The J-Curve Theory in High Blood Pressure: A Really Dangerous Curve

By Cristiano Antonino On Oct 31, 2022

The J-curve hypothesis represents one of the most debated issues in the recent, but also current, treatment of arterial hypertension. I.M. Stewart, in an article published in 1979 in the Lancet, proposed the existence of a relationship between the reduction of blood pressure and myocardial infarction in patients with severe hypertension under treatment (1)

J-curve of arterial hypertension, the Stewart hypothesis

Stewart referred to a first myocardial infarction in hypertensive subjects, and with severe hypertension, in drug therapy, thus circumscribing the area under study in a very precise way.

It should be remembered immediately that Stewart’s observation emerged from a “case series” study, that is, an observational study, not prospective, nor specifically designed to test this hypothesis.

The study design represents the key methodological element for the researcher interested in studying a clinical hypothesis; a question formulated in rigorous terms, focused on examining a weighty clinical problem is, as the method of Evidence-Based Medicine teaches, the very basis of the actions of the researcher, clinician and epidemiologist (2, 3).

In the case of blood pressure control, and the J-curve hypothesis in particular, the clinical question of interest is whether an aggressive reduction (compared to a less aggressive reduction) in blood pressure results in more unfavorable clinical outcomes in hypertensive patients, and not if the excessive reduction in blood pressure is harmful to the hypertensive patient.

In fact, in the field of hypertension therapy, the risk of excess treatment (even involuntary) does not seem to be a particularly significant problem; the real problems are those of insufficient, if not incorrect, treatment and poor adherence to therapy for the hypertensive patient.

To have, or at least to try to find, a rational, comprehensive and evidence-based answer, it is necessary to identify the appropriate clinical studies

The association between low diastolic blood pressure values and unfavorable clinical outcomes must be verified and dimensioned, evaluating their strength and consistency in different, but all of high quality, studies.

Strength and constancy of the association are therefore essential requirements to speak with full knowledge of the cause of association.

To then define a causal link, minimum criteria must be met such as the presence of a risk gradient with increasing duration and intensity of exposure; the occurrence of a constant and quantitatively comparable association from study to study analyzed; the existence of a sequential temporal association, meaning that the exposure must precede the observed outcome; the essential need for an appropriate pathophysiological explanation, that is to trace a biological plausibility for the observed association.

The reader of Evidence-Based Medicine will not miss that those listed above are nothing more than the requisites necessary to transform an element of risk into a risk factor in the full sense.

Indeed, if one compares the event frequencies in patients treated with diastolic blood pressure levels below or equal to 90 mmHg with the event frequencies in patients with diastolic blood pressure values above 90, as emerged from the studies available in the literature, one can appreciate how the relative risk (RR) is higher than 3 only in Stewart’s study, which dates back to more than 20 years ago (4, 5, 6).

Three, it should be remembered, represents the threshold above which we can speak of a real association between exposure and increased risk for the outcome studied in observational studies (“case series”, case-control, cohort) (2).

In the other studies (5, 6) the RR is around 1, thus indicating a very limited force in the association between “low” diastolic pressure, unfavorable clinical outcomes and long-term prognosis.

In particular, the RR calculated for the study by J. Merlo (6), a population-based cohort study including 484 male patients, was approximately 1.7 for the comparison between the group of patients with lower or lower diastolic blood pressure. equal to 90 mmHg (267 subjects), and that with blood pressure higher than 90 mmHg (217 subjects), thus suggesting a weak and insignificant association between “low” diastolic pressure and poor prognosis.

The hypothesis of the J-curve, as pointed out at the beginning, therefore emerged from a “case series”, therefore, a type of study often subject to selection and measurement errors (“bias”) (7)

Randomized controlled trials provide more solid evidence of causality, although they are rarely designed to assess harm (more often than not they are designed to assess the benefit of a therapeutic intervention).

It is true that the subgroup analysis can be used to explore the extent of the damage or harm, but in this case there are not a few “caveats” that this analysis requires.

And in fact, as regards in particular the hypothesis of the J-curve in arterial hypertension, there is a clear discrepancy between the trend found in the subgroups, a trend that would somehow indicate the presence of an increased cardiovascular risk in subjects with “low” diastolic blood pressure values, and the trend in the populations considered in full in the same studies (6, 7, 8, 9), a trend that instead strongly argues against the existence of a J-curve.

Today, in 2000, the integrated efforts of doctors and patients are aimed at achieving optimal blood pressure control.

The current state of the art, and the latest WHO guidelines on the global management of the hypertensive patient are confirmation of this (10), therefore suggests directing the strength of the doctor-patient therapeutic alliance towards improving blood pressure control, not only for the purpose of preventing cerebral stroke, but more generally of all cardiovascular diseases and in particular of myocardial infarction.

This objective must be pursued bearing in mind that the patient who meets daily in the clinic or in the outpatient clinic rarely follows the profile of the patient enrolled in the clinical study and that in any case the intervention strategies must be widely shared and tailored to the individual patient we face. (11, 12, 13).

This approach is a guarantee of ethical and methodological correctness, and contributes to shedding full light on clinical controversies, such as that of the J-curve, which, on the basis of the best evidence available today, seem to have no reason to exist.

J-curve, bibliography:

Stewart IM. Relation of reduction in pressure to first myocardial infarction in patients receiving treatment for severe hypertension. Lancet 1979;1:861-5.
Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence-Based Medicine. How to practice and teach EBM. London, Churchill Livingstone, 1997.
Gensini GF, Galanti G, Conti AA. La medicina basata sulle evidenze: prospettive, applicazioni e confini. Il Policlinico 1998;105:592-601.
McAlister FA. Using evidence to resolve clinical controversies: is aggressive antihypertensive therapy harmful? Evidence-Based Medicine 1999;4:4-6.
Cooper SP, Hardy RJ, Labarthe DR, et al. The relation between degree of blood pressure reduction and mortality among hypertensives in the hypertension detection and follow-up program. Am J Epidemiol 1988;127:387-403.
Merlo J, Ranstam J, Liedholm H, et al. Incidence of myocardial infarction in elderly men being treated with antihypertensive drugs: population based cohort study. BMJ 1996;313: 457-61.
Collins R, Peto R, MacMahon S, et al. Blood pressure, stroke, and coronary heart disease. Part 2, Short-term reductions in blood pressure: overview of randomised drug trials in their epidemiological context. Lancet 1990;335:827-38.
U.K. Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. BMJ 1998;317: 703-13.
Hansson L, Zanchetti A, Carruthers SG, et al. Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment (HOT) randomised trial. HOT Study Group. Lancet 1998;351:1755-62.
1999 World Health Organization – International Society of Hypertension Guidelines for the Management of Hypertension. Guidelines Subcommittee. J Hypertens 1999;17:151-83.
Gensini GF, Conti AA. EBM e pratica clinica: quanto è simile il paziente dello studio clinico al paziente della realtà quotidiana? Evidence-Based Medicine (edizione italiana) 1999;3(2):3-4.
Gensini GF, Conti AA. Il valore prognostico della definizione diagnostica. Considerazioni metodologiche sull’impatto prognostico dei nuovi criteri diagnostici di diabete mellito (DM) dell’American Diabetes Association. Evidence-Based Medicine (edizione italiana) 1999;3(3):3-4.
Smulyan H, Safar ME. The diastolic blood pressure in systolic hypertension. Ann Intern Med 2000;132:233-7.