Journal of Human Hypertension (2005) 19, 565–567 & 2005 Nature Publishing Group All rights reserved 0950-9240/05 $30.00 www.nature.com/jhh
ORIGINAL ARTICLE
Normal ambulatory blood pressure: a clinical practice-based comparison of two recently published definitions IZ Ben-Dov, L Ben-Arie, J Mekler and M Bursztyn Department of Internal Medicine, Hadassah University Hospital, Jerusalem, Israel
The European Society of Hypertension (ESH) has issued guidelines for the detection and treatment of hypertension. According to these guidelines, normal 24-h ambulatory blood pressure (ABP) is defined as lower than 125/80 mmHg. Another publication of ESH recommendations for blood pressure (BP) measurement defines normal awake and asleep blood pressure as lower than 135/85 and 120/70 mmHg, respectively. Our aim was to investigate the compatibility of these two recently proposed ABP cutoffs in clinical practice. We analysed 1495 consecutive ABP measurements. In all, 56% of the subjects were female; age 58716 years; body mass index 2774 kg/m2; clinic BP 151722/ 84713 mmHg. Two-thirds were treated for hypertension, and 11% for diabetes. Subjects were classified as having normal 24-h BP if the corresponding value was o125/ 80 mmHg. Normal awake–sleep BP was diagnosed if
awake BP was o135/85 mmHg and sleep BP was o120/ 70 mmHg. Concordance between the cutoffs was found in 93% of the subjects. Among the 7% discordant subjects, 4.5% were hypertensive applying the 24 h, but not awake–sleep, BP values, whereas only 2.5% were hypertensive according to awake–sleep, but not 24 h, BP values (Po0.005). In Conclusion, in real-life ABP measurement, a good agreement was found between two recently issued ABP normality definitions. However, some subjects are classified as hypertensive only according to one of these methods, more often by the 24-h cutoff of 125/80. This discordance may be significant in large-scale clinical BP monitoring. Journal of Human Hypertension (2005) 19, 565–567. doi:10.1038/sj.jhh.1001866 Published online 14 April 2005
Keywords: ambulatory blood pressure monitoring; normality; guidelines
Introduction The upper limit of normal ambulatory blood pressure (ABP) is under debate. Based on correlations with clinic blood pressure (BP) of untreated subjects in the PAMELA study, Mancia et al1 calculated the cutoff point of normal ABP. They suggest a 24-h average of 120–130/75–81 mmHg as the limit. The corresponding value in elderly participants2 was 120/76 mmHg. In the Ohasama study, Ohkubo et al3 prospectively evaluated the prognostic implications of ABP. Twenty-four-hour values above 134/79 mmHg were associated with increased cardiovascular mortality. The seventh report of the joint national committee on prevention, detection, evaluation, and treatment of high BP4 did not define normal ABP, but stated that ‘awake hypertensive individuals have a mean BP of more than 135/85 mmHg and during sleep, more than 120/ 75 mmHg’. Highlighting the uncertainty regarding Correspondence: Dr IZ Ben-Dov, PO Box 1544, Mevasseret-Zion 90805, Israel. E-mail:
[email protected] Conflicts of interest: none. Received 25 November 2004; revised 12 February 2005; accepted 22 February 2005; published online 14 April 2005
normality is the fact that the European Society of Hypertension (ESH) has recently published two different normality definitions. In May 2003, the ESH issued recommendations for conventional, ambulatory and home BP measurement. According to these recommendations,5 normal ABP should be diagnosed if both awake and sleep BP are lower than 135/85 mmHg and 120/70 mmHg, respectively. On the other hand, in June 2003, the ESH–European Society of Cardiology (ESC) guidelines for the management of arterial hypertension were published. According to these guidelines,6 normal 24-h ABP is defined as lower than 125/80 mmHg. Table 1 summarizes these definitions. Are these two cutoffs, namely, the 24-h and the awake/sleep definitions, compatible? According to which of these cutoffs are more subjects classified as hypertensive? Our aim was to investigate the compatibility of these two recently proposed ABP cutoffs in real-life clinical practice.
Methods We analysed 1495 consecutive ambulatory blood pressure monitoring (ABPM) sessions of nonpregnant adults, performed during a 4-year period. Subjects were referred to our laboratory by their primary
Normal ambulatory BP definitions IZ Ben-Dov et al 566
Table 1 Summary of the recent ESH publications regarding normal ABP ESH/ESC guidelines6
ESH recommendations5
24-h BP o125/80 mmHg
Awake BP o135/85 mmHg, and sleep BP o120/70 mmHg J Hypertens 2003;21:821–848
J Hypertens 2003;21:1011–1053
ESH, European Society of Hypertension; ESC, European Society of Cardiology; BP, blood pressure.
Table 2 Characteristics of the study population Age (years) Gender (female:male) BMI (kg/m2) Clinic BP (mmHg) Diabetes mellitus (%) Treated hypertension (%)
58716 56:44 2774 151722/84713 11 67
BMI, body mass index; BP, blood pressure.
physicians for conventional clinical indications, such as evaluation of resistant hypertension, treatment effect, and suspected isolated clinic hypertension. The subjects’ characteristics are displayed in Table 2. Patients undergoing a repeat measurement were excluded. Twenty-four-hour ABPM was performed with Spacelab 90207 (Redmond, WA, USA), as previously described.7 The monitor was mounted on the nondominant arm between 0800 and 1000, and removed 24 h later. A mercury sphygmomanometer was initially attached to the monitor through a Y-connector to ensure agreement between the two modes of measurement; ABPM was not performed when disagreement exceeded 5 mmHg. These initial measurements, taken by a trained technician, were considered as the subject’s clinic BP.8 BP was measured every 20 min during the day, and every 30 min at night. Sleep, including afternoon naps, was recorded in a diary. Measurements recorded during sleep, including afternoon naps (as logged in a diary) were analysed as sleep BP. Night-time arousals were noted, and analysed along with awake BP. Subjects were classified as having normal 24-h BP if the corresponding value was o125/80 mmHg. Likewise, subjects were classified as having normal awake– sleep BP if their awake BP was o135/85 mmHg and their sleep BP was o120/70 mmHg. Differences between subgroups were analysed using ANOVA (with Bonferroni’s corrections) or w2, as appropriate. Compatibility between the cutoffs in classification of subjects as hypertensive or normotensive was assessed using McNemar change test and inter-rater agreement (kappa) test. Two-sided P-values o0.05 were considered significant.
Results Concordance between the two cutoffs was found in 93% of the subjects. As shown in Figure 1, among Journal of Human Hypertension
the 7% discordant subjects, 4.5% were defined as hypertensive applying the 24-h ABP definition, but not awake–sleep ABP definition, whereas 2.5% were hypertensive according to the awake–sleep ABP definition but not 24-h ABP definition. Thus, significantly more subjects were classified as hypertensive according to the 24-h classification (Po0.005). Concordance between the two distinct cutoffs was also analysed by an inter-rater agreement test; one ‘examiner’ applied the 24-h definition, while the other employed the awake–sleep definition. We found a kappa value of 0.77, which implies a good agreement. Age and gender varied throughout the four subgroups (namely, normotensive by both methods, hypertensive by both methods, and the two discordant subgroups), Table 3, as did the proportion of treated hypertensive subjects. Clinic and awake blood pressures were similar in the discordant subgroups. Sleep diastolic BP (DBP) (but not systolic (SBP)) was lower in the normotensive subgroup solely according to awake–sleep definitions than in those normotensive only according to the 24-h definition. As expected, 24-h SBP was higher in the subgroup with an abnormal 24-h BP but normal awake–sleep BP than in the discordant subgroup (though significance was marginal). The BP dip (both absolute and relative) was significantly deeper in this subgroup, as well.
Discussion ABPM practice is in need of normality definitions. In accordance with this necessity, the ESH has
Figure 1 Classification of subjects according to the 24-h definition (outer doughnut), and according to the awake/asleep definition (inner doughnut). The asterisks mark the discordant subgroups (large asterisks, 4.5%; small asterisks, 2.5%). HTN, hypertensive subjects; normal, normotensive subjects.
Normal ambulatory BP definitions IZ Ben-Dov et al 567
Table 3 Characteristics of the four distinct BP subgroups (the italicized columns indicate the two discordant subgroups)* Awake-sleep BP 24-h BP Number (%) Age (years) Gender (% female) Treated HTN (%) Clinic BP (mmHg) 24-h BP (mmHg) Awake BP (mmHg) Sleep BP (mmHg) BP dip (mmHg) BP dip (%)
Normal HTN
Normal Normal
HTN Normal
HTN HTN
68 (4.5) 57717ab 71 63 140714a/7878d 12771a/7175d 13272a/7575d 11374a/6076d 1875ab/1475d 1474a/1976d
221 (15) 54720a 67 56 132719b/77710d 11776b/6875d 12277/7376d 10677/5975d 1678abc/1475de 1376a/1977d
37 (2.5) 59718ab 62 70 137727ab/78711d 12173ab/7276d 134718a/7778d 11578a/6677e 10711c/1079e 878b/12711e
1169 (78) 59715b 52 70 155721/86713 142712/81710 146712/84710 130714/71710e 17712b/1378de 1178b/1679e
BP, blood pressure; HTN, hypertension; BMI, body mass index. *Comparisons between the four subgroups were conducted by ANOVA or w2 as appropriate. All P-values were o0.0001, except percentage of treated hypertensive subjects (P ¼ 0.001), and absolute DBP dip (Po0.05). Superscript letters denote subgroups with values that do not differ significantly, as determined by Bonferroni’s pairwise comparisons. Generally, a–c are used for SBP, while d–f for DBP values.
recently issued relevant guidelines. It is noteworthy, however, that two different sets of recommendations were published. In this study, we showed that these two recommendations are not similar, although their compatibility is fair. Significantly, more subjects are classified as hypertensive according to the 24-h BP definition. Nevertheless, a subgroup of subjects with normal 24-h BP but abnormal awake–sleep BP also exists. This subgroup has a somewhat lower 24-h SBP, a higher sleep DBP, and a smaller dip. In conclusion, the clinical significance of the discordance between the two cutoff definitions remains to be determined in larger-scaled trials. The ESH is requested to settle the discrepancy between the two recent recommendations.
What is known on this topic Validated and time-honored ambulatory BP normality definitions are absent. The ESH has recently issued two different recommendations. While one paper addresses 24-h BP values, the other considers awake and sleep BP separately What this study adds We examined the concordance between the two recommendations. Among 1495 consecutive referred subjects, 7% discordance was found, and the agreement between definitions was good. However, the clinical significance of the discordance cannot be determined at this stage
References 1 Mancia G et al. Ambulatory blood pressure normality: results from the PAMELA study. J Hypertens 1995; 13: 1377–1390. 2 Sega R et al. Ambulatory and home blood pressure normality in the elderly: data from the PAMELA population. Hypertension 1997; 30: 1–6. 3 Ohkubo T et al. Reference values for 24-hour ambulatory blood pressure monitoring based on a prognostic criterion: the Ohasama Study. Hypertension 1998; 32: 255–259. 4 Chobanian AV et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003; 289: 2560–2572. 5 O’Brien E et al. European Society of Hypertension recommendations for conventional, ambulatory and home blood pressure measurement. J Hypertens 2003; 21: 821–848. 6 2003 European Society of Hypertension–European Society of Cardiology. Guidelines for the management of arterial hypertension. J Hypertens 2003; 21: 1011–1053. 7 Perk G, Mekler J, Bursztyn M. Ambulatory pulse pressure is a relatively sleep independent variable. J Hypertens 2003; 21: 723–728. 8 Owens P, Atkins N, O’Brien E. Diagnosis of white coat hypertension by ambulatory blood pressure monitoring. Hypertension 1999; 34: 267–272.
Journal of Human Hypertension
