Ambulatory Blood Pressure Monitoring and Postprandial Hypotension in Elderly Patients with Isolated Systolic Hypertension

Home , Jagiellonian University

ORIGINAL ARTICLE Ambulatory blood pressure monitoring and postprandial hypotension in elderly patients with isolated systolic hypertension

T Grodzicki1, M Rajzer2, R Fagard3, ET O’Brien4, L Thijs3, D Clement5, C Davidson6, P Palatini7, G Parati8, J Kocemba1 and JA Staessen3, on behalf of the Systolic Hypertension in Europe (SYST-EUR) Trial Investigators 1Department of Gerontology and Family Medicine, Jagiellonian University, Cracow, Poland; 21st Department of Cardiology, Jagiellonian University, Cracow, Poland; 3Syst-Eur Coordinating Ofﬁce, Universitaire Ziekenhuizon UZ Gasthuisberg, Leuven, Belgium; 4Blood Pressure Unit, Beamont Hospital, Dublin, Ireland; 5Department of Cardiology and Angiology, University Hospital, Gent, Belgium; 6Royal County Hospital, Brighton, UK; 7Policlinico, Padua, Italy; 8Centro di Fisiologia Clinica e Ipertensione, Ospedale Maggiore, Milan, Italy

The present analysis was undertaken to evaluate post- in the 2 h before and after the meal, both SBP and DBP prandial (PP) changes in blood pressure (BP) assessed decreased significantly (respectively −6.6 mm Hg, −5.4 with ambulatory BP monitoring (ABPM) in elderly sub- mm Hg; P Ͻ 0.001). In 67.6% of all patients a decrease jects with isolated systolic hypertension (ISH) on con- in SBP was observed and in 24.1% it exceeded 16 ventional measurement. A total of 530 patients (335 mm Hg. The corresponding values for DBP were 71.3% women and 195 men, aged 60–100 years, median 70 and 24.5% (DBP decreased more than 12 mm Hg). A years) who performed an ABPM during the placebo run- greater fall in DBP was associated with a greater in period of the Syst-Eur trial were included into the decrease in HR (r = 0.20, P Ͻ 0.001), while changes in analysis. The PP changes in BP and heart rate (HR) were SBP and HR were not interrelated. Regression analysis calculated by subtracting the mean systolic BP (SBP), did not identify any significant covariate of PPH. Group diastolic BP (DBP) and HR in the 2 h preceding the main means of PPH could be reproduced without significant meal from the corresponding means covering the 2 h changes in their values, but the within-subject repro- after the meal. The reproducibility of the postprandial ducibility of the PP changes was low. There were no dif- fall in BP and heart rate (PPH) was assessed by con- ferences in PPH according to the place of residence of trasting the first and second ABPM in a subgroup of 147 the patients. In conclusion, the descriptive analysis of patients who performed two ABPM’s during the placebo the meal-induced changes in ABPM in elderly subjects run-in period. The mean SBP and DBP decreased and with ISH showed that in every day circumstances most reached the nadir 2 h after the main meal while HR did of them experience falls in both SBP and DBP within 2 h not change. When PPH was assessed by comparing BP after the meal.

Keywords: postprandial hypotension; isolated systolic hypertension; elderly; ambulatory blood pressure monitoring; SYST- EUR

Introduction suggested as a reason of this phenomenon.1,4 Age- associated cardiovascular changes such as arterial Significant reductions in blood pressure (BP) after stiffness and left ventricular diastolic dysfunction mixed-meals or oral glucose have been observed in may cumulatively produce alterations in BP regulat- elderly healthy, hypertensive or diabetic patients, ory mechanisms that impair an older person’s and in patients with Parkinson’s disease or auto- 5 1–3 ability to adapt to hypotensive stress. Moreover, the nomic insufficiency. Although the underlying blunted baroreceptor reaction in elderly hyper- mechanisms are not well established, the increase tensive patients may be partially responsible for in concentration of vasoactive gastrointestinal pep- profound postprandial hypotention. It has been tides or insulin-related vasodilatation have been demonstrated that patients with a history of falls or syncope present with a more distinct BP decrease after meals.6 Correspondence: Dr Tomasz Grodzicki, Dept of Gerontology and Family Medicine, School of Medicine, Jagiellonian University, Isolated systolic hypertension (ISH) remains an Wielicka 267, 30-663 Cracow, Poland important feature of aging associated with a dimin- Received 3 May 1997; revised and accepted 13 October 1997 ished baroreceptor function and susceptibility to ABPM and postprandial BP in isolated systolic hypertension T Grodzicki et al 162 orthostatic hypotention.7 Nowadays, 24-h ambulatory BP monitoring (ABPM) is used widely in research and everyday clinical practice and remains the method of choice to evaluate fluctuations in BP in hypertensive and normotensive subjects. As to the best of our knowledge the literature describing the influence of meals on the 24-h BP profile is scarce, it seemed interesting to undertake the present analysis to evaluate changes in BP in patients with ISH in relation to the main meal.

Patients and methods The protocol of the multicentre SYST-EUR trial and its side-project on ABPM has been published else- where.8,9 After discontinuation of all anti-hypertensive drugs, the participants (у60 years) first entered a single-blind period on placebo, during which eli- gible patients maintained a sitting systolic BP (SBP) of 160–219 mm Hg and a diastolic BP (DBP) below 95 mm Hg on conventional measurement. Ambulat- ory pressure was recorded non-invasively at inter- vals not greater than 30 min. A total of 578 patients included into the SYST-EUR (on August 1995) had their ABPM measured at least once during the placebo run-in period phase. Of these 48 were excluded from the analysis because their ABPM recordings were incomplete (n = 26), the time of the main meal was missing (n = 18) or BP measurements during 1 h Figure 1 Means of SBP, DBP and heart rate before and after preceding or after meal were missing (n = 4). The main meal. ABPM recordings of the remaining 530 patients were included into analysis. In a subgroup of 147 Results patients ABPM had been repeated during the base- line period on placebo. The ABPM recordings of 530 patients were included The postprandial change in BP and heart rate (HR) into analysis. This group consisted of 335 women was calculated by subtracting the mean SBP, DBP and 195 men, aged 60–100 years (median 70 years), and HR in the 2 h preceding the main meal with mean body mass index of 26.9 ± 4.2 kg/m2 (according to the patients’ diaries) from the corre- (women) and 26.2 ± 3.4 kg/m2 (men). Cardiovascular sponding means from the 2 h after the main meal. complications were present in 27% of the patients. All analyses were performed in the entire group of The main meal was taken between 10.30 am and patients (n = 530) and separately in subjects taking 11.00 pm, but 95% of patients took their main meal their main meal between 11.00 am and 2.00 pm (n between 11.45 am and 7.00 pm. = 296), and between 5.00 pm and 8.00 pm (n = 130). In the entire group SBP and DBP decreased after the main meal reaching a minimum 1–2 h after the main meal. Heart rate was similar before and after Statistical analysis the main meal (Figure 1). When the postprandial change in BP was assessed by comparing the average Database management and statistical analyses were BP in the 2 h before and after the meal both SBP performed with the SAS software (The SAS institute and DBP decreased significantly but HR remained Inc, Cary, NC, USA). Unless otherwise indicated the unchanged (Table 1). The analysis of these changes standard deviation (s.d.) was used to report the according to the time of the main meal did not reveal spread of the results. In order to identify factors that any differences between subjects who had their may contribute to the postprandial BP fall, a sub- main meal in the afternoon or in the evening. The group and single and multiple linear regression analyses were performed. For multiple means comparisons Ducan’s test was used. The reproducibility Table 1 Postprandial changes in systolic (SBP), diastolic (DBP) of the postprandial changes in BP was assessed by blood pressure and heart rate (HR) contrasting the first and second BP monitoring. The repeatability coefficients were calculated as 2 s.d. of Mean ± s.d. Confidence the differences between repeated measurements. To interval allow comparisons between various measurements SBP (mm Hg) −6.6 ± 13.8*** −7.77; −5.43 the repeatibility coefficients were expressed as per- DBP (mm Hg) −5.4 ± 10.9*** −6.33; −4.47 centages of nearly maximal biological variation, ie, Pulse rate (bpm) −0.1 ± 8.7 −0.84; +0.64 four times the s.d. of the averaged duplicate measurements. ***P Ͻ 0.001. ABPM and postprandial BP in isolated systolic hypertension T Grodzicki et al 163 Table 2 Reproducibility of the postprandial changes in blood pressure and pulse rate

Change Consistency Repeatability (mm Hg) (mm Hg) (mm Hg) (%)

SBP −0.5 9.3 (0.6–31.3) 29 (68%) DBP −1.0 7.2 (0.7–28.1) 25 (70%) Heart rate −1.6 5.7 (0.8–23.1) 24 (79%)

Change = mean difference between duplicate recordings taking into account the sign of the difference. Consistency = median difference between duplicate recordings, disregarding the sign of the difference (5th and 95th percentile range in parentheses). Repeatability = twice the standard deviation of the changes between repeated recordings (percent of maximal variation in parentheses).

litus, R-wave amplitude in aVL and the Sokolov + Figure 2 Histogram of the postprandial changes in SBP. index (SV1 RV5), and anti-hypertensive treatment in the 6 months preceeding enrolment in the study. Group means of postprandial changes in SBP, DBP histogram of postprandial fluctuations demonstrates and HR could be reproduced without significant that in 67.6% of the patients a decrease in SBP was changes in their values (Table 2). However, the observed and 24.1% showed a decrease of more than within-subject reproducibility of the postprandial 16 mm Hg (Figure 2). The analysis of the DBP changes was rather low as expressed by the high changes shows that in 71.3% DBP decreased and in repeatability coefficients. The reproducibility of the 24.5% this decrease exceeded 12 mm Hg (Figure 3). postprandial BP and HR changes did not differ Single correlations between changes in BP and HR between patients who took their meal in the after- in the entire group demonstrated that a greater fall noon or evening. The postprandial changes in BP in DBP was associated with a greater decrease in HR and HR were also analysed according to the place (r = 0.20, P Ͻ 0.001). In contrast, changes in SBP of residence (Table 3). Although patients from and HR were not interrelated. Southern Europe (n = 166) showed a higher decrease Subgroup and regression analysis did not identify in SBP (−8.1 ± mm Hg) and DBP (−6.4 ± mm Hg) in any significant covariate of the postprandial changes comparison with patients living in Northern Europe in BP and pulse rate. In both analyses the following (n = 340) (respectively, −5.9 ± mm Hg, −5.1 ± variables were included: age, gender, body mass mm Hg), the differences between these falls did not index, level of BP (and pulse pressure) on conven- achieve statistical significance. tional measurement, the orthostatic change in BP, the nocturnal fall in BP, the 24-h and daytime s.d. of BP (with exclusion of the 2 h before and after the Discussion main meal), the presence of cardiovascular compli- The analysis of ambulatory BP recordings in 530 cations, drinking and smoking habits (discrete and elderly patients with isolated systolic hypertension continuous), log serum activity of gamma- glutamyltranspeptydase, presence of diabetes mel- Table 3 Average postprandial fall in systolic (SBP), diastolic (DBP) blood pressure and heart rate (HR) according to the place of residence

Country n SBP DBP HR (mm Hg) (mm Hg) (bpm)

Finland 142 −4.1 −4.9 −1.1 Netherlands 15 −5.8 −5.9 −1.4 Slovenia 18 −6.4 −4.2 −4.3 Ireland 34 −6.6 −4.1 +3.4 UK 76 −6.9 −6.1 +1.7 Belgium 45 −7.2 −4.5 +1.0 Italy 109 −7.4 −5.2 −0.2 Greece 9 −8.1 −8.7 −3.1 Germany 3 −8.9 −6.6 −2.0 Poland 25 −9.3 −4.7 +0.7 Spain 44 −9.7 −9.0 −1.7 Portugal 4 −12.0 −2.9 +6.4 Northern Europe 340 −5.9 −5.1 +0.3 Southern Europe 166 −8.1 −6.4 −0.6

Southern Europe = Italy, Portugal, Spain, Greece. Northern Europe = Finland, Netherlands, Ireland, UK, Belgium, Figure 3 Histogram of the postprandial changes in DBP. Germany, Poland. ABPM and postprandial BP in isolated systolic hypertension T Grodzicki et al 164 showed that within 2 h after the main meal about Acknowledgements 70% of subjects experienced a decrease in BP. In nearly one-quarter of them this decrease exceeded The SYST-EUR trial is a concerted action of the 16 mm Hg for systolic or 12 mm Hg for diastolic BP. European Union’s BIOMED Research Program. It is At the same time, an average heart rate did not show carried out in consultation with the World Health significant changes. Organization, the International Society of Hyperten- The epidemiological prevalence and magnitude of sion, the European Society of Hypertension and the World Hypertension League. The trial is supported meal-induced changes in BP remains unknown.10 As by Bayer AG (Wuppertal, Germany) and the it has been shown in clinical studies BP reaches a National Fund for Scientific Research (Brussels, nadir between 30 and 60 min after the meal, but it Belgium). Study medication is donated by Bayer AG may appear as early as after 15 min or as late as after 1,10 and Merck Sharpe and Dohme Inc (West Point, 75 min. BP was taken every 30 min, but the time PA, USA). of first measurement did not necessarily coincide with the beginning of the main meal. Therefore we chose the 2-h period after the meal as a postprandial Participating centres period and compared the latter with the 2-h period On 15 May, 1996, the following centres with ran- before the meal. domised patients participated in the project on 24- Diurnal profile of BP according to the age has been h ambulatory blood pressure monitoring: recently studied in the population-based studies such as the Belgian Population Study,11 PAMELA Belgium study12 and Ohasama study.13 Unfortunately, none Leszek Bieniaszewski, Hilde Celis, Robert Fagard, of these trials addressed the problem of meal- Jan A Staessen, Roger Van Hoof (Leuven), Paul De induced changes in BP profile nor evaluated Cort (Kumtich), Dirk Staessen, Jan A Staessen (Mechelen) subjects with isolated systolic hypertension as a separate group. Nevertheless all these studies have Bulgaria shown that older, in comparison with younger sub- Boyan Shahov (Sofia) jects, presented with higher BP variability. Further- Estonia more the analysis of the BP curves from the Belgian Tovio Laks (Tallinn) Study and the PAMELA survey demonstrated that a Finland ¨ dip in both systolic and diastolic BP was observed Kari Halonen (Kunsankoski), Tapio Hakamaki, Aapo 11,12 ¨ ¨ in the early afternoon hours. However, data pro- Lehtonen (Turku), Matti Jaaskivi, Liisa Tiito-Wiht, vided in both papers does not allow us to conclude Cinzia Sarti (Vantaa), Pavla Kivinen (Kuopio), Erkki ¨ with certainty that it was related to the mid-day Lehtomaki (Tampere), Reijo Tilvis, Hannu Van- meal or to the other factors. Nakajima et al14 hanen (Helsinki), Hannu Wallinhiemo (Hensinki) reported changes in the ABPM related to the stan- Germany dardised meal. The authors in a ‘Letter to the Editor’ Eberhard Ritz (Heidelberg) demonstrated that a postprandial fall occurred in 18 (36%) of 50 middle-age hypertensive subjects. Greece The reproducibility of the postprandial decrease Aris D Efstratopoulos (Athens) in BP in the present patients with ISH was high in Ireland the group analysis but much worse for individual Neil Atkins, Eoin T O’Brien (Dublin) patients. In a study by Jansen et al15 who evaluated Israel ´ the reproducibility of postprandial BP changes in 22 Joseph Rosenfeld, Jose Zabludowski (Petha Tiqva), nursing home residents after a standardised meal, Serge Zerapha (Givataim) on two occasions the mean intra-class correlation Italy was significant for both SBP and DBP. In our group Alfredo Bossini (Rome), Roberto Fogari (Pavia), Giu- ABPM was performed in the everyday life. The tim- liana Ginocchio, Paolo Palatini (Padova), Salvatore ing, composition and temperature of the meals etc Lattuada (Milan), Giuseppe Maiorano (Bari), Anna might therefore influence the intra-individual repro- Pirrelli (Bari), Antonio Salvetti (Pisa), Laura Terzoli ducibility of BP postprandial changes. As we were (Milan), Alvaro Vaccarella (Casatenovo) concerned that nutritional habits might differ Netherlands between countries, we analysed the postprandial fall Peter de Leeuw (Maastricht), Anton H van den Meir- according to the place of residence, but the mag- acker (Rotterdam), Arend J J Woittiez (Almelo) nitude of the meal-induced fall was similar across Europe. Poland In conclusion, about 70% of elderly patients with Kalina Kawecka-Jaszcz (Cracow), Jozef Kocemba ISH experience a postprandial fall in BP measured (Cracow) in everyday circumstances, and in one-quarter of Portugal them a decrease of more than 16 mm Hg in SBP or Gago Leira (Faro) 12 mm Hg in DBP occurred. Follow-up of these Slovenia patients in the SYST-EUR trial may demonstrate Jurij Dobovisek (Ljubljana) whether those with the more pronounced postpran- Spain dial decrease in BP are at higher cardiovascular risk. Blas Gil-Extremera (Granada), Rafael Marin ABPM and postprandial BP in isolated systolic hypertension T Grodzicki et al ´ 165 (Oviedo), Joan Ocon-Pujadas (Barcelona), Josep Coordinator of General Practices ´ ´ Redon (Sagunto), Jose L Rodicio (Madrid) Hilde Celis UK ` Garreth Beevers, Michele Beevers (Birmingham), References Christopher J Bulpitt (London), Christopher David- 1 Mathias CJ. Postprandial hypotension. Pathophysio- son (Brighton), Pandita Gunawardena (London), logical mechanisms and clinical implications in differ- John Webster (Aberdeen) ent disorders. Hypertension 1991; 18: 694–704. 2 Haigh RA et al. Possible impairment of the sympath- etic nervous system response to postprandial hypotension in elderly hypertensive patients. J Hum Hypertens 1991; 5: 83–89. Committees and Coordination 3 Lipsitz LA, Nyquist RP, Wei JY, Rowe JW. Postpran- Data Monitoring Committee dial reduction in blood pressure in the elderly. N Engl Christopher J Bulpitt, Astrid E Fletcher, Jan A J Med 1983; 309: 81–81. 4 Jansen RWMM, Hoefnagels WHL. Hormonal mech- Staessen, Lutgarde Thijs anisms of postprandial hypotension. J Am Geriatr Soc Endpoint Committee 1991; 39: 1201–1207. Peter de Leeuw, Robert Fagard, Gastone Leonetti, 5 Messerli FH, Grodzicki T. Hypertension in the elderly. James C Petrie In: Messerli FH (ed) Cardiovascular Disease in the Eld- erly. Kluwer Academic Publishers: Boston, 1993, Ethics Committee ¨ pp 121–144. Willem H Birkenhager, Collin T Dollery, Robert 6 Jansen RWMM et al. Postprandial hypotension in eld- Fagard erly patients with unexplained syncope. Arch Intern EU SYST-EUR Liaison Committee Med 1995; 155: 945–952. ¨ Willem H Birkenhager, Fernando De Padua, Collin 7 Vanhanen H et al. Associations of orthostatic blood pressure fall in older patients with isolated systolic T Dollery, Aris D Efstratopoulos, Robert Fagard, hypertension. J Hypertens 1996; 14: 943–949. Françoise Forette, Detlev Ganten, Eoin T O’Brien, ´ 8 Amery A et al. Syst-Eur. A multicentre trial on the Kevin O’Malley, Jose L Rodicio, Jaakko Tuomilehto, treatment of isolated systolic hypertension in the eld- Christian van Ypersele, Alberto Zanchetti erly: objectives, protocol and organization. Aging 1991; Publlication Committee 3: 287–302. ¨ Willem H Birkenhager, Christopher J Bulpitt, Jan A 9 Staessen J et al. Twenty-four hour blood pressure monitoring in the Syst-Eur trial. Aging 1992; 4: 85–91. Staessen, Alberto Zanchetti 10 Jansen RWMM, Lipsitz LA. Postprandial hypotension: Steering Committee epidemiology, pathophysiology, and clinical manage- Paul De Cort, Robert Fagard, Françoise Forette, Kal- ment. Ann Intern Med 1995; 122: 286–295. ina Kawecka-Jaszcz, Gastone Leonetti, Choudomir 11 Staessen JA et al. An epidemiological approach to ´ Nachev, Eoin T O’Brien, Jose L Rodico, Joseph ambulatory blood pressure monitoring: the Belgian Rosenfeld, Jaakko Tuomilehto, John Webster, Yair Population Study. Blood Press Monit 1996; 1: 13–26. 12 Mancia G et al. Ambulatory blood pressure normality: Yodfat results from the PAMELA study. J Hypertens 1995; 13: Trial Coordinators 1377–1390. Robert Fagard, Jan A Staessen 13 Imai Y et al. Ambulatory blood pressure of adults in Ohasama, Japan. Hypertension 1993; 22: 900–912. Coordinators of the Project on Ambulatory Blood 14 Nakajima S et al. Ambulatory blood pressure and post- Pressure Monitoring prandial hypotension. Am Heart J 1992; 124: 1669– Denis Clement, Eoin T O’Brien, Giuseppe Mancia, 1671. Gianfranco Parati, Jan A Staessen, Lutgarde Thijs 15 Jansen RWMM, Kelley-Gagnon MM, Lipsitz LA. Coordinator of the Project on Vascular Dementia Intraindividual reproducibility of postprandial and Françoise Forette orthostatic blood pressure changes in older nursing- home patients: relationship with chronic use of cardio- Coordinators of the Project on Quality of Life vascular medications. J Am Geriatr Soc 1996; 44: Christopher J Bulpitt, Astrid E Fletcher 383–389.