The importance of hemodynamic considerations jn essential hypertension

Hypertension is a major risk factor for cardiovascular morbidity and mortality. Antihypertensive therapy consistently reduces complications from stroke and congestive heart failure, whereas beneiits from the treatment of ischemic heart disease events are variable. Several plausible explanations, including hemodynamic hypotheses, have been put forth to account for the failure of treatment to more favorably influence mortality from ischemic heart disease. The effect of hypertension on coronary heart disease is probably much more complex than a simple elevation of arterial pressure. Some of ihese complexities include the potential separate risks of high total peripheral resistance, hsgh cardiac output, increased myocardial power that reflects pressure t&s flow, and several structural and functional vascular changes. These factors may act in concert to unfavorably alter the balance between myocardial oxygen supply and demand. Several of these factors will be highlighted in an attempt to offer alternative or adjunctive pathophyslologic examlnations for the high-risk subgroups of obesity and the failure of antihypertensive therapy to normalize the rate of coronary heart disease events. (AM HEART J 1988;116:594.)

Brent Egan, MD, and Robert Schmouder, MD Ann Arbor, Mich.

Epidemiologic data provide evidence that elevated (MAP), on morbidity and mortality from coronary arterial blood pressure is associated with a signifi- heart disease. Although the review is based on a cantly higher incidence of cardiovascular complica- wealth of physiologically sound experimental obser- tions, including myocardial infarction.‘s2 Pharmaco- vations, some less recognized pathophysiologic con- logic lowering of arterial pressure reduces total cepts are explored in an attempt to more fully cardiovascular morbidity and mortality in hyperten- account for incompletely explained epidemiologic sive patients.3-6 Although the overall event rate is data. The data include not only the failure of reduced, specific ischemic cardiac events are not antihypertensive therapy to normalize the excessive consistently prevented.’ Numerous plausible expla- incidence of coronary artery disease, but also the nations have been provided by various experts to excessive cardiac risk accompanying borderline account for this major shortcoming in several treat- hypertension and obesity, particularly in younger ment trials (Table I). men. This review focuses predominantly on the poten- tial role of both early and late hemodynamic vari- HEMODYNAMIC CONSIDERATIONS ables, including total peripheral (systemic) vascular Peripheral vascular resistance. Several observations resistance* (TPR) and cardiac output (CO), as well suggest the predominant importance of TPR in as their product, mean arterial (blood) pressure? explaining the cardiac complications associated with hypertension. First, increased resistance to blood flow is the hemodynamic hallmark of hyperten- From the Division of Hypertension, University of Michigan Medical Center. sion.Q*lo This conclusion is based on a continuously Reprint requests: Brent Egan, MD, Division of Hypertension, University of fixed pressure and constant flow model of circula- Michigan Medical Center, 3918 Taubman Center, Ann Arbor, MI 48109- tion in which MAP = CO x TPR. However, the 0356. power generator for the cardiovascular system, the “Total systemic resistance (TSR) in arbitrary units is equivalent to mean blood pressure divided by cardiac output (CO). Total peripheral resistance heart, is pulsatile. In a pulsatile system, the forces (TPR) equals mean arterial pressure minus mean right atria1 pressure opposing ventricular ejection of blood are more divided by CO. Since right atrial pressure is generally 0 to 4 mm Hg in most .,I I. appropriately termed impedance.” Unfortunately, :>-L‘L.L.,; : “,,<

594 Volume 116 Number 2, Part 2 Hemodynamic considerations in hypertension 595 provided a useful conceptual framework for inter- Table I. Seven recommendations for improving the out- preting innumerable hemodynamic studies and will come of coronary artery disease in hypertensive patients serve as the predominant model in this discussion. Lower BP more As noted, in most persons with elevated arterial Treat BP longer blood pressure, CO is within the normal range, Reduce BP less whereas calculated resistance in absolute units is Multiple risk factor intervention high. In hypertensive subjects with increased CO, Avoid electrolyte imbalances which occurs in a substantial proportion of border- Minimize SNS and RAS activity Decrease TPR and not CO line hypertensive and obese persons,12* l3 the absolute resistance is normal. However, the resistance in SNS = sympathetic nervous system; RAS = renin-angiotensin system. relative terms is higher at any specsed level of CO for those with higher arterial pressure.14 The importance of this apparently semantic argu- repudiate either the longer-term hemodynamic ment is highlighted by hemodynamic data from effects of diuretics or b-blockers, which, in fact, studies on predominantly young men conducted in include reduction of TPR,‘*p ls or the beneficial Ann Arbor.15 Subjects were divided into two groups effects of these compounds on stroke rate,6s6 but on the basis of percentage over ideal body weight rather to suggest that pharmacologic approaches to (IBW). One group was <20% over IBW, whereas the reducing ischemic heart disease may be more hemo- other (obese) was >20% over IBW. The obese dynamically complex than simply reducing arterial subjects were characterized by higher blood pressure pressure. (BP) and CO but similar values for peripheral Finally, a study of elderly men observed signifi- resistance in comparison to lean participants. Fur- cantly lower cardiovascular mortality in overweight thermore, in the overweight men BP was positively compared with lean hypertensive subjects. The low- correlated to CO, but BP was not related to TPR. er absolute TPR in overweight hypertensive sub- However, in the lean normotensive men, TPR jects was cited as one potential explanation for their decreased appropriately as CO increased so that BP more favorable cardiovascular outcome.” remained normaLl These data are consistent with Cardiac factors in determining the cardiovascular previously reported information that CO is elevated complication rate. If cardiovascular mortality were similarly in overweight normotensive vs both bor- simply an inverse function of absolute TPR, then derline and overweight mild hypertensive subjects.13 obese normotensive subjects would have the lowest Therefore the fundamental hemodynamic mecha- cardiovascular complication rate, which is not the nism of elevated BP in both the lean borderline case.21.22However, there is substantial evidence that hypertensive and overweight borderline-mild sub- obesity, which is accompanied by lower absolute jects is an inappropriately high resistance for the values for TPR at any given level of arterial pres- prevailing level of blood flow. Additional studies sure, is an independent risk factor for higher rather performed in our laboratory indicate that an than lower cardiovascular complication rates.23 Fur- increase in vascular a-adrenergic tone is an impor- thermore, borderline hypertension, which is charac- tant factor in the relatively high resistance to blood terized by relatively normal absolute values for flow in overweight people.16 These studies were TPR,= I3 is associated with a substantial increase in conducted in young men who, as will be discussed, cardiovascular complications despite the minimal may be the group at greatest relative hemodynamic elevation of blood pressure.%, 25 risk from being overweight. Hemodynamically, borderline hypertension and Second, CO remains essentially unchanged across obesity are both characterized by elevated absolute the spectrum of hypertension, from mild to severe values for CO.26 In other words, at any given level of levels. Consequently, a successive increase in TPR is afterload, which is roughly reflected by higher MAP, the predominant explanation for the progression of venous return or preload is also higher.n-29 Conse- hypertension. Therefore one could argue that the quently, the heart is exposed to the double burden of progressive rise in TPR is an important component both increased preload (venous return) and after- of the higher cardiovascular complication rate in load (arterial pressure).2s more severely hypertensive patients. The double hemodynamic load in overweight Third, antihypertensive compounds that initially patients may explain some of the excessive incidence reduce CO, namely, diuretics and P-blockers, have of heart disease. In the Framingham Study, persons not consistently afforded primary prevention >20% over ideal weight, when compared with sub- against coronary events,34 although exceptions have jects <20 % overweight, were at threefold higher risk been noted.17 These comments are not intended to for subsequent hypertension.30 Moreover, obese sub- 596 Egan and Schmouder

Table II. Hemodynamic power in various patient sub- MAP.‘l Therefore cardiac power reflects compo- groups nents of both preload and afterload. Cardiac power Data = Lean Lean is also a more comprehensive term than the two mean + SE NT BHT Obese HT more commonly used measures: stroke work and pressure-rate product. Stroke work does not account No. 236 129 70 39 for the frequency (heart rate) of work, whereas Age 25 +- 1 26 & 1 30 i 1 40 c 1 pressure-rate product does not include stroke vol- % >IBW Ok1 4+-l 23 + 1 1.5 f 2 SAP 118 -+ 1 132 a 1 134 f 12 150 r+_ 13 ume. DAP 63 k 1 73 +- 1 76 + 1 85 -+ 2 An index of cardiac power was calculated from CO (L/min) 5.6 k 0.1 6.7 f 0.1 6.7 f 0.2 5.8 f 0.1 baseline data obtained in supine volunteers who Power 74 f 1 101 -e 3 103 * 34 98 zt 3 participated in hemodynamic studies conducted in Results of ANOVA: > p < 0.05; = p > 0.05 the Hypertension Division at Ann Arbor. One Age: HT > obese > BHT = NT % IBW: Obese > HT > BHT > NT adjustment was made in the constant pressure-flow SAP: HT > Obese = BHT > NT model in the calculations. Since the left ventricle DAP: HT > Obese = BHT > NT generates CO only during systole, calculated mean CO: Obese = BHT > HT = NT systolic arterial pressure rather than MAP was used Power: Obese = BHT = HT > HT in the cardiac power calculation. Because left atrial SAP = systolic arterial pressure, mm Hg; HT = hypertensive; DAF’ = dia- pressure data were unavailable, no adjustment was stolic arterial pressure, mm Hg; lean cardiovascular disease (HCVD), which results are summarized in Table II. As shown, is defined as left ventricular hypertrophy by either cardiac power was substantially higher in lean bor- chest radiography or electrocardiography. derline hypertensive and obese subjects compared Data from the Australian Heart Foundation with lean normotensive persons. In fact, power in Study attributed as much as 30 % of hypertension in these two conditions was virtually identical to levels men and women between 25 and 64 years of age to calculated in older subjects with established essen- being overweight.31 However, within the subgroup of tial hypertension. men less than 45 years old, as much as 60% of Consequently, cardiac power rather than BP may hypertension was explained by weight. Therefore serve as a more useful predictor of age-adjusted from a hemodynamic viewpoint, overweight younger cardiac complications in borderline hypertension men may be at the greatest relative risk. Conse- and obesity. This hypothesis requires confirmation quently, an analysis of hemodynamic data in this but may provide a partial hemodynamic basis for high risk subgroup, as will be discussed, may prove the independent contributions of overweight to car- enlightening. diovascular disease. 21 Indirect confirmation for the The potential importance of cardiac power. As previ- potential use of cardiac power is provided by data ously noted, the incidence of HCVD in overweight obtained in children from Muscatine, Iowa.% Chil- subjects substantially exceeds that predicted from dren with BP levels in the upper quintile had a the higher rates of hypertension. Consequently, it is higher left ventricular mass index compared with possible that in obesity the double cardiac burden of their cohorts with BP levels in the lowest quintile. both increased preload and afterload offers a more Within the upper quintile of BP, children with satisfactory explanation than pressure levels alone higher COs (greater cardiac power) had the highest for the excessive HCVD. A prerequisite for initial values for left ventricular mass. examination of this hypothesis is a quantitative Vascular Factors (structural vascular abnormalltles). expression of that dual cardiac load. In the development of hypertension in the spontane- Preload or venous return is reflected quite pre- ously hypertensive rat, a progression of structural cisely by CO, since over time venous return and CO vascular changes occurs. In 6- to &week-old sponta- must be equal. Afterload is approximated by higher neousiy hyperisrlsive rats, capillary rarefaction is MAP. In the constant pressure-flow model of the followed by functional arteriolar rarefaction at 12 to circulatory system, the power the heart generates to 14 weeks and finally structural arteriolar rarefaction overcome its load equals the product of CO and at 16 to 18 weeks.34 Similarly, capillary rarefaction volume 116 Number 2, Part 2 Hemodynamic considerations in hypertension 597

Table Ill. Known and postulated effects of the hyperten- sive process on the balance between myocardial oxygen CARDIAC supply and demand t OUTPUT Increased demand Decreased supply t I tBP tHR Y _ I tHR Functional vascular I changes #ALPHA-TONE &/OR VASCULAR : tcontractility (early) &sponse to hypoxia ) MYOGENIC TONE RAREFACTtON ! ~EDV i 1 I I 1 ! I TCardiacpower iFTY;:ic me ,I I tLVMM i I Increased Supply Structural vascularchanges : ! Rarefaction : i tWa&lumen : + TPR __------_--__---J tDBP Diastolicdysfunction I Coronaryatherosclerosis I: (acceleratedby HTN) I: HR = hem-t rate; DBP = diastolic blood preesure; EDV = ends diastolic I volume; HTN = hypertension; LVMM = left ventricular muscle mess. : I LUMEN Ii has been reported in borderline hypertensive I: humans;= evidence for arteriolar rarefaction has I I: been observed in established human hypertensives.36 L,--,..------Interestingly, in borderline hypertensive subjects, the degree of rarefaction was not significantly re- lated to arterial pressure, but rather was inversely related to CO. In other words, borderline hyperten- Fig. 1. Proposed causesand effects of vascular changes in hypertension. sive patients with the highest CO displayed the greatest degree of capillary rarefaction. In contrast, elevated BP induces medial hypertrophy, 35 which ished coronary vasodilator reserve has been demon- predominantly involves arteries and larger arteri- strated in the hearts of hypertensive men and 01es.37 animals subjected to pressure overload from either Functional vascular abnormalltles. One of the most systemic hypertension or aortic stenosis.@ The widely reported and recognized abnormalities of the impaired vasodilator reserve in hypertensive sub- arterial circulation in hypertension is an impaired jects likely reflects some of the functional and vasodilator response to ischemia.38-40 The reduced structural vascular changes previously noted. More vasodilator response to 10 minutes of regional is- specifically, vascular rarefaction and an increased chemia partially reflects structural vascular wall-lumen ratio (concentric vascular hypertrophy) changes.38-40 However, hypertensive vessels may could provide a structural component to the dimin- maintain higher tone under some conditions, in part ished flow reserve. Rarefaction, even in the presence through decreased sensitivity to hypoxemia or its of normal total coronary flow levels, would increase metabolic consequences. 41 Enhanced vasodilator the mean diffusion distance from individual capil- responses to both calcium channel blockade42*43 (an laries to adjacent myocardial cells.47Thus the myo- index of myogenic tone) and a-receptor antago- cardial cells at greatest distance from nutrient sup- nists&,45 (an index of vascular a-tone) comprise ply would be at increased risk for ischemia. other well-documented functional abnormalities of Impaired vasodilator responses to ischemia”’ would hypertensive vessels. Explanations for these func- further exacerbate the effects of the structural and tional abnormalities are numerous, but agreement functional vascular changes. An increased vascular on specific mechanisms is not uniform. A hypothetic wall-lumen ratio,37 although not documented in the schematic depicting the causes and effects of vascu- coronary circulation,@ would accentuate vascular lar changes in the development and progression of responses to all vasoconstrictor stimuli.37, 3s,4g, 5a In hypertension is shown in Fig. 1. other words, when compared with a normal vessel, potential relationship of the structural and functional equivalent degrees of smooth muscle contraction in vascular changes to ischemlc heart disease. Dimin- a vessel characterized by greater wall thickness in 598 Egan and Schmouder relationship to luminal radius produce greater com- habit, relative weight, and ECG abnormalities to incidence of major coronary events: final report of the pooling project, promise of the lumen. 3’ Since resistance varies pooling project research group. J Chronic Dis 1978;31:201. inversely with the fourth power of the radius,51 the 3. Veterans Administration Cooperative Study Group on Anti- increased wall-lumen ratio would enhance vascular hypertensive Agents. Effects of treatment on morbidity in hypertension. II. Results in patients with diastolic blood resistance responses to all vasoactive compounds.38 pressure averaging SO through 114 mm Hg. JAMA 1970; The enhanced vasoconstrictor responsiveness would 213:1143-52. include that reflecting impaired endothelium- 4. Hypertension Detection and Follow-up Program Cooperative Group. Five-year findings of the hypertension detection and dependent vasodilation in hypertension.52 More spe- follow-up program. I. Reduction in mortality of persons with cifically, the vasoconstrictor effects of several com- high blood pressure, including mild hwertension.-- JAMA pounds, including thrombin, serotonin, and cate- 1979;242:2562-71. 5. Report of the Management Committee. The Australian ther- cholamines acting as a2-receptor agonists, are par- apeutic trial in mild hvpertenison.__ Lancet 198O;I: tially counterbalanced by endothelium-dependent 1261-7. vasodilation.53-55 6. Medical Research Council Working Party. MRC trial of treatment of mild hypertension: principal results. Br Med J The microvascular problems discussed previously 1985;291:97-104. would compound the adverse effects on myocardial 7. Fries ED. Should mild hypertension be treated? N Engl J perfusion of coronary macrovascular atherosclerosis, Med 1982;309:306-9. 8. McDonald DA. Blood flow in arteries. 2nd ed. London: which is accelerated by hypertension. If the struc- Arnold, 1974. tural and functional abnormalities interact diaboli- 9. Freis ED. Hemodynamics of hypertension. Physiol Rev cally in the coronary circulation as postulated, then 1960,40:727-54. 10. Conway J. Hemodynamic aspects of essential hypertension in one can envision how these factors would, over time, humans. Physiol Rev 1984;64:617-60. produce ischemia-related events, including angina, 11. Yin CP. Ventricular/vascular coupling. New York: Springer- infarction, arrhythmias, and cardiac decompensa- Verlag, 1986. 12. JuliusS,EslerM.Autonomicnervouscardiovascularregulation tion. The multiple effects of the hypertensive pro- in borderline hypertension. Am J Cardiol 1975;3&685-96. cess and the vascular changes in the balance 13. Messerli FH, Christie B, DeCarvalho JGR, et al. Obesity and between myocardial oxygen supply and demand are essential hypertension. Arch Int Med 1981:141:81-5. 14. Julius S, Conway J. Hemodynamic studies in patients with summarized in Table III. An extensive discussion of borderline blood pressure elevation. Circulation 1968;38: the separate risks of left ventricular hypertrophy 282-8. exceeds the scope of this discussion, but the topic 15. Egan BM, Schork NA, Julius S. Mechanism of inappropriate- ly normal peripheral resistance in overweight men. Circula- has been succinctly reviewed.56 tion 1987;76(suppl2):209 16. Egan BM. Schork NA. Increased vascular alpha-tone contrib- CONCLUSION utes to the inappropriately normal peripheral resistance in overweight men. Clin Res 1987;35:83OA. The hemodynamic dysregulation that occurs in 17. Hypertension Detection and Follow-up Program Cooperative the development and progression of hypertension Group. Effect of stepped-care treatment on the incidence of may contribute not only to elevated arterial blood myocardial infarction and angina pectoris. Five-year findings of the hypertension detection and follow-up program. Hyper- pressure, but may also act in concert with the higher tension 1984g(supplI):I198-1206. pressure levels to induce structural and functional 18. Conway J, Lauwers P. Hemodynamic and hypotensive effects cardiovascular changes. The anatomic and physio- of long-term therapy with chlorothiazide. Circulation 1960; 21:21. logic cardiovascular changes, which are not fully 19. Colfer HT, Cottier C, Sanchez R, Julius S. Role of cardiac reversed by antihypertensive therapy, provide the factors in the initial hypotensive action by beta-adrenorecep- substrate for subsequent complications. A greater tor blocking agents. Hypertension 1984;6:145-51. 20. Barret-Connor E, Khaw KT. Is hypertension more benign understanding of the sequential hemodynamic when associated with obesity? Circulation 1985;72:53. pathophysiology of hypertension may permit the 21. Lew EA, Garfinkel L. Variations in mortality by weight development of more effective treatment strategies. among 750,000 men and women. J Chronic Dis 1979;32:563. 22. McCue H Jr. The 1979 build end blood pressure study. In: Although unproved, physiologically based treatment Bostrom H, Ljungstedt N, eds. Medical aspects of mortality vs conventional approaches, which are targeted at statistics. Stockholm: Almauist & Wiksell International, correcting the underlying hemodynamic abnormali- 1981:182. 23. Hubert HB, Feinleib M, McNamara PM, Castelh WP. Obesi- ties in hypertension, may more effectively interrupt ty as an independent risk factor for cardiovascular disease: a the sequence of events leading to cardiac morbidity 26-year follow-up of participants in the Framingham Heart and mortality. Study. Circulation 1983;67968-77. 24. Smith WM. Epidemiology of hypertension. Med Clm North Am 1977;61:467. REFERENCtS 25. Kannel WB, Sorlie MS, Gordon T. Labile hypertension: a 1. Kannel WB, Dawber TR. Hypertension as an ingredient of a faulty concept? The Framingham Study. Circulation cardiovascular risk profile. Br J Hosp Med 1974;2:508. 1980;61:1183-7. 2. Relationship of blood pressure, serum cholesterol, smoking 26. Messerli FH, Ventura HO, Reisin E, E, et al. Borderline Volume 118 Number 2, Part 2 Hemodynamic considerations in hypertension 599

hypertension and obesity: two prehypertensive states with 42. Robinson BF, Dobbs RJ, Bayley S. Response of forearm elevated cardiac output. Circulation 1982;66:55-60. resistance vessels to verapamil and sodium nitroprusside in 27. Ellis CN, Julius S. Role of central blood volume in hyperki- normotensive and hypertensive men: evidence for a function- netic borderline hypertension. Br Heart J 1973;35:450-5. al abnormality of vascular smooth muscle in primary hyper- 28. Safar ME, Weiss YA, London CM, Frackowiak RD, Millie2 tension. Clin Sci 1982;63:33-42. PL. Cardiopulmonary blood volume in borderline hyperten- 43. Hulthen UL, Bolli P, Amann FW, Kiowski W, Buhler FR. sion. Clin Sci 1974;47:153-64. Enhanced vasodilatation in essential hypertension by calci- 29. Messerli FH, Sundgaard-Riise K, Reisin E, Dreslinski G, um channel blockade with verapamil. Hypertension 1982; Dunn FG. Frohlich E. Disnarate cardiovascular effects of obe- 4(suppl 11):1126-1131. sity and arterial hypertension. Am 3 Med 1982;74:808-12. 44. Amann FW, Bolli P, Kiowski W, Buhler FR. Enhanced 30. Kannel WB, Brand N, Skinner JJ Jr, Dawber TR, McNamara alpha-adrenoreceptor-mediated vasoconstriction in essential PM. The relation of adiposity to blood pressure and develop- hypertension. Hypertension 1981;3(suppl I):I119-23. ment of hypertension. The Framingham study. Ann Intern 45. Egan BM, Panis R, Hinderliter A, Schork N, Julius S. Med 1967;48-59. Mechanism of increased alpha-adrenergic vasoconstriction in 31. MacMabon SW, Blacket RB, Macdonald GJ, Hall W. Obesi- human essential hypertension. J Clin Invest 1987;80:812-7. ty, alcohol consumption, and blood pressure in Australian 46. Marcus ML, Harrison DG, Chilian WM, et al. Alterations in men and women. The National Heart Foundation of Austra- the coronary circulation in hypertrophied ventricles. Circula- lia risk factor prevalence study. J Hypertens 1984;2:85-91. tion 1987;74(suppl 1):1119-I!%. 32. Sokolow M. Clinical cardiology, 3rd ed. Los Altos, California: 47. Anversa P, Ricci R, Olivetti G. Effects of exercise on the Lange Medical Publishers, 1981. capillary vasculature of the rat heart. Circulation 1987; 33. Schieken RM, Clarke WR, Laur RM. Left ventricular hyper- 75&3uppl 1):112-8. trophy in children with blood ‘pressures in the upper quintile 48. Kamal MF, Campbell ACP. Medical hypertrophy in the of the distribution. The Muscatine study. Hypertension small intestinal arteries in systemic hypertension, renal and 1981;3:669. essential. J Path01 1979;129:99-110. 34. Prewitt RL, Chen IIH, Dowell R. Development of microvas- 49. Folkow B. Cardiovascular structural adautation: its role in cular rarefaction in the spontaneously hyperactive rat. Am J the initiation and maintenance of primary*hypert&sion. Clin Physiol 1982;243:H243-51. Sci 1978;55:3s-22s. 35. Sullivan JM, Prewitt RL, Joseph JA. Attenuation of the 50. Egan BM, Panis R, Hinderliter A, Schork N. Vascular microcirculation in young patientswith high-output border- structure enhances regional resistance responses in mild line hypertension. Hypertension 1983;5:844-51. hypertension. J Hype&s 1988;6:41-8. 36. Harper RN, Moore MA, Marr MC, Walls LE, Hutchins PM. 51. Poiseiulle JLM. Recherches exnerimentales surle mouvement Arteriolar rarefaction in the conjunctiva of human essential des liquides dans les tubes de tres petits diameters. Mem hypertensives. Microvasc Res 197&l&369-72. Savant Etragners 1846;9:433. 37. Folkow B. Grimbv G. Thulesius 0. Adantive structural 52. Locke& W. Otsuka Y. Carretero 0. The loss of endothelium- changes of the vascular walls in hypertension and their dependent vascular relaxation in hypertension. Hypertension__ relation to the control of the peripheral resistance. Acta 1986;8(suppl 11):1161-6. Physiol Stand 1958;44:255-72. 53. Cohen RA. Shenherd JT. VanHoutte PM. Endothelium and 38. Sivertsson R. The hemodynamic importance of structural asymmetrical responses ‘of the coronary arterial wall. Am vascular changes in essential hypertension. Acta Physiol Physiol Sot 247 (Heart Circ Physiol 16):1984;H403-8. Stand Suppl 1970;343:1-56. 54. Ku DD. Coronary vascular reactivity after acute myocardial 39. Conway J. A vascular abnormality in hypertension. A study ischemia. Science 1982;218:576. of blood flow in the forearm. Circulation 1986;27:520-9. 55. Matsuda H, Kuon E, Holtz J, Busse R. Endothelium- 49. Egan BM, Julius S. Vascular hypertrophy in borderline mediated dilations contribute to the polarity of the arterial hypertension: relationship to blood pressure and sympathetic wall in vasomotion induced by cu,-adrenergic agonists. J drive. Clin Exp Hypertens 1985,A7:243-55. Cardiovasc Pharmacol 1985;7:680-8. 41. Walsh GM, Tsuchyia M, Gox AC, Tobia AJ, Frohlich ED. 56. Frohlich ED. Potential mechanisms explaining the risk of left Altered hemodynamic response to acute hypoxemia in spon- ventricular hypertrophy. Am J Cardiol 1987;59:91A-7A. taneously hypertensive rata. Am J Physiol 1978;234:H275.