Comparing Office, Central, Home and Ambulatory Blood Pressure in Predicting Left Ventricular Mass

Comparing Office, Central, Home and Ambulatory Blood Pressure in Predicting Left Ventricular Mass

Hipertens Riesgo Vasc. 2019;36(1):5---13 www.elsevier.es/hipertension ORIGINAL Comparing Office, Central, Home and Ambulatory Blood Pressure in Pred...

397KB Sizes 0 Downloads 27 Views

Hipertens Riesgo Vasc. 2019;36(1):5---13

www.elsevier.es/hipertension

ORIGINAL

Comparing Office, Central, Home and Ambulatory Blood Pressure in Predicting Left Ventricular Mass Lucas Sebastian Aparicio ∗ , Jessica Barochiner, Veronica A. Peuchot, Diego H. Giunta, Rocío Martínez, Margarita S. Morales, Paula E. Cuffaro, Gabriel D. Waisman Hospital Italiano de Buenos Aires, Buenos Aires, Argentina Received 15 May 2018; accepted 16 September 2018 Available online 19 October 2018

KEYWORDS Hypertension; Ventricular mass; Office blood pressure; Home blood pressure; Central blood pressure



Abstract The blood pressure measurement method that more accurately predicts a left ventricular mass is controversial, and the evidence suggesting superiority of central over brachial measurements is contradictory. The aim of this study was to compare the relationship between the different clinic and out-of-clinic blood pressure measurements methods with left ventricular mass in patients who attended a specialised hypertension centre for a central blood pressure measurement. An analysis was performed on the correlations between left ventricular mass and central and brachial blood pressure measurements made in the clinic, and home, as well as 24-h systolic blood pressure measurements. A linear regression analysis was then performed to assess the independent relationship of each blood pressure measurement with left ventricular mass. The results on 824 treated and 123 untreated patients showed no significant differences between correlations, although home readings tended to have the best correlations. In regression adjusted models, for each 10 mmHg increase in systolic home blood pressure the left ventricular mass increased 10 g/m2 (95% CI; 3.7-27, P=.01, adj R2 0.38), and for 24-h ambulatory systolic blood pressure it increased 2.3 g/m2 (95% CI 0.76-3.9, P<.01, adj R2 0.15) in treated and untreated patients, respectively. The association of systolic blood pressure with left ventricular mass was better explained by home and 24-h ambulatory monitoring than to clinic-based measurements in treated and untreated patients, respectively. In the clinic, however, the central measurement was not superior to brachial blood pressure. © 2018 SEH-LELHA. Published by Elsevier Espa˜ na, S.L.U. All rights reserved.

Corresponding author. E-mail address: [email protected] (L.S. Aparicio).

https://doi.org/10.1016/j.hipert.2018.09.001 1889-1837/© 2018 SEH-LELHA. Published by Elsevier Espa˜ na, S.L.U. All rights reserved.

6

L.S. Aparicio et al.

PALABRAS CLAVE Hipertensión; Masa ventricular; Presión arterial consultorio; Presión arterial domiciliaria; Presión arterial central

Comparación entre presión arterial de consultorio, central, domiciliaria y ambulatoria para predecir masa ventricular izquierda Resumen Existe controversia sobre qué método de medición de presión arterial predice más precisamente la masa ventricular izquierda. La evidencia que sugiere superioridad de las mediciones centrales sobre las braquiales resulta contradictoria. Nuestro objetivo fue comparar la asociación de diferentes formas de medir la presión dentro y fuera del consultorio con masa ventricular izquierda en pacientes que asistieron a un centro especializado en hipertensión a medirse la presión central. Analizamos las correlaciones entre masa ventricular izquierda y presión sistólica a nivel central y braquial en consultorio, en el domicilio y ambulatoria de 24 h. Luego realizamos un análisis de regresión lineal para evaluar la asociación independiente de cada método con la masa ventricular izquierda. Como resultado, en 824 pacientes tratados y 123 no tratados las diferencias entre correlaciones no fueron significativas, aunque las lecturas tomadas fuera del consultorio tuvieron mejores asociaciones. En los modelos ajustados, por cada 10 mmHg de aumento en la presión sistólica domiciliaria la masa ventricular aumentó 10 g/m2 (IC 95%: 3,7-27; p = 0,01; R2 aj : 0,38), y para la presión sistólica ambulatoria de 24 h aumentó 2,3 g/m2 (IC 95%: 0,76-3,9; p < 0,01; R2 aj : 0,15) en pacientes tratados y no tratados, respectivamente. La asociación de la presión arterial sistólica con masa ventricular izquierda fue explicada mejor por el monitoreo domiciliario y ambulatorio de 24 h, más que con las mediciones de consultorio en pacientes tratados y no tratados, respectivamente. En el consultorio, sin embargo, la presión central no fue superior a la braquial. © 2018 SEH-LELHA. Publicado por Elsevier Espa˜ na, S.L.U. Todos los derechos reservados.

Introduction Hypertension is a major cardiovascular risk factor with a long term effect on cardiac left ventricular structure and function.1 The resulting increase in ventricular mass is an important intermediate phenotype in the progression of heart disease, 2 which is associated with adverse outcomes.3 Blood pressure (BP) is such a powerful tool for predicting these cardiac changes that no matter which technique we use or how we measure it, all the available methods such as brachial BP in the office,4 central BP in the office,5,6 selfmeasured BP at home, 7,8 24-h central ambulatory BP, 9,10 or 24-h brachial ambulatory BP, 11 are to some extent predictive of it. However, the question of which BP-measuring method reflects more accurately the hemodynamic stress on the heart and its subsequent changes in mass is still a matter of debate. Some data from clinical studies and a systematic review and meta-analysis suggest that central compared with brachial BP seems to be more strongly associated with most of the investigated indices of preclinical organ damage, including left ventricular mass.12---14 However, recent studies do not support this hypothesis 15---18 and the current evidence lacks homogeneity in terms of central pressure measurement technique, calibration and validation.19 On the other hand, up to this date the standardized home BP measurements have not been included in these analyses. Therefore, our objective was to find how standard available BP-measuring methods (including office central, office brachial, home and 24-h ambulatory measurements) associate with left ventricular mass in adult patients referred for hypertension, and perform a head-to-head comparison

between them in order to establish which method was the most predictive.

Methods In this pragmatic cross-sectional study we included patients 18 years or older, who were referred by their treating physician to the Hypertension Section of the Hospital Italiano de Buenos Aires, Argentina, between June 8th, 2011 and April 14th 2015, in order to assess hypertension control status through a central BP (CBP) assessment and who had an echocardiography performed within a 1-year period. (Fig. 1) The design of the study complied with the Code of Ethics of the World Medical Association (Declaration of Helsinki, 1964 and Declaration of Tokyo, 1975, as revised in 2008). We recorded demographic and laboratory variables available in the electronic medical records (Table 1). Central BP assessment comprised two analyzable parameters: First, the standardized brachial BP (St.Br), which is the BP measurement performed after a 5-minute rest in the right arm immediately prior to the central tonometric BP measurement. For this purpose, we used either the first or, when available, the mean of up to 3 recordings measured with ® a validated OMRON HEM-7130 device. Second, the aortic ® CBP, which was recorded using Sphygmocor (AtCor Medical, Sydney, Australia), through the applanation method applied in the right radial artery, using the built-in transfer function calibrated with the systolic and diastolic BP components. Valid recordings were considered for those with an operator index ≥ 70. When several CBP measurements where available we used the one with highest operator index. We also included, when available, the office blood

Blood pressure measurement and ventricular mass

7

1431 patients with central BP measurement

Excluded: 347 without echocardiogram 40 without LV mass measurement 87 operator index <70 10 no treatment data 1 <18 years old

947 patients with central BP measurement

824 Treated

123 Untreated

32 with all studies 113 with office BP 92 with 24-h Ambulatory BP 48 with home BP

368 with all studies 783 with office BP 612 with 24-h Ambulatory BP 513 with home BP

Figure 1

pressure (OBP), the home blood pressure (HBP), the 24-hour ambulatory blood pressure (ABP) and the left ventricular mass index (LVMI, g/m2 ), which was calculated according to Devereux´s modified formula 20 using the transthoracic method with a Philips iE33 (Phillips Healthcare, Andover, Massachusetts). The OBP was one conventional seated blood pressure registered by the referring physician, using a wall® mounted Welch-Allyn CE0297 sphygmomanometer. The HBP was recorded with a validated Omron HEM-705CP-II. In this case, after receiving appropriate training on its use, the patients returned home with the device and registered duplicate sitting BP readings (one minute apart) in the nondominant arm, during fixed hours in the morning (8-12 a.m), afternoon (14-18 p.m.) and evening (20-24 p.m.), for four ® days. 21 ABP was recorded with a Spacelabs 90207, set to perform measurements every 15 minutes during daytime and 20 minutes during nighttime for 24 hours.22

Statistical Analysis Patients were stratified in treated or untreated for hypertension in order to avoid a drug exposure related bias in cardiac remodelling. Due to non-normality within the variable distributions, we used Spearman´s correlation coefficient to assess the association between left ventricular mass and BP values measured with all available BP methods. The 95% confidence intervals were measured for each correlation coefficient based on Fisher´s z transformation. Thereafter, in order to compare correlations, confidence intervals were analyzed with the asymptotic method proposed by Zou,23 taking into account existing dependency and overlapping, and using the BP method with the highest association as reference method. For each BP method, we performed a linear regression analysis comparing models with left ventricular mass by adjusted R2 . We adjusted by the variables found significant in the bivariate analyses. A significant difference was considered if the p-value from a two-tailed test was

<0.05. Database management and analysis was performed with Stata statistical software. Release 15. College Station, TX: StataCorp LLC. The study protocol was approved by a local ethics committee and all patients who accepted to participate gave informed consent.

Results Of all 1431 patients with central BP measurements, 947 were finally analyzed after excluding those lacking an echocardiogram or left ventricular measurement, those with a low operator index (<70) or missing treatment data.

Untreated Patients In 123 untreated patients, the best correlation was found between home systolic blood pressure and left ventricular mass, followed by systolic 24-h ABP (Table 2). Thereafter, with the exception of office systolic blood pressure (Coefficient difference 0.44 [95%CI 0.04-0.75]), we did not find any differences in how the BP-measuring methods associate with left ventricular mass after comparing correlation coefficient´s with Zou´s method using home blood pressure as reference. (Table 3). On the other hand, after performing a linear regression analysis between systolic blood pressure and left ventricular mass, the model which better explained the association between the different BP-measuring methods and left ventricular mass was the one that included home blood pressure as the explanatory variable (Table 2). In the unadjusted model, for each 10 mmHg increase in home systolic blood pressure the left ventricular mass significantly increased 10 g/m2 (95%CI 3.3-17, p<0.01, adj R2 0.20). After adjusting for age, sex and glycemia this association significantly persisted (10 g/m2 [95% CI 3.7-17, p<0.01, adj R2 0.38]). 24 h ambulatory and Standardized Brachial systolic blood pressures were also significant in both the unadjusted and the multi-adjusted models, but with lower adjusted R2 values.

8

L.S. Aparicio et al. Table 1

Baseline Characteristics.

Variable

Untreated (n = 123)

Treated (n = 824)

Demographic Age,y Women, n (%) Height, cm Weight, kg BMI, kg/m2 Current smokers, n/total (%)

54 (35-67) 64 (52) 167 ± 11.6 72 (62-85) 25.65 (23.1-28.1) 15/102 (15)

68 (58-76) 510 (62) 163 ± 10.3 74 (63-87) 27.94 (25.1-30.8) 91/707 (13)

Laboratory Glucose (mg/dL) Total cholesterol (mg/dL) LDL cholesterol (mg/dL) HDL cholesterol (mg/dL) Creatinine (mg/dL)

95 (89-103) 194 (164-218) 118.5 (88.5-140) 50 (43-60) 0.84 (0.73-1.01)

98 (91-108) 189 (165-215) 110.5 (88-134) 53 (45-63) 0.88 (0.75-1.04)

Systolic Diastolic Heart rate Systolic Diastolic Heart rate Systolic Diastolic Systolic Diastolic Systolic Diastolic

129 (118.5-134.5) 76 (71-81.5) 71.5 (65-79) 124 (118.5-132.5) 75 (68-78.5) 70.5 (64.5-77) 132 (122-142) 80 (70-90) 118 (110-130) 80 (75-88) 132 (123-144) 80 (73-87)

128 (120-136) 75 (67.5-81) 70 (63-76) 130 (122-140) 73 (67-79) 68 (62-74) 136 (124-150) 80 (70-84) 127 (117-140) 79 (71-86.5) 139 (128-151) 78 (72-85)

All Men Women

88.9 (73.4-105.1) 95.0 (81.0-113.0) 84.0 (70.0-97.0) 55 (55-62.8) 0.95 (0.87-1.06) 1.01 (0.94-1.14) 3.80 (3.41-4.08) 3.14 (2.90-3.37)

100.4 (85.5-118.7) 104.0 (90.0-122.0) 98.0 (82.0-116.0) 55 (55-62.3) 1.05 (0.93-1.15) 1.11 (1.00-1.22) 4.1 (3.72-4.47) 3.23 (2.97-3.50)

Blood pressure ABP

Home

Office Central St.Br.¶ Echocardiographic Left ventricular mass, g/m2

Ejection fraction, % Posterior wall thickness, cm Septum, cm Left atrium, cm Aortic root, cm

Values are represented as n (%), mean ± SD, or median (25th percentile---75th percentile). All blood pressure values are expressed in mmHg. BMI = body mass index. LDL = low-density lipoprotein. HDL = high-density lipoprotein. ABP = 24-h ambulatory blood pressure. ¶ Standardized brachial blood pressure

Neither office nor central systolic blood pressures showed a significant association. Table 4 shows the bivariate associations with left ventricular mass.

Treated Patients In 824 treated patients, all the BP-measuring methods showed significant correlations with left ventricular mass, and the highest association was found for HBP (Table 2). However, we did not find any differences in how the BPmeasuring methods associate with left ventricular mass after comparing correlation coefficient´s with Zou´s method using HBP as reference (Table 3). The model which better explained the association between the different BP-measuring methods and left ventricular mass was with the one that included systolic 24-h

ABP as the explanatory variable. In the unadjusted model, for each 10 mmHg increase in systolic 24-h ABP the left ventricular mass increased 3.5 g/m2 (95% CI 1.8-5.3, p<0.01, 2 adj R 0.02), and in the model adjusted by age, sex, diuretics, Angiotensin II receptor blockers, Calcium Channel Blockers and alpha-blockers it increased 2.3 g/m2 (95% CI 0.76-3.9, p<0.01, adj R2 0.15). The rest of the methods, with the exception of office systolic BP in the fully adjusted model (p = 0.056), were also significantly associated in both the unadjusted and the multi-adjusted models (Table 2).

Discussion The results of our current study may be summarized as follows: (1) In untreated patients, only out-of-office such as home and 24-h ambulatory systolic blood pressure were

Association between systolic blood pressure and left ventricular mass in treated and untreated patients.

BP-measuring Method. (n)a

Correlation Spearman’s rho (95%CI)

Adjusted linear regressionc

Simple Linear Regression p

Coefficientb

p

Adj-R2

Coefficientb

p

Adj-R2

Untreated patients (n = 123) ABP (n = 92) Home (n = 32) Office (n = 86) Central (n = 92) St.Br.d (n = 92)

0.32 0.46 0.02 0.11 0.19

(0.12---0.49) (0.13---0.69) (−0.18---0.23) (−0.09---0.31) (−0.02---0.38)

<0.01 <0.01 0.81 0.29 0.07

4.8 (1---8.6) 10 (3.3---17) 4 (−2.8---3.7) 3 (−0.2---0.62) 3.7(0.6---0.68)

0.01 <0.01 0.78 0.06 0.02

0.05 0.20 0 0.03 0.05

5.9(2.3---9.6) 10 (3.7---17) 1.5(−0.7---0.38) 1 (−0.17---3.8) 2.6 (0.2---5)

<0.01 <0.01 0.18 0.45 0.03

0.11 0.38 0.04 0.04 0.07

Treated patients (n = 824) ABP (n = 632) Home (n = 377) Office (n = 601) Central (n = 632) St.Br.d (n = 632)

0.17 0.26 0.13 0.18 0.18

(0.09---0.24) (0.13---0.39) (0.05---0.21) (0.11---0.26) (0.11---0.26)

<0.01 <0.01 <0.01 <0.01 <0.01

3.5 (1.8---5.3) 46 (10---83) 1.7(0.6---2.9) 2.6 (1.5---3.7) 2.8 (1.7---3.9)

<0.01 0.01 <0.01 <0.01 <0.01

0.02 0.01 0.01 0.03 0.04

2.3(0.76---3.9) 50 (10---89) 1 (−0.002---0.21) 1.2(0.3---2.1) 1.3 (0.4---2.2)

<0.01 0.01 0.056 <0.01 <0.01

0.15 0.01 0.14 0.15 0.15

a b c d

Blood pressure measurement and ventricular mass

Table 2

n varies because multiple imputation for missing data in the adjusted model was not used and incomplete cases were dropped. Linear regression coefficient for each 10 mmHg increase in SBP (95%CI). Adjusted by age, sex and glycemia in untreated patients, and by age, sex, diuretics, Angiotensin II receptor blockers, Calcium Channel Blockers and alpha-blockers in treated patients. Standardized brachial blood pressure. ABP = 24-h ambulatory blood pressure.

9

10

L.S. Aparicio et al.

Table 3

Comparison of dependent correlations with overlapping.a

Untreated patients

Home correlation (95%CI)

Correlation 2 (95%CI)

Difference (95%CI)

Home Home Home Home

0.46 (0.13---0.69)

0.32 0.02 0.11 0.19

0.14 0.44 0.35 0.27

vs. vs. vs. vs.

ABP office central St.Br.b

(0.12---0.49) (−0.18---0.23) (−0.09---0.31) (−0.02---0.38)

(−0.19---0.40) (0.04---0.75) (−0.04---0.66) (−0.10---0.57)

Treated patients

Home correlation (95%CI)

Correlation 2 (95%CI)

Difference (95%CI)

Home Home Home Home

0.26 (0.13---0.39)

0.17 0.13 0.18 0.18

0.09 0.13 0.08 0.08

vs. vs. vs. vs.

ABP office central St.Br.b

(0.09---0.24) (0.05---0.21) (0.11---0.26) (0.11---0.26)

(−0.04---0.23) (−0.01---0.27) (−0.05---0.21) (−0.05---0.21)

a

With Zou’s modified asymptotic method.23 Except for Home versus Office BP in untreated patients, there were no differences in the correlations of all BP-measuring methods and left ventricular mass in patients with or without treatment. b Standardized brachial blood pressure. ABP = 24-h systolic ambulatory blood pressure.

Table 4

Bivariate associations with left ventricular mass. Coefficient (95%CI)

p-value

Coefficient (95%CI)

Untreated patients (N = 123) Age Sex Body mass index Smoking status (n = 74) Tot. cholesterol (n = 81) Glycemia (n = 82) Diuretics Beta-blockers ACEIs ARBs CCBs ABs Other drugs for HT

0.25 (0.005---0.49) −10.25 (−19.43---1.06) −0.21 (−1.18---0.76) 8.05 (−6---22.10) −0.15 (−0.23---0.02) 0.52 (0.11---0.93) ---------------

0.04 0.02 0.66 0.25 0.11 0.01 ---------------

p-value

Treated patients (N = 824) 0.47 −7.16 0.29 −1.8 −0.06 0.07 5.5 3.58 −1.30 7.32 7.38 17.59 14.34

(0.35---0.60) (−11.38---2.93) (−0.07---0.65) (−7.98---4.36) (−0.11---0.006) (−0.02---0.17) (1.15---9.99) (−0.54---7.72) (−5.7---3.00) (3.23---11.41) (3.18---11.58) (3.90---31.29) (3.35---25.34)

<0.01 <0.01 0.12 0.56 0.02 0.14 0.01 0.08 0.56 <0.01 <0.01 0.01 0.01

ACEIs = angiotensin-converting-enzyme inhibitors; ARBs = angiotensin II receptor blockers; CCBs = calcium channel blockers; ABs = alphablockers; HT = hypertension.

associated with left ventricular mass, whereas in treated patients all of the blood pressure methods were associated; (2) The comparison of correlation coefficients, however, did not show significant differences in how the BP-measuring methods associate with left ventricular mass, with the exception of office versus home systolic blood pressure in the untreated group; (3) In regression analysis, the models which better associated with left ventricular mass were home and 24-h ambulatory systolic blood pressure, for the untreated and treated patients, respectively; (4) Central office systolic blood pressure was not associated in untreated patients, and in treated patients it performed lower than the 24-h systolic ambulatory blood pressure measurements; (5) the standardized brachial systolic blood pressure was better associated with left ventricular mass than the actual central systolic blood pressure recording and the conventional office blood pressure measured by the treating physician. Our results regarding the superiority of out-of-office blood pressure are not surprising and confirm what is already

accepted. Namely, that although clinical guidelines still advocate for the use of office blood pressure as the main aid for treatment management in hypertensive patients,24 there is enough evidence of the added value of home and 24-hour ambulatory blood pressure monitoring as a better predictor of target organ damage and cardiovascular prognosis. 8,25---28 On the other hand, our results suggesting that office central systolic blood pressure is a weaker stimulus to the left ventricle than brachial systolic pressure are challenging in the midst of conflicting results from other cross-sectional studies. For instance, favouring central pressure are the results from the Strong Heart Study,5 where in 2585 middleaged American Indians (32% hypertensives, 60% medicated) the central systolic blood pressure was somewhat more closely related to left ventricular mass than office blood pressure. Also, a meta-analysis published in 2016 by Kollias et al.12 showed that central compared with brachial systolic blood pressure was more closely associated with left ventricular mass index (12 studies, n = 6431; weighted age 49.9 years, 51% hypertensives). Like in our study,

Blood pressure measurement and ventricular mass central blood pressure was also measured with applanation tonometry, although in diverse sites (radial n = 8; carotid n = 4). The pooled correlation coefficient for central was r = 0.30 (95%CI, 0.23---0.37); and for brachial systolic BP was r = 0.26 (95% CI, 0.19---0.33; P<0.01 for difference, z-statistic). This review was somehow limited by the heterogeneity of the central blood pressure estimations, the analysis of summary statistics rather than raw data, and the difference between the population characteristics, but it was a large sample. Regarding prospective studies, they have shown contradictory messages as well, with some favouring central blood pressure 14,29,30 and a meta-analysis showing no superiority of central over brachial. 31 With respect to ambulatory central blood pressure, as measured with an oscillometric cuff, Weber et al.9 found a strong trend toward a closer association with left ventricular mass and hypertrophy for the central recordings. Instead, other cross-sectional studies with small samples agree with our results. A Finnish study15 with 246 participants failed to find a significant difference in diagnostic accuracy for left ventricular damage between brachial and central measurements performed with an oscillometric central device in the office, and de la Sierra et al., using 24-h ambulatory central blood pressure could not find a better association than with peripheral blood pressure. 16 To our knowledge, only one study 17 compared different BP-measuring methods to echocardiographically measured left ventricular mass in a similar fashion as ours. The sample included a Spanish middle-aged population of 392 never treated hypertensives (mean age 49 ± 12 years) and they included office blood pressure, 24-h brachial ambulatory blood pressure, and central office blood pressure with applanation tonometry, but not self-measured monitoring at home. In this case they found that left ventricular mass index was more closely related to 24-h ambulatory systolic brachial blood pressure than to office or central blood pressure. An increase of 10 mmHg in mean systolic 24-h brachial blood pressure corresponded to an increase of 5.3 g/m2 (95%CI, 3.5-7.1,p<0.01) in left ventricular mass. These findings are in accordance with our results in the untreated group, but the difference lies in the fact that we included also home blood pressure in the analysis, which in our case performed better than 24-h systolic ambulatory blood pressure in this subgroup of untreated subjects. Another interesting finding in our study was the difference between the standardized BP method taken in the office before the Sphygmocor reading and both the conventional office and the actual central BP reading. These are the baseline brachial BP values which are used to calibrate the central BP device. In our study, the peripheral reading performed better than the central readings both in treated and untreated patients. However, interpreting these results may be challenging. During the time-lapse between the peripheral and central reading, beat-to-beat BP variability may result on discrepancies, and also an alarm reaction during this time frame cannot be ruled out. All in all, the brachial and central readings do not account for the exact pressure in a precise point in time with this method, so that a bias may be introduced. With respect to the standardized versus the conventional blood pressure method, we found that the former performed

11 better, being significantly predictive of left ventricular mass in the regression analysis, whereas office was not, both in treated and in untreated patients. Moreover, taking home blood pressure as reference method, conventional office blood pressure was the only significantly different method in the comparison of correlations. This poor association of the blood pressure measured by treating physicians when the patients attend an ambulatory consultation is illustrative of the heterogeneous and biased nature of these kind of measurements, many times performed recklessly with non-calibrated devices, rounding of numbers, using inadequate cuffs, etc., and emphasizes the need for standardized methods which are less operator-dependent such as the automated office blood pressure.32 The strong points of our current study are the inclusion of home blood pressure monitoring to the analysis, a method which has been elusive in the studies performed up to this date and also the analysis of the standardized office blood pressure. On the other hand, the results of this study should be interpreted within the confines of its limitations: First, this was a pragmatic cross-sectional study, less effective than a prospective study for assessing clinical value whereas a causal relationship with left ventricular mass cannot be proved. Second, the sample was a selected group of middle-class individuals of European descent living in Argentina, who may not be representative of other populations elsewhere. Third, we cannot guarantee the accuracy of the conventional non-standardized office blood pressure measurements because they were obtained from medical records as reported from their respective physicians or nurses, being a reflection of real-life, daily practice (i.e. different levels of technique, different devices, no calibration control). Fourth, we measured central blood pressure under static -not ambulatory- conditions, which prevented a better comparison between all ambulatory blood pressure methods. Fifth, the sample size was relatively small in the untreated group. Sixth, some of the variables included in the multivariable analysis may be a source for reverse causality bias which weight in favour of not seeing differences. However, when the associations still persist in multivariable analysis in these conditions, the association conclusions are strengthened. Seventh, the 24-h systolic BP taken as whole without discriminating day or night-time values, might confound the interpretation given the well known superiority of night-time ambulatory BP as a predictor of target organ damage. Finally, this intended head-to-head comparison between blood pressure methods did not include 24-h central blood pressure recordings. It would have been very interesting to see how the central 24-h blood pressure monitoring compared with the rest of the out-of-office methods available in this study. In conclusion, our findings suggest that the best performance in predicting left ventricular mass was with brachial out-of-office blood pressure measurements (home in untreated and 24-h systolic ambulatory in treated patients, respectively). Office-based central blood pressure might not be more informative than the corresponding brachial measurements. More information is needed in order to determine if daily practice brachial blood pressure measurements should be replaced or potentiated by an added central blood pressure assessment.

12

Conflicts of interest The authors have no conflicts of interest to declare.

Acknowledgements We gratefully acknowledge Ms. Silvia Yeha, Ms. Ayelen Margarella and Ms. Cynthia Medina for clerical assistance.

References 1. Stamler J. Blood pressure, systolic and diastolic, and cardiovascular risks. US population data. Arch Intern Med. 1993;153:598---615. 2. Drazner MH. The Progression of Hypertensive Heart Disease. Circulation. 2011;123:327---34. 3. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med. 1990;322:1561---6. 4. Devereux RB, Pickering TG, Harshfield GA, Kleinert HD, Denby L, Clark L, et al. Left ventricular hypertrophy in patients with hypertension: importance of blood pressure response to regularly recurring stress. Circulation. 1983;68:470---6. 5. Roman MJ, Okin PM, Kizer JR, Lee ET, Howard BV, Devereux RB. Relations of central and brachial blood pressure to left ventricular hypertrophy and geometry: the Strong Heart Study. J Hypertens. 2010;28:384---8. 6. Roman MJ, Devereux RB. Association of central and peripheral blood pressures with intermediate cardiovascular phenotypes. Hypertension. 2014;63:1148---53. 7. Sivén SSE, Niiranen TJ, Langén VLJ, Puukka PJ, Kantola IM, Jula AM. Home versus office blood pressure: longitudinal relations with left ventricular hypertrophy: the Finn-Home study. J Hypertens. 2017;35:266---71. 8. Bliziotis IA, Destounis A, Stergiou GS. Home versus ambulatory and office blood pressure in predicting target organ damage in hypertension: a systematic review and meta-analysis. J Hypertens. 2012;30:1289---99. 9. Weber T, Wassertheurer S, Schmidt-Trucksäss A, Rodilla E, Ablasser C, Jankowski P, et al. Relationship Between 24-Hour Ambulatory Central Systolic Blood Pressure and Left Ventricular Mass: A Prospective Multicenter Study. Hypertension. 2017;70:1157---64. 10. Protogerou AD, Argyris AA, Papaioannou TG, Kollias GE, Konstantonis GD, Nasothimiou E, et al. Left-ventricular hypertrophy is associated better with 24-h aortic pressure than 24-h brachial pressure in hypertensive patients: the SAFAR study. J Hypertens. 2014;32:1805---14. 11. Verdecchia P, Schillaci G, Boldrini F, Guerrieri M, Gatteschi C, Benemio G, et al. Risk stratification of left ventricular hypertrophy in systemic hypertension using noninvasive ambulatory blood pressure monitoring. Am J Cardiol. 1990;66:583---90. 12. Kollias A, Lagou S, Zeniodi ME, Boubouchairopoulou N, Stergiou GS. Association of Central Versus Brachial Blood Pressure With Target-Organ Damage: Systematic Review and Meta-Analysis. Hypertension. 2016;67:183---90. 13. Chi C, Yu X, Auckle R, Lu Y, Fan X, Yu S, et al. Hypertensive target organ damage is better associated with central than brachial blood pressure: The Northern Shanghai Study. J Clin Hypertens. 2017;19:1269---75. 14. Pini R, Chiara Cavallini M, Palmieri V, Marchionni N, Di Bari M, Devereux RB, et al. Central But Not Brachial Blood Pressure Predicts Cardiovascular Events in an Unselected Geriatric Population. J Am Coll Cardiol. 2008;51:2432---9.

L.S. Aparicio et al. 15. Lindroos AS, Langén VL, Kantola I, Salomaa V, Juhanoja EP, Sivén SS, et al. Relation of blood pressure and organ damage: comparison between feasible, noninvasive central hemodynamic measures and conventional brachial measures. J Hypertens.(epub ahead of print 20 February 2018; doi:10.1097/HJH. 0000000000001688). 16. de la Sierra A, Pareja J, Fernández-Llama P, Armario P, Yun S, Acosta E, et al. Twenty-four-hour central blood pressure is not better associated with hypertensive target organ damage than 24-h peripheral blood pressure. J Hypertens. 2017;35:2000---5. 17. Pérez-Lahiguera FJ, Rodilla E, Costa JA, Gonzalez C, Martín J, Pascual JM. Relationship of central and peripheral blood pressure to left ventricular mass in hypertensive patients. Rev Esp Cardiol. 2012;65:1094---100. 18. Yang W-Y, Mujaj B, Efremov L, Zhang Z-Y, Thijs L, Wei F-F, et al. ECG Voltage in Relation to Peripheral and Central Ambulatory Blood Pressure. Am J Hypertens. 2018;31:178---87. 19. Sharman JE, Avolio AP, Baulmann J, Benetos A, Blacher J, Blizzard CL, et al. Validation of non-invasive central blood pressure devices: ARTERY Society task force consensus statement on protocol standardization. Eur Heart J. 2017;38:2805---12. 20. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al. American Society of Echocardiography’s Nomenclature and Standards Committee, Task Force on Chamber Quantification, American College of Cardiology Echocardiography Committee, American Heart Association, European Association of Echocardiography. European Society of Cardiology. Recommendations for chamber quantification. Eur J Echocardiogr. 2006;7:79---108. 21. Barochiner J, Cuffaro PE, Aparicio LS, Elizondo CM, Giunta DH, Rada MA, et al. [Reproducibility and reliability of a 4-day HBPM protocol with and without first day measurements]. Rev Fac Cien Med Univ Nac Cordoba. 2011;68:149---53. 22. O’Brien E, Asmar R, Beilin L, Imai Y, Mallion J-M, Mancia G, et al. European Society of Hypertension recommendations for conventional, ambulatory and home blood pressure measurement. J Hypertens. 2003;21:821---48. 23. Zou GY. Toward using confidence intervals to compare correlations. Psychol Methods. 2007;12:399---413. 24. Mancia G, Fagard R, Narkiewicz K, Redón J, Zanchetti A, Böhm M, et al. Task Force Members. 2013 ESH/ESC Guidelines for the management of arterial hypertension: the Task Force for the management of arterial hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). J Hypertens. 2013;31:1281---357. 25. Asayama K, Thijs L, Brguljan-Hitij J, Niiranen TJ, Hozawa A, Boggia J, et al. International Database of Home Blood Pressure in Relation to Cardiovascular Outcome (IDHOCO) investigators. Risk stratification by self-measured home blood pressure across categories of conventional blood pressure: a participant-level meta-analysis. PLoS Med. 2014;11:e1001591. 26. Imai Y, Hosaka M, Elnagar N, Satoh M. Clinical significance of home blood pressure measurements for the prevention and management of high blood pressure. Clin Exp Pharmacol Physiol. 2014;41:37---45. 27. Li Y, Wei F-F, Thijs L, Boggia J, Asayama K, Hansen TW, et al. International Database on Ambulatory blood pressure in relation to Cardiovascular Outcomes (IDACO) Investigators. Ambulatory hypertension subtypes and 24-hour systolic and diastolic blood pressure as distinct outcome predictors in 8341 untreated people recruited from 12 populations. Circulation. 2014;130:466---74. 28. Sega R, Facchetti R, Bombelli M, Cesana G, Corrao G, Grassi G, et al. Prognostic value of ambulatory and home blood pressures compared with office blood pressure in the general population: follow-up results from the Pressioni Arteriose Monitorate e Loro Associazioni (PAMELA) study. Circulation. 2005;111: 1777---83.

Blood pressure measurement and ventricular mass 29. Wang K-L, Cheng H-M, Chuang S-Y, Spurgeon HA, Ting C-T, Lakatta EG, et al. Central or peripheral systolic or pulse pressure: which best relates to target organs and future mortality? J Hypertens. 2009;27:461---7. 30. Williams B, Lacy PS, Thom SM, Cruickshank K, Stanton A, Collier D, et al. CAFE Investigators, Anglo-Scandinavian Cardiac Outcomes Trial Investigators. CAFE Steering Committee and Writing Committee. Differential impact of blood pressurelowering drugs on central aortic pressure and clinical outcomes:

13 principal results of the Conduit Artery Function Evaluation (CAFE) study. Circulation. 2006;113:1213---25. 31. Vlachopoulos C, Aznaouridis K, O’Rourke MF, Safar ME, Baou K, Stefanadis C. Prediction of cardiovascular events and all-cause mortality with central haemodynamics: a systematic review and meta-analysis. Eur Heart J. 2010;31: 1865---71. 32. Myers MG. The great myth of office blood pressure measurement. J Hypertens. 2012;30:1894---8.