sign in sign up
<<
Home , Hypertension, Pseudohypertension, Systolic hypertension

Journal of Human (2009) 23, 182–187 & 2009 Macmillan Publishers Limited All rights reserved 0950-9240/09 $32.00 www.nature.com/jhh ORIGINAL ARTICLE Peripheral arterial and isolated : The ATTEST Study

ME Safar1, P Priollet2, F Luizy1, J-J Mourad3, P Cacoub4, H Levesque5, J Benelbaz6, P Michon7, M-A Herrmann8 and J Blacher1 1Faculte´ de Me´decine, Hoˆtel-Dieu hospital, Universite´ Paris Descartes, Paris, France; 2Saint Joseph hospital, Paris, France; 3Avicenne hospital, AP-HP and EA 3412, University Paris, Bobigny, France; 4Pitie´-Salpe´trie`re hospital, AP-HP, University Paris, Paris, France; 5Department of Internal , Rouen University hospital, 76031 Rouen Cedex, France; 6Issy les Moulineaux, France; 7Sanofi Aventis, Paris, France and 8Bristol-Myers Squibb, Rueil-Malmaison, France

Hypertension is a for cardiovascular (CV) subjects with systolic hypertension were characterized , either coronary disease (CAD), periph- by the presence of PAD, with little or no presence of eral artery disease (PAD) or CAD and/or CVD. Subjects with systolic–diastolic (CVD). The relationships between those different locali- hypertension were characterized by the presence of zations of CV disease and the haemodynamic features CAD and/or CVD, but without PAD. Although the of hypertension have been poorly evaluated in the past. former was only influenced by age, dyslipidaemia and In the ATTEST study, a geographically representative mellitus influenced the latter. This study panel of 3020 general practitioners recruited 8316 confirms the high prevalence of hypertension (80%) in consecutive patients with CV diseases (PAD, CAD or a large population of patients with CV diseases selected CVD, alone or in association). pressure, which in . Analysis of different features of was not an inclusion criterion, was then measured and hypertension revealed that isolated systolic hyperten- related to the different forms of CV diseases. Blood sion was the most prevalent form of hypertension in pressure classification involved 20% normotensive this treated population. Finally, one of the predominant subjects, 24% subjects with controlled hypertension, goals of secondary prevention in subjects with 42% subjects with isolated systolic hypertension and PAD should be the treatment of isolated systolic 14% subjects with systolic–diastolic hypertension, all hypertension. hypertensives with or without antihypertensive therapy. Journal of Human Hypertension (2009) 23, 182–187; From multiple regression analysis, it appeared that doi:10.1038/jhh.2008.121; published online 2 October 2008

Keywords: ; peripheral; hypertension; systolic

Introduction differ according to the dominant site of CV events.2 For instance, blockers of the –angiotensin being the first cause of system might be more efficient in preventing cardiac mortality worldwide, and hypertension being the 1 complications whereas entry blockers could most prevalent cardiovascular risk factor, it is of be more efficient in preventing cerebral complica- paramount importance to study the precise relation- tions.3 Whether the haemodynamic features of ships existing between different forms of hyperten- hypertension, which may either be isolated systolic, sion and different localizations of . isolated diastolic or systolic–diastolic, have a part in During therapy in patients the therapeutic indications, has not yet been clearly with previous cardiovascular (CV) event, the goal of investigated. treatment is to avoid the recurrence of CV events, When the clinical aspects of end- damage particularly those observed in the cerebral, cardiac are considered in hypertensive subjects, an in- or renal circulations as well as in the lower limbs. creased incidence of systolic hypertension is The results of meta analysis indicate nowadays that particularly observed in subjects with cerebrovas- the prescriptions of antihypertensive agents might cular complications and, more recently, in subjects with chronic renal disease.4 In patients with peripheral artery disease (PAD), systolic–diastolic Correspondence: Professor J Blacher, Centre de Diagnostic et de hypertension is not frequently observed but the The´rapeutique, Hoˆpital Hoˆtel-Dieu, AP-HP; 1 Place du Parvis repartition of subjects between normotensives and Notre-Dame 75004 Paris, France. isolated systolic hypertensives has not been exten- E-mail: [email protected] 2 Received 11 June 2008; revised 4 August 2008; accepted 24 sively investigated. In coronary disease, it is August 2008; published online 2 October 2008 classically observed that, following myocardial Arterial disease and hypertension ME Safar et al 183 , the incidence of increased systolic blood circulation and abdominal aortic ). CAD pressure (SBP) is markedly reduced.2,4 Thus, it is was defined as any history of pectoris (chest important, in a large population of hypertensive pain precipitated by exertion and relieved by rest or subjects with CV end-organ damage, to evaluate the nitrates) confirmed by coronary angiography, incidence of normal (BP), systolic– , typical sequelae on electro- diastolic hypertension and isolated systolic hyper- cardiography, coronary percutaneous transluminal tension according to the various sites of CV events. , or coronary artery bypass surgery or a The objectives of the ATTEST study (‘prise en combination. CVD was defined by history of charge de l’ArTe´riopaThie oblitErante des membreS ischaemic confirmed by computed tomo- infe´rieurs chez les paTients en me´decine ge´ne´rale’)5 graphy or magnetic resonance imaging. were to compare pharmacological treatment (pri- A physician-completed inclusion questionnaire mary objective) and medical management, including contained the following data: gender, age, weight, cardiovascular tests and physician’s assessment of height, personal history of diabetes mellitus, dysli- future cardiovascular and amputation risks (second- pidaemia, or hypertension, current and previous ary objectives) of patients with established PAD to habit, CV tests previously performed, and patients with other vascular site(s) involved in current use of antiplatelet, antihypertensive, lipid- atherothrombosis. The objective of the analysis of lowering and antidiabetic drugs. These medical data the ATTEST study presented in the present paper is were not obtained from a formal testing protocol, to compare the prevalence of the different varieties and the ATTEST study was not planned to add any of hypertension as systolic–diastolic hypertension tests to the management of the patients. Rather, and isolated systolic hypertension and to determine physicians were asked to give the most precise which category of atherosclerotic CV events may information present in their medical files. Diabetes affect the heart, the brain and the lower limbs, mellitus was diagnosed on the basis of plasma respectively with regard to these varieties. glucose 47.0 mmol lÀ1 and/or antidiabetic treat- ment. Dyslipidaemia was defined on the basis of guidelines that prevailed at the time of inclusion 6 Materials and methods and/or presence of hypolipidaemic treatment. Population The design of the ATTEST study has been reported BP classification elsewhere.5 Briefly, this investigation was an ob- Blood pressure was measured after a 5-min rest. servational, cross-sectional, epidemiologic study Different devices were used according to the device that took place in France between April and generally used by the practitioner, and the mean of November 2003. Patients were selected by a geo- three determinations was considered as systolic (S) graphically representative panel of 3020 general and diastolic (D) BP. pressure was calculated practitioners who agreed to participate. Each physi- as PP ¼ SBP–DBP. Mean blood pressure was cian was asked to include the first three patients MBP ¼ DBP þ 1/3 PP. identified from his or her practice to fulfill the Based on these measurements and on question- inclusion criteria. These criteria were patients (1) naires (previous diagnosis of hypertension), the BP 418 years old, (2) with at least one site of proven classification was composed of four groups (Table 1): atherosclerotic CV disease: PAD of the lower limbs for two patients, and proven  Normotensives (potentially receiving cardio- (CAD) or cerebrovascular disease (CVD) for the third vascular drugs with antihypertensive effects patient and who were (3) willing to participate in such as b-blockersy): SBPo140 mm Hg and DBP the study and give consent after receiving written o90 mm Hg (n ¼ 1687) and oral information. A total of 8316 patients were  Controlled hypertensives (receiving antihyper- included (all data completed) and analyzed. The tensive drugs): SBPo140 mm Hg and DBPo protocol was approved by the institutional review 90 mm Hg (n ¼ 2018) committees ‘Conseil National de l’Ordre des Me´decins’,  Isolated systolic hypertension, with SBPX ‘Comite´ Consultatif sur le Traitement de l’Information 140 mm Hg and DBPo90 mm Hg (n ¼ 3480) en Matie`re de Recherche dans le domaine de la Sante´’,  Systolic–diastolic hypertension, with SBPX ‘Commission Nationale Informatique et Liberte´’, Paris, 140 mm Hg and DBPX90 mm Hg (n ¼ 1131). France. Atherosclerotic diseases were defined using the Subjects with systolic and systolic–diastolic International Classification of Diseases (10th edi- hypertension were either treated or untreated by tion). PAD was defined from typical symptoms and antihypertensive drug therapy (see Table 1). severe (470%) at arterial duplex or arter- iography (99% of patients had arterial duplex examinations), and/or a history of surgical or Statistical analysis percutaneous transluminal treatment for lower limb The total population was divided into three cate- arterial disease (excluding the renal and splanchnic gories of atherosclerotic diseases: patients with

Journal of Human Hypertension Arterial disease and hypertension ME Safar et al 184 Table 1 Mean±s.d. of cardiovascular risk factors, blood pressure measurements and pharmacological treatment of patients included in the ATTEST study according to the BP classification

Normotensives Controlled Systolic Systolic+diastolic P* (n ¼ 1687) hypertensives hypertension hypertension (n ¼ 2018) (n ¼ 3480) (n ¼ 1131)

Age (years) 61±12 68±11 68±11 67±11 o0.0001 Gender (% male) 84 78 78 79 o0.0001

Cardiovascular risk factors (kg/m2) 24.8±3.6 26.1±4.0 26.3±3.8 27.2±3.9 o0.0001 Dyslipidaemia (%) 56 73 68 70 o0.0001 Diabetes mellitus (%) 14 27 27 31 o0.0001 Current smoker (%) 30 15 19 24 o0.0001 LDL (mmol lÀ1) 3.23±0.90 3.13±0.90 3.26±0.88 3.53±1.00 o0.0001

Blood pressure parameters Systolic blood pressure (mm Hg) 125±8 128±7 145±7 152±14 o0.0001 Diastolic blood pressure (mm Hg) 74±775±779±593±6 o0.0001 Mean blood pressure (mmHg) 91±693±6 101±5 112±7 o0.0001 Pulse blood pressure (mmHg) 51±853±766±959±13 o0.0001 PP460 mm Hg (%) 6 6 75 15 o0.0001

Pharmacological treatment Cardiovascular agents (%) 0 37 32 34 o0.0001 b-blockers (%) 29 38 29 30 o0.0001 ACE inhibitors (%) 17 44 36 37 o0.0001 Angiotensin II RA (%) 2 23 19 23 o0.0001 antagonist (%) 0 42 35 33 o0.0001

Hypolipidemic agents (%) 60 69 62 61 o0.0001 Other agents (%) 1 3 2 2 0.0291

Other treatments Non- agents (%) 2 7 6 7 o0.0001 Aspirin (%) 39 40 37 43 0.0131 Clopidogrel/ticlopidine (%) 66 62 64 58 0.0008 No antiplatelet drug (%) 8 8 10 12 0.0015 Vasoactive agents (%) 45 47 54 48 o0.0001

Abbreviations: n, number of subjects; LDL, low-density lipoprotein; ACE, angiotensin-converting enzyme; RA, receptor antagonist. *P-values are adjusted on age and sex.

isolated PAD (n ¼ 3127), patients with PAD asso- ‘normotension’ to ‘controlled hypertension’, ‘systo- ciated with atherosclerotic event in at least one lic hypertension’ and ‘systolic þ diastolic hyperten- other location, CAD or CVD (n ¼ 3128) and patients sion’, there was a positive and significant with CAD or CVD, or both, but without PAD (Po0.0001) increase of almost all the studied (n ¼ 2061). Quantitative parameters were described variables. Note that, whatever the antihypertensive by using means and s.d., and qualitative parameters agent studied, the proportions of treated subjects by number and percent. Qualitative parameter with hypertension were relatively low (generally distributions within the three groups were analyzed between 30 and 40%) as well in controlled and non- using the w2-test. Analysis of variance was used for controlled subjects. The lower proportions were quantitative parameter comparisons among the two observed for angiotensin II receptor antagonists: hypertensive groups; no post hoc test was per- between 2 and 23%. formed. A Po0.05 was considered significant. All Figure 1 illustrates in percent the different sites of analyses were performed using SAS 9.1 software CV events as function of the BP classification. In the (SAS Institute, Cary, NC, USA). totality of the population, patients with normal BP represented 44% of subjects, including 20% normo- tensives and 24% controlled hypertensives; ‘systolic Results hypertension’ 42% and ‘systolic–diastolic hyperten- sion’, 14%: Po0.0001 for intergroup comparisons. Table 1 indicates, in the total population, CV risk From intragroup comparisons, it appears that the factors, BP parameters and pharmacological treat- PAD groups, that is, with or without associated CAD ment according to the BP classification. From or CVD, represented 70 to 80% of subjects, the

Journal of Human Hypertension Arterial disease and hypertension ME Safar et al 185 remaining population (20 to 30%) representing because of the relative presence of antihypertensive subjects with CAD and/or CVD without PAD. All drug therapy. The most important difficulty of the intragroup comparisons were statistically sig- diagnosis is that, in atherosclerotic subjects, non- nificant (Po0.0001). invasive determination of BP gives usually adequate Table 2 shows the multiple regression analysis of isolated systolic and systolic–diastolic hyperten- sion. For isolated systolic hypertension, there were 50 three characteristics: (i) a positive association 47 with SBP contrasting with a negative association 45 with DBP, thus indicating the specific role of PP, (ii) 42 the presence of PAD, in contrast with the negative 41 40 association of CAD and/or CVD and (iii) the role of 40 39 age. For systolic–diastolic hypertension, the pattern 37 differed markedly: (i) positive association with SBP and DBP, (ii) presence of CAD and/or CVD, but 35 negative association with isolated PAD, (iii) role of metabolic disorders linked to plasma glucose and 30 29 29 lipids. 28

24

% 25 22 Discussion 21 20 This study was performed in a large population of subjects with at least one site of proven athero- sclerotic organ damage, represented by PAD and/or 15 CAD and/or CVD. Approximately half had normal BP and half had hypertension. From the hyperten- 10 sive subjects, the majority was characterized by isolated systolic, and not systolic–diastolic, hyper- tension. This haemodynamic pattern was primarily 5 associated with PAD, whether CAD or CVD were present or not. Taken together, these results suggest 0 that isolated systolic hypertension is one of the Non hypertensives Controlled Systolic hypertensives Systolic + diastolic (n=1687, 20%) hypertensives (n=3480, 42%) hypertension dominant problems for patients in secondary cardi- (n=2018, 24%) (n=1131, 14%) ovascular prevention, namely those with PAD. On account of the classical difficulties of BP Isolated PAD (n=3127) PAD associated with CAD or CVD (n=3128) measurements, the present results should be ana- CAD and/or CVD without PAD (n=2061) lyzed with caution. Because the population is mainly composed of old people, increased varia- Figure 1 Percentage (%) of the different sites of CV events (PAD isolated, PAD associated with CAD or CVD, CAD and/or CVD bility of BP is present, making difficult the diagnosis without PAD) according to the BP classification. Differences in 7 of systolic hypertension. More specifically, the the groups are statistically significant (Po0.0001). Percentages are diagnosis of normal BP may be overestimated, indicated on the top of each histogram.

Table 2 Multiple regression analysis of isolated systolic hypertension and systolic+diastolic hypertension

Systolic hypertension Systolic+diastolic hypertension

Step Variable Coefficient Partial Model P Step Variable Coefficient Partial R2 Model R2 P entered R2 R2 entered

1 SBP (mm Hg) 0.026 0.28 0.2838 o0.0001 1 DBP (mm Hg) 0.021 0.36 0.3599 o0.0001 2 DBP (mm Hg) À0.018 0.067 0.3513 o0.0001 2 SBP (mm Hg) 0.0025 0.0069 0.3668 o0.0001 3 Isolated PAD 0.042 0.0023 0.3536 o0.0001 3 CAD and/or CVD 0.023 0.0017 0.3685 0.0001 without PAD 4 Age 0.0011 0.0007 0.3543 0.0152 4 Isolated PAD À0.018 0.0006 0.3691 0.0219 5 LDL-cholesterol À0.013 0.0006 0.3548 0.0247 5 Diabetes mellitus 0.016 0.0005 0.3696 0.0393 (mmol lÀ1) 6 CAD and/or CVD À0.030 0.0006 0.3554 0.0303 6 LDL-cholesterol 0.0084 0.0005 0.3701 0.0476 without PAD (mmol lÀ1) 7 Gender À0.016 0.0004 0.3705 0.0658

Forward stepwise regression. Criterion for entry into the model is 0.10.

Journal of Human Hypertension Arterial disease and hypertension ME Safar et al 186 measurements for SBP, but not for DBP.8,9 Compar- subjects with PAD and favour the presence of ison between non-invasive and intra-arterial BP systolic–diastolic hypertension.19 Finally, clinical measurements have shown that the frequency of studies have shown that increased increased DBP is often overestimated,8 and in and disturbed wave reflections are not only frequent consequence, that the incidence of isolated systolic in subjects with PAD but also are significantly hypertension in the elderly is frequently under- associated with reduced distance and estimated.9,10 In addition, the absence of a single impaired vasodilating arteriolar properties.20–22 device for BP measurements has probably increased Mostly, all these disorders may be attenuated by the heterogeneity of BP measurements, leading to chronic ACEI treatment.23,24 Thus, such results increased misclassification in BP groups and then taken together suggest that a more active drug to decreased power of statistical analysis. Finally, it therapy should be quite relevant to consider in is important to note that the ATTEST study is not an subjects with systolic hypertension and target organ epidemiological survey; as the physicians had to damage, particularly those with PAD. Other studies recruit two subjects with PAD and one with CAD/ report bad blood pressure control in patients in CVD, the resulting prevalence of isolated systolic secondary prevention; after a major cardiovascular hypertension is somewhat biased. A different selec- event, the EUROASPIRE II25 survey showed that tion of PAD versus CAD/CVD could have resulted in ‘only’ half of the population (including normoten- a lower prevalence of systolic hypertension in the sives and hypertensives) had normal BP. Similarly, ATTEST study. in our study, 45% of the patients had normal BP and An important result of this study was the 30% of the hypertensives were controlled. significant association that we observed between In the PARTNERS program,26 it has been widely systolic hypertension and the presence of PAD. In reported that participants with PAD are treated less most subjects with PAD, ejection fraction and intensively with antihypertensive and antiplatelet ventricular ejection are preserved, giving no cardiac therapy and cholesterol-lowering agents, than those explanation to the frequency of systolic hyperten- with CAD. This study indicates a quite similar trend sion in this population.10,11 Increased SBP in (Table 1). Thus an improvement of PAD manage- subjects with PAD is mainly because of increased ment should be nowadays of primary importance for arterial stiffness and disturbed timing and ampli- the development of CV prevention in subjects with tude of central wave reflections.12–14 More specifi- hypertension, particularly those with isolated sys- cally, reflection sites are closer from the heart than tolic hypertension. in normal subjects, because of the importance of structural alterations of the aortic bifurcations, as extensively shown from arteriography.12 This inter- pretation is strengthened by the presence of an What is known about the topic K Hypertension is a risk factor for cardiovascular diseases, increased frequency of systolic hypertension around either coronary artery disease, peripheral artery disease or 50 years of age in subjects with traumatic amputa- cerebrovascular disease. tion of the lower limbs.15 In these patients the height K Patients with peripheral artery disease are treated less of SBP is proportional to the number of lower limb intensively with antihypertensive, antiplatelet therapy and cholesterol-lowering agents, than those with coronary amputations and therefore to the number and artery disease. diffusion of reflection sites, which all together K The relationships between the different localizations of become closer from the heart. In contrast, in this cardiovascular disease and the haemodynamic varieties of study, subjects with CAD and/or CVD, but without hypertension have been poorly evaluated in the past PAD, had a lower incidence of systolic hyperten- What this study adds sion, but a relatively high incidence of systolic– K This study confirms the high prevalence of hypertension diastolic hypertension. (80%) in a large population of patients with cardiovascular In this study, other explanations may be proposed diseases selected in primary care. for the significant association between haemody- K Analysis of different features of hypertension revealed that isolated systolic hypertension was the most prevalent form namic features of hypertension and sites of CV of hypertension in this treated population. events. First, in subjects with increased arterial K Our results suggest that isolated systolic hypertension is stiffness, increased SBP may be attenuated or even one of the dominant problems for patients in secondary disappear in the presence of reduced ventricular cardiovascular prevention, namely those with peripheral ejection.2 This haemodynamic pattern is frequently artery disease. observed in subjects with CAD, particularly follow- ing myocardial infarction. Second, drug treatment may interfere with BP level in several subjects of the studied population. excess and b-blocking Acknowledgements agents may increase SBP through increase in arterial stiffness and disturbed wave reflections, whereas This work has been made possible by Bristol–Myers nitrates and angiotensin-converting enzyme inhibi- Squibb and Aventis, France. Members of the tors (ACEI) have an opposite effect.12,16–18 Third, scientific committee would like to acknowledge the stenoses of renal artery are relatively frequent in contribution of each general practitioner. This

Journal of Human Hypertension Arterial disease and hypertension ME Safar et al 187 analysis was performed with the help of GPH-CV lower limbs; Observation of the effect of intravenous (Groupe de Pharmacologie et d0He´modynamique . 1982; 2: 266–271. Cardiovasculaire), Paris. We thank Benjamin Re- 13 Levenson JA, Simon ACH, Safar ME, Fiessinger JN, gnault for statistical analysis and Dr Anne Safar for Housset EM. Systolic hypertension in patients with helpful and stimulating discussions. arteriosclerosis of the lower limbs. Clin Exper Hyper Theory Practice 1982; A4(7): 1059–1072. 14 Van Popele NM, Grobbee DE, Bots ML, Asmar R, References Topouchian J, Reneman RS et al. Association between arterial stiffness and atherosclerosis. The Rotterdam 1 Lawes CM, Vander Hoorn S, Rodgers A. International Study. Stroke 2001; 32: 454–460. Society of Hypertension. Global burden of blood- 15 Labouret G, Achimastos A, Benetos A, Safar M, pressure-related disease, 2001. Lancet 2008; 371: Housset E. L’hypertension arte´rielle des ampute´s 1513–1518. traumatiques. Press Med 1983; 21: 1349–1354. 2 Arterial stiffness in hypertension. In: Safar ME 16 Levenson JA, Simon AC, Maarek BE, Gitelman RJ, and O’Rourke ME (eds). Handbook of Hypertension, Fiessinger JN, Safar ME. Regional compliance of vol. 23. Elsevier: Edinburgh, UK, 2006, pp 3–62, 75–136, and saline infusion in patients with 459–501. arteriosclerosis obliterans. Arteriosclerosis 1985; 5: 3 Verdecchia P, Reboldi G, Angeli F, Gattobigio R, 80–87. Bentivoglio M, Thijs L et al. Angiotensin-converting 17 Achimastos A, Brahimi M, Raison J, Billaud E, Ayache enzyme inhibitors and calcium channel blockers for M, Moatti N et al. Plasma insulin, plasminogen coronary heart disease and stroke prevention. Hyper- activator inhibitor, and ankle-brachial systolic blood tension 2005; 46: 386–392. pressure ratio in hypertensive subjects. 4 Mancia G, De Backer G, Dominiczak A, Cifkova R, J Hum Hypertens 1999; 13: 329–335. Fagard R, Germano G et al. Guidelines for the Manage- 18 Simon AC, Levenson JA, Levy BI, Bouthier JE, ment of Arterial Hypertension: The Task Force for the Peronneau PP, Safar ME. Effects of nitroglycerine on Management of Arterial Hypertension of the European peripheral large arteries in hypertension. Br J Clin Society of Hypertension (ESH) and of the European Pharmacol 1982; 14: 241–245. Society of (ESC). J Hypertens 2007; 25: 19 Louie J, Isaacson JA, Zierler RE, Bergelin RO, Strand- 1105–1187. ness DE. Prevalence of carotid and lower extremity 5 Blacher J, Cacoub P, Luizy F, Mourad JJ, Levesque H, arterial disease in patients with . Benelbaz J et al. Peripheral arterial disease versus other Am J Hypertens 1994; 7: 436–439. localizations of : the ATTEST study. 20 Brewer LC, Chai HS, Bailey KR, Kullo IJ. Measures of J Vasc Surg 2006; 44: 314–318. arterial stiffness and wave reflection are associated 6 Expert Panel on Detection, Evaluation, and Treatment with walking distance in patients with peripheral of High Blood Cholesterol in Adults. Executive arterial disease. Atherosclerosis 2007; 191: 384–390. Summary of The Third Report of The National 21 Safar ME, Totomoukouo JJ, Asmar RA, Laurent ST. Cholesterol Education Program (NCEP) Expert Panel Increased in patients with arteriosclero- on Detection, Evaluation, And Treatment of High sis obliterans of the lower limb. Arteriosclerosis 1987; Blood Cholesterol In Adults (Adult Treatment Panel 7: 232–237. III). JAMA 2001; 285: 2486–2497. 22 Farrar DJ J, Malindzak GS, Johnson G. Large Vessel 7 Safar ME, Smulyan H. Systolic versus diastolic blood Impedance in Peripheral Atherosclerosis. Circulation pressure, 2000. In: Bulpitt CJ (ed). Handbook 1977; 56(3 Suppl II): 171–178. of Hypertension. Elsevier Science BV: Amsterdam, pp 23 Yusuf S, Sleight P, Pogue J, Boch J, Davies R, Dagenais 73–85. G. Effects of an angiotensin-converting-enzyme inhi- 8 Spence JD, Sibbald WJ, Cape RD. bitor, , on cardiovascular events in high-risk in the ederly. Clin Sci Molec Med 1978; 55: 399s–402s. patients. The Heart Outcomes Prevention Evaluation 9 Messerli FH, Ventura HO, Amodeo C. Osler’s maneu- Study Investigators. N Engl J Med 2000; 342: 145–153. ver and pseudohypertension. New Engl J Med 1985; 24 Achimastos AA, Natoli AK, Lawler A, Blombery PA, 312: 1348–1351. Kingwell BA. Ramipril reduces large-artery stiffness in 10 Safar ME, Simon AC. Hemodynamics in systolic peripheral arterial disease and promotes elastogenic hypertension, 1986. In: Zanchetti A, Tarazi RC (eds). remodeling in cell culture. Hypertension 2005; 45: Handbook of Hypertension, vol 7: Physiopathology of 1194–1199. Hypertension, Cardiovascular Aspects. Elsevier 25 EUROASPIRE I and II Group. European Action on Science: Amsterdam, pp 70–84. Secondary Prevention by Intervention to Reduce 11 Safar ME, Laurent ST, Asmar RE, Safavian A, London Events. Clinical reality of coronary prevention guide- GM. Systolic hypertension in patients with arterio- lines: a comparison of EUROASPIRE I and II in nine sclerosis obliterans of the lower limbs. 1987; countries. Lancet 2001; 357: 995–1001. 28: 287–295. 26 Hirsch AT, Criqui MH, Treat-Jacobson D, Regensteiner 12 Levenson JA, Simon AC, Fiessinger JN, Safar ME, JG, Creager MA, Olin JW et al. Peripheral arterial London GM, Housset EM. Systemic arterial compli- disease detection, awareness, and treatment in primary ance in patients with arteriosclerosis obliterans of the care. JAMA 2001; 286: 1317–1324.

Journal of Human Hypertension