Prognostic Value of Reading-to-Reading Blood Pressure Variability Over 24 Hours in 8938 Subjects From 11 Populations
In previous studies, of which several were underpowered, the relation between cardiovascular outcome and blood pressure (BP) variability was inconsistent. We followed health outcomes in 8938 subjects (mean age: 53.0 years; 46.8% women) randomly recruited from 11 populations. At baseline, we assessed BP variability from the SD and average real variability in 24-hour ambulatory BP recordings. We computed standardized hazard ratios (HRs) while stratifying by cohort and adjusting for 24-hour BP and other risk factors. Over 11.3 years (median), 1242 deaths (487 cardiovascular) occurred, and 1049, 577, 421, and 457 participants experienced a fatal or nonfatal cardiovascular, cardiac, or coronary event or a stroke. Higher diastolic average real variability in 24-hour ambulatory BP recordings predicted (P≤0.03) total (HR: 1.14) and cardiovascular (HR: 1.21) mortality and all types of fatal combined with nonfatal end points (HR: ≥1.07) with the exception of cardiac and coronary events (HR: ≤1.02; P≥0.58). Higher systolic average real variability in 24-hour ambulatory BP recordings predicted (P<0.05) total (HR: 1.11) and cardiovascular (HR: 1.16) mortality and all fatal combined with nonfatal end points (HR: ≥1.07), with the exception of cardiac and coronary events (HR: ≤1.03; P≥0.54). SD predicted only total and cardiovascular mortality. While accounting for the 24-hour BP level, average real variability in 24-hour ambulatory BP recordings added <1% to the prediction of a cardiovascular event. Sensitivity analyses considering ethnicity, sex, age, previous cardiovascular disease, antihypertensive treatment, number of BP readings per recording, or the night:day BP ratio were confirmatory. In conclusion, in a large population cohort, which provided sufficient statistical power, BP variability assessed from 24-hour ambulatory recordings did not contribute much to risk stratification over and beyond 24-hour BP.
Ambulatory blood pressure monitoring not only provides information on the blood pressure level but on the diurnal changes in blood pressure as well. Blood pressure variability includes both short-term and circadian components, which can be estimated by the SD of the blood pressure values over a defined period of the day or by the night:day blood pressure ratio, respectively. We recently reported in >7000 subjects recruited from 6 populations on the prognostic accuracy of long-term blood pressure variability.1 Both daytime and nighttime blood pressure consistently predicted the composite end point of all cardiovascular events. Adjusted for the 24-hour blood pressure, the night:day blood pressure ratio predicted mortality but not fatal combined with nonfatal events.
Although the aforementioned analyses shed light on the association between outcome and long-term blood pressure variability, the predictive value of short-term reading-to-reading blood pressure variability remains uncertain. Possible limitations of previous studies were a lack of statistical power,2–5 selection of specific groups of patients,5–7 categorization of variability by arbitrary cutoff points,2,4,7–9 and sole reliance on fatal end points.10,11 Moreover, various parameters can capture short-term blood pressure variability over 24 hours, but most studies only considered the SD of systolic4,6,12 or diastolic blood pressure or both.8–10 To address the prognostic value of short-term blood pressure variability, we expanded, updated, and analyzed the International Database on Ambulatory Blood Pressure in Relation to Cardiovascular Outcome.
Previous publications described the construction of the International Database on Ambulatory Blood Pressure in Relation to Cardiovascular Outcome.1,13–15 Studies were eligible for inclusion if they involved a random population sample, if baseline information on ambulatory blood pressure and cardiovascular risk factors was available, and if the subsequent follow-up included fatal and nonfatal outcomes. At the time of writing this report, the International Database on Ambulatory Blood Pressure in Relation to Cardiovascular Outcome included prospective studies from 11 centers (11 785 subjects). All studies received ethical approval and have been reported in peer-reviewed publications. In line with previous reports,1,13–15 we excluded 252 participants because they were <18 years of age and 1892 participants because they had <10 daytime or <5 nighttime blood pressure readings. For the analyses of the variability, we additionally disregarded 703 subjects because they had missing readings during 3 consecutive hours. The 8938 analyzed participants were 2018 residents from Copenhagen, Denmark16; 1086 subjects from Noorderkempen, Belgium17; 1069 older men from Uppsala, Sweden18; 226 subjects from Novosibirsk, the Russian Federation19,20; 1430 inhabitants from Ohasama, Japan21; 346 villagers from the JingNing county, China22; 1093 subjects from Montevideo, Uruguay23; 161 subjects from Pilsen, the Czech Republic20; 900 subjects from Dublin, Ireland24; 303 subjects from Padova, Italy20; and 306 subjects from Kraków, Poland.20 All participants gave informed written consent.
Blood Pressure Measurements
Conventional blood pressure was measured by trained observers with a mercury sphygmomanometer16–20,22,24; with validated auscultatory21 (USM-700F, UEDA Electronic Works) or oscillometric23 (OMRON HEM-705CP, Omron Corporation) devices, using the appropriate cuff size; and with participants in the sitting16,17,19–24 or supine18 position. Conventional blood pressure was the average of 2 consecutive readings obtained either at the person’s home17,19,20,22,23 or at an examination center.16,18,21,24 Hypertension was a conventional blood pressure of ≥140 mm Hg systolic or ≥90 mm Hg diastolic or the use of antihypertensive drugs.
We programmed portable monitors to obtain ambulatory blood pressure readings at 30-minute intervals throughout the whole day21,24 or at intervals ranging from 1516 to 3018 minutes during daytime and from 3016 to 6018 minutes at night. The devices implemented an auscultatory algorithm (Accutracker II) in Uppsala18 or an oscillometric technique (SpaceLabs 90202 and 90207, Nippon Colin, and ABPM-630) in the other cohorts.17–24 While accounting for the daily pattern of activities of the participants, we defined daytime as the interval from 10:00 am to 8:00 pm in Europeans16–20,24 and South Americans23 and from 8:00 am to 6:00 pm in Asians.21,22 The corresponding nighttime intervals ranged from 12:00 pm to 6:00 am16–20,23,24 and from 10:00 pm to 4:00 am,21,22 respectively. In dichotomous analyses, we defined systolic and diastolic nondipping as a night:day blood pressure ratio of ≥0.90.1
As measures of short-term reading-to-reading blood pressure variability, we used the SD over 24 hours weighted for the time interval between consecutive readings (SD24), the average of the daytime and nighttime SDs weighted for the duration of the daytime and nighttime interval (SDdn),25 and the average real variability weighted for the time interval between consecutive readings (average real variability in 24-hour ambulatory BP recordings; ARV24).4 The SDdn is the mean of day and night SD values corrected for the number of hours included in each of these 2 periods (Figure 1A), according to the following formula25: SDdn=[(day SD×hours included in the daytime)+(night SD×hours included in the nighttime)]/(hours included in daytime+nighttime). This method removes the influence of the day-night blood pressure difference from the estimate of blood pressure variability. The ARV24 averages the absolute differences of consecutive measurements and accounts in this manner for the order in which the blood pressure measurements are obtained (Figure 1B). It is calculated by the following formula: equation
where k ranges from 1 to N−1 and w is the time interval between BPk and BPk+1. N is the number of blood pressure readings.4
We used the questionnaires originally administered in each cohort to obtain information on each subject’s medical history and smoking and drinking habits. Body mass index was body weight in kilograms divided by height in meters squared. We measured serum cholesterol and blood glucose by automated enzymatic methods. Diabetes mellitus was the use of antidiabetic drugs, a fasting blood glucose concentration of ≥7.0 mmol/L,16–21,23 a random blood glucose concentration of ≥11.1 mmol/L,17,21,22 a self-reported diagnosis,17,22,23 or diabetes mellitus documented in practice or hospital records.23
Ascertainment of Events
In each cohort, outcomes were adjudicated against source documents described in previous publications.13,18,21,22,26–28 The adjudication process was the same in the Belgium study28 and in all of the other studies included in the European Project on Genes in Hypertension (Novosibirsk, Pilsen, Padova, and Kraków).29 Outcomes were coded according to the international classification of diseases (ICD), as tabulated in the online Data Supplement (Table S1), available at http://hyper.ahajournals.org.
Fatal and nonfatal stroke (ICD8/9 430 to 434 and 436, and ICD10 I60 to I64, I67 and I68) did not include transient ischemic attacks. Coronary events encompassed death from ischemic heart disease (ICD8 411 and 412, ICD9 411 and 414, and ICD10 I20, I24 and I25), sudden death (CD8 427.2 and 795, ICD9 427.5 and 798, and ICD10 I46 and R96), nonfatal myocardial infarction (ICD8/9 410, and ICD10 I21 and I22), and coronary revascularization. Cardiac events were composed of coronary end points and fatal and nonfatal heart failure (ICD8 428, 427.1, 427.2 and 429, ICD9 429, and ICD10 I50 and J81). Hospitalizations for unstable angina were coded as ischemic heart disease. In the Danish and Swedish cohorts, the diagnosis of heart failure required admission to the hospital. In the other cohorts, heart failure was either a clinical diagnosis or the diagnosis on the death certificate, but in all cases it was validated against hospital files or the records held by family doctors. The composite cardiovascular end point included all aforementioned end points plus cardiovascular mortality (ICD8 390 to 448, ICD9 390.0 to 459.9, and ICD10 I00 to I79 and R96). In all outcome analyses, we only considered the first event within each category.
For database management and statistical analysis, we used SAS software, version 9.1.3 (SAS Institute). For comparison of means and proportions, we applied the large-sample z test and the χ2 statistic, respectively. We used a Pearson correlation coefficient to assess the correlations among the 3 measures of short-term blood pressure variability. Statistical significance was an α-level of <0.05 on 2-sided tests. After stratification for cohort and sex, we computed missing values of body mass index (n=916) and serum cholesterol (n=598) from the regression slope on age. In subjects with unknown smoking status (n=38) or drinking habits (n=435 among Swedish men and n=316 among the other cohorts), we set the design variable to the cohort- and sex-specific mean of the codes (0 and 1).
We used Cox regression to compute standardized hazard ratios (HRs). We checked the proportional hazards assumption by the Kolmogorov-type supremum test, as implemented in the PROC PHREG procedure of the SAS package and by testing the interaction terms between follow-up duration and the variable of interest. We first plotted incidence rates by fifths of the distributions of systolic and diastolic blood pressure variability, while standardizing by the direct method for cohort, sex, and age (≤40, 40 to 60, and ≥60 years). We computed HRs while stratifying for cohort and adjusting for sex and baseline characteristics, including age (used as a continuous variable), 24-hour heart rate, body mass index, smoking (0 and 1) and drinking (0 and 1), serum cholesterol, history of cardiovascular disease (0 and 1), diabetes mellitus (0 and 1), and treatment with antihypertensive drugs (0 and 1). In fully adjusted models, we additionally adjusted for the 24-hour systolic or diastolic blood pressure. We tested heterogeneity in the HRs across subgroups by introducing the appropriate interaction term in the Cox model. Finally, we applied the generalized R2 statistic to assess the risks explained in Cox regression30 by consecutively entering the 24-hour blood pressure and ARV24 as predictor variables into the models for the composite cardiovascular end point.
The study population consisted of 6069 Europeans (67.9%), 1093 Asians (12.2%), and 1176 South Americans (19.9%). The 8938 participants included 4785 women (46.8%) and 3664 patients with hypertension (41.0%), of whom 1749 (47.7%) were taking blood pressure–lowering drugs. Mean (±SD) age was 53.0±15.8 years. At enrollment, 2558 participants (28.7%) were current smokers, and 4351 (53.1%) reported intake of alcohol.
Table 1 shows the baseline characteristics by quartiles of diastolic ARV24. Across quartiles, all characteristics were significantly different (P<0.05). Participants with a higher blood pressure variability were older, had higher blood pressure, were more likely to be male, and were more likely to have diabetes mellitus (Table 1). The ARV24, SD24, and SDdn were highly correlated with one another; the correlation coefficients ranged from 0.75 to 0.81 (P≤0.001) for systolic blood pressure and from 0.71 to 0.79 (P≤0.001) for diastolic blood pressure.
Incidence of Events
In the overall study population, median follow-up was 11.3 years (fifth to 95th percentile interval: 2.5 to 17.6 years). Across cohorts, median follow-up ranged from 2.5 years (fifth to 95th percentile interval: 2.3 to 2.6) in JingNing to 17.6 years (fifth to 95th percentile interval: 16.4 to 18.2 years) in Dublin. During 96 041 person-years of follow-up, 1242 participants died (12.9 per 1000 person-years), and 1049 experienced a fatal or nonfatal cardiovascular complication (11.3 per 1000 person-years). Mortality included 487 cardiovascular and 713 noncardiovascular deaths and 42 deaths from unknown causes (Table 2). Considering cause-specific first cardiovascular events, the incidence of fatal and nonfatal stroke amounted to 138 and 371, respectively. Cardiac events consisted of 172 fatal and 405 nonfatal events, including 72 fatal and 204 nonfatal cases of acute myocardial infarction, 50 deaths from ischemic heart diseases, 13 sudden deaths, 37 fatal and 151 nonfatal cases of heart failure, and 50 cases of surgical or percutaneous coronary revascularization. For comparison, cohort-specific mortality data and country-specific mortality statistics published by the World Health Organization are presented in Table S2.
Risk Associated With Blood Pressure Variability
Figure 2 shows the cohort-, sex-, and age-standardized rates of mortality and fatal combined with nonfatal outcomes across quintiles of systolic and diastolic ARV24. The multivariable-adjusted and standardized HRs associated with systolic and diastolic blood pressure variability for mortality and for all fatal combined with nonfatal cardiovascular events appear in Table 2.
In adjusted models not including the 24-hour blood pressure level, systolic blood pressure variability predicted both total and cardiovascular mortality (P≤0.04), with the exception of SD24 in relation to total mortality (P=0.17). We obtained similar results after additional adjustment for the 24-hour systolic blood pressure, with the exception of SD24 and SDdn, which no longer predicted cardiovascular mortality (P≥0.71). Diastolic blood pressure variability predicted total and cardiovascular mortality both in adjusted and fully adjusted models (P≤0.002; Table 2). Blood pressure variability did not predict noncardiovascular mortality (0.14≤P≤0.75).
Fatal and Nonfatal Cardiovascular Events
In adjusted analyses not including the 24-hour blood pressure level, systolic blood pressure variability predicted all of the fatal combined with nonfatal outcomes (P≤0.03) with the exception of coronary events (P≥0.07). However, in fully adjusted analyses, systolic blood pressure variability lost its predictive value with the exception of ARV24 in relation to all cardiovascular events combined and stroke (Table 2).
Diastolic blood pressure variability was predictive of all of the combined end points (P≤0.03), with the exception of coronary events (P≥0.15). In fully adjusted models, diastolic blood pressure variability only predicted all cardiovascular events combined (ARV24 and SDdn) and fatal plus nonfatal stroke (ARV24). Figure 3 shows the absolute risk of a combined cardiovascular event in relation to the ARV24 at different levels of systolic and diastolic 24-hour blood pressure (Figure 3A and 3B) and in relation to 24-hour blood pressure at different levels of the systolic and diastolic ARV24 (Figure 3C and 3D). The analyses were standardized to the distributions (mean or ratio) of cohort, sex, age, 24-hour heart rate, body mass index, smoking and drinking, serum cholesterol, history of cardiovascular disease, diabetes mellitus, and treatment with antihypertensive drugs. Absolute risk increased with both the 24-hour blood pressure (P<0.001) and ARV24 (P≤0.04). However, with the 24-hour blood pressure in the model, ARV24 added only 0.1% to the explained risk of a composite cardiovascular event (Table 3).
In sensitivity analyses, we considered total mortality and all cardiovascular events combined in relation to diastolic and systolic ARV24 (Tables S3 and S4). We stratified the study population according to sex; median age (60 years); antihypertensive treatment; the presence or absence of hypertension; European, Asian, or South American origin; numbers of blood pressure readings in individual blood pressure recordings; and dipping status dichotomized by a night:day blood pressure ratio of 0.9. HRs were not statistically different across strata (0.07<P<0.93) with 2 exceptions. First, the HR for diastolic ARV24 in relation to total mortality was higher in treated than untreated subjects (1.20 versus 1.04; P=0.02). Second, the HR for diastolic ARV24 in relation to all cardiovascular events was higher in Asians than in Europeans (1.36 versus 1.04; P=0.009) but similar in Europeans and South Americans (P=0.20). The predictive value of systolic ARV24 was similar across all strata (Table S4; 0.07≤P≤0.88).
Analyses on the basis of daytime and nighttime ARV were confirmative but did not attain the significance levels of ARV24 because of the fewer readings included in the daytime and nighttime ARV (data not shown). For SD24 and SDdn, none of the interaction terms with the strata shown in Tables S3 and S4 reached significance (0.08<P<0.97).
Our current meta-analysis of individual data included >8000 people randomly recruited from 11 populations and covered, on average, 11 years of follow-up, during which 1242 people died and 1049 experienced a major cardiovascular complication. The key finding was that, while accounting for the 24-hour blood pressure level and other covariables, blood pressure variability was a significant and independent predictor of mortality and of cardiovascular and stroke events. However, the proportion of the risk explained by the variability index is low.
For most outcomes, ARV24 was a better predictor than SD24 and SDdn, probably because, as illustrated in Figure 1, subjects with different blood pressure profiles might have similar SDs but different ARV24s. Thus, ARV24 might be a more specific measure of blood pressure variability than SD.
Several prospective studies in populations4,10–12,31 and hypertensive patients2,3,5–9 searched for association between cardiovascular outcomes and blood pressure variability but reported inconsistent results. This might be because of insufficient sample size, too few events, varying definitions of the outcomes of interest, or the use of different indices of blood pressure variability. To assess blood pressure variability, most studies used ambulatory blood pressure monitoring with intermittent readings at intervals ranging from 158 to 305 minutes throughout 24 hours. In the Northwick Park Study,3 the investigators performed continuous intra-arterial recordings but did not fully exploit the potential of this recording technique. Instead of analyzing variability in the frequency domain, they computed hourly means of blood pressure and the within-subject SD of the hourly means as a measure of each participant’s blood pressure variability. In the Ohasama Study, investigators used the self-measured blood pressure31 in addition to ambulatory blood pressure monitoring.10 In all but 2 studies,4,7 the researchers used the SD of daytime, nighttime, or 24-hour blood pressure as an index of variability. Four studies5,8–10 deliberately did not report on the predictive value of the variability in the 24-hour blood pressure, because the diurnal blood pressure profile also includes long-term variability, which is captured by the night:day blood pressure ratio. To address this potential concern, we computed SDdn and ARV24 as measures of variability. Only 2 other prospective studies, one in a small general Venezuelan population (312 subjects with 31 composite cardiovascular end points),4 and one in a hypertensive population,7 implemented ARV24. Bilo et al25 were the first to propose SDdn, but to our knowledge there is no prospective study that has used this index of variability.
Diastolic blood pressure variability tended to be a stronger predictor of outcome than systolic blood pressure variability. We can only speculate about the mechanisms underlying this finding, but arterial stiffness might be involved. In normal conditions, systolic and diastolic blood pressures change in parallel in response to physiological stimuli, such as excise or arousal. However, in subjects with stiff arteries, when systolic blood pressure increases, often diastolic blood pressure increases less or even falls,32,33 giving rise to larger variability. On the other hand, a chance finding cannot be excluded.
From a clinical point of view, our current findings suggest that, although statistically significant, the clinical applicability of blood pressure variability for risk stratification might be limited. First, antihypertensive drug treatment is bound to influence blood pressure variability. Second, the reproducibility of blood pressure variability is poor. In 97 normotensive subjects,34 the relative repeatability coefficient of the SD of the 24-hour blood pressure in individual recordings, expressed as a percentage of the fifth to 95th percentile interval in all recordings, was 13% systolic and 16% diastolic, whereas for the 24-hour blood pressure these coefficients were 4% and 5%, respectively, lower values, indicating better reproducibility.34 Finally, the added value in terms of absolute risk was modest in our population. For example, in adjusted analyses (Figure 3), the increase in the 10-year absolute risk of a composite cardiovascular event associated with an increase from the median to the 75th percentile was 0.21% for systolic ARV24 (1.5 mm Hg) and 1.23% for the 24-hour systolic blood pressure (9.8 mm Hg). The corresponding estimates for diastolic ARV24 and for 24-hour diastolic blood pressure were 0.16% (2.3 mm Hg) and 1.05% (5.8 mm Hg), respectively.
Notwithstanding the statistical power and the consideration of fatal and nonfatal events, our study has potential limitations. First, the International Database on Ambulatory Blood Pressure in Relation to Cardiovascular Outcome is currently composed of 11 population-based cohorts from 3 continents, but our results might not yet be generally applicable, in particular to Africans of black ancestry or African Americans. Second, we and most other investigators applied intermittent techniques of ambulatory blood pressure monitoring, which compared with continuous blood pressure recording, is a less precise technique to capture short-term blood pressure variability. However, intra-arterial recordings or continuous recordings of the arterial signal at the finger are difficult, if not impossible, to implement in large epidemiological studies. Third, in the current meta-analysis of individual data, blood pressure variability turned out to provide independent risk information, and this finding was consistent in stratified sensitivity analyses. However, even in large cohort studies with numerous events, the power to detect heterogeneity across strata is generally low. For example, considering a 2-sided α-level of 0.05, we had only 46% power to detect a 0.24 difference between normotensive and hypertensive subjects in the log-transformed HR of all cardiovascular events.
In line with several6–8,10,11 but not all9 previous studies, our current report established that short-term reading-to-reading blood pressure variability is an independent risk factor, but moreover it also highlighted that the level of the 24-hour blood pressure remains the primary blood pressure–related risk factor to account for in clinical practice. This caveat also applies to the morning surge in blood pressure, as described in the companion article. Notwithstanding these limitations, in the setting of clinical research, studies of blood pressure variability will continue to generate meaningful information. For research making use of intermittent techniques of ambulatory blood pressure monitoring, our current findings suggest that both SDdn and ARV24 might be useful measures, but not the SD computed over the whole day, which also includes the day-night blood pressure difference.
We gratefully acknowledge the expert assistance of Sandra Covens and Ya Zhu (Studies Coordinating Centre, Leuven, Belgium). The International Database of Ambulatory Blood Pressure in Relation to Cardiovascular Outcome investigators are listed in the online Data Supplement available at http://hyper.ahajournals.org.
Sources of Funding
The European Union (grants IC15-CT98-0329-EPOGH, LSHM-CT-2006-037093, and HEALTH-F4-2007-201550), the Fonds voor Wetenschappelijk Onderzoek Vlaanderen (Ministry of the Flemish Community, Brussels, Belgium) (grants G.0575.06 and G.0734.09), and the Katholieke Universiteit Leuven (grants OT/00/25 and OT/05/49) gave support to the Studies Coordinating Centre in Leuven. The European Union (grants LSHM-CT-2006-037093 and HEALTH-F4-2007-201550) also supported the research groups in Shanghai, Kraków, Padova, and Novosibirsk. The Bilateral Scientific and Technological Collaboration between China and Flanders, Ministry of the Flemish Community, Brussels (grant BIL02/10), supported the fellowship of Y.L. in Leuven. The Danish Heart Foundation (grant 01-2-9-9A-22914) and the Lundbeck Fonden (grant R32-A2740) supported the studies in Copenhagen. The Ministries of Education, Culture, Sports, Science, and Technology (grants 15790293, 16590433, 17790381, 18390192, 18590587, 19590929, and 19790423) and of Health, Labor, and Welfare (Health Science Research grants, Medical Technology Evaluation Research grants, H17-Kenkou-007, H18-Junkankitou[Seishuu]-Ippan-012, and H20-Junkankitou[Seishuu]-Ippan-009, 013), a Grant-in-Aid from the Japanese Society for the Promotion of Science (16.54041, 18.54042, 19.7152, 20.7198, 20.7477, and 20.54043), the Japan Atherosclerosis Prevention Fund, the Uehara Memorial Foundation, the Takeda Medical Research Foundation, the National Cardiovascular Research grants, and the Biomedical Innovation grants supported research in Japan. The National Natural Science Foundation of China (grants 30871360 and 30871081), Beijing, China, the Shanghai Commission of Science and Technology (grant 07JC14047 and the “Rising Star” program, grant 06QA14043), and the Shanghai Commission of Education (grant 07ZZ32 and the “Dawn” project, grant 08SG20) supported the JingNing Study in China.
Correspondence to Jan A. Staessen, Studies Coordinating Centre, Division of Hypertension and Cardiovascular Rehabilitation, Department of Cardiovascular Diseases, University of Leuven, Campus Sint Rafaël, Kapucijnenvoer 35, Block d, Level 00, Box 7001, B-3000 Leuven, Belgium. E-mail email@example.com; firstname.lastname@example.org
- Received August 13, 2009.
- Revision received August 31, 2009.
- Accepted January 14, 2010.
Boggia J, Yan L, Thijs L, Hansen TW, Kikuya M, Björklund-Bodegård K, Richart T, Ohkubo T, Kuznetsova T, Torp-Pedersen C, Lind L, Ibsen H, Imai K, Wang J, Sandoya E, O'Brien E, Staessen JA, on behalf of the IDACO investigators. Prognostic accuracy of day versus night ambulatory blood pressure: a cohort study. Lancet. 2007; 370: 1219–1229.
Khattar RS, Swales JD, Dore C, Senior R, Lahiri A. Effect of aging on the prognostic significance of ambulatory systolic, diastolic, and pulse pressure in essential hypertension. Circulation. 2001; 104: 783–789.
Mena L, Pintos S, Queipo NV, Aizpúrua JA, Maestre G, Sulbarán T. A reliable index for the prognostic significance of blood pressure variability. J Hypertens. 2003; 23: 505–511.
Eguchi K, Ishikawa J, Hoshide S, Pickering TG, Schwartz JE, Shimada K, Kario K. Nighttime blood pressure variability is a strong predictor for cardiovascular events in patients with type 2 diabetes. Am J Hypertens. 2009; 22: 46–51.
Pringle E, Phillips C, Thijs L, Davidson C, Staessen JA, de Leeuw PW, Jaaskivi M, Nachev C, Parati G, O'Brien ET, Tuomilehto J, Webster J, Bulpitt CJ, Fagard RH, on behalf of the Syst-Eur Investigators. Systolic blood pressure variability as a risk factor for stroke and cardiovascular mortality in the elderly hypertensive population. J Hypertens. 2003; 21: 2251–2257.
Pierdomenico SD, Nicola MD, Esposito AL, Mascio RD, Ballone E, Lapenna D, Cuccurullo F. Prognostic value of different indices of blood pressure variability in hypertensive patients. Am J Hypertens. 2009; 22: 842–847.
Verdecchia P, Angeli F, Gattobigio R, Rapicetta C, Reboldi G. Impact of blood pressure variability on cardiac and cerebrovascular complications in hypertension. Am J Hypertens. 2007; 20: 154–161.
Pierdomenico SD, Lapenna D, Di Tommaso R, Di Carlo S, Esposito AL, Di Mascio R, Ballone E, Cuccurullo F, Mezzetti A. Blood pressure variability and cardiovascular risk in treated hypertensive patients. Am J Hypertens. 2006; 19: 991–997.
Kikuya M, Hozawa A, Ohokubo T, Tsuji I, Michimata M, Matsubara M, Ota M, Nagai K, Araki T, Satoh H, Ito S, Hisamichi S, Imai Y. Prognostic significance of blood pressure and heart rate variabilities: the Ohasama Study. Hypertension. 2000; 36: 901–906.
Mancia G, Bombelli M, Faccheti R, Madotto F, Corrao G, Trevano FQ, Grassi G, Sega R. Long-term prognostic value of blood pressure variability in the general population: results of the Pressioni Arteriose Monitorate e Loro Associazioni study. Hypertension. 2007; 49: 1265–1270.
Thijs L, Hansen TW, Kikuya M, Björklund-Bodegård K, Li Y, Dolan E, Tikhonoff V, Seidlerová J, Kuznetsova T, Stolarz K, Bianchi M, Richart T, Casiglia E, Malyutina S, Filipovský J, Kawecka-Jaszcz K, Nikitin Y, Ohkubo T, Sandoya E, Wang J, Torp-Pedersen C, Lind L, Ibsen H, Imai Y, Staessen JA, on behalf of the IDACO investigators. The International Database of Ambulatory Blood Pressure in Relation to Cardiovascular Outcome (IDACO): protocol and research perspectives. Blood Press Monit. 2007; 12: 255–262.
Kikuya M, Hansen TW, Thijs L, Björklund-Bodegård K, Kuznetsova T, Ohkubo T, Richart T, Torp-Pedersen C, Lind L, Ibsen H, Imai K, Staessen JA, on behalf of the IDACO investigators. Diagnostic thresholds for ambulatory blood pressure monitoring based on 10-year cardiovascular risk. Circulation. 2007; 115: 2145–2152.
Hansen TW, Thijs L, Boggia J, Kikuya K, Björklund-Bodegård K, Richart T, Ohkubo T, Jeppesen J, Torp-Pedersen C, Lind L, Sandoya E, Imai K, Wang J, Ibsen H, O'Brien E, Staessen JA, on behalf of the IDACO investigators. Prognostic value of ambulatory heart rate revisited in 6928 subjects from 6 populations. Hypertension. 2008; 52: 229–235.
Hansen TW, Jeppesen J, Rasmussen S, Ibsen H, Torp-Pedersen C. Ambulatory blood pressure and mortality: a population based study. Hypertension. 2005; 45: 499–504.
Kuznetsova T, Staessen JA, Kawecka-Jaszcz K, Babeanu S, Casiglia E, Filipovsky J, Nachev C, Nikitin Y, Peleskã J, O'Brien E, for the EPOGH Investigators. Quality control of the blood pressure phenotype in the European Project on Genes in Hypertension. Blood Press Monit. 2002; 7: 215–224.
Ohkubo T, Hozawa A, Yamaguchi J, Kikuya M, Ohmori K, Michimata M, Matsubara M, Hashimoto J, Hoshi H, Araki T, Tsuji I, Satoh H, Hisamichi S, Imai Y. Prognostic significance of the nocturnal decline in blood pressure in individuals with and without high 24-h blood pressure: the Ohasama Study. J Hypertens. 2002; 20: 2183–2189.
Schettini C, Bianchi M, Nieto F, Sandoya E, Senra H, for the Hypertension Working Group. Ambulatory blood pressure: normality and comparison with other measurements. Hypertension. 1999; 34: 818–825.
Hansen TW, Jeppesen J, Rasmussen S, Ibsen H, Torp-Pedersen C. Ambulatory blood pressure monitoring and risk of cardiovascular disease: a population based study. Am J Hypertens. 2006; 19: 243–250.
Dolan E, Stanton A, Thijs L, Hinedi K, Atkins N, McClory S, Den Hond E, McCormack P, Staessen JA, O'Brien E. Superiority of ambulatory over clinic blood pressure measurement in predicting mortality: the Dublin Outcome Study. Hypertension. 2005; 46: 156–161.
Tikhonoff V, Staessen JA, Kuznetsova T, Thijs L, Hasenkamp S, Baumer V, Stolarz K, Seidlerová J, Filipovský J, Nikitin Y, Peleška J, Kawecka-Jaszcz K, Casiglia E, Brand-Herrmann SM, Brand E, for the European Project On Genes in Hypertension (EPOGH) Investigators. SAH gene variants revisited in the European Project on Genes in Hypertension. J Hypertens. 2008; 26: 244–250.
Gillespie BW. Use of generalized R-squared in Cox regression APHA Scientific Session and Event Listing. 2006. http://apha.confex.com/apha/134am/techprogram/paper_135906.htm.
Kikuya M, Ohkubo T, Metoki H, Asayama K, Hara H, Obara T, Inoue R, Hoshi H, Hashimoto J, Totsune K, Satoh H, Imai K. Day-by-day variability of blood pressure and heart rate at home as a novel predictor of prognosis: the Ohasama Study. Hypertension. 2008; 52: 1045–1050.
MacWilliam JA, Melvin GS. Systolic and diastolic blood pressure estimation with special reference to the auditory method. Br Med J. 1914; March 28: 693–697.
Li Y, Wang JG, Dolan E, Gao PJ, Guo HF, Nawrot T, Stanton AV, Zhu DL, O'Brien E, Staessen JA. Ambulatory arterial stiffness index derived from 24-hour ambulatory blood pressure monitoring. Hypertension. 2006; 47: 359–364.