
(n = 298)

Appropriate studies to
be included in the
review - No echocardiographic data on left
(n = 94) ventricular mass index (n = 16)
- WCH not defined by ambulatory BP
criteria (n = 6)
- Review or editorial articles (n = 21)
- Lack of clinical data (n = 8)
- Antihypertensive treatment (n = 6)
- Study population <15 patients (n = 5)
- Double or serial publications (n = 4)
- Miscellaneous reasons (n = 3)
Studies included in
the final review
(n = 25)

FIGURE 1 Schematic flowchart for the selection of studies. BP, blood pressure; WCH, white-coat hypertension.

Journal of Hypertension www.jhypertension.com 25

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Cuspidi et al.

occurrence of LVH as an event rate. Demographic and in population-based samples [27,29,39], and one in a
clinical data provided by the selected studies are expressed primary care setting [26].
as absolute numbers, percentage, mean  SD, mean 
standard error (SE), or median and interquartile range. Characteristics of white-coat hypertension
Meta-regression analysis was used to determine the impact patients
of office and ambulatory BP upon left ventricular mass index. Mean age range was 33–70 years [24,29]; 50.9% of the
The limit of statistical significance was set at P value less participants were men (n ¼ 846, data provided by 24 stud-
than 0.05. ies including 1663 patients). Average office SBP ranged
Heterogeneity was estimated using the I-square test; from 141  13 [35] to 176  12 mmHg [18], and DBP from
random-effect models were applied when the heterogen- 86  9 [39] to 105  10 mmHg [23]. Average daytime SBP
eity across studies was high (I2 >75). Publication bias was varied from 115  12 [34] to 137  8 mmHg [18], and DBP
assessed using the funnel plot method. from 73  6 [36] to 88  6 mmHg [16] (24 studies, 1527
patients). Average BMI ranged from 24.9  2.5 [19] to
28.7  5.2 kg/m2 [37] (18 studies, 1384 patients). All
RESULTS examined patients were free from previous or overt
Characteristics of the studies cardiovascular disease.
Table 1 shows the main characteristics of the analyzed Normal ABPM thresholds were defined according to 11
studies, including the year of publication, sample size, different criteria (five based on average 24-h BP and six
mean age, sex distribution, mean BMI, office BP, mean on average daytime values). The most frequently used
24-h and/or daytime ABPM values, and criteria defining criterion for defining normal out-of office BP was average
normal ABPM values. daytime SBP/DBP lower than 135/85 mmHg [24–26,29,
Overall, 7382 untreated patients (2493 normotensive, 31–35,38–40].
1705 WCH, and 3184 hypertensive individuals) of both
sexes were included in the 25 studies performed in different Echocardiographic findings in white-coat
geographical areas (Europe 18; Asia 6; North America 1). hypertension, true normotensive and
The assessment of left ventricular structural and/or func- hypertensive patients
tional changes associated to WCH was the primary aim of all In all selected studies, left ventricular mass was normalized
studies but two [33,40]. Most of these studies examined to body surface area (BSA). In the pooled study population,
patients recruited in out-patient hypertension clinics, three mean left ventricular mass index ranged from 70 [38] to

TABLE 1. Summary of 25 studies reporting data on left ventricular structure and function in white-coat hypertensive patients
Mean Mean Normal
Year Sample Office 24-h daytime ambulatory
Author of size Age Men BMI SBP/DBP SBP/DBP SBP/DBP BP values
(reference) publication (n) (years) (%) (kg/m2) (mmHg) (mmHg) (mmHg) (mmHg)
Cardillo et al. [16] 1993 18 43  5 55 NA 148  13/98 6 121  5/82  5 126  7/88  6 Day SBP/DBP <134/90
Hoegholm et al.[17] 1993 53 46  13 36 25.1 þ 3.4 158 þ 16/102 þ 7 NA 133  13/84  5 Day DBP <90
Kuwajiama et al. [18] 1993 17 74  6 18 NA 176 þ 12/91 þ 8 133  6/74  6 137  8/78  7 24-h SBP <140
Cavallini et al. [19] 1995 24 61  9 33 24.9  2.5 158 þ 10/92 þ 4 128  5/77  5 130  5/79  6 Day SBP/DBP <134/90
Cuspidi et al. [20] 1995 31 35  12 65 NA 144  18/97  4 127  6/79  4 132  7/83  5 24-h SBP/DBP <132/85
Pierdomenico 1995 25 46  11 52 26.3 þ 2.8 149 þ 5/96 þ 2 123  7/75  5 129  6/78  6 24-h SBP/DBP <135/85
et al. [21]
Rizzo et al. [22] 1996 22 69  3 55 NA 159  15/102 þ 7 NA 133  13/84  5 Day SBP/DBP <142/90
Glen et al. [23] 1996 22 58  8 64 NA 160  19/105  10 NA 135  10/83  6 Day DBP <95
Palatini et al. [24] 1998 260 33  8 69 25.1  3.7 143  10/93  5 121  7/76  6 124  8/88  6 Day SBP/DBP <135/85
Owens et al. [25] 1999 33 40 27 NA 162/102 8 NA 125/78 Day SBP/DBP <135/85
Martinez et al. [26] 1999 71 54  11 35 28.0 þ 4.0 146 þ 15/95 þ 6 NA 124  8/80  7 Day SBP/DBP <135/85
Sega et al. [27] 2001 178 58  11 50 NA 149 þ 9/93 þ 4 119  6/74  4 NA 24-h SBP/DBP <125/80
Grandi et al. [28] 2001 42 42  7 NA 25.3  2.7 154  16/93  14 120  5/70  5 126  5/74  6 Day SBP/DBP <130/80
Bjorklund et al. [29] 2002 49 70 100 25.3  3.0 150  17/85  5 NA 128  6/75  5 Day SBP/DBP <135/85
Silveira et al. [30] 2002 57 46  2a 47 26.0  1.0a 148  3/89  2a NA 124  2/79  1a Day SBP/DBP <130/84
Pose-Reino et al. [31] 2002 27 46  12 44 27.5  3.2 148  11/96  5 120  8/71  6 125  9/75  7 Day SBP/DBP <135/85
Karter et al. [32] 2003 24 50  11 46 29.0  4.0 156  21/98  11 NA 121  6/74  5 Day SBP/DBP <135/85
Curgunlu et al. [33] 2005 33 49  2 48 25.0  0.9 152  7/88  3 122  5/75  3 131  3/82  6 Day SBP/DBP <135/85
Erdogan et al. [34] 2006 35 47  6 46 28.1  1.9 147  8/93  4 113  10/73  11 115  12/75  11 Day SBP/DBP <135/85
Cuspidi et al. [35] 2007 43 46  12 53 25.7  3.6 141  13/95  7 121  5/77  4 125  4/81  3 Day SBP/DBP <135/85
Mulè et al. [36] 2007 145 43 þ 12 48 27.5  4.0 150  17/92  10 117  7/70  6 121  7/7  6 Day SBP/DBP <130/80
Kotsis et al. [37] 2008 274 52  14 37 28.7  5.2 154  13/95  10 120  7/77  6 121  6/73  6 24-h SBP/DBP <125/80
Ihm et al. [38] 2009 30 48  9 33 24.0  3.0 145  16/95  12 NA 124  7/76  6 Day SBP/DBP <135/85
Sung et al. [39] 2013 153 58  13 51 25.0  3.0 145  13/86  9 122  7/76  5 126  8/77  8 Day SBP/DBP <135/85
Caliskan et al. [40] 2013 40 45  7 45 28.1  2.2 146  7/93  4 116  11/74  10 123  7/78  5 Day SBP/DBP <135/85

NA, not available.

a
Standard error; confidence intervals.

26 www.jhypertension.com Volume 33  Number 1  January 2015

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 White-coat hypertension and cardiac damage

124 g/m2 [29] in the normotensive controls, from 70 [30] to findings of the meta-analysis from 22 studies, left ven-
132 g/m2 [31] in the WCH patients, and from 84 [30] tricular mass index was significantly higher in sustained
to 142 g/m2 [31] in the sustained hypertensive patients. hypertensive (n ¼ 3184) as compared to WCH patients
As shown in Fig. 2a, mean left ventricular mass index (n ¼ 1505), with a SDM of 0.42  0.03 (95% CI 0.35–0.48,
was lowest in normotensive (88.05  2.5 g/m2), inter- P < 0.01).
mediate in WCH (95.72  1.8 g/m2) and highest Data on LVH prevalence in normotensive, WCH, and
(109.2  2.5 g/m2) in the sustained hypertensive patients. sustained hypertensive patients were provided by a limited
Figure 3a reports the results of the meta-analysis from fraction of the reports included in the meta-analysis, namely
20 studies providing data on average left ventricular by two [27,31], eight [17,19–22,27,31,35], and seven [17,19–
mass indexed to BSA and SD in 1355 WCH patients 21,27,31,35] studies, respectively. Overall, 58 out of the 408
and 2493 normotensive controls. The standardized differ- WCH patients (15%) and 209 out of the 1028 sustained
ence in means (SDM) was positive in favor of the WCH hypertensive patients (21%) were found to have LVH
individuals [0.50  0.10, 95% confidence interval (CI) according to the different criteria provided by the authors.
0.31–0.70, P < 0.01]. As shown in Fig. 3b, reporting the LVH prevalence consistently varied among studies, ranging
from 4 to 59% in WCH and from 13 to 75% in sustained
hypertensive patients.
As for left ventricular diastolic function, as assessed by
(a)
120 the ratio of early (E) to late (A) peak of mitral inflow velocity
(E/A ratio), the average value from pooled data of eight
110 studies [16,18,23,24,29,34,38,40] was 1.17  0.07 in normo-
tensive (n ¼ 337), 1.07  0.07 in WCH (n ¼ 471), and
100 0.99  0.11 in sustained hypertensive patients (n ¼ 852)
LVMI (g/m2)

(Fig. 2b). The ratio was higher in normotensive than in

90 WCH patients (SDM 0.27  0.07, 95% CI 0.12–0.41,
P < 0.01) (Fig. 4a) and in WCH compared to sustained
80 hypertensive patients (SDM 0.20  0.06, 95% CI 0.09–
0.31, P ¼ 0.01) (Fig. 4b).
70 Finally, pooled data from five studies [17,18,31,34,40]
2493 1705 3184 documented that left atrium diameter was 3.26  0.03 cm in
60
normotensive (n ¼ 161), 3.31  0.11 cm in WCH (n ¼ 193),
(b) and 3.44  0.13 cm in sustained hypertensive patients
1.40 (n ¼ 261) (Fig. 2c). Left atrium diameter was greater in
WCH than in 141 normotensive patients (SDM
0.29  0.13, 95% CI 0.04–0.54, P < 0.05) (Fig. 5a) and in
1.20 sustained hypertensive than in WCH patients (SDM
0.27  010, 95% CI 0.07–0.46, P < 0.01) (Fig. 5b).
E/A ratio

1.00 Correlation analyses

A meta-regression analysis performed on data from all WCH
(Fig. 6a) and sustained hypertensive patients (Fig. 6b)
0.80
included in the 25 studies showed a direct correlation
337 471 852
between left ventricular mass index and office SBP
(r ¼ 0.42, P < 0.01; and r ¼ 0.64, P < 0.05, respectively).
0.60
The relation between left ventricular mass index and
daytime SBP in WCH patients (meta-regression data from 24
(c)
3.75 studies, 1527 WCH individuals) was not significant
(r ¼ 0.18, P ¼ 0.17). This was also the case for sustained
3.50 hypertensive patients (meta-analysis from 21 studies, 3163
individuals), as left ventricular mass index showed a weak
correlation with daytime BP (P ¼ 0.26, P ¼ 0.44). Further-
LA (cm)

3.25
more, we failed to demonstrate in WCH patients a signifi-
3.00 cant relation between left ventricular mass index and 24-h
SBP (P ¼ 0.32) in the 16 studies providing this kind of
2.75 information. Similar findings were observed in sustained
161 196 261 hypertensive patients.
2.50 A funnel plot excluded the presence of publication bias
NT WCH SH of studies comparing left ventricular mass index in WCH
FIGURE 2 (a) Left ventricular mass index (LVMI), (b) E/A ratio, and (c) left atrium and sustained hypertensive patients (23 studies), as well in
diameter in normotensive (NT), white-coat hypertensive (WCH), and sustained WCH and normotensive controls (20 studies). A sensitivity
hypertensive (SH) patients. Meta-analysis from 25 echocardiographic studies.
Means  SE; number of patients in each group are reported in the histograms. SE, analysis showed that the final result was not substantially
standard error. affected by a single study effect.

Journal of Hypertension www.jhypertension.com 27

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Cuspidi et al.

(a)
Left ventricular mass index
SDM and 95% CI
StudyRef. % weight
Cardillo16 3.53
Owens25 4.18
Mule36 5.06
Kuwajiama18 3.69
Pose-Reino31 4.92
Karter32 4.43
Curgunlu33 4.84
Grandi25 5.20
Sega27 6.61
Ihm38 4.74
Palatini24 6.30
Sung39 6.45
Pierdomenico21 4.51
Cavallini19 4.46
Rizzo22 4.26
Caliskan40 5.19
Bjorklund29 5.69
Erdogan34 5.03
Kotsis37 6.66
Glen23 4.26
Total 100
–2.0 –1.0 0.0 1.0 2.0
Favours Favours
NT WCH

(b) Left ventricular mass index

SDM and 95% CI
StudyRef. % weight
Pierdomenico21 1.45
Glen23 0.99
Ihm38 1.44
Cavallini19 1.23
Cuspidi20 2.00
Silveira30 2.83
Grandi25 2.25
Cardillo16 1.30
Palatini24 I8.05
Sung39 13.00
Caliskan40 2.65
Karter32 1.22
Curgunlu33 1.86
Kuwajiama18 1.24
Bjorklund29 4.12
Martinez26 1.94
Hoelgholm17 3.67
Kotsis37 19.70
Erdogan34 2.28
Pose-Reino31 1.41
Cuspidi35 4.45
Sega27 10.91
Total 100
–2.0 –1.0 0.0 1.0 2.0
Favours Favours
WCH SH
FIGURE 3 Forest plots for standardized difference in mean (SDM) of left ventricular mass index (LVMI, g/m2) in: (a) white-coat hypertensive (WCH, n ¼ 1335) and
normotensive (NT, n ¼ 2493) patients; (b) WCH (n ¼ 1505) and sustained hypertensive (SH, n ¼ 3184) patients. CI, confidence interval.

28 www.jhypertension.com Volume 33  Number 1  January 2015

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 White-coat hypertension and cardiac damage

(a) (a) Left atrium diameter

E/A ratio SDM and 95%CI
StudyRef. SDM and 95% CI %Weight Study Ref.
%Weight
Cardillo16 4.67
Kuwajiama18 12.64
Glen23 5.99
8.07 Pose-Reino31 27.34
Ihm38
Kuwajiama18 4.57 Erdogan34 27.99
Palatini24 38.79 Caliskan40 32.02
Bjorklund29 17.01 Total 100
Caliskan40 11.08
Erdogan34 9.81 –2.0 –1.0 0.0 1.0 2.0
Favours Favours
Total 100
NT WCH
–2.0 –1.0 0.0 1.0 2.0
Favours Favours (b) StudyRef. %Weight
WCH NT
(b) StudyRef. Hoelghom17 32.47
%Weight
Caliskan40 7.79 Kuwajiama18 10.90
Ihm38 4.84
Pose-Reino31 12.45
Glen23 3.48
Bjorklund29 12.57 Erdogan34 20.24
Erdogan34 6.92 Caliskan40 23.93
Kuwajiama18 3.84
Palatini24 56.51 Total 100
Cardillo16 4.07 –2.0 –1.0 0.0 1.0 2.0

Total 100 Favours Favours

WCH SH
–2.0 –1.0 0.0 1.0 2.0
FIGURE 5 Forest plot for standardized difference in mean (SDM) of left atrial
Favours Favours diameter in: (a) white-coat hypertensive (WCH, n ¼ 118) and normotensive patients
SH WCH (NT, n ¼ 141); (b) WCH (n ¼ 171) and sustained hypertensive patients (SH, n ¼ 261).
FIGURE 4 Forest plot for standardized difference in mean (SDM) of E/A ratio in: (a)
white-coat hypertensive (WCH, n ¼ 471) and normotensive patients (NT, n ¼ 337); reduced in WCH as compared to the normotensive patients
(b) WCH (n ¼ 471) and sustained hypertensive patients (SH, n ¼ 852). and in sustained hypertensive as compared to the WCH
patients; left atrial diameter was greater in WCH as com-
Study quality evaluation pared to the normotensive controls, and in sustained hyper-
Two reviewers graded each study independently evaluat- tensive as compared to the WCH patients; office but not
ing four items: office BP measurements taken in two or ambulatory BP showed a direct, significant correlation with
more sessions versus a single session; use of validated left ventricular mass index in both WCH and sustained
ABPM devices; blind assessment of echocardiographic hypertensive patients. Several aspects of our results deserve
examinations; and sample size including at least 35 cases to be further commented.
and controls, allowing to detect a 10% or greater difference Our meta-analysis clearly shows that WCH is a risk factor
among groups in echocardiographic variables. The score of for LVH development. Patients with WCH, indeed, exhib-
each article can range from 0 (lowest quality) to 4 (highest ited an average left ventricular mass index intermediate
quality). Nine studies showed a score of 4 points (excellent between true normotensive and sustained hypertensive
quality), seven a score of 3 points (good quality), and nine a patients. This finding supports the view that transient BP
score of 2 points (fair quality). rises, such as those triggered by stress-related sympathetic
activation during clinical visit, may modulate cardiac
DISCUSSION growth. This hypothesis is also supported by the meta-
regression including all selected studies that shows a sig-
The present meta-analysis of 25 studies published in the nificant, direct association between office BP and left ven-
past two decades provides an updated information on tricular mass index. Thus, transient BP elevations in the
subclinical markers of cardiac damage, as assessed by office setting tend to increase the left ventricular mass,
echocardiographic left ventricular mass index, E/A ratio, although this parameter remains below the conventional
and left atrial diameter in a pooled population of 1705 WCH cut-off values for hypertrophy, as in our pooled population,
patients from different clinical settings characterized by only a limited fraction of the WCH patients (15%) fulfilled
various degrees of office BP levels as compared to true the diagnostic criteria for LVH.
normotensive and hypertensive counterparts. The relation between left ventricular mass and risk of
The principal findings of our analysis are the following: cardiovascular events has been shown to be extended over
left ventricular mass index showed a graded, signifi- a wide range of left ventricular mass values, and to persist
cant increase from normotensive, WCH to sustained below upper normal limits currently accepted by the
hypertensive patients; the E/A ratio was significantly authoritative international guidelines [3].

Journal of Hypertension www.jhypertension.com 29

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Cuspidi et al.

(a) adjustment for traditional risk factors, as well as new ones,

160 such as carotid–femoral pulse wave velocity.
r = 0.42
P < 0.01 The increased left ventricular mass is the major mani-
140 festation of hypertensive heart disease, and reflects the
response of the heart to chronic BP elevation aimed to
LVMI (g/m2)

120 counterbalance the left ventricular wall stress [43,44].

Additional alterations of cardiac structure and function
100 may also occur in the hypertensive patients, including left
atrial enlargement and left ventricular systolic/diastolic
80 dysfunction [45,46].
Most of the attention has been focused on left ventricular
60 diastolic dysfunction because of the high prevalence of this
130 140 150 160 170 180 alteration and, more importantly, for its association with an
Office systolic BP (mmHg)
increased risk of heart failure and cardiovascular mortality.
A large body of evidence points to LVH as a key factor in the
(b) pathogenesis of diastolic dysfunction, although other
160
r = 0.64 observations indicate that an impairment of left ventricular
P < 0.01 compliance may precede LVH development [46]. Among
140
conventional Doppler indexes assessing diastolic function,
LVMI (g/m2)

120
the E/A ratio has been widely used in clinical research and
current practice [47]. The relation of this index with the
cardiovascular outcomes has a U-shaped form, the lowest
100
and highest values being associated with poor cardiovas-
cular prognosis in different clinical settings [48]. In our
80
meta-analysis, the E/A ratio progressively decreased from
normotensive, WCH, and sustained hypertensive patients:
60 this finding indicates that a subtle impairment of left ven-
130 140 150 160 170 180
tricular diastolic function, as expressed by a reduction in the
Office systolic BP (mmHg) E/A ratio, may be detected not only in sustained hyper-
FIGURE 6 Meta-regression of left ventricular mass index (LVMI) with office SBP in: tensive but also in WCH patients.
(a) white-coat hypertensive (n ¼ 1431) and (b) sustained hypertensive patients Left atrial enlargement is also regarded as an independ-
(n ¼ 3164).
ent marker of increased cardiovascular risk in the general
population, as well as in patients with hypertension and
chronic heart disease [49]. Cross-sectional studies have
An observational study by Schillaci et al. [41], including a shown that left atrial enlargement is associated with a
total of 1925 uncomplicated essential hypertensive patients, variety of pathogical entities such as hypertensive heart
showed that the relative risk of developing cardiovascular disease, diabetes, obesity, metabolic syndrome, and sleep
events progressively increased from the first to the fifth apnea. Longitudinal investigations have consistently dem-
quintile of the left ventricular mass index, after adjusting onstrated that left atrial enlargement, as documented by a
for several risk factors and 24-h ambulatory BP values. Similar single left atrial diameter or the more accurate left atrial
results have been recently reported in the Pressioni Moni- volume measurement, is a strong predictor of cardiovascular
torate E Loro Associazioni (PAMELA) study, a population- outcomes [50,51]. In spite of the clinical and prognostic
based study comprising 1716 patients [42]. After adjusting for relevance of this issue, only a few studies included in the
age, sex, office or ambulatory BP, blood glucose, total present meta-analysis provided data on left atrial size, in
cholesterol, and use of antihypertensive drugs, the patients particular, less than 10% of the total population of WCH and
stratified in the two highest quintiles of the left ventricular less than 5% of normotensive controls. In spite of this
mass indexed to BSA or height2.7 exhibited a greater like- limitation, the significant difference in left atrial diameter
lihood of incident cardiovascular disease, the relative risk between the normotensive and WCH patients indicates
being 2.69 (95% CI1.05–6.96, P ¼ 0.04) and 4.62 (95% CI that left atrial morphology is also not preserved in
1.42–15.02, P ¼ 0.01), respectively, as compared to the first WCH patients.
reference quintile. These results support the view that Some other points of our study, including the strengths
patients with left ventricular mass index in the high-normal and limitations, deserve to be briefly discussed. First, the
range are at a higher risk than their counterparts in the meta-regression analyses demonstrated a direct, significant
lower range. relation between office SBP and left ventricular mass index
In a prospective study, Sung et al. [39]. showed that in both WCH and sustained hypertensive patients. This
WCH patients with intermediate values of left ventricular finding is in keeping with the general notion that elevated
mass index (98  25 g/m2) between true normotensive SBP values measured in the office are associated with
(89  21 g/m2) and sustained hypertensive patients increased risk of subclinical cardiac damage. Our meta-
(111  28 g/m2) displayed a higher risk for cardiovascular analysis adds a new piece of information on this topic by
mortality than their normotensive counterparts. The greater showing that transient elevation in BP in the setting of WCH
hazard ratio in the WCH patients remained significant after may impact on the cardiac structure.

30 www.jhypertension.com Volume 33  Number 1  January 2015

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 White-coat hypertension and cardiac damage

In contrast, we found no significant relation between REFERENCES

ambulatory daytime systolic or 24-h SBP and left ventricular 1. Pickering TG, James GD, Boddie G, Harshfield GA, Blank S, Laragh JH.
mass index in either group. The closer association between How common is white-coat hypertension? JAMA 1988; 259:225–228.
left ventricular mass index and office as compared to 2. Owens P, Atkins N, O’Brien E. Diagnosis of white coat hypertension by
ambulatory blood pressure monitoring. Hypertension 1999; 34:267–
ambulatory BP observed in this study, not only in WCH 272.
but also in sustained hypertensive patients, appears in 3. Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Bohm M, et al.
contrast with the view that multiple measurements of BP 2013 ESH/ESC guidelines for the management of hypertension.
during routine life reflect BP load more accurately than a J Hypertens 2013; 31:1281–1357.
single BP measurement in the doctor’s office. 4. Niiranen TJ, Jula AM, Kantola IM, Reunanen A. Prevalence and deter-
minants of isolated clinic hypertension in the Finnish population.
This may be related to the previous observations that in J Hypertens 2006; 24:463–470.
sustained hypertensive patients, nocturnal BP load is of 5. Cuspidi C, Parati G. Isolated clinic hypertension in a general popu-
major relevance in the development of LVH, as suggested lation. J Hypertens 2006; 24:437–440.
by a recent study of our group [52]; unfortunately, nocturnal 6. Mancia G. White coat hypertension. Innnocuous or adverse phenom-
enon? Eur Heart J 2000; 21:1647–1648.
BP values were not available in all studies of the present 7. Mancia G, Bombelli M, Seravalle G, Grassi G. Diagnosis and manage-
analysis. In WCH, on the contrary, diurnal BP values were, ment of patients with white coat and masked hypertension. Nat Rev
by definition, in the normal range. Cardiol 2011; 8:686–693.
Second, as reported in the ‘Methods’ section, we selected 8. Zakopoulos N, Papamichael C, Papaconstantinou H, Dubbins PA,
only studies in which normal BP outside the medical Burrell CJ, Lekakis J, et al. Isolated clinic hypertension is not an
innocent phenomenon: effect on the carotid artery. Am J Hypertens
environment was based on ABPM. This is because we 1999; 12:245–250.
included only studies conducted with comparable BP 9. Di Mauro S, Spallina G, Scalia G, Insolia G, Borzi S, Distefano A, Torrisi
measurement tools, and this methodological approach G. Urinary albumin excretion in elderly patients with white coat
for the diagnosis of WCH phenotype has been used by hypertension. Arch Gerontol Geriatric 1999; 28:23–29.
10. Triantafyllou A, Doumas M, Anyfanti P, Gkaliagkousi E, Zabulis X,
the vast majority of reports published in the literature. Petidis K, et al. Divergent retinal vascular abnormalities in normoten-
The left ventricular mass was indexed to BSA in all sive persons and patients with never-treated, masked, with coat
selected studies: this allowed us to compare the echocar- hypertension. Am J Hypertens 2013; 26:318–325.
diographic data concerning the left ventricular structure of 11. Khattar RS, Senior R, Lahiri A. Cardiovascular outcome in white-coat
all cases and controls examined across the 25 studies, versus sustained mild hypertension. A 10-year follow-up study.
Circulation 1998; 98:1892–1897.
without potential confounders related to heterogeous 12. Strandberg TE, Saloma V. White coat effect: blood pressure and
methods of left ventricular mass indexation. It should be, mortality in men: prospective cohort study. Eur Heart J 2000;
however, noted that left ventricular mass normalized to 21:1714–1718.
different body size measures such as height or height to 13. Gustavsen PH, Hoegholm A, Bang LE, Kristensen KS. White coat
hypertension is a cardiovascular risk factor: a 10-year follow-up study.
allometric power of 2.7 was not an exclusion criterion in J Hum Hypertens 2003; 17:811–817.
our study selection. 14. Pierdomenico SD, Cuccurullo F. Prognostic value of white coat and
In early studies, upper normal limits of ambulatory BP masked hypertension diagnosed by ambulatory monitoring in initially
were higher [16–19,22,23] than those recommended by untreated subjects: an updated meta-analysis. Am J Hypertens 2011;
contemporary guidelines. Thus, a fraction of the patients 24:52–58.
15. Franklin SS, Thijs L, Hansen TW, Li Y, Boggia J, Kikuya M, et al.
classified as having WCH in the previous studies nowadays Significance of white coat hypertension in older persons with isolated
would be classified as sustained hypertensive patients; systolic hypertension: a meta-analysis using the International Database
however, a meta-analysis restricted to studies based on on Ambulatory Blood Pressure Monitoring in relation to Cardiovascular
ambulatory BP updated normal thresholds and confirmed Outcomes Population. Hypertension 2012; 59:564–571.
16. Cardillo C, De Felice F, Campia U, Folli G. Psychophysiological
the differences in left ventricular mass index across the reactivity and cardiac end-organ changes in white coat hypertension.
three groups (data not shown). Finally, as WCH patients Hypertension 1993; 21:836–844.
were classified in all studies according to the normal day- 17. Hoegholm A, Kristensen KS, Bang LE, Nielsen JW, Nielsen WB, Madsen
time or 24-h ambulatory BP levels, without taking in to NH. Left ventricular mass and geometry in patients with established
account the night-time BP, it was possible that individuals hypertension and white coat hypertension. Am J Hypertens 1993;
6:282–286.
with nocturnal hypertension were misclassified as 18. Kuwajiama I, Suzuki Y, Fujisawa A, Kuramoto K. Is white coat hyper-
having WCH. tension innocent? Structure and function of the heart in the elderly.
In conclusion, our meta-analysis shows that echo- Hypertension 1993; 22:826–831.
graphic alterations in cardiac structure and function in 19. Cavallini MC, Roman MJ, Pickering TG, Schwartz JE, Pini R, Devereux
RB. Is white coat hypertension associated with arterial disease or left
WCH patients, as defined by ABPM, are intermediate ventricular hypertrophy? Hypertension 1995; 26:413–419.
between sustained hypertensive patients and normoten- 20. Cuspidi C, Marabini M, Lonati L, Sampieri L, Comerio G, Pelizzoli S,
sive controls; this finding supports the view that WCH et al. Cardiac and carotid structure in patients with established hyper-
should be no longer considered an entirely benign entity. tension and white-coat hypertension. J Hypertens 1995; 13:1707–1711.
In a practical perspective, patients with WCH should be 21. Pierdomenico SD, Lapenna D, Guglielmi MD, Antidormi T, Schiavone
C, Cuccurullo F, Mezzetti A. Target organ status and serum lipids in
carefully evaluated in order to provide appropriate patients with white coat hypertension. Hypertension 1995; 26:801–
prevention strategies. 807.
22. Rizzo V, Cicconetti P, Bianchi A, Lorido A, Morelli S, Vetta F, et al.
White-coat hypertension and cardiac organ damage in elderly subjects.
ACKNOWLEDGEMENTS J Hum Hypertens 1996; 10:293–298.
23. Glen SK, Eliott HL, Curzio JL, Lees KR, Reid JL. White-coat hypertension
Conflicts of interest as a cause of cardiovasacular dysfunction. Lancet 1996; 348:654–
The authors report no conflicts of interest. 657.

Journal of Hypertension www.jhypertension.com 31

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Cuspidi et al.

24. Palatini P, Mormino P, Santonastaso M, Mos L, Dal Follo M, Zanata G, 37. Kotsis V, Stabouli S, Toumanidis S, Papamichael C, Lekakis J, Germa-
et al. Target-organ damage in stage I hypertensive subjects with white nidis G, et al. Target organ damage in ‘white coat hypertension’ and
coat and sustained hypertension. Results from the HARVEST study. ‘masked hypertension’. Am J Hypertens 2008; 21:393–399.
Hypertension 1998; 31:57–63. 38. Ihm SH, Youn HJ, Park CS, Kim HY, Chang K, Seung KB, et al. Target
25. Owens PE, Lyons SP, Rodriguez SA, O’Brien ET. Is elevation of clinic organ status in white-coat hypertensives, uselfulness of serum procol-
blood pressure in patients wth white coat hypertension who have lagen type I propeptide in the respect of left ventricular diastolic
normal ambulatory blood pressure associated with target organ dysfunction. Circ J 2009; 73:100–105.
changes? J Hum Hypertens 1999; 12:743–748. 39. Sung SH, Cheng HM, Wang KL, Yu WC, Chuang SY, Ting CT, et al.
26. Martinez MA, Garcia-Puig J, Martin JC, Guallar-Castillon P, Aguirre de White coat hypertension is more risky than prehypertension: important
Carcer A, Torre A, et al. Frequency and determinants of white-coat role of arterial wave reflections. Hypertension 2013; 61:1346–1353.
hypertension in mild to moderate hypertension. A primary care-based 40. Caliskan M, Ciftci O, Gullu H, Kaliskan Z, Guven A, Erdogan D,
study. Am J Hypertens 1999; 12:251–259. Muderrisoglu H. Effect of masked, white-coat, and sustained hyper-
27. Sega R, Trocino G, Lanzarotti A, Carugo S, Cesana G, Schiavina R, et al. tension on coronary flow reserve and peripheral endothelial functions.
Alterations of cardiac structure in patients with isolated office, ambu- Clin Exp Hypertens 2013; 35:183–191.
latory, or home hypertension. Data from the general population 41. Schillaci G, Verdecchia P, Porcellati C, Cuccurullo O, Cosco C, Perti-
[Pressioni Arteriose Monitorate E Loro Associazioni (PAMELA) Study]. cone F. Continuous relation between left ventricular mass and cardi-
Circulation 2001; 104:1385–1392. ovascular risk in essential hypertension. Hypertension 2000; 35:580–
28. Grandi AM, Broggi R, Colombo S, Santillo R, Imperiale D, Bertolini A, 586.
et al. Left ventricular changes in isolated office hypertension. A blood 42. Cuspidi C, Facchetti R, Bombelli M, Sala C, Grassi G, Mancia G.
pressure-matched comparison with normotension and sustained hy- Different prognostic value of left ventricular mass index and wall
pertension. Arch Intern Med 2001; 161:2677–2681. thickness in predicting cardiovascular prognosis: data from the
29. Bjorklund K, Lind L, Vessby B, Andrèn B, Lithell H. Different metabolic PAMELA population. Am J Hypertens 2014; 8:1079–1086.
predictors of white coat hypertension and sustained hypertension over 43. Ruilope LM, Schmieder RE. Left ventricular hypertrophy and clinical
a 20-year follow-up period: a population-based study of elderly men. outcomes in hypertensive patients. Am J Hypertens 2008; 21:500–506.
Circulation 2002; 106:63–68. 44. Diez J, Frohlich E. A translational approach to hypertensive heart
30. Silveira A, Mesquita A, Maldonado J, Silva JA, Polonia J. White coat disease. Hypertension 2010; 55:1–8.
effect in treated and untreated patients with high office blood pressure. 45. Raman SV. The hypertensive heart. An integrated understanding
Relationship with pulse wave velocity and left ventricular mass index. informed by imaging. J Am Coll Cardiol 2010; 55:91–96.
Rev Port Cardiol 2002; 21:517–530. 46. Frohlich E, Gonzales A, Diez J. Hypertensive left ventricular hyper-
31. Pose-Reino A, Rodriguez-Fernandez M, Lopez-Barreiro L, Coleman IC, trophy risk: beyond adaptive cardiomyocytic hypertrophy. J Hypertens
Estevez-Nunez JC, Mendez-Naya I. Diagnostic criteria of white coat 2011; 29:17–26.
hypertension (WCH): consequences for the implications of WCH for 47. Schirmer H, Lunde P, Rasmussen K. Mitral flow derived Doppler
target organs. Blood Press 2002; 11:144–150. indices of left ventricular diastolic function in a general population.
32. Karter Y, Curgunlu A, Altnisik S, Erturk N, Vehid S, Mihmani I, et al. The Tromso study. Eur Heart J 2000; 21:1376–1386.
Target organ damage and changes in arterial compliance in white coat 48. Bella JN, Palmieri V, Roman MJ, Liu JE, Welty TK, Lee ET, et al. Mitral
hypertension. Is white coat innocent ? Blood Press 2003; 12:307– ratio of peak early to late diastolic filling velocity as a predictor of
313. mortality in middle-aged and elderly adults. The Strong Heart Study.
33. Curgunlu A, Karter Y, Uzun H, Aydin S, Erturk N, Vehid S, et al. Hyper- Circulation 2002; 105:1928–1933.
homocysteinemia: an additional risk factor in white coat hypertension. 49. Douglas PS. The left atrium. A biomarker of chronic diastolic dysfunc-
Int Heart J 2005; 46:245–254. tion and cardiovascular disease risk. J Am Coll Cardiol 2003; 44:1206–
34. Erdogan D, Caliskan M, Gullu H, Yildirim I, Ozer I, Ozgul A, Muderri- 1207.
soglu H. Aortic elastic properties and left ventricular diastolic function 50. Kizer JR, Bella JN, Palmieri V, Liu JE, Best LG, Lee ET, et al. Left atrial
in white-coat hypertensive indviduals. Blood Press Monit 2006; 11:191– diameter as an independent predictor of first cardiovascular events in
198. middle-aged and elderly adults: the Strong Heart Study (SHS). Am
35. Cuspidi C, Meani S, Fusi V, Sala C, Zanchetti A, Mancia G. How reliable Heart J 2006; 151:412–418.
is isolated clinic hypertension defined by a single ambulatory blood 51. Leung DY, Chi C, Allman C, Boyd D, Ng AC, Kadappu KK, et al.
pressure monitoring? J Hypertens 2007; 25:315–320. Prognostic implications of left atrial volume index in patients with sinus
36. Mulè G, Nardi E, Cottone S, Cusimano P, Incalcaterra F, Palermo A, rhythm. Am J Cardiol 2010; 105:1635–1639.
et al. Metabolic syndrome in subjects with white-coat hypertension: 52. Cuspidi C, Facchetti R, Bombelli M, Sala C, Negri F, Grassi G, Mancia G.
impact on left ventricular structure and function. J Hum Hypertens Night-time blood pressure and new onset left ventricular hypertrophy;
2007; 21:854–860. findings from the PAMELA population. Hypertension 2013; 62:78–84.

Reviewers’ Summary Evaluations Reviewer 2

A valuable meta-analysis suggesting that white coat hyper-
Reviewer 1 tension (WCH) is not benign, but an intermediate pheno-
The authors performed a meta-analysis to provide a quan- type between normotension and hypertension. There are
titative estimation of cardiac structure and function in no data for subjects aged <30 years or for WCH detected by
normotensive subjects, patients with white-coat hyperten- home blood pressure (BP) monitoring, both of which might
sion (WCH) and hypertensive patients. Alterations found in give different findings. The significant association of office
patients with WCH were intermediary between those found BP, but not of daytime or 24-h ambulatory BP, with left
in sustained hypertensive patients and normotensive sub- ventricular mass index in WCH and sustained hypertensives
jects. This meta-analysis sheds new light on a controversial is problematic. Finally, it might be argued that studies
and clinically relevant topic, i.e. target organ damage reporting only daytime ambulatory BP might be misleading
associated with WCH. It suggests that WCH is not an because the most important component of the BP profile
innocuous condition, and that patients with WCH may (nocturnal BP) is ignored.
benefit from preventive measures and a strict follow-up.

32 www.jhypertension.com Volume 33  Number 1  January 2015

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.