Aggressive Blood Pressure Lowering Is Dangerous: The J-Curve
2013; Lippincott Williams & Wilkins; Volume: 63; Issue: 1 Linguagem: Inglês
10.1161/hypertensionaha.113.01922
ISSN1524-4563
AutoresGiuseppe Mancia, Guıdo Grassı,
Tópico(s)Heart Rate Variability and Autonomic Control
ResumoHomeHypertensionVol. 63, No. 1Aggressive Blood Pressure Lowering Is Dangerous: The J-Curve Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBAggressive Blood Pressure Lowering Is Dangerous: The J-CurvePro Side of the Argument Giuseppe Mancia and Guido Grassi Giuseppe ManciaGiuseppe Mancia From the Centro Interuniversitario di Fisiologia Clinica e Ipertensione, Università Milano-Bicocca and Istituto Auxologico Italiano, Milano, Italy (G.M.); Clinica Medica,Dipartimento di Scienze della Salute, Università Milano-Bicocca, Milano, Italy (G.G.); and Multimedica, Sesto SanGiovanni (Milano), Italy (G.G.). and Guido GrassiGuido Grassi From the Centro Interuniversitario di Fisiologia Clinica e Ipertensione, Università Milano-Bicocca and Istituto Auxologico Italiano, Milano, Italy (G.M.); Clinica Medica,Dipartimento di Scienze della Salute, Università Milano-Bicocca, Milano, Italy (G.G.); and Multimedica, Sesto SanGiovanni (Milano), Italy (G.G.). Originally published18 Nov 2013https://doi.org/10.1161/HYPERTENSIONAHA.113.01922Hypertension. 2014;63:29–36Other version(s) of this articleYou are viewing the most recent version of this article. Previous versions: January 1, 2013: Previous Version 1 IntroductionHistorically, the possibility that reducing blood pressure (BP) might have harmful effects has gone through 3 temporal steps. In the first half of the last century, it was a widespread belief that an elevated BP provided the necessary hydraulic gradient to preserve organ perfusion in the face of an increase in systemic vascular resistance, the implication being that hypertension was a compensatory mechanism and no BP-lowering intervention was justified.1 This was largely forgotten in the subsequent 2 to 3 decades where the results of the antihypertensive treatment trials dominated the scene and emphasized the beneficial effects of BP reductions at most if not all degrees of BP elevation.2–6 The possibility that antihypertensive treatment might produce harm rather than benefit resurfaced in the late 70s and 80s,7,8 however, due in particular to a report that in patients with a high cardiovascular risk the incidence of myocardial infarction was diminished by reducing diastolic BP to 85 to 90 mm Hg but increased when a more pronounced reduction occurred.8 Despite its important limitations (small number of patients, few cardiac events, and retrospective nature of the observations), the report was widely referred to and revived the popularity that has since characterized the J-curve phenomenon, that is, the possible J-shaped rather than linear relationship of BP reductions by treatment with incident cardiovascular events.The present article will argue in favor of this hypothesis. It will be acknowledged that at present this is not supported by undisputable evidence such as that provided by randomized trials. It will be emphasized, however, that the J-curve hypothesis is supported by common sense, physiological data, and results from large-scale observational studies, often derived from antihypertensive treatment trials on individuals at high or very high cardiovascular risk. This makes it not just a remote but a real possibility, which deserves to be addressed by future studies.9 Its important implications for clinical practice should make investigation on the J-curve phenomenon a priority for cardiovascular medicine.Background Considerations in Favor of the J-CurveAs discussed in a recent editorial,10 the hypothesis that the BP achieved with treatment bears a J-shaped relationship with incident cardiovascular events owes an important portion of its persistent popularity to common sense considerations and physiological data. Common sense tells that a BP value below which organ perfusion and function are compromised ought to exist, and that a too low BP target must have an adverse effect on morbidity and mortality. This is confirmed by physiological studies that have shown that in vital organs such as the brain, the heart, and the kidneys, autoregulatory mechanisms cause vasodilatation and maintain perfusion constant when BP is progressively reduced.11–13 They have further shown, however, that there is a BP value below which a further progressive reduction is accompanied by a steep progressive blood flow fall, and that this BP threshold or nadir is reset to higher values in hypertension14 and is increased also in the presence of asymptomatic organ damage or overt disease, presumably because in a damaged organ alterations of structure and function impair the vascular smooth muscle ability to relax, compromise the vasodilatatory reserve, and make perfusion more critically dependant on an adequate arteriovenous BP gradient. The top panel of Figure 1 shows that a diastolic BP reduction to <90 mm Hg, obtained via intravenous injection of nitroprussiate, led to a marked reduction of coronary blood flow in hypertensive patients with left ventricular hypertrophy, whereas in hypertensive patients without left ventricular hypertrophy, no alterations of coronary perfusion occurred down to diastolic values of ≈70 mm Hg.15 The bottom panel shows that reducing diastolic BP to <80 mm Hg markedly increased the incidence of cardiovascular (mainly cardiac) events in patients with coronary disease, the effect being much less pronounced in those who had undergone a revascularization procedure.16 The available physiological data have several important implications. One, the J-curve hypothesis needs a reformulation. The question should not be whether a J-curve phenomenon exists: it obviously does, with an ascending limb that will reach a 100% mortality at zero BP. It should rather be whether the ascending limb can become manifest at the BP values achieved with drugs that lower BP. Two, there is not just a single but probably many nadir BP values to refer to in order to avoid organ underperfusion and increased risk, depending on the characteristics of the patients. Three, although antihypertensive treatment may move blood flow autoregulation back toward the normotensive range,14 in hypertension as well as in cardiovascular disease the risk of reducing BP below the range over which perfusion is preserved may be greater.Download figureDownload PowerPointFigure 1. A, The effects of reducing coronary perfusion pressure by intravenous infusion of nitroprusside on coronary blood flow (measured in the great cardiac vein) in hypertensive patients with and without left ventricular hypertrophy (LVH). B, The cardiovascular event incidence at different achieved diastolic blood pressure (DBP) levels in patients with coronary artery disease (CAD) who did not undergo coronary revascularization compared with those who had the procedure. HT indicates hypertension. Adapted with permission from Polese et al15 and Messerli et al.16 Authorization for this adaptation has been obtained both from the owner of the copyright in the original work and from the owner of copyright in the translation or adaptation.Evidence in Favor of the J-CurveA large number of observational studies, many of which obtained by post hoc analyses of event-based trials, have observed a J-shaped relationship between BP reductions and some or all measured cardiovascular events, that is, no additional but rather an attenuation or a disappearance of the BP-dependant cardiovascular protective effects with greater compared with more modest BP reductions from the initially higher levels.16–37 The results can be summarized as follows. One, an increased incidence of cardiovascular outcomes has been seen for more versus less pronounced reductions of both systolic and diastolic BP. Two, in line with the view that nadir BP values differ in different subjects, the systolic and diastolic BP below which the cardiovascular risk increased has shown marked between-study differences.10 Three, the J-curve phenomenon has been observed for fatal and nonfatal events as well as major cardiovascular (heart failure, myocardial infarction, and stroke) and renal complications (serious deterioration of renal function or end-stage renal disease). Four, the above observations have been made in widely different patient populations, although in most instances individuals with a high or very high cardiovascular risk (because of conditions such as a marked initial systodiastolic BP elevation, isolated systolic hypertension, old age, multiple cardiovascular risk factors, organ damage, or clinical cardiovascular disease) have been involved. To quote some representative examples from recent large-scale studies with several years of follow-up, in the 22 576 hypertensive patients with coronary disease recruited for the International Verapamil SR/T Trandolapril trial (INVEST), the incidence of cardiovascular morbidity and mortality showed a reduction in patients in whom diastolic BP was reduced to between 80 and 89 mm Hg, with a clearcut progressive increase, however, in those in whom values <80 and 140 to between 130 and 139 mm Hg. The incidence increased, however, when BP values <130 mm Hg were achieved, with a risk that became more than double that of patients in the higher BP category when on-treatment BP fell to 110 mm Hg (Figure 3). Finally, in the 20 332 patients with a history of stroke of the Prevention Regimen for Effectively Avoiding Second Strokes trial (PRoFESS), reducing systolic BP from ≥150 to 140 to 149 and to 130 to 139 mm Hg was found to reduce progressively and markedly the risk of cardiovascular morbidity and mortality. A further reduction to <130 and <120 mm Hg increased the risk again (+16% and +31%, respectively), however, with values that were even greater than those exhibited by patients remaining at the 140 to 149 mm Hg on-treatment BP range37 (Figure 4, left).Download figureDownload PowerPointFigure 2. Incidence of cardiovascular outcomes (primary end point) according to the percentage of visits in which blood pressure (BP) was reduced to <130/80 mm Hg with resulting different on-treatment mean BP values (left). The 3-fold end point to cardiovascular morbidity and mortality except for incident heart failure. D indicates diastolic; and S, systolic. Data from the ONTARGET trial. Adapted with permission from Mancia et al.28 Authorization for this adaptation has been obtained both from the owner of the copyright in the original work and from the owner of copyright in the translation or adaptation.Download figureDownload PowerPointFigure 3. The cumulative event incidence of primary outcome (nonfatal myocardial infarction, nonfatal stroke, and all-cause mortality in 6400 patients with diabetes mellitus of the INVEST trial in whom systolic blood pressure (SBP) remained ≥140 mm Hg or was reduced to between 130 and 139 mm Hg and <130 mm Hg (left). The risk of all-cause mortality in subgroups of patients achieving different systolic BP values 2 groups of patients are available. This limits the contribution to the J-curve issue of both the UKPDS46 and the ACCORD trial.47 In the ACCORD trial,47 the cardiovascular event incidence was similar at the achieved systolic BP of 133 and 119 mm Hg, but this cannot exclude the possibility of a lower risk of events (and thus a J-curve) in a third group with an on-treatment BP value in-between. At present, the effects of different achieved BP values have been compared in 3 randomized groups only in the Hypertension Optimal Treatment trial (HOT), with no evidence of a J-curve either in the entire low cardiovascular risk trial population and in the small diabetic subgroup.48 Unfortunately, in both instances the treated BP exhibited only small between-group differences as well as mean diastolic and systolic values, respectively >80 and 140 mm Hg, that is, above the BP range where a J-curve is more likely to occur. Thus, although randomized trials do not provide evidence in favor of the J-curve phenomenon, they do not disprove it either.EpidemiologyA widely used argument against the J-curve phenomenon is that in a large number of epidemiological studies on a total of ≈1 million subjects, the relationship between BP and cardiovascular events seems to be linear from BP values of ≈180/110 mm Hg to BP values of ≈110/70 mm Hg, this being the case in both younger and older subjects, including those aged ≥80 years.49,50 This has greatly contributed to the belief that BP reductions exert a clearcut protective effect at any initial BP level, and treatment strategies should be based on the principle that the lower the BP the better is for the patient. However, a close look at the above epidemiological data allows to identify an upward deviation from linearity at the lowest diastolic BP values. Furthermore, the lowest explored BP values have been associated with an increased incidence of cardiovascular morbid and fatal events in several individual epidemiological studies, particularly in elderly subjects.51–53 A recent example is provided by the large database obtained by the MOnica, Risk, Genetics, Archiving and Monograph (MORGAM) Project, in which the risk of stroke showed a progressive reduction at progressively lower diastolic BP values, with a J-shaped increase, however, <71 mm Hg in the elderly fraction of the study population.53 Finally, the described linearity between BP and cardiovascular events has been recognized to be somewhat misleading because it originates from the use of a logarithmic scale to quantify the event incidence, which changes little in the lower portion of the BP range when absolute values are used. More in general, epidemiological evidence should never be considered a proxy for the effects of active treatment because (1) the risk may be totally or in part irreversible; (2) the treatment strategies that are used may have unforeseen effects that modify or mask the expected benefit; and (3) subjects recruited for epidemiological studies may differ from those in whom treatment is tested. This has been common in hypertension research in which the effects of BP-lowering interventions have mostly been studied in longstanding hypertensive patients, often with a history of cardiovascular events and almost invariably with a cardiovascular risk much higher than that of subjects recruited for epidemiological purposes. That this can make a difference for the risk of an event is supported by the observation that the relationship between BP and incident myocardial infarction is linear in subjects without and J-shaped in those with a history of coronary disease.54Organ HeterogeneityIn recent years, the hypothesis has been made that vital organs may respond differently to BP-lowering interventions. That is, BP reductions may have a limited protective effect on coronary events, with an ultimate flattening or even a reversal at low values, whereas they may progressively reduce the incidence of stroke and end-stage renal disease throughout the BP range achievable by antihypertensive treatment.55,56 As reviewed in some recent articles,57–59 the evidence is the following. One, in the ACCORD trial a systolic BP reduction to <120 mm Hg did not have a significant effect on myocardial infarction, but it reduced substantially the risk of stroke.47 Two, in the Perindopril Protection against Recurrent Stroke Study (PROGRESS), a reduction in the incidence of stroke recurrence was seen also at a baseline systolic BP between 120 and 139 mm Hg, the treatment-induced BP fall leading to on-treatment systolic values well <130 mm Hg.60,61 Three, in post hoc analyses of some trials (eg, INVEST, ONTARGET), reducing systolic BP to <130 or even <120 mm Hg was accompanied by a reduction of stroke with an increased incidence of myocardial infarction.16,27,28 Four, similarly marked BP reductions have been shown to (1) more markedly reduce urinary protein excretion28,62,63; (2) delay incipient nephropathy as identified by new onset microalbuminuria or proteinuria63; and (3) attenuate the progressive decline of renal function accompanying renal disease.64These intriguing observations might find their explanation in a more effective blood flow autoregulation of the brain and the kidney compared with the heart, possibly because in extracardiac organs perfusion is not impaired during the systolic phase. It should be emphasized, however, that not all studies are consistent with the hypothesis that the brain and the kidney are not exposed to the J-curve phenomenon, within the BP range at which it seems to occur in the heart because (1) the ability of urinary protein excretion to predict renal outcomes is controversial,9 and (2) recent reports rather seem to indicate that reducing BP to <130 or <120 mm Hg may worsen renal function and increase the risk of end-stage renal disease,35,65 adding support to the common experience that a marked BP reduction may lead to a marked increase in serum creatinine, with thus an impairment of renal function. Furthermore, some epidemiological data have shown an increase in the risk of stroke at the lower BP values exhibited by the study population,53 and the relationship between BP reductions and stroke observed in trials seems to be much more complex than linear, an important role being played by the clinical condition of the patient, the timing of antihypertensive treatment initiation, the speed of the BP change, and the type of cerebrovascular event considered. It is well known that, if obtained few hours after a stroke, BP reductions may have deleterious effects, and a J-shaped increase in the incidence of stroke recurrence when BP was reduced to <130 or <120 mm Hg (+10% and +29%, respectively; Figure 4, right) has recently been reported also when antihypertensive treatment was started at a few weeks distance from the acute cerebral event.37 To date, a linear relationship with BP reductions down to low absolute levels seems to more consistently hold only for hemorrhagic stroke, which in the PROGRESS trial has exhibited a striking reduction in patients in whom chronic antihypertensive treatment led to a systolic BP <120 mm Hg,60,66 these low achieved BP values being associated also with the smallest incidence of intracranial bleeding in patients under anticoagulant treatment.66–68 However, hemorrhagic stroke accounts for only a minor fraction of the overall stroke prevalence.68 Furthermore, a progressive risk reduction as BP falls to low values may not invariably occur even for this event. This is the implication of a recent trial in which an early systolic BP fall to 2 groups of patients randomized to different BP targets will be needed; (2) the degree of BP reduction by treatment can be unpredictable; (3) different BP-lowering effects can only be obtained through different numbers and doses of antihypertensive drugs, which means that the role of BP versus the treatment differences may remain uncertain; and (4) the BP threshold below which underperfusion and damage occur may exhibit wide interpatient differences (as suggested by the marked discrepancies in the nadir BP values below which cardiovascular event rate increased as found by different J-curve–positive studies) and perhaps change over time and progression of the disease even within a given individual. This can make the prevailing type of effect offered by trials of limited help for the decision to take in any single patient.At practical level, physicians will have to decide by the principle of primum non nocere, which means that the possible existence of a J-curve should never be overlooked and represent a primary concern in conditions such as coronary disease, pronounced degrees of organ damage, a recent cardiovascular event, or an old age, that is, when vital organs may be particularly exposed to the risk of underperfusion by treatment-induced BP falls. This concern should be shared by guidelines that should avoid recommending aggressive BP targets in the above conditions and rather opt for safer, more conservative strategies. This has been done in the recent European hypertension guidelines, which have elevated the BP values to be reached with treatment in patients with diabetes mellitus, a previous cardiovascular disease, or an advanced renal disease from <130/90 to <140/90 mm Hg (140/85 mm Hg in diabetes mellitus), the systolic values to pursue in the elderly being between 140 and 150 mm Hg9DisclosuresNone.FootnotesThe opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.This paper was sent to Takayoshi Ohkubo, Guest editor, for review by expert referees, editorial decision, and final disposition.Correspondence to Giuseppe Mancia, Centro di Fisiologia Clinica e Ipertensione, Via Francesco Sforza, 35, 20121 Milano, Italy. E-mail [email protected]References1. Pickering G. High Blood Pressure. 2nd ed. London: J&A Churchill;1968.Google Scholar2. Veterans Administration Cooperative Study Group on Antihypertensive Agents. Effects of treatment on morbidity in hypertension. II. Results in patients with diastolic blood pressure averaging 90 through 114 mm Hg.JAMA. 1970; 213:1143–1152.CrossrefMedlineGoogle Scholar3. Medical Research Council Working Party. MRC trial of treatment of mild hypertension: principal results.Br Med J. 1985; 291:97–104.CrossrefMedlineGoogle Scholar4. Management Committee. The Australian therapeutic trial in mild hypertension.Lancet. 1980; 1:1261–1267.MedlineGoogle Scholar5. Amery A, Birkenhäger W, Brixko P, Bulpitt C, Clement D, Deruyttere M, De Schaepdryver A, Dollery C, Fagard R, Forette F. Mortality and morbidity results from the European Working Party on High Blood Pressure in the Elderly trial.Lancet. 1985; 1:1349–1354.CrossrefMedlineGoogle Scholar6. Hypertension Detection and Follow-up Program Cooperative Group. Five-year findings of the hypertension, detection and follow-up program. II. Reduction in stroke incidence among persons with high blood pressure.JAMA. 1982; 247:633–638.CrossrefMedlineGoogle Scholar7. Stewart IM. Relation of reduction in pressure to first myocardial infarction in patients receiving treatment for severe hypertension.Lancet. 1979; 1:861–865.CrossrefMedlineGoogle Scholar8. Cruickshank JM, Thorp JM, Zacharias FJ. Benefits and potential harm of lowering high blood pressure.Lancet. 1987; 1:581–584.CrossrefMedlineGoogle Scholar9. Mancia G, Fagard R, Narkiewicz K, et al; Task Force Members. 2013 ESH/ESC Guidelines for the management of arterial hypertension: the Task Force for the management of arterial hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC).J Hypertens. 2013; 31:1281–1357.CrossrefMedlineGoogle Scholar10. Zanchetti A. Blood pressure targets of antihypertensive treatment: up and down the J-shaped curve.Eur Heart J. 2010; 31:2837–2840.CrossrefMedlineGoogle Scholar11. Lucas SJ, Tzeng YC, Galvin SD, Thomas KN, Ogoh S, Ainslie PN. Influence of changes in blood pressure on cerebral perfusion and oxygenation.Hypertension. 2010; 55:698–705.LinkGoogle Scholar12. Hall JE, Guyton AC, Brands MW. Control of sodium excretion and arterial pressure by intra-renal mechanisms and the renin-angiotensin system in hypertension.In:, Laragh JH, Brenner BM, eds. H
Referência(s)