Correlation between visit-to-visit and short-term blood ... - Nature

Journal of Human Hypertension (2017) 31, 132–137 © 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved 0950-9240/17 www.nature.com/jhh

ORIGINAL ARTICLE

Correlation between visit-to-visit and short-term blood pressure variability calculated using different methods and glomerular filtration rate J Wang1, B Jiang2, L Song3, C Yang4, Y Wu5, S Chen5, C Li5, H Zhao5, F Wang6 and S Wu5 The aim of this study was to explore the correlation between visit-to-visit and short-term blood pressure variability (BPV), including systolic BPV (SBPV) and diastolic BPV (DBPV), calculated using different methods, and the glomerular filtration rate (GFR) in a late, middle-aged population. Using cluster sampling, we randomly selected retired employees of the Kailuan Group who were ⩾ 60 years and participated in a third health examination for 24-h ambulatory blood pressure monitoring and inspection. Among the 3064 randomly selected observation subjects, 2464 were included based on the criteria. BPV was calculated using s.d., coefficient of variation (CV, s.d./Mean), variability independent of mean (VIM, s.d./Meanx) and BPV ratio (BPVR, s.d. (SBPV)/s.d. (DBPV)). Multivariate linear regression was used to analyse the correlation between estimated GFR (eGFR) and BPV calculated using different methods. The mean age of 2464 subjects was 67.4 ± 6.1 years, with 1667 male subjects (67.7%). A total of 2104 cases were included in the visit-to-visit BPV group, and 1382 in the short-term BPV group. SBPV calculated using different methods showed statistically significant increasing trends for the SBP versus all s.d. and short-term BPVR. There was a significant, positive correlation between the visit-to-visit and short-term BPV calculated using different methods, which were all negatively correlated with eGFR (P o0.05). Multivariate linear regression analysis showed that, with correction for possible confounding factors, SBPV (24-h s.d., CV and VIM, and daytime CV and night time CV) and all DBPV demonstrated negative linear relationships with eGFR (P o 0.05). Journal of Human Hypertension (2017) 31, 132–137; doi:10.1038/jhh.2016.51; published online 4 August 2016 INTRODUCTION Blood pressure variability (BPV) refers to the degree of fluctuation in blood pressure over a certain period of time, and can be divided into long-term BPV (between days, weeks, months and years BPV) and short-term BPV, including the s.d.,1 coefficient of variation (CV) variability independent of mean (VIM)2 and BPV ratio (BPVR)3 and so on. BPV is closely associated with renal function injuries,4 and the glomerular filtration rate (GFR) is an important indicator of renal function. Manios et al.5 found that systolic BPV (SBPV) was correlated with estimated GFR (eGFR). However, it is unknown whether visit-to-visit BPV and eGFR are correlated, and if so, whether this correlation is consistent when BPV is calculated using different methods. There are few studies on this topic; therefore, we analysed the correlation between visit-to-visit and short-term BPV calculated using different methods and eGFR in a late, middle-aged population based on the data from the Kailuan study (registration number: ChiCTR-TNC-1100 1489).

MATERIALS AND METHODS Observation subjects From 2006 to 2007, data were collected from the health examinations of current and retired employees of the Kailuan

Group at 11 hospitals responsible for the health care of the Kailuan community. The same group of employees underwent three more examinations conducted by the same medical staff in 2008–2009, 2010–2011 and 2012–2013, at the same locations and with the same time sequence, questionnaire, and anthropometric and biochemical analysis methods as the first health examination. During the third health examination, retired employees of the Kailuan Group (⩾60 years old), who underwent their examinations at the Kailuan General Hospital, Kailuan Linxi Hospital or Kailuan Zhaogezhuang Hospital, were selected as study candidates by cluster sampling, with 25% of the population being randomly selected. The study subjects were physically examined and, with their consent, rescheduled a new appointment for 24-h ambulatory blood pressure monitoring (the third+ health examination). The study was approved by the ethics committees of the Kailuan Group Hospital. Inclusion and exclusion criteria Inclusion criteria. (1) ⩾ 60 years of age and undergoing examinations at the designated hospitals; (2) without severe disabilities; (3) without defects in cognitive ability; and (4) providing informed consent. Exclusion criteria. (1) valvular heart disease or cardiomyopathy; (2) frequent premature beats (all kinds of beats 46 times min− 1);

1 Medical Informatics Center, Peking University, Beijing, People’s Republic of China; 2Department of Orthopedics and Trauma, Peking University People’s Hospital, Beijing, People’s Republic of China; 3North China University of Science and Technology, Tangshan, People’s Republic of China; 4Department of Epidemiology and Biostatistics, Peking University, Beijing, People’s Republic of China; 5Department of Cardiology, Kailuan Hospital, North China University of Science and Technology, Tangshan, People’s Republic of China and 6 Sixth Hospital of Peking University, Beijing, People’s Republic of China. Correspondence: Dr F Wang, Sixth Hospital of Peking University, No. 51, Huayuan Road, Haidian District, Beijing 100191, People’s Republic of China. E-mail: [email protected] or Professor S Wu, Department of Cardiology, Kailuan Hospital, North China University of Science and Technology, Tangshan 063000, People’s Republic of China. E-mail: [email protected] Received 14 March 2016; revised 31 May 2016; accepted 10 June 2016; published online 4 August 2016

Visit-to-visit/short-term blood pressure variability and glomerular filtration rate J Wang et al

(3) atrial fibrillation or atrioventricular block or intraventricular block; and (4) exposure to antipsychotic drugs, anti-Parkinson drugs, antidepressants or sedative drugs within previous 2 weeks.

Data collection The content of the epidemiological survey, anthropometric indicators and biochemical testing methods used in the current study can be found in our previously published study.6 A calibrated mercury sphygmomanometer was used to measure the blood pressure at the right brachial artery following the standard recommended procedures. The blood pressure was measured three times consecutively, with an interval of 1–2 min between each reading, and the mean value was recorded. Smoking was defined as at least one cigarette per day on average for at least 1 year; drinking alcohol was considered when at least 50 ml of alcohol was consumed daily

Table 1.

133 within the past 12 months; and exercise was defined as working out at least three times a week for at least 30 min.

Determination of serum creatinine During the third health examinations, after 8 h of fasting, 5 ml of venous blood was collected from the subjects into a vacuum tube, and then centrifuged at 3000 g for 10 min at room temperature (24 °C), with the upper serum collected for measurement within 4 h. Professional laboratories measured the serum creatinine levels by the colorimetric method using a liquid double dosage from the Creatinine Assay Kit manufactured by Beikong Biotechnology Co. Ltd (Changping, China; production batch number: 150203; batch variation coefficient o 5%). The blood creatinine analysis was conducted on a Hitachi automatic analyser (7600 Automatic Analyser; Hitachi, Tokyo, Japan).

General characteristics of observation subjects

Variable Age (year) Heart rate (beats per min) SBP (mm Hg) BMI (kg m 2) eGFR (ml min − 1 1.73 m −2) FBG (mmol l − 1) TC (mmol l − 1) HDL-C (mmol l − 1) LDL-C (mmol l − 1) lgTG lgCRP Smoking (%) Drinking (%) Exercise (%) Exposure to antihypertensive drugs (%)

Total population (n = 2464)

Male (n = 1667)

Female (n = 797)

P-value

67.4 ± 6.1 72.1 ± 10.7 138.1 ± 20.6 25.4 ± 3.4 83.0 ± 16.3 5.9 ± 1.7 5.3 ± 2.2 1.5 ± 0.5 2.8 ± 0.8 0.1 ± 0.2 0.1 ± 0.5 408 (16.6) 459 (18.6) 553 (22.4) 545 (22.1)

67.8 ± 6.1 71.7 ± 10.7 139.5 ± 20.4 25.4 ± 3.3 82.9 ± 16.9 5.9 ± 1.6 5.1 ± 1.6 1.5 ± 0.4 2.8 ± 0.8 0.1 ± 0.2 0.1 ± 0.5 396 (23.8) 456 (27.4) 398 (23.9) 351 (21.1)

66.7 ± 5.9 72.9 ± 10.6 135.0 ± 20.6 25.3 ± 3.6 83.3 ± 15.2 5.9 ± 1.8 5.7 ± 3.0 1.6 ± 0.5 3.0 ± 0.8 0.2 ± 0.2 0.1 ± 0.5 12 (1.5) 3 (0.4) 155 (19.4) 194 (24.3)

o0.001 0.036 o0.001 0.754 0.678 0.629 o0.001 o0.001 o0.001 o0.001 0.333 o0.001 o0.001 0.014 0.066

Abbreviations: BMI, body mass index; eGFR, glomerular filtration rate; FBG, fasting blood glucose; HDL-C, high-density lipoprotein cholesterol; lghsCRP, logtransformed high-sensitivity C-reactive protein; lgTG, log-transformed triglycerides; LDL-C, low-density lipoprotein cholesterol; SBP, systolic blood pressure; TC, total cholesterol. 1 mmHg = 0.133 kPa.

Figure 1. SBPV in the different SBP groups. BPVR, blood pressure variability ratio; CV, coefficient of variation; SBP, systolic blood pressure; SBPV, systolic blood pressure variability; s.d., standard deviation; VIM, variability independent of mean. © 2017 Macmillan Publishers Limited, part of Springer Nature.

Journal of Human Hypertension (2017) 132 – 137


134 Measurement of ambulatory blood pressure During the third+ health examination, ambulatory blood pressure measurements were carried out using the European CE-, FDA- and SFDA-certified SunTech Oscar 2 Ambulatory Blood Pressure Monitor (SunTech, Morrisville, NC, USA) and the Spacelabs Ultralite Dynamic 90217 Blood Pressure Monitor (Spacelabs Healthcare, Issaquah, WA, USA), with strict inspection of the monitoring device before measurement. All the measurements followed the standard procedures at the correct position. Daytime and night time was defined as 0600 to 1000 hours and 2200 to 0600 hours, respectively.7 Ambulatory blood pressure requires one measurement every 15 min during the daytime, and one measurement every 30 min during the night time.

Relative definitions 1. Visit-to-visit s.d.: The visit-to-visit BP was defined as the mean of three BP measurements at each of the first, second, third, third + and fourth health examinations. 2. The short-term s.d. was represented by the s.d. generated from the 24-h ambulatory blood pressure monitoring, with the s.d. of Table 2. The correlation analysis between visit-to-visit and short-term BPV calculated using different methods and eGFR s.d.

CV

VIM

BPVR

Visit-to-visit (sample 1) SBPV − 0.05* DBPV − 0.06*

− 0.02 − 0.04

− 0.01 − 0.04

0.00

24-h (sample 2) SBPV − 0.11** DBPV − 0.05

− 0.07** − 0.08**

− 0.07** − 0.07*

− 0.10**

Daytime (sample 2) SBPV − 0.11** DBPV − 0.05

− 0.08** − 0.08**

− 0.08** − 0.07*

− 0.10**

Night time (sample 2) SBPV − 0.08** DBPV − 0.06*

− 0.05 − 0.08**

− 0.06** − 0.07**

− 0.04

Abbreviations: BPVR, blood pressure variability ratio; CV, coefficient of variation; DBPV, diastolic blood pressure variability; eGFR, estimated glomerular filtration rate; SBPV, systolic blood pressure variability; s.d., standard deviation; VIM, variability independent of mean. *Po 0.05; **Po0.001.

Table 3.

3. 4. 5. 6.

24-h, daytime and night time BP representing 24-h s.d., daytime s.d. and night time s.d., respectively. CV: CV = 100 × s.d./Mean, where s.d. is the standard deviation of BP and Mean the average BP. s:d: VIM2: VIM ¼ Mean x , where s.d. is the standard deviation of BP, Mean is the average BP and x derived from curve fitting. BPVR3 was defined as (systolic BP variability)/(diastolic BP variability), namely s.d.(SBP)/s.d.(DBP). Calculation of eGFR: eGFR was calculated using the CKD-EPI method.8 Female: when creatinine is ⩽ 62 μmol l − 1, eGFR = 144 × (SCR/0.7)–0.329 × (0.993)age; when creatinine 462 μmol l − 1, eGFR = 144 × (SCR/0.7)–1.209 × (0.993)age. Male: eGFR = 141 × when creatinine is ⩽ 80 μmol l − 1, (SCR/0.9)–0.411 × (0.993)age; when creatinine 480 μmol l − 1, eGFR = 141 × (SCR/0.9)–1.209 × (0.993)age.

Statistical analysis Statistical analyses were performed using SPSS 13.0 (SPSS Inc., Chicago, IL, USA). Normally distributed data (including nonnormally distributed data that was converted to a normal distribution by base-e logarithmic transformation) were presented as x ± s, with independent-samples t-tests used for between-group comparisons. The count data were presented as percentages (%), and were using χ2 tests. The correlations between the visit-to-visit/ short-term BPV calculated using different methods and eGFR were analysed by Pearson’s correlation analyses. The effects of visit-tovisit/short-term BPV calculated using different methods on eGFR were analysed by a multivariate linear regression. P o0.05 was considered statistically significant. RESULTS In total, 12 257 retired employees (⩾60 years old) of the Kailuan Group underwent physical examination at the three abovementioned hospitals (accounting for 52.67% of all the retired Kailuan employees who were ⩾ 60 years and underwent the third health examination). In accordance with the sampling proportion of 25%, 3064 cases were selected, with 2860 cases agreeing to participate in the study (response rate: 93.34%). However, 46 cases failed to participate (actual response rate: 91.84%) and 350 cases met the exclusion criteria. Finally, 2464 cases were included in this current study cohort. Among the 2464 subjects, more than two measurements of the visit-to-visit blood pressure value were missing for 152 cases and eGFR data were missing for 208 cases; finally, 2104 cases were included in sample 1 (visit-to-visit BPV group). A total of 1838

Multivariate linear regression analysis of visit-to-visit BPV calculated using different methods and eGFR

Variables

SBPV

DBPV

BPVR

Unstandardized β-value (95% CI)



Visit-to-visita s.d. CV VIM

0.12 (−0.22 to − 0.02)* − 7.04 (−21.99 to 7.93) − 0.03 (−0.13 to 0.08)

− 0.26 (−0.45 to − 0.07)** − 16.02 (−32.27 to 0.23) − 0.18 (−0.37 to 0.02)

− 0.001 (−0.17 to 0.17)

Visit-to-visitb s.d. CV VIM

0.01 (−0.09 to 0.11) 2.17 (−11.99 to 16.33) 0.02 (−0.09 to 0.12)

− 0.18 (−0.37 to 0.00) − 13.28 (−28.76 to 2.20) − 0.15 (−0.34 to 0.03)

0.002 (−0.20 to 0.20)

Abbreviations: BPVR, blood pressure variability ratio; CI, confidence interval; CV, coefficient of variation; DBPV, diastolic blood pressure variability; eGFR, estimated glomerular filtration rate; SBPV, systolic blood pressure variability; s.d., standard deviation; VIM, variability independent of mean. a Unadjusted model. bAdjustment for age and gender mean SBP/DBP, fasting blood glucose, total cholesterol, antihypertensive drugs, smoking and drinking.


© 2017 Macmillan Publishers Limited, part of Springer Nature.


135 Table 4.

Multivariate linear regression analysis of short-term BPV calculated using different methods and eGFR

Variables

SBPV

DBPV

BPVR




24-h s.d. CV VIM

− 0.46 (−0.69 to − 0.23)** − 43.35 (−76.46 to − 10.23)* − 0.35 (−0.61 to − 0.09)**

− 0.31 (−0.63 to 0.01) − 36.92 (−60.82 to − 13.02)** − 0.43 (−0.76 to − 0.10)*

− 2.96 (−6.26 to 0.34)

24-hb s.d. CV VIM

− 0.25 (−0.50 to − 0.01)* − 33.72 (−65.80 to − 1.64)* − 0.26 (−0.51 to − 0.01)*

− 0.40 (−0.71 to − 0.01)* − 32.37 (−56.05 to − 8.69)** − 0.42 (−0.73 to − 0.10)*

− 2.96 (−6.26 to 0.34)

Daytimea s.d. CV VIM

− 0.43 (−0.65 to − 0.22)** − 46.97 (−78.27 to − 15.67)** − 0.35 (−0.59 to − 0.12)**

− 0.26 (−0.55 to 0.02) − 32.72 (−54.19 to − 11.25)** − 0.36 (−0.65 to − 0.07)*

− 2.52 (−5.29 to 0.26)

Daytimeb s.d. CV VIM

− 0.22 (−0.45 to 0.004) − 30.28 (−60.42 to − 0.13)* − 0.23 (−0.46 to 0.003)

− 0.31 (−0.59 to − 0.03)* − 27.00 (−48.18 to − 5.82)* − 0.33 (−0.61 to − 0.05)*

− 2.52 (−5.29 to 0.26)

Night timea s.d. CV VIM

− 0.31 (−0.50 to − 0.11)** − 24.99 (−50.75 to 0.76) − 0.23 (−0.44 to − 0.02)*

− 0.38 (−0.70 to − 0.07)* − 32.07 (−53.65 to − 10.48)** − 0.43 (−0.75 to − 0.11)**

0.14 (−1.64 to 1.92)

Night timeb s.d. CV VIM

− 0.16 (−0.36 ~ 0.03) − 25.37 (−50.33 to − 0.40)* − 0.20 (−0.40 to 0.004)

− 0.44 (−0.75 to − 0.14)** − 31.91 (−53.50 to − 10.31)** − 0.45 (−0.76 to −0.14)**

0.14 (−1.64 to 1.92)

a

Abbreviations: BPVR, blood pressure variability ratio; CI, confidence interval; CV, coefficient of variation; DBPV, diastolic blood pressure variability; eGFR, estimated glomerular filtration rate; SBPV, systolic blood pressure variability; VIM, variability independent of mean. *Po0.05; **Po 0.01. aUnadjusted model. b Adjustment for age and gender mean SBP/DBP, fasting blood glucose, total cholesterol, antihypertensive drugs, smoking and drinking.

cases completed the ambulatory blood pressure monitoring. Noncompliance was observed in 40 cases for the daytime ambulatory blood pressure data and 220 cases for the night time data, as well as 196 cases for the eGFR data; finally, 1382 cases were included in sample 2 (short-term BPV group). All the samples were used to analyse the impact of SBPV calculated using different methods on eGFR. General characteristics of subjects The mean age of the 2464 cases was 67.4 ± 6.1 years, with 1667 males (67.7%) and 797 females (32.3%). Statistically significant differences were found among the female and male groups in terms of age, heart rate, SBP, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, logtransformed triglycerides, smoking, drinking and physical exercise (P o 0.05) (Table 1). The ambulatory blood pressure parameters (mean 24-h, daytime, night time and BPV indices) were shown in Supplementary 1. SBPV of the different groups The mean visit-to-visit, 24-h, daytime and night time SBP were divided into four categories (o 120, 120–139, 140–159, ⩾ 160 mm Hg). The SBP values of the X axis refererd to the mean visit-to-visit, 24-h, daytime and night time SBP. Analysis of the SBPV calculated using different methods showed statistically significant increasing trends for the SBP versus all s.d. (visit-to-visit, © 2017 Macmillan Publishers Limited, part of Springer Nature.

24-h, daytime and night time) and short-term BPVR (24-h, daytime and night time) (P o 0.05; Figure 1). The correlation analysis between visit-to-visit and short-term BPV calculated using different methods and eGFR A significantly positive correlation was found between the visit-tovisit and short-term SBPV calculated using different methods; the visit-to-visit SD and short-term BPV (except for 24-h and daytime SD (DBPV), night time CV (SBPV) and night time BPVR) calculated using different methods were all negatively correlated with eGFR (P o 0.05) (Table 2). Multivariate linear regression analysis of visit-to-visit BPV calculated using different methods and eGFR The visit-to-visit s.d., CV, VIM and BPVR demonstrated no linear correlations with eGFR (P 40.05) (correction for gender, age, mean SBP/DBP, fasting blood glucose, total cholesterol, antihypertensive drugs, smoking and drinking) (Table 3). Multivariate linear regression analysis of short-term BPV calculated using different methods and eGFR SBPV (24-h s.d., CV and VIM, and daytime and night time CV) and all DBPV demonstrated negative linear correlations with eGFR (P o 0.05) (correction for gender, age, mean SBP, fasting blood glucose, total cholesterol, antihypertensive drugs, smoking and Journal of Human Hypertension (2017) 132 – 137


136 drinking); all BPVR showed no linear correlations with eGFR (Table 4).

temperature and environmental changes and so on, that were not corrected.

DISCUSSION Increased BPV is not only an independent predictor of cardiovascular disease9,10 but also is associated with renal function damage as indicated by eGFR.4 Other studies have found that the higher the daytime eGFR, the lower the BPV.11 The mechanism of elevated BPV leading to decreased eGFR is rather complicated, associated with a number of nervous and humoral regulatory mechanisms, among which the increase of daytime BPV is related with sympathetic excitement,12 leading to renal dysfunction and consequently reduced eGFR. The correlation analysis showed a significantly positive correlation between the visit-to-visit and short-term SBPV calculated using different methods; the visit-to-visit s.d. and short-term BPV (except for 24-h and daytime s.d. (DBPV), night time CV(SBPV) and night time BPVR) calculated using different methods were all negatively correlated with eGFR. The data also showed that the higher the 24-h and daytime SBPV,11 the lower the eGFR. However, 24-h BPV was negatively correlated with eGFR; this could be because of the influence of daytime BPV,13 which may affect the results of the multiple regression analysis of eGFR. When the other confounding factors were corrected in the multivariate linear regression analysis, SBPV (24-h s.d., CV and VIM, and daytime CV and night time CV) and all DBPV demonstrated negative linear correlations with eGFR. Kawai et al.11 found that daytime s.d. and the CV (SBPV) were negatively correlated with eGFR. In the multiple linear regression analysis, when other confounding factors were corrected for, all visit-to-visit BPV calculated using different methods showed no linear correlation with eGFR. Yokota et al.14 found that, after a mean follow-up of 83 months, among nondiabetic patients with chronic kidney disease, neither s.d. nor the CV during the follow-up were correlated with eGFR. Other studies have also found that night time CV11 are not correlated with eGFR. It is thought that the mechanism of elevated 24-h and daytime SBPV resulting in decreased eGFR may function through the following pathways: first of all, increased daytime SBPV leads to increased sympathetic activity,12 followed by renal vasoconstriction, leading to decreased renal blood flow and GFR, which may even result in renal failure.15 Second, increased daytime SBPV leads to changes in renal vascular pathology. In sinoaortic denervated rats,16 although the mean 24-h SBP was normal, SBPV was significantly increased, which led to glomerular sclerosis, glass cell degeneration and thickening of the basement membrane. These changes led to thickening, hardening and lumen stenosis of the kidney arterioles and microvascular walls, causing renal ischaemia, and consequently resulting in reduced eGFR. Although negative linear correlations were found between eGFR and short-term BPV generated using different calculation methods, our study had some limitations: (1) since the current study is cross-sectional, the results only suggest close correlations between them, without identifying a causal relationship; (2) one of the important indicators of renal damage, urinary protein excretion rate, was not included in the current study; (3) the daytime and night time intervals did not refer to the intervals recommended by the ESH Guidelines on the ambulatory blood pressure monitoring (2013),17 as we performed the monitorings between 2010 and 2012; and (4) when evaluating the relationship between BPV and eGFR, although we corrected as many confounding factors as possible, there may still be other confounding factors, such as

CONCLUSION SBPV (24-h s.d., CV and VIM, and daytime CV and night time CV) and all DBPV demonstrated negative, linear correlations with eGFR.


What is known about this topic? ● BPV is closely associated with renal function injuries, and the GFR is an important indicator of renal function. ● SBPV is correlated with eGFR. ● Little information is available on the correlation between visit-tovisit BPV and eGFR, and whether this correlation is consistent when BPV is calculated using different methods. What this study adds? ● There was a significant, positive correlation between the visit-tovisit and short-term BPV calculated using different methods, which were all negatively correlated with eGFR. ● SBPV (24- h s.d., CV and VIM, and daytime CV and night time CV) and all DBPV all demonstrated negative, linear relationships with eGFR.

CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS We thank all participants and staff of the Kailuan study for their important contributions.

REFERENCES 1 Muntner P, Shimbo D, Tonelli M, Reynolds K, Arnett DK, Oparil S. The relationship between annual variability in systolic blood pressure and all-cause mortality in the general population: findings from NHANES III, 1988–1994. Hypertension 2011; 57(2): 160–166. 2 Rothwell PM, Howard SC, Dolan E, O’Brien E, Dobson JE, Dahlöf B et al. Prognostic significance of annual variability, maximum systolic blood pressure, and episodic hypertension. Lancet 2010; 375(9718): 895–905. 3 Gavish B, Ben-Dov IZ, Kark JD, Mekler J, Bursztyn M. The association of a simple blood pressure-independent parameter derived from ambulatory blood pressure variability with short-term mortality. Hypertens Res 2009; 32(6): 488–495. 4 Veerman DP, de Blok K, vnn Montfrans A. Relationship of steady state and ambulatory blood pressure variability to left ventricular mass and urinary albumin excretion in essential hypertension. Am J Hypertens 1996; 9(5): 455–460. 5 Manios E, Tsagalis G, Tsivgoulis G, Barlas G, Koroboki E, Michas F et al. Time rate of blood pressure variation is associated with impaired renal function in hypertensive patients. J Hypertens 2009; 27(11): 2244–2248. 6 Wu SL, Huang ZR, Yang XC, Zhou Y, Wang AX, Chen L et al. Prevalence of ideal cardiovascular health and its relationship with the 4-year cardiovascular events in a northern Chinese industrial city. Circ Cardiovasc Qual Outcomes 2012; 5(4): 487–493. 7 Farhan H, Al M. Comparative study of ambulatory blood pressure monitoring and clinic blood pressure measurement in the risk assessment and management of hypertension. Sultan Qaboos Univ Med J 2010; 10(3): 370–376. 8 Su C, Zhang GX, Wang RF, Yuan L, Zhang JJ, Wang XR et al. Application of the CKD-EPI equation for estimation of glomerular filtration rate in Chinese patients with chronic kidney disease. Chin J Dis Control Prev 2013; 17(7): 621–624. 9 Okada H, Fukui M, Tanaka M, Inada S, Mineoka Y, Nakanishi N et al. Annual variability in systolic blood pressure is correlated with diabetic nephropathy and atherosclerosis in patients with type 2 diabetes. Atherosclerosis 2012; 220(1): 155–159. 10 Hata J, Arima H, Rothwell PM, Woodward M, Zounqas S, Anderson C et al. Effects of annual variability in systolic blood pressure on macrovascular and microvascular complications in patients with type 2 diabetes mellitus: the ADVANCE trial. Circulation 2013; 128(12): 1325–1334.



137 11 Kawai T, Ohishi M, Kamide K, Nakama C, Onishi M, Ito N et al. Differences between daytime and nighttime blood pressure variability regarding systemic atherosclerotic change and renal function. Hypertens Res 2013; 36(3): 232–239. 12 Narkiewicz K, Winnicki M, Schroeder K, Phillips BG, Kato M, Cwalina E et al. Relationship between muscle sympathetic nerve activity and diurnal blood pressure profile. Hypertension 2002; 39(1): 168–172. 13 Kawai T, Ohishi M, Ito N, Onishi M, Takeya Y, Yamamoto K et al. Alteration of vascular function is an important factor in the correlation between annual blood pressure variability and cardiovascular disease. J Hypertens 2013; 31(7): 1387–1395. 14 Yokota K, Fukuda M, Matsui Y, Hoshide S, Shimada K, Kario K. Impact of annual variability of blood pressure on deterioration of renal function

in patients with nondiabetic chronic kidney disease. Hypertens Res 2013; 36(2): 151–157. 15 Grassi G, Quarti-Trevano F, Seravalle G, Arenare F, Volpe M, Furiani S et al. Early sympathetic activation in the initial clinical stages of chronic renal failure. Hypertension 2011; 57(4): 846–851. 16 Su DF, Miao CY. Blood pressure variability and organ damage. Clin Exp Pharmacol Physiol 2001; 28(9): 709–715. 17 Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Böhm M et al. 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). Eur Heart J 2013; 34(28): 2159–2219.

Supplementary Information accompanies this paper on the Journal of Human Hypertension website (http://www.nature.com/jhh)

