Early morning—Best time window of hourly 24‐hour ambulatory blood pressure in relation to hypertensive organ damage: The Japan Morning Surge‐Home Blood Pressure study

Abstract

The correlations between organ damage and hourly ambulatory blood pressure (BP) have not been established. The patients were 1464 participants of the Japan Morning Surge‐Home Blood Pressure (J‐HOP) study participants who underwent ambulatory BP monitoring. The hourly systolic BP (SBP) at x o'clock was defined as the average of SBP values measured at times x − 30 minutes, x, and x + 30 minutes. The mean age was 64.8 ± 11.6 years. The percentage of male participants was 47.8%. The left ventricular mass index (LVMI) was significantly associated with SBP at 6 o'clock (r = 0.166, P < 0.001). The carotid intima‐media thickness was significantly associated with SBP at 5 o'clock (r = 0.196, P < 0.001). After adjustment for age, sex, smoking, hyperlipidemia, diabetes mellitus, antihypertensive drug use, clinic SBP, and 24‐hour ambulatory SBP, the correlations of the LVMI and hourly SBP at 6 o'clock remained significant (beta coefficient = 0.125, P < 0.01). In conclusion, morning ambulatory systolic BP especially at 5 and 6 o'clock was independently associated with organ damage.

1. INTRODUCTION

The prevalence of hypertension has increased substantially worldwide and has led to high numbers of hypertension‐related deaths, such as death from both ischemic and hemorrhagic stroke or ischemic heart disease.1 Asymptomatic hypertensive organ damages such as left ventricular hypertrophy (LVH), increased carotid intima‐media thickness (IMT), increased pulse wave velocity (PWV), and microalbuminuria are independent predictors of cardiovascular mortality, and thorough investigations of these organ damages are recommended for better cardiovascular risk estimations.2 The use of ABPM in hypertension management is endorsed by many practice guidelines.2, 3, 4 In routine clinical practice, the most frequently used parameters of ABPM data are average daytime, nighttime, and 24‐hour blood pressure (BP) values and the BP dipping status,2 all of which have been demonstrated to be better predictors of cardiovascular outcomes than clinic BP.5, 6, 7 However, there were only a few studies that examined the importance of specific time points of ABPM values. Morning BP (defined as mean BP of the 2 hours after waking up8 and as the mean BP obtained between 05:00 am and 07:59 am 9) was the strongest independent predictor of stroke8 and was associated with asymptomatic intracranial arterial stenosis.9 Furthermore, ABPM can assess BP variability including morning BP surge which is a well‐known index of BP variability that was demonstrated to be associated with organ damage and to be an independent predictor of cardiovascular diseases.10, 11, 12, 13, 14 The importance of the hourly BP surge (defined as the difference between hourly BP values and the lowest BP values) has never been studied.

In this study, we examined the correlation between hourly systolic BP values and hypertensive organ damage, including the left ventricular mass index (LVMI), brachial‐ankle PWV (baPWV), N‐terminal pro‐B‐type natriuretic peptide (NT‐proBNP), the urine albumin‐to‐creatinine ratio (UACR), and IMT. We also evaluated the association of organ damage and hourly SBP surge.

2. METHODS

2.1. Study population

The 1464 participants who enrolled in the Japan Morning Surge‐Home Blood Pressure (J‐HOP) study and underwent ABPM were included in the present analyses. The J‐HOP study was a prospective observational study (University Hospital Medical Information Network Clinical Trials Registry, UMIN000000894) of 4310 patients with at least one risk factor for CVD or a history of CVD or both who were enrolled between 2005 and 2012 by 75 physicians at 71 institutions (45 primary practices, 22 hospital‐based outpatient clinics, and four university hospitals). Details of the J‐HOP study design and methods have been published.15 Written informed consent for participation in the J‐HOP study was obtained from each of the patients, and the J‐HOP study was approved by the Institutional Review Board of Jichi Medical University.

2.2. Blood pressure measurements

The clinic BP was measured by a physician or nurse using the HEM‐5001 (Omron Healthcare, Kyoto, Japan), an upper arm cuff oscillometric BP device, and was taken three times at 15‐second intervals during two clinic visits; the mean values were used as the patient's clinic BP.

ABPM was carried out on a weekday and outpatient setting. Ambulatory BP was monitored every 30 minutes for 24 hours using a validated automatic upper arm device (TM‐2421 or TM‐2425; A&D, Tokyo, Japan), according to the Guidelines for the Clinical Use of 24 hour ABPM.16

The moving average hourly systolic BP at x o'clock was defined as the average of the systolic BP values measured at the three times: x − 30 minutes, x, and x + 30 minutes.

The moving lowest nighttime systolic BP (ie, the trough systolic BP) was defined as the lowest 1‐hour moving average of consecutive systolic BP values obtained from the time point at which the patient went to bed until the time point at which he or she got out of bed in the morning. The hourly systolic BP surge was defined as the hourly systolic BP minus the moving lowest nighttime systolic BP.

2.3. Organ damage assessment

Echocardiography, ultrasonic measurements of the carotid artery, and measurement of the baPWV were performed at each participating institution. Two‐dimensional M‐mode or B‐mode echocardiography was performed. The left ventricular mass (LVM) was derived from the American Society of Echocardiography validated formula: LVM  =  0.8(1.04([LV internal dimension in diastole (LVIDD)  +  diastolic posterior wall diameter +  diastolic interventricular septal diameter]3 − [LVIDD]3))  +  0.6 gram. The LVM index (LVMI) was defined as the product of LVM divided by body surface area. The carotid IMT was assessed by B‐mode ultrasonography and was measured at three points proximal to the bilateral carotid bulb at end diastole and in a plaque‐free segment. The baPWV was measured by volume plethysmographic method (Form PWV/ABI; Omron Healthcare). The mean of the right and left IMT values and the mean of the right and left baPWV values were used for the analyses. Blood and spot urine samples were collected once in the morning after a fasting state, and all assays were performed at a single commercial laboratory center. The details of organ damage assessment in the J‐HOP study have been described.15

Hypertension was defined as a clinic systolic BP of ≥140 mm Hg and/or diastolic BP ≥90 mm Hg, or current use of antihypertensive medication. Diabetes mellitus (DM) was defined as self‐report of physician's diagnosis, antihyperglycemic therapy use, fasting plasma glucose ≥126 mg/dL, or random plasma glucose ≥200 mg/dL. Hyperlipidemia was defined as a total cholesterol ≥240 mg/dL or treated hyperlipidemia.

2.4. Statistical analyses

Data are expressed as the mean ± SD, median (25th percentile, 75th percentile), or percentages as appropriate. The differences in prevalence rate were evaluated by chi‐square test. We used unpaired t tests to compare the mean values of two groups. Pearson correlation coefficients and linear regressions were calculated for hourly systolic BP, the hourly systolic BP surge, and organ damage markers, that is, LVMI, baPWV, NT‐proBNP, UACR, and IMT. Multiple regression analyses of organ damage (as the dependent variable) were performed using the following covariates: age, sex, smoking, hyperlipidemia, DM, antihypertensive drug use, clinic systolic BP, and 24‐hour ambulatory systolic BP. The statistical analyses were performed using SPSS software ver. 21 (IBM, Armonk, NY, USA). Differences with a two‐tailed P‐value <0.05 were considered significant.

3. RESULTS

3.1. Participant characteristics

Of the 4310 participants in the J‐HOP study, 1464 patients who underwent ABPM were included in the present analysis. Table 1 summarizes the characteristics of the participants in this study. The following characteristics of the present study's patient series were not significantly different from those of the J‐HOP participants who were excluded from this study: age (64.8 ± 11.6 vs 65.0 ± 10.6 years), the percentage of male patients (47.8% vs 46.6%), body mass index (24.5 ± 3.5 vs 24.2 ± 3.5 kg/m²), and percentage of current smokers (11.7% vs 12%).

Table 1.

Baseline characteristics of participants (n = 1464)

Variable
Age, y	64.8 ± 11.6
Male, %	47.8
BMI, kg/m2	24.5 ± 3.5
Current smokers, %	11.7
Hypertension, %	92.9
Diabetes mellitus, %	26.2
Hyperlipidemia, %	34.8
Past history of angina, acute myocardial infarction, or stroke, %	16.5
Antihypertensive medication, %	82.6
Calcium channel blockers, %	53.1
ACE inhibitors, %	7.3
Angiotensin receptor blockers, %	53.5
Beta blockers, %	15.4
Alpha blockers, %	6.7
Diuretics, %	28.9
Evening or bedtime dosing of at least 1 antihypertensive Medication, %	29.7
Statins, %	24.1
Aspirin, %	18.6
Cholesterol, mg/dL	199.7 ± 32.5
Triglyceride, mg/dL	127.2 ± 86.4
HDL‐cholesterol, mg/dL	56.2 ± 15.2
LDL‐cholesterol, mg/dL	118.1 ± 29.9
Fasting plasma glucose, mg/dL	109.7 ± 29.2
Hemoglobin A1c, %	5.9 ± 0.8
Creatinine, mg/dL	0.77 ± 0.26
Estimated glomerular filtration rate, mL/min/1.73 m2	72.95 ± 18.59
Clinic SBP, mm Hg	139.9 ± 15.9
Clinic DBP, mm Hg	80.3 + 11.1
Ambulatory BP
24‐h SBP, mm Hg	130.8 ± 13.0
24‐h DBP, mm Hg	77.2 + 9.2
Daytime SBP, mm Hg	135.4 ± 13.6
Daytime DBP, mm Hg	80.2 + 9.8
Nighttime SBP, mm Hg	120.1 ± 15.0
Nighttime DBP, mm Hg	70.4 + 9.4
LVMI, g/m2 (n = 1304)	98.5 ± 25.4
baPWV, m/s (n = 1386)	16.46 ± 3.28
NT‐proBNP, pg/mL (n = 1114)	[Median(P25, P75)] 55.9(28.0, 118.4)
UACR, mg/gCr (n = 1461)	[Median(P25, P75)] 13.3(7.2, 32.2)
Mean carotid IMT, mm (n = 584)	0.79 ± 0.18

ACE, angiotensin‐converting enzyme; baPWV, brachial‐ankle pulse wave velocity; BMI, body mass index; BP, blood pressure; DBP, diastolic blood pressure; HDL, high‐density lipoprotein; IMT, intima‐media thickness; LDL, low‐density lipoprotein; LVMI, left ventricular mass index; NT‐proBNP, N‐terminal pro‐B‐type natriuretic peptide; SBP, systolic blood pressure; UACR, urine albumin:creatinine ratio.

3.2. Ambulatory systolic BP values and hourly systolic BP surge

The moving average hourly systolic BP was demonstrated in Figure 1A. The highest and lowest systolic BP values were measured at 9 o'clock and 2 o'clock, respectively.

Figure 1.

(A) The moving average hourly systolic blood pressure (SBP) measured at x o'clock =  the average of the SBPs measured at x − 30 min, x, and x + 30 min. (B) Hourly SBP surge = moving average hourly SBP −  moving lowest nighttime systolic BP

According to the patients' diaries, the mean awakening time was 06:15 am and the mean bedtime was 10:25 pm The BP measured at 5 o'clock was preawakening BP in 88.8% of the patients, and the BP measured at 6 o'clock was preawakening BP of 53.5% of the patients.

The hourly systolic BP surge was first defined in this study as the hourly systolic BP value minus the moving lowest nighttime systolic BP (the trough systolic BP). In our patients, the trough systolic BP value was 107.8 ± 15.3 mm Hg. The hourly systolic BP surge value at 9 o'clock was highest. (Figure 1B).

3.3. Correlations between hourly systolic BP, hourly systolic BP surge, and organ damage

Most of the hourly systolic BP values correlated significantly with the LVMI. The hourly systolic BP at 6 o'clock had the highest correlation coefficient with LVMI (r = 0.166, P < 0.001). Only the early‐morning hourly systolic BP surge had a significant correlation with the LVMI. The hourly systolic BP surge at 6 o'clock had the highest correlation coefficient (r = 0.099, P < 0.01). After adjustment for age, sex, smoking, hyperlipidemia, DM, antihypertensive drug use, clinic systolic BP, and 24‐hour ambulatory systolic BP, only the correlations of the LVMI and early‐morning hourly systolic BP (hourly SBP at 6 o'clock, beta coefficient = 0.125, P < 0.01) and early‐morning hourly systolic BP surge remained significant (hourly SBP surge at 6 o'clock, beta coefficient = 0.008, P < 0.01) (Figure 2).

Figure 2.

(A) Linear regression analysis (B value) of the left ventricular mass index (LVMI) and hourly systolic blood pressure (SBP). (B) Linear regression analysis (B value) of LVMI and hourly SBP surge. (C) Multivariate analysis of LVMI and hourly SBP, after adjusting for age, sex, smoking, hyperlipidemia, diabetes mellitus, antihypertensive drug use, clinic SBP, and 24‐h SBP. (D) Multivariate analysis of LVMI and hourly SBP surge after adjusting for age, sex, smoking, hyperlipidemia, diabetes mellitus, antihypertensive drug use, clinic SBP, and 24‐h SBP. Shaded areas indicate statistical significance at P < 0.05

Almost all of the nighttime hourly systolic BP values correlated significantly with the IMT. The correlation coefficients between IMT and hourly systolic BP at 5 o'clock were highest (r = 0.196, P < 0.001). (Figure 3) Only hourly systolic BP surge at 4 and 5 o'clock had significant correlations with IMT. The correlation coefficient between IMT and hourly systolic BP surge at 5 o'clock was highest (r = 0.149, P < 0.01; Figure S1). After adjusting for covariates including clinic systolic BP and 24‐hour ambulatory systolic BP, these associations were no longer significant.

Figure 3.

Linear regression analysis (B value) of mean left and right carotid intima‐media thickness (IMT) (A) urinary albumin:creatinine ratio (UACR) (B) brachial‐ankle pulse wave velocity (baPWV) (C) N‐terminal pro‐B‐type natriuretic peptide (NT‐proBNP), (D) and hourly systolic blood pressure (SBP). Shaded areas indicate statistical significance at P < 0.05

The associations between other types of organ damage (UACR, baPWV, and NT‐proBNP) and hourly systolic BP or hourly systolic BP surge are shown in Figure 3 and Figure S1.

3.4. Correlations between organ damage and blood pressure values

Correlations between target organ damages and the BP values or BP surge indices are summarized in Table 2, and correlations between target organ damage and BP values (diastolic BP or pulse pressure) are shown in Tables S1 and S2.

Table 2.

Correlations (r) between target organ damage and blood pressure (BP) values, and BP surge indices

	BP values							BP surge indices
	Clinic SBP	24‐h SBP	Daytime SBP	Nighttime SBP	Hourly SBP at 4 o'clocka, *	Hourly SBP at 5 o'clockb, **	Hourly SBP at 6 o'clockc, ***	Sleep‐trough morning surged	Preawakening morning surgee	Hourly SBP surge at 4 o'clockf	Hourly SBP surge at 5 o'clockg	Hourly SBP surge at 6 o'clockh
LVMI	0.148***	0.129***	0.127***	0.125***	0.130***	0.159***	0.166***	0.066*	0.009	0.056	0.091**	0.099***
IMT	0.076	0.075	0.041	0.155***	0.162***	0.196***	0.078	0.031	−0.082	0.128**	0.149**	0.016
baPWV	0.290***	0.111***	0.094**	0.173***	0.167***	0.182***	0.155***	0.021	−0.058	0.079**	0.089**	0.041
UACR	0.138***	0.099***	0.098***	0.096***	0.099***	0.088**	0.121***	0.028	−0.008	0.066*	0.048	0.049
NT‐proBNP	0.034	0.047	0.030	0.090**	0.076*	0.092**	0.095**	0.030	−0.024	0.032	0.068*	0.057

baPWV, brachial‐ankle pulse wave velocity; IMT, carotid intima‐media thickness; LVMI, left ventricular mass index; NT‐proBNP, N‐terminal pro‐B‐type natriuretic peptide; SBP, systolic blood pressure; UACR, urinary albumin:creatinine ratio.

^aMean of SBP measured at 03:30, 04:00, and 04:30 am.

^bMean of SBP measured at 04:30, 05:00, and 05:30 am.

^cMean of SBP measured at 05:30, 06:00, and 06:30 am.

^dAverage morning SBP − moving lowest nighttime SBP.

^eAverage morning SBP − preawakening morning SBP.

^fHourly SBP at 4 o'clock − moving lowest nighttime SBP.

^gHourly SBP at 5 o'clock − moving lowest nighttime SBP.

^hHourly SBP at 6 o'clock ‐ moving lowest nighttime SBP.

^* P < 0.05.

^** P < 0.01.

^*** P < 0.001.

4. DISCUSSION

This study is the first to show the correlations of multiple asymptomatic hypertensive organ damage and hourly ambulatory systolic BP in a nationwide general practice‐based patient series with at least one CVD risk factor or a history of CVD or both. The moving average hourly systolic BP values measured at 5 and 6 o'clock exhibited significant correlations with organ damage, independent of clinic systolic BP and 24‐hour ambulatory systolic BP. Our analyses are also the first to describe a new BP variability index, that is, the hourly systolic BP surge, which is defined as the hourly systolic BP value minus the moving lowest nighttime systolic BP. Only early‐morning hourly systolic BP surge was significantly associated with organ damage.

In line with previous studies, we found that the BP values measured in the early morning had higher correlations with hypertensive organ damage. An earlier study from our institution showed that morning BP defined as the average of four systolic BP values measured during the first 2 hours after awakening was the strongest independent predictor of stroke events in 519 elderly hypertensive patients.8 Chen et al9 showed that in 757 high‐risk hypertensive Chinese patients, early‐morning systolic BP (average of systolic BP values measured between 5:00 and 7:59 am) had a significant correlation with intracranial arterial stenosis and that this correlation remained significant after adjustment for other cardiovascular risk factors.

There were some differences between the present and previous studies. In this study, systolic BP measured at 5 and 6 o'clock (which had the highest correlation with organ damage) was the preawakening BP according to the patients' diaries. In contrast, in Chen's study,9 a clock‐based definition was used and it was assumed that the early‐morning period (5:00‐7:59 am) was the post‐awakening period. In our earlier study,8 we used the patients' diary‐based definition of awakening, but the exact times of rising were not examined.

The systolic BP measured at 5 and 6 o'clock in this study was measured before waking in most participants and was lower than the systolic BP measured at other hours of the day, but it nonetheless exhibited the strongest correlation with organ damage. The preawakening period (5 and 6 o'clock in this study) may be a rapid eye movement (REM) period of the sleep cycle, which was demonstrated by Somers et al17 to have higher BP than non‐REM sleep. During the preawakening period, there are no other external factors which affect BP, such as physical activity, exercise, or psychological stress. The elevated preawakening BP may be associated with autonomic dysfunction18, 19, 20 and may be the most important period of elevated BP as a chronic risk factor for organ damage,18 whereas elevated post‐awakening BP may play a more important role as a trigger of acute CV events.

Our findings revealed that almost all of the hourly systolic BP values had significant correlations with the LVMI, baPWV, and UACR. The IMT and NT‐proBNP had significant correlations with hourly systolic BP during sleep, whereas only the hourly systolic BP surges of the early‐morning period had significant correlations with organ damage. Our results pointed out the importance of BP measured in the early‐morning period which had independent correlation with organ damage. Our results indicate that early‐morning SBP constitutes a residual risk for patients in addition to conventional clinic BP. In order to have an additional patient's risk assessment and minimize organ damage and CVD, the early‐morning BP should be assessed and controlled.21, 22 Furthermore, the correlations between hourly systolic BP and organ damage were higher than those between the hourly systolic BP surges and organ damage. This finding indicates that the BP level itself may be more important than BP variability, as was shown in many earlier studies.10 However, the morning period is the most sensitive period to detect the impact of BP variability on cardiovascular risk.

The strength of this study is this is the first study to (a) evaluate the association between multiple markers of subclinical hypertensive organ damage and ambulatory hourly systolic BP and hourly systolic BP surge in participants with cardiovascular risk; and (b) describe a new BP variability index, that is, the hourly BP surge. However, there are limitations. First, it was a cross‐sectional study so the causality of findings cannot be determined. Second, other variables which may affect the diurnal change in BP. For example, sleep quality were not evaluated. Third, both echocardiography and carotid ultrasonography were performed in each institute, so there may have been some differences among observers.

5. CONCLUSIONS

This hourly ambulatory BP analysis demonstrates that only early‐morning BP values and hourly BP surge were significantly associated with organ damage even after adjusting for clinic SBP and 24‐hour ambulatory SBP. Therefore, the results of our hourly ambulatory BP analysis might be useful to assess the residual risk for patients in addition to the risk assessed by conventional clinic BP measurement. However, further longitudinal study is necessary to investigate the association between early‐morning hourly BP and cardiovascular events.

CONFLICT OF INTEREST

KK received research funding from Teijin Pharma Limited, Omron Healthcare Co., Ltd., Fukuda Denshi, Bayer Yakuhin Ltd., A &D Co., Ltd., Daiichi Sankyo Company, Limited, Mochida Pharmaceutical Co., Ltd, EA pharma, Boehringer Ingelheim Japan Inc, Tanabe Mitsubishi Pharma Corporation, Shionogi & Co., Ltd., MSD KK, Sanwa Kagaku Kenkyusho Co., Ltd., Bristol‐Myers Squibb KK, and honoraria from Takeda Pharmaceutical Company Limited and Omron Healthcare Co., Ltd.

Supporting information

 

 

Wanthong S, Kabutoya T, Hoshide S, Buranakitjaroen P, Kario K. Early morning—Best time window of hourly 24‐hour ambulatory blood pressure in relation to hypertensive organ damage: The Japan Morning Surge‐Home Blood Pressure study. J Clin Hypertens. 2019;21:579–586. 10.1111/jch.13498