Journal of Human Hypertension (2000) 14, 435–440  2000 Macmillan Publishers Ltd All rights reserved 0950-9240/00 $15.00 www.nature.com/jhh ORIGINAL ARTICLE Practice vs ambulatory blood pressure measurement under treatment with ramipril (PLUR Study): a randomised, prospective long-term study to evaluate the benefits of ABPM in patients on antihypertensive treatment

J Schrader, S Lu¨ ders, C Zu¨ chner, M Herbold and G Schrandt St. Josefs-Hospital, Ritterstr. 17, D-49661 Cloppenburg, Germany

The usefulness of ambulatory blood pressure monitor- reason for discontinuation in 55 of these patients (20 ing (ABPM) vs casual blood pressure measurement in from group 1 and 35 from group 2) was that they the physicians practice (PM) for the routine manage- reached the main endpoint (total mortality/morbidity and ment of patients with hypertension concerning total cardio- and cerebrovascular events) of the study. The -in fa (0.037 ؍ mortality and morbidity has been compared in a pro- difference was statistically significant (P spective, randomised, open multicentre study with a 5- vour of group 1. Cardio- and cerebrovascular events in (14 ؍ year follow-up. The study was performed in general also occurred in a lower number of patients (n The difference .(24 ؍ practitioners offices in Germany from 1991 to 1997. A group 1, as compared to group 2 (n A difference in .(0.097 ؍ total of 1298 patients with essential hypertension were however was not significant (P included. Cardio- and cerebrovascular events, total a direct drug-saving effect could not be observed mortality/morbidity and drug-saving effects in hyperten- between the two groups but 22% of the initially screened sives treated according to two different methods of patients were detected with the ABPM to have white blood pressure measurement were the primary and sec- coat hypertension and thus these patients did not ondary endpoints of the study. A total of 239 patients receive antihypertensive treatment. and 208 from group 2 Journal of Human Hypertension (2000) 14, 435–440 ,(651 ؍ from group 1 (ABPM, n prematurely discontinued the study. The (647 ؍ PM, n)

Keywords: ambulatory blood pressure monitoring; ramipril; PLUR; office blood pressure; reverse white coat hypertension

Introduction number of blood pressure values measured during patients normal daily activity, which results in an In recent years, ambulatory blood pressure (ABPM) improved evaluation of the patients blood pressure has been increasingly used in hospitals and in the under stresses of everyday life.4,5 physician’s office for diagnosis, clinical trials, and Furthermore the potency and duration of action in particular for monitoring treatment of hyperten- of an antihypertensive agent can be recorded more sion in selected patients. This method is considered effectively over the entire dosage interval.6 Thereby, to be a better predictor of cardiovascular risk in the 1,2 unnecessary intensified antihypertensive treatment individual patient. Moreover this procedure helps can be avoided in patients who have higher casual to better target patients who will benefit from anti- in-office measured blood pressure than with ABPM. hypertensive drug treatment, and at the same time Alternatively, it offers the possibility to intensify avoiding unnecessary antihypertensive treatment in 3 antihypertensive therapy in those patients who have patients with white coat hypertension. ABPM also higher ABPM values as compared to blood pressure offers advantages for monitoring antihypertensive levels measured in the physicians practice. treatment. The possible superiority of ABPM in To date little information exists in the literature patients undergoing treatment is based on a large concerning the prognostic benefits of ABPM. A number of studies have indicated that ABPM as compared to casual office blood pressure measure- Correspondence: Prof Dr Joachim Schrader, St.-Josefs-Hospital, Ritterstr. 17, D-49661 Cloppenburg, Germany ment shows a better correlation to hypertensive tar- E-mail: slclpȰt-online.de get organ damage and to cardiovascular events.7–13 Received 13 September 1999; revised 29 February 2000; accepted Nevertheless, there is a lack of prospective, long- 30 March 2000 term studies, which demonstrate the value of ABPM Practice vs ABPM under treatment with ramipril J Schrader et al 436 as compared to blood pressure measured occasion- and/or diastolic Ͼ85 mm Hg or average 24 h blood ally in the physician’s office. The important ques- pressure systolic Ͼ130 and/or diastolic Ͼ80 mm Hg, tion, whether consistent use of ABPM will result in according to the guidelines of the German Hyperten- an improved prognosis of hypertensives and to a sion League16 these patients were included in group reduction in the use of antihypertensive medication 1 of the study, and those with lower blood pressure as compared to in-office casual blood pressure values than mentioned above were defined as ‘white measurement, still remains unanswered. coat hypertensives’ and were not included. Therefore the following prospective, randomised, Scheduled visits in both groups were after 2–4 multicentre study was performed to answer the weeks, 3, 6, 9, 12 months and 2, 3, 4, 5 years after question, whether hypertensives managed on the inclusion. ABPM in group 1 was performed once a basis of ABPM have a better prognosis and require a year or when change of treatment seemed to be lesser number of antihypertensive drugs than those necessary according to office blood pressure (Ͼ140 treated according to casual office blood pressure and/or Ͼ90 mm Hg) or clinical reasons. Office blood measurement. pressure was documented at each visit in both groups. Materials and methods In group 1 office blood pressure was documented, but during the entire study period all required This prospective, open, randomised, multicentre changes in antihypertensive medication were car- study was approved by the Ethics Committee of the ried out on the basis of ABPM; ie, the antihyperten- University of Go¨ttingen, Germany. The study was sive treatment was to be intensified if the average conducted according to the Declaration of Helsinki.14 day blood pressure exceeded 135 and/or 85 mm Hg, There were 50 general practitioners who participated or the average 24-h blood pressure was above 130 in this trial, and an independent Clinical Event Com- and/or 80 mm Hg. In group 2, the decision to adjust mittee was used for evaluation of events. A total of the dosage of the antihypertensive medication was 1500 patients were planned to be included in the exclusively taken according to the blood pressure study. According to the definition of the Deutsche values measured in the physician’s practice. The Hochdruckliga (German High Blood Pressure League) treatment was to be intensified if office blood press- the patient must present hypertension requiring treat- ure values were above 140 and/or 90 mm Hg (mean ment (after 5 min sitting), clinic blood pressure (Riva- of three measurements). In patients of this group Rocci method) Ͼ140 and/or Ͼ90 mm Hg in three ABPM was not performed. The patients were measurements on two different days. Patients had to included in this study from 1991 to 1993, and moni- be between 35 and 65 years old. tored for an average period of 4–7 years (maximum Exclusion criteria were contraindication for rami- 6 years, minimum 4 years). The study was termin- pril treatment, manifest cardiac insufficiency ated in July 1997. (NYHY III–IV), higher-grade aortic or mitral sten- To ensure a comparable antihypertensive treat- osis, pregnant or lactating women, women of child- ment in both groups, the ACE inhibitor ramipril was bearing potential, patients with a progressive, poten- used as baseline drug starting with 1.25 mg o.d. in tially fatal disease (life expectancy Ͻ1 year), both the groups. The dosage was titrated up to 5 mg patients with alcohol or drug abuse, known hyper- o.d. if necessary for normalising blood pressure dur- sensitivity against angiotensin-converting enzyme ing the first 4–6 weeks. In the event of undesirable (ACE)-inhibitor, history of Quincke’s oedema, reactions or the need for a combination treatment patients who received other investigational sub- according to the guidelines of the German Hyperten- stances within 30 days or five half-lives of this drug sion League, a calcium antagonist (felodipine 5 mg (the longer will apply) before entry to this protocol, o.d. or nifedipine ret 10 mg t.d.), or a diuretic patients who have been included in a previous (hydrochlorothiazide 12.5 mg o.d.), or a beta-blocker ongoing study, legally incapable patients or patients (metoprolol 100 mg o.d.) could be administered unable to give consent. alternatively or added to ramipril on physicians Patients receiving antihypertensive medication preference. had a washout period of 4 weeks (when blood press- ure increased to Ͼ165/95 mm Hg patients could be randomised before the end of these 4 weeks). The Definition of events patients were then randomised into two groups: Cardiovascular events were defined as all (fatal and group 1 ‘ABPM’ and group 2 casual office blood non-fatal) myocardial infarctions, all (fatal and non- pressure measurement (‘PM’). The patients belong- fatal) strokes, and all other cardiovascular deaths. ing to group 2 were directly included into the study, The diagnosis of these events required electrocardio- and in the patients of group 1 an ABPM was perfor- grams or hospital records or necropsy reports. med in addition by using Space Labs equipment consisting of validated oscillometric 90207 moni- 15 tors. The ambulatory readings were programmed Drug score at 15-min intervals from 6 am to 10 pm and at 30- min intervals from 10 pm to 6 am according to the For comparison of antihypertensive therapy all recommendations of the German Hypertension medication was estimated by a drug-score which League.16 If a hypertensive blood pressure profile was based on the daily recommended dosage of each was also documented during ABPM (average day drug (ϭ 1 pt, half of recommended dosage ϭ 0.5 pts, blood pressure from 6 am to 10 pm) systolic Ͼ135 double-dosage ϭ 2 pts).

Journal of Human Hypertension Practice vs ABPM under treatment with ramipril J Schrader et al 437 Table 1 Patient’s characteristics

Total Group 1 (ABPM) Group 2 (PM)

Patients Patients total (n) 1298 651 647 Male (n) 593 289 304 Female (n) 705 362 343 Anamnesis (n) Diabetes mellitus 158 72 (11.1%) 86 (13.3%) (NS) Hyperlipoproteinemiol 527 263 (40.4%) 264 (40.8%) (NS) Nicotine 239 125 (19.2%) 114 (17.6%) (NS) Alcohol 296 152 (23.3%) 144 (22.3%) (NS) Pos. family history (cardiovascular disease) (n) 358 169 (26.0%) 189 (29.2%) (NS) Previous vascular diseases 91 39 (6.0%) 52 (8%) (NS) Laboratory values Creatinine (mg/dl) 0.94 ± 0.29 0.93 ± 0.22 (NS) Glucose (mg/dl) 101.4 ± 32.9 102.9 ± 32.9 (NS) Cholesterol (mg/dl) 231.8 ± 44.4 231.1 ± 41.4 (NS) LDL (mg/dl) 156.8 ± 47.3 160.0 ± 40.2 (NS) HDL (mg/dl) 50.1 ± 23.4 50.6 ± 21.9 (NS) Triglycerides (mg/dl) 175.0 ± 116.2 182.9 ± 108.2 (NS)

NS, Not significant. LDL, low-density lipoprotein; HDL, high-density lipoprotein.

Statistical methods Of the total 1298 patients included and treated in the study, 705 were female, and 593 male, the aver- Statistical analysis was completed on an intention- age age was 54.3 ± 9.4 years. Groups 1 and 2 were to-treat basis for all treated patients (with exception comparable, and no differences in age, sex, body of patients with white coat hypertension). For evalu- mass index, nicotine and alcohol consumption, con- ation of the demographic data with respect to the comitant use of statins or aspirin, family history of homogeneity of the two groups, the Student’s t-test cardiovascular diseases, concomitant diseases was used for age, body mass index, and baseline including previous cardiovascular events, average blood pressure, and a chi-squared test for sex. casual office blood pressure values and biochemical The size of the study population, the concomitant variables could be observed. The patients were fol- diseases, and reasons for premature study discon- lowed up for an average of 4.7 years. Concomitant tinuation were evaluated on the basis of frequency diseases which prevailed in the two groups are sum- comparisons (chi-squared test, or if numbers of marised in Table 1. cases were small, with Fischer’s exact test). A correction of multiple t-tests was used for the comparison of average blood pressure values at indi- Patients discontinuing the study vidual measured time points. The Student’s t-test From the total of 1298 patients who entered the was used for comparing the mean values of bio- study, 851 patients completed the study according chemical variables between the groups. The Wil- to the protocol. The number of patients discontinu- coxon-Mann-Whitney test was used for testing for ing the study in each group over the 5 years of fol- differences in the dosage of the baseline drug rami- low-up is shown in Figure 1. pril between the two groups at the conclusion of Although a higher number of patients in group 1, the study.

Results Patient population A total of 1506 patients were screened for the study. Of these 1298 were finally enrolled to the two groups. Group 1, ABPM (n ϭ 651); group 2, casual practice blood pressure measurement (PM) (n ϭ 647). For inclusion of 651 patients in group 1, ABPM was performed in 859 patients. In 189 (22%) of these patients, white coat hypertension was detected, with increased casual blood pressure values (Ͼ140/90 mm Hg) measured in the physician’s practice, but normotensive 24-h blood pressure values (Ͻ130/80 mm Hg) with ABPM. These 189 patients were not enrolled to the long- term-study. In 19 patients ABPM was either not car- ried out correctly or the patients refused to repeat Figure 1 Number of patients continuing the study in the two ABPM, and therefore were not enrolled to the study. groups over the 5 years of follow-up period.

Journal of Human Hypertension Practice vs ABPM under treatment with ramipril J Schrader et al 438 as compared to group 2, discontinued the study the difference was however not significant. Of the total 447 patients prematurely dropping out of the study, 239 were in group 1 and 208 in group 2. A total of 308 patients out of the 447 discontinued the study due to protocol violation. There were 174 patients from group 1 and 134 from group 2. The difference between the groups was significant (P ϭ 0.0108). The reasons for discontinuation were proto- col infringements, non-compliance of the patient to antihypertensive medication, patient shifting resi- dence (n ϭ 62), centres discontinuing their partici- pation in the study (five centres ended their co-oper- ation for personal reasons (two practitioners died (n ϭ ϭ 52 patients), three collegues retired (n 81 Figure 2 Mean values of casual systolic and diastolic in-office patients)). By telephone contact it was possible to blood pressure in both groups during the study period. get the main information of 113 patients. They were still alive at least to the end of the study period and none of them suffered from a major cardiovascular mm Hg at baseline to 130.4 ± 11.9/79.6 ± 8.4 mm Hg event. The reason for the higher number of drop- at the end of the study. The baseline 24 h blood outs in group 1 was because of repeat ABPM, which pressure values, as well as the values achieved dur- was not tolerated by some patients. The number of ing the study period were as expected, lower with lost for follow-up patients with no further infor- ABPM than the values determined by casual office mation is 195 (15%). blood pressure measurement. A total of 53.4% of Forty-five patients in group 1 and 39 in group 2 patients achieved target blood pressure in terms of discontinued antihypertensive treatment due to ABPM. drug-related adverse effects, these being mainly No difference in the mean values of biochemical cough and allergic reactions. A significant difference variables such as serum-cholesterol, serum-triglycer- between the groups could not be observed. ides and renal function could be detected between Reaching the combined study endpoint (total the two groups during the study period. mortality/morbidity and cardiovascular events) was the reason for 55 patients (20 in group 1 vs 35 in Consumption of antihypertensive drugs group 2) to discontinue the study. This difference was significant (P ϭ 0.0366) in favour of group 1. A total of 68.7% of patients in group 1 and 68.3% Also a lower number of cardio- and cerebrovascu- of those in group 2 required ramipril as monotreat- lar events were observed in group 1 in comparison ment. A combination of two antihypertensive agents to group 2 (14 vs 24 events). Due to the small num- was needed in 23.8% and 25.1% of the patients of ber of these events in total, the difference was sig- group 1 and 2 respectively, and only 7.5% of the nificant on 10% level only. patients in group 1 and 6.6% in group 2 had to be The cardio- and cerebravascular mortality was treated with a combination of more than two antihy- comparable with six cases each in the two groups. pertensive drugs. Non-vascular endpoints did not differ significantly Also no significant difference could be observed in the two groups, nor did total mortality (11 vs 14). in the number of ramipril dose-titration steps required in the two groups. Only in group 2 (PM) was there a slightly higher number of prescriptions Blood pressure and heart rate of the 5 mg ramipril dosage, and in group 1 (ABPM), The changes in mean values of systolic and diastolic a slightly higher number of prescriptions for the blood pressure measured in the physician’s office lower dose of 1.25 mg ramipril. This indicates a during the course of the treatment period in the trend to a lower ramipril dose requirement in group group 1 (ABPM) and group 2 (PM) are depicted in 1, as compared to group 2. Figure 2. Although the baseline systolic blood press- When the total antihypertensive drug prescrip- ure levels at the start of the study (group 1: 165.9 ± tions were evaluated with the help of a drug scoring 17.3/100 ± 10.1 mm Hg, group 2: 167.6 ± 16.9/99.5 system, no difference could be found in the number ± 10 mm Hg) were 1.7 mm Hg higher in group 2, the or dosage of drugs prescribed in the two groups. lowered blood pressure values achieved by active Therefore it could be concluded that there was no treatment during the entire study period are compa- noticeable direct antihypertensive drug-saving effect rable in both the groups, and show no difference at in both of the groups. In order to determine the rea- any measured time-point over the average study per- son for the absence of differences in the drug-saving iod of 4.7 years. In group 2 59.7% of patients effect the values of blood pressure achieved using achieved target blood pressure. Heart rate was com- the in-office blood pressure measurement and parable as well (76.3 ± 9.8 vs 75.7 ± 10.9 per min) ABPM were additionally evaluated. in both groups. Forty-eight percent of the total blood pressure The mean systolic and diastolic blood pressure values measured at any visit were found to be in the measured by ABPM in group 1 declined during the normal range according to each of the two measure- course of treatment from 143.7 ± 13.1/89.3 ± 9.2 ment methods employed. Thus both these modalities

Journal of Human Hypertension Practice vs ABPM under treatment with ramipril J Schrader et al 439 were in good agreement in nearly half of the total on the basis of in-office blood pressure measure- blood pressure values measured in the study. In ment. All these patients have been classified differ- 21.1% of the blood pressure values measured, they ently, if only in-office blood pressure measurement were found to be elevated with both the methods would have been performed. The antihypertensive (ABPM and in-office measurement) employed. A treatment in these patients was changed according total of 17.7% of the blood pressure measurements to values measured by ABPM, this would also not performed in the office were elevated (Ͼ140/90 have happened if only in-office blood pressure mm Hg), but in the normal range when measured by measurement had been done. Assuming that a simi- ABPM (Ͻ135/85 mm Hg). This was defined as ‘white lar percentage of patients also exists in group 2, the coat hypertension’ and in these patients a drug-sav- better blood pressure control could then be the rea- ing effect was possible. On the other hand 13.2% of son for the prognostic difference observed in favour blood pressure values measured in the physician’s of group 1. office were in the normal range (Ͻ140/90 mm Hg), Moreover it should be noted that in group 1, 22% but elevated (Ͼ135/85 mm Hg) when measured by of screened patients identified as having white coat ABPM. This was defined as ‘reverse white coat hypertension were not included in the study, while hypertension’, and in these patients intensified drug in group 2 probably a similar number of these treatment was necessary, inspite of normotensive in- patients with low risk2,17 were treated. It can be office blood pressure values. The number of patients speculated that the observed differences in main in whom antihypertensive drugs could be reduced endpoints would have been more pronounced, if was therefore slightly higher than the number of patients with white coat hypertension in group 2 patients requiring intensified drug treatment. This were also excluded from the study. Furthermore, probably explains why overall no drug-saving effect various studies have shown that blood pressure was achieved in this study. One can also conclude measured by ABPM correlates better with the occur- that on the basis of only in-office blood pressure rence of hypertensive end-organ damage—in parti- measurement in group 1, 30.9% of the patients cular with left ventricular hypertrophy and micro- (17.7% ‘white coat hypertension’ and 13.2% ‘reverse albuminuria—than with casual blood pressure white coat hypertension’) would have been classi- values measured in the physicians practice.8,11 fied differently. The reason for the superiority of ABPM is prob- ably due to the greater number of blood pressure Discussion readings taken under the actual situation of daily life, which represents a true blood pressure profile This prospective randomised study, is to our know- of the patient. ABPM allows a more accurate selec- ledge the first long-term trial investigating the bene- tion of patients for treatment and detects patients fit of consistent use of ABPM as compared to casual with white coat hypertension, in whom treatment is blood pressure measurement in the physician’s not necessary.3 office, during 5 years of antihypertensive treatment. A further explanation for better prognosis The results show, that in comparison to antihy- observed in group 1 could be that due to consistent pertensive treatment based only on office blood use of ABPM, the patient was closely attended to, pressure measurement, the consistent use of ABPM and cared for more intensively by the patient’s phys- for monitoring treatment resulted in a significantly ician. As a result, patient compliance, detection of improved prognosis, with a reduction in cardio- other risk factors, and treatment may have been vascular morbidity and total mortality. This success improved. Although this could not be recorded by observed in the group with consistent ABPM was statistical analysis, this possibility however cannot accompanied, however, by a significantly higher be ruled out with certainty. number of patients rejecting this procedure due to The observed reduction in morbidity and mor- inconvenience caused. A direct drug-saving effect tality in group 1 in this study is therefore probably could not be observed during long-term treatment due to a combination of a number of factors, such based on ABPM. as improved selection of patients for treatment, more The question that arises is whether the observed intensive care of the patients, resulting in a better reduction in endpoints in group 1 can be explained and reliable control of patients blood pressure, and by better and reliable blood pressure control. Taking improved prognosis. into consideration the in-office blood pressure A drug-saving effect could not be observed in the measurement which was performed in both groups, two groups. A comparable number of patients in no difference in blood pressure behaviour whatso- both the groups received monotherapy (ramipril) or ever in the two groups could be detected. Differ- a combination of two or three antihypertensive ences in blood pressure control by means of in-office drugs. The quality of blood pressure control seems blood pressure measurement cannot therefore be the to have been better in the ABPM group, 30.9% of explanation for the observed difference in prognosis. the patients in group 1, who were ‘falsely’ classified However during evaluation of in-office blood press- on the basis of in-office blood pressure measurement ure values in group 1, as compared to values were ‘correctly’ reclassified and adequately treated. obtained by ABPM, it was found that due to the lim- In order to find possible reasons for the absence its of blood pressure values defined by the German of drug-saving effect of treated patients in the study, Hypertension League, 30.9% of the patients in group the blood pressure values measured by ABPM and 1 had their antihypertensive treatment changed on in the physicians office of group 1 were compared. the basis of ABPM, which would not have occurred According to in-office blood pressure measurement

Journal of Human Hypertension Practice vs ABPM under treatment with ramipril J Schrader et al 440 17.7% of the patients would have received antihy- References pertensive treatment, which was not necessary by using ABPM. On the other hand 13.2% of cases in 1 Pickering TG, Devereux RB. Ambulatory monotoring group 1 had normotensive blood pressure values on of blood pressure as a predictor of cardiovascular risk. the basis of in-office measurement, but elevated Am Heart J 1987; 114: 925–928. blood pressure values using ABPM, and therefore 2 Verdecchia P et al. Ambulatory blood pressure—an received more effective antihypertensive treatment. independent predictor of prognosis in essential hyper- Since the difference between patients requiring tension. Hypertension 1994; 24: 793–801. increased antihypertensive drugs (13.2%) vs 3 Pickering TG et al. How common is white coat hyper- reduction of drugs (17.7%) is rather small, no sig- tension? JAMA 1988; 259: 225–228. 4 Staessen JA et al. For the ambulatory blood pressure nificant drug-saving effect could be found using the monitoring and treatment of hypertension investi- two different methods of blood pressure measure- gators. Antihypertensive treatment based on conven- ment in group 1. However a trend to use lower dos- tional or ambulatory blood pressure measurement. ages of ramipril could be observed in group 1 JAMA 1997; 278: 1065–1072. (ABPM) as compared to group 2 (PM). 5 Schrader J, Schoel G. Benefits of noninvasive ambulat- An important aspect which should be taken into ory blood pressure monitoring during antihypertensive account when discussing the drug-saving effect is therapy. J Ambulat Monitoring 1990; 3: 203–214. that antihypertensive treatment in patients with 6 Schrader J et al. Comparison of the antihypertensive white coat hypertension could be avoided in 22% efficiency of nitrendipine, metroprolol, mepindolol of the initially screened patients. This percentage is and enalapril using ambulatory 24-hour blood press- in accordance with dates from different studies. If ure monitoring. Am J Cardiol 1990; 66: 967–972. 7 Devereux RB et al. Left ventricular hypertrophy in this percentage would be included in the evaluation, patients with hypertension: importance of blood press- a considerable drug-saving effect resulted. Such a ure response to regularly stress. Circulation 1983; 68: drug-saving effect has been shown in a randomised 470–476. control study by Staessen and co-workers.4 8 Verdecchia P et al. Circadian blood pressure changes On the other hand it is important to control patients and left ventricular hypertrophy in essential hyperten- with white coat hypertension carefully because many sion. Circulation 1990; 81: 528–536. of them develop a manifest hypertension even within 9 Perloff D, Sokolow M, Cowan R. The prognostic value a short period.18–20 In conclusion, this prospective, of ambulatory blood pressures. JAMA 1983; 249: randomised study shows that consistent use of ABPM 2792–2798. in hypertensives results in an improved patient prog- 10 Doutheil A et al.Ha¨ufigkeit und Bedeutung einer nosis compared to casual blood pressure measurement na¨chtlichen Hypertonie bei Patienten nach cerebralen in the physician’s office. Insulten. Nieren- und Hochdruckkrankheiten 1992; 21: 492–494. 11 Benhamou PY et al. Early disturbances of ambulatory Acknowledgements blood pressure load in normotensive type I diabetic Participating investigators (listed alphabetically): patient with microalbumin-uria. Diabetes-Care 1992; 15/11: 1614–1619. Al-Ansari, MD (Einbeck); U Bartlau, MD 12 Clement DL, De Buyzere M, Duprez D. Prognostic (Bovenden); V Bartz, MD (Herzberg); CR Behme, MD value of ambulatory blood pressure monitoring. J Hy- (Braunschweig); R Brokate, MD (Walkenried); A pertens 1994; 12: 857–864. Czech, MD (Dassel); E Eberl, MD (No¨rten- 13 Redon J et al. Prognostic value of ambulatory blood Hardenberg); G Eickhoff, MD (Wanfried); U Epp- pressure monitoring in refractory hypertension: a pro- enstein, MD (Echte); S Erhard-Schmelzer MD (Groß- spective study. Hypertension 1998, 31: 712–718. Ellershausen); KW Fraulborn, MD (Northeim); J Gan- 14 The 41st World Medical Assembly. Declaration of Hel- asinski, MD (Kassel), J Gatz, MD (Go¨ttingen); W sinki: recommendations guiding physicians in biom- Gertjegerdes, MD (Apen); W Hamel, MD (Verna); H edical research involving human subjects. Bull Pan Am Health Organ 1990; 24: 606–609. Hartmann, MD (Gieboldehausen); W Heesch, MD 15 O’Brien E, Mee F, Atkins N, O’Malley K. Accuracy Of (Vellmar); D Heidbreder, MD (Go¨ttingen); HU the Space Labs 90207 determined by the British Heynen, MD (Karlstadt); K Kalis, MD (Kalefeld); P Hypertension Society Protocol. J Hypertens 1991; 9: Kathmann, MD (Westerstede); D Kellermann, MD 573–574. (Vellmar); KD Kurzke, MD (Sulingen); C Lang, MD 16 Blood Pressure Measurement Sectio of the Deutsche (Marsberg); M Lindstedt, MD (Bremke); E Lu¨ tzen, Liga zur Beka¨mpfung des hohen Blutdrucks e. V. MD (Westerstede); R Mahrt, MD (Reyershausen); R (German Hypertension League). Statement on ambulat- Metze; PD MD (Frankfurt/Oder); H Mu¨schen, MD ory blood pressure monitoring by the German Hyper- (Waldau); W Pfleiderer, MD (Northeim); HP Rohne, tension League. J Hum Hypertens 1995; 9: 777–779. MD (Ilsede); M Rothe, MD (Lenglern); H Ro¨ver, MD 17 Cavallini MC et al. Is white coat hypertension associa- (Celle); C Scha¨fer, MD (Holzminden), P Schankath, ted with arterial disease or left ventricular hypertro- MD (Wunsdorf), WR Schankath, MD (Wunsdorf); D phy? Hypertension 1995; 26: 413–419. 18 Witte M et al. Entwicklung einer manifesten Hyper- Scheitz, MD (Ochold); A Schober, MD Prof tonie bei Patienten mit Praxishypertonie. Nieren- und (Go¨ttingen); W Schro¨der, MD (Go¨ttingen); C Hochdruckkrankheiten 1996; 25: 214–216. Schro¨ter, MD (Elliehausen); P Segebrecht, MD 19 Bidlingmeyer I et al. Isolated office hypertension: A (Eddelak); C Slotty, MD (Go¨ttingen), JH Struck- prehypertensive state? J Hypertens 1996; 14: 327–332. meyer, MD (Holzminden), R To¨le, MD 20 Middeke M. Ist die Praxishypertonie die fru¨ hste Mani- (Rittmarshausen); P Wand, MD (Go¨ttingen); R festation der Hochdruckkrankheit? Nieren-und Hoch- Ziegler, MD (Wendeburg). druckkrankheiten 1994; 8: 386.

Journal of Human Hypertension