(2001) 15, 413 417 2001 Nature Publishing Group All rights reserved 0950-9240/01 $15.00 www.nature.com/jhh ORIGINAL ARTICLE Prognostic significance of blood pressure measured on rising P Gosse, C Cipriano, L Bemurat, D Mas, P Lemétayer, G N Tela and J Clementy Service de Cardiologie-Hypertension artérielle, Hospital Saint André, 1 rue Jean Burguet, 33075 Bordeaux Cedex, France Previous works using ambulatory blood pressure (BP) monitoring demonstrated that independently of the mean level of BP, the variability in BP, or the day night range, could have prognostic significance. We have also found that the value of BP on rising in the morning is strongly correlated with left ventricular mass of hypertensive individuals independently of the 24-h value. In the present study, we sought its predictive value for cardiovascular complications in a cohort of hypertensive patients. The population studied belongs to a cohort of initially untreated hypertensive patients recruited since 1983 and followed for more than 5 years. Patients were then treated and followed by their family doctor. At entry, all patients were equipped with a device to measure ambulatory BP. They were requested to trigger a measurement manually on rising in the morning (arising BP). The data on their outcome were collected by a physician unaware of the initial state of the patients. A total of 256 patients have been followed up for 5 years or more, 19 were lost to follow-up. The mean follow-up period was 84 29 months. Cardiovascular complications were recorded in 23 individuals. The arising systolic BP (SBP) was significantly higher in the group who presented a complication. In a stepwise discriminant analysis including age, office, fitting, arising and 24-h average SBPs only age and arising SBP entered the equation. In conclusion, the single BP value measured by an ambulatory device on rising in the morning seems more discriminant of future cardiovascular events than the value of BP measured on fitting the device or the average of three measurements taken under standardised conditions in the hospital or office. (2001) 15, 413 417 Keywords: blood pressure on rising; cardiovascular events; ambulatory blood pressure monitoring Introduction High blood pressure (BP) as measured in the hospital or office is accompanied by an increased risk of cardiovascular complications. In view of the wide variability in BP, recommendations on the diagnosis and the treatment of high BP 1 insist on repetition of measurements under standardised conditions. Over recent years, ambulatory measurement of BP has seen considerable development as it enables repeated measurements over a 24-h period during the normal living conditions of the patient. Measurements can also be made in physiological conditions, such as during sleep, which have been inaccessible by conventional methods. The nocturnal value of BP is thought to have particular significance. 2,3 We have been interested in the significance of the value of the BP on rising in the morning, and have found that this value is strongly correlated with left ventricular Correspondence: Dr Philippe Gosse, Service de Cardiologie Hypertension artérielle, Hôpital St André, 1 rue Jean Burguet, 33075 Bordeaux Cedex, France E-mail: philippe.gosse chu-bordeaux.fr Received 27 September 2000; revised and accepted 11 January 2001 mass of hypertensive individuals independently of the 24-h value. 4 In the present study, we sought its predictive value for cardiovascular complications in a cohort of hypertensive patients followed up for more than 5 years. Our hypothesis is that this daily sharp elevation in BP may influence target organ damage and the incidence of cardiovascular events. Methods Population The population studied belongs to a cohort of hypertensive patients recruited since 1983 on the following criteria: office BP 140/90 mm Hg on at least two occasions, having never presented cardiovascular complications and devoid of associated pathology affecting prognosis, patients assigned to our department before administration of any antihypertensive treatment and patients for whom we had good quality echocardiographic determinations of left ventricular mass (LVM). Diabetes mellitus type 1 not type 2 was an exclusion criteria. BP was measured under standardised conditions (average of three measurements with a mercury sphygmomanometer after 10 min of rest with the patient supine,
414 diastolic pressure measured in phase V) by the same physician (PG). All patients were equipped with a device for ambulatory measurement of BP. Antihypertensive treatment was administered to most patients, which was supervised by their family doctors. The patients were seen in our department at different times as a function of their needs, but they were contacted by telephone at regular intervals to check on their outcome. The last contact was in the first half of 1999, and we included in the present study patients who had been followed up for 5 years or more. In the event of a cardiovascular complication, details were obtained from the family doctor or specialist. Complications included: angina confirmed by coronarography, myocardial infarction, cardiac insufficiency, cerebrovascular accident, renal insufficiency requiring dialysis, symptomatic arteritis of lower limbs confirmed at least by echo- Doppler examination. The data were collected by a cardiologist unaware of the initial state of the patients (CC). Echocardiographic measurement of LVM In all cases, LVM was measured from 2D guided M mode tracings of the left ventricle from the left parasternal view. The tracings were recorded on a high quality chart running at 100 mm/sec, and were read according to the PENN convention by echographists who were unaware of the clinical findings. LVM was calculated from Devereux s formula 5 and indexed to height to the power of 2.7. 6 Ambulatory measurement of BP All patients benefited from 24-h recordings of BP during their normal routine. We employed two different validated 7,8 auscultatory devices; a Spacelabs 5200 up to 1986 and a Diasys 200 (Novacor, Rueil- Malmaison, France) thereafter. At the time of fitting the device, BP was simultaneously measured by the physician or the nurse and by the device (fitting BP) on opposite arms. If the difference between the two values of systolic pressure exceeded 20 mm Hg, the patient was not fitted with the device. Measurements were programmed every 15 min during the day and every 15 min or 30 min during the night. All patients were requested to trigger a measurement manually immediately on rising in the morning (arising BP). We only retained patients in this study who had good quality recordings (at least 80% of the programmed measurements) with values triggered on rising in the morning. Daytime BP was calculated as the average BP between getting up in the morning and lying down at night. Statistics SPSS software was employed for the statistical analysis. Groups were compared for the quantitative variables by analysis of variance with Neumann Keuls test when the analysis of variance was significant (P 0.05), and for the qualitative variables by the 2 test. A stepwise discriminant analysis between the two groups with and without complications was performed with age, office, fitting, arising and 24-h average systolic blood pressures. Results Our cohort included 741 hypertensive patients up to 1 June 1999. We obtained good quality ambulatory measurements of BP over 24 h with measurements triggered on rising in 413 patients of which 256 have been followed up for 5 years or more. We were unable to contact 19 patients at the end-point of the study and they were considered lost to follow-up. There was no significant difference between this group and the rest of the studied population (Table 1) and the 237 patients followed were not significantly different from the whole cohort. The mean follow-up period of the population was 84 ± 29 months. Cardiovascular complications were recorded in 23 individuals: coronary disease in nine cases (three myocardial infarction, six with coronary angioplasty or bypass), cerebral vascular accident in five cases, arteritis of lower limbs in five cases, sudden death in two cases, renal insufficiency in one case and cardiac insufficiency in one case. Table 2 compares the patients with and without complications. It can be seen that the arising systolic BP (SBP) differed significantly between the two groups. The other measures of SBP (office, mean daytime, night time and 24-h) were also significantly higher in the group with complications. There was no significant difference in diastolic BP between the two groups. Table 3 describes the population according to quartiles in arising SBP. Higher arising SBP is associated with higher values of office, average 24- h, day and night SBP, older age, higher LVM and a Table 1 Comparison of patients followed up and those lost to follow-up Followed Lost to P up follow-up No. 237 19 M/F 160/77 15/4 NS Age (years) 50 ± 12 45 ± 13 NS Office SBP (mm Hg) 160 ± 19 155 ± 14 NS Office DBP (mm Hg) 98 ± 10 96 ± 12 NS 24-h SBP (mm Hg) 134 ± 16 134 ± 13 NS 24-h DBP (mm Hg) 88 ± 10 86 ± 11 NS Arising SBP (mm Hg) 139 ± 23 135 ± 21 NS Arising DBP (mm Hg) 95 ± 15 88 ± 14 NS Arising HR (bpm) 81 ± 15 84 ± 19 NS Weight (kg) 73 ± 14 77 ± 15 NS Smokers (%) 23% 37% NS Hypercholesteroaemia (%) 14% 21% NS LVM/H 2.7 51 ± 18 49 ± 20 NS M, males; F, females; SBP, systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate; LVM, left ventricular mass; H 2.7, height to the power 2.7.
Table 2 Comparison of patients with and without cardiovascular complications Without With P complications complications No. 214 23 M/F 140/74 20/3 0.04 Age (years) 49 ± 12 57 ± 11 0.002 Office SBP (mm Hg) 159 ± 18 169 ± 17 0.008 Office DBP (mm Hg) 98 ± 10 100 ± 9 NS 24-h SBP (mm Hg) 133 ± 16 143 ± 14 0.001 24-h DBP (mm Hg) 87 ± 10 91 ± 9 NS Day SBP (mm Hg) 138 ± 16 149 ± 15 0.002 Day DBP (mm Hg) 92 ± 11 96 ± 11 NS Night SBP (mm Hg) 121 ± 17 129 ± 14 0.03 Night DBP (mm Hg) 78 ± 12 80 ± 10 NS Arising SBP (mm Hg) 137 ± 22 156 ± 26 0.001 Arising DBP (mm Hg) 95 ± 15 100 ± 15 NS Arising HR (bpm) 81 ± 15 83 ± 20 NS SBP fitting (mm Hg) 152 ± 20 160 ± 24 NS DBP fitting (mm Hg) 100 ± 13 104 ± 18 NS Weight (kg) 73 ± 14 73 ± 10 NS Smokers (%) 22% 30% NS Hypercholesterolaemia (%) 13% 22% NS Diabetes mellitus (%) 6% 9% NS LVM/H 2.7 53 ± 15 63 ± 13 NS M, males; F, females; SBP, systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate; LVM, left ventricular mass; H 2.7, height to the power 2.7. Table 3 Main characteristics of patients according to quartiles in arising systolic blood pressure 1st 2nd 3rd 4th P 98 121 122 136 137 152 153 231 M/F 32/27 37/23 43/15 48/12 0.05 Age 48 ± 12 47 ± 12 50 ± 11 55 ± 11 0.001 Weight 71 ± 14 72 ± 14 73 ± 12 75 ± 14 NS Office SBP 149 ± 17 155 ± 13 158 ± 15 176 ± 17 0.001 24-h SBP 119 ± 11 129 ± 11 135 ± 9 150 ± 15 0.001 Day SBP 124 ± 10 135 ± 12 141 ± 10 156 ± 15 0.001 Night SBP 108 ± 13 117 ± 10 123 ± 14 138 ± 18 0.001 LVM/H 2.7 43 ± 18 51 ± 15 51 ± 16 61 ± 19 0.001 CV events 3 4 2 14 0.01 M, males; F, females; SBP, systolic blood pressure; LVM, left ventricular mass; H 2.7, height to the power 2.7. higher proportion of males. The incidence of cardiovascular events is significantly increased in the fourth quartile with arising SBP above 152 mm Hg. We entered in the discriminant analysis age, office, fitting 24-h and arising SBP. The only significant variables emerging from the analysis were age and arising SBP allowing 71% of cases to be adequately classified. Fourteen patients died, but a cardiovascular cause of death was only formally established in five cases. Only the arising SBP was significantly higher in the group who died of cardiovascular disease as compared to the group alive or died or noncardiovascular causes (175 ± 23 mm Hg vs 137 ± 23 and 152 ± 23 respectively, P = 0.001). Discussion Our study shows the interest of the measurement of BP on arising in the morning for prediction of cardiovascular complications in the hypertensive patient. This single value measured by an ambulatory device had better discriminant value than the BP measured on fitting the device or the average of three measurements taken under standardised conditions in hospital or office. It also appeared to have higher predictive power than the average BP during 24 h representing 84 96 measurements. However, in view of the limited number of complications in our cohort, the respective predictive values of arising and mean 24-h BP should be interpreted with due caution. It is certainly premature to conclude that this single measurement performed when the patient arose in the morning gives more information on future risk for cardiovascular event than 24-h average BP but this preliminary study underlines the interest of measuring BP at this very peculiar moment. Another weakness of our study is that we have no systematic evaluation of ambulatory and rising BP on treatment as most of these patients were treated after the initial evaluation. The risk of future event may be more closely related to BP on treatment The interest of the ambulatory measurement of BP is two-fold: since measurements can be taken at frequent intervals (15 min in the present study) it can provide a more reliable baseline than the mean of a few measurements taken during consultation and on which the decision to treat is usually made. The other advantage is that it can assess variability of BP in physiological conditions. For example, several studies have shown the significance of nocturnal variations (dippers vs non-dippers). Rising in the morning is a physiological condition giving rise to a sharp elevation in BP. The average elevation of BP was found to be 12 mm Hg for the systolic and 10 mm Hg for the diastolic when measured on arising, as in the present study, in a population of 233 normal subjects. 9 A much larger elevation is found in hypertensive patients; 4 on average, 18 mm Hg for systolic and 16 mm Hg for diastolic BP with extreme values of around 100 mm Hg for SBP (illustrated in Figure 1). This elevation of BP on rising is of parti- Figure 1 Ambulatory values of BP in a hypertensive patient exhibiting marked elevation of BP and HR on rising. 415
416 cular significance as it is a daily occurrence and eventually may have an influence on the cardiovascular system. For example, it has been shown in animals that transitory elevations in BP (6 h per day) can induce LV hypertrophy (LVH). 10 The elevation in BP on rising may last several hours, as BP tends to be higher in the morning. 11 We showed in a population of 181 untreated hypertensive patients that the value of BP on rising is as well correlated with LVM of hypertensives as the mean 24-h value. 4 Furthermore, multivariate analysis showed that its contribution was independent of the 24-h value. This has been corroborated by another study 12 showing a strong correlation between the elevation of the BP on rising and LVM. The elevation of BP on rising in the morning stems essentially from the standing up action rather than from the transition from sleep to wakefulness. 9,13 The elevation of BP on rising is accompanied by a sharp rise in plasma noradrenaline. 14 Although our findings do not provide any information on the link between a single measure of BP and cardiovascular complications, several hypotheses can be put forward: the mechanical effect of this daily elevation of BP can affect target organs such as the heart, as previously found, but probably also on blood vessels. This would be consistent with the observation of a higher incidence of cardiovascular complications between 7 am and 11 am in the morning. 15,16 Another possibility is that this value reflects the activity of the sympathetic system where hyperactivity is known to increase the risk of cardiovascular events. 17,18 However, in this case, the link with cardiovascular complications should be at least as strong with heart rate as with arising BP, which we did not observe in our population. Our results may seem at variance with those of Staessen 19 in older patients with systolic hypertension. They found that a 1 mm Hg per hour steeper increase in the morning SBP was associated with an 8% lower incidence of cardiovascular end points. The morning increase in BP was computed by fitting a regression line through each patient s BP readings between 4 am and 10 am. As this relationship was adjusted for the daytime BP, a steeper increase in the morning BP may correspond to a lower night time BP which was shown to be associated with lower risk. The significance of the single BP measured on rising may be different as it is linked to BP reactivity on rising but also to daytime BP and even to night BP in non-dipper patients who keep elevated BP when getting up in the morning. However the link to night BP in non-dipper patients is unlikely to explain our findings as in patients with complications there is an average 27 mm Hg increase in BP when getting up. The major finding in our study is that this single measurement appears to predict cardiovascular events as strongly as the average of many measurement performed during daytime or even 24 h. In conclusion, we found a strong relation between arising BP measured by an ambulatory device and the occurrence of cardiovascular complications in a population of previously untreated hypertensive patients. This may have considerable significance in assessing the risk of complications in the hypertensive as well as in the decision to treat, but will need to be confirmed in larger series of patients. References 1 The sixth report of the Joint National Committee on prevention, detection, evaluation and treatment of high blood pressure. Arch Intern Med 1997; 157: 2413 2445. 2 O Brien E, Sheridan J, O Malley K. Dippers and non dippers. Lancet 1988; 2: 397. 3 Ohkubo T et al. Relation between nocturnal decline in blood pressure and mortality. The Ohasama Study. Am J Hypertens 1997; 10: 1201 1207. 4 Gosse P, Ansoborlo P, Lemetayer P, Clementy J. Left ventricular mass is better correlated with arising blood pressure than with office or occasional blood pressure. Am J Hypertens 1997; 10: 505 510. 5 Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in man: anatomic validation of the method. Circulation 1977; 5: 613 618. 6 De Simone G, et al. Effects of growth on variability of left ventricular mass: assessment of allometric signals in adults and children and their capacity to predict cardiovascular risk. J Am Coll Cardiol 1995; 25: 1056 1062. 7 Graettinger WR, Lipson HL, Cheung DG, Weber MA. Validation of portable non invasive blood pressure monitoring devices: comparisons with intra-arterial and sphygmomanometer measurements. Am Heart J 1998; 116: 1155 1160. 8 O Brien E, Mee F, Atkins N, O Malley K. Accuracy of the Novacor DIASYS 200 determined by the British Hypertension Society protocol. J Hypertens 1991; 9: 569 570. 9 Gosse P, Lamaison C, Roudaut R, Dallocchio M. Ambulatory blood pressure monitoring, values in normotensive patients and suggestions for interpretation. Therapie 1991; 46: 305 309. 10 Julius S, Brant D, Krause L, Buda A. Neurogenic pressor episodes fail to cause hypertension but do induce cardiac hypertrophy. Hypertension 1989; 13: 422 429. 11 Pickering TG. Diurnal, ultradian and seasonal rhythm in blood pressure. In: Pickering TG (ed). Ambulatory Monitoring and Blood Pressure Variability. Science Press: London, 1991, pp. 5.1 5.15. 12 Kuwajima I, Mitani K, Miyao M, Suzuki Y, Kuramoto K, Ozawa T. Cardiac implications of the morning surge in blood pressure in elderly hypertensive patients: relation to arising time. Am J Hypertens 1995; 8: 29 33. 13 Khoury AF, Sunderajan P, Kaplan NM. The early morning rise in blood pressure is related mainly to ambulation. Am J Hypertens 1992; 5: 339 344. 14 Dodt C et al. Plasma epinephrine and norepinephrine concentrations of healthy humans associated with nightime sleep and morning arousal. Hypertension 1997; 30: 71 76. 15 Muller J et al. Circadian variation in the frequency of onset of acute myocardial infarction. N Engl J Med 1985; 313: 1315 1322.
16 Muller JE et al, and the MILIS study group. Circadian variation in the frequency of sudden cardiac death. Circulation 1987; 75: 131 138. 17 Julius S. Sympathetic hyperactivity and coronary risk in hypertension. Corcoran lecture. Hypertension 1993; 21: 886 893. 18 Palatini P, Julius S. Heart rate and the cardiovascular risk, J Hypertens 1997; 15: 3 17. 19 Staessen JA et al. Predicting cardiovascular risk using conventional versus ambulatory blood pressure in older patients with systolic hypertension. JAMA 1999; 282: 539 546. 417