We assessed whether there is a paradoxical increase in cardiovascular events with lower blood pressure values among patients with hypertension and coronary artery disease (CAD) who were enrolled in the International Verapamil-Trandolapril Study (INVEST). The relationship between systolic and diastolic pressure and the risk of primary outcome followed a J-curve pattern, with the relationship being relatively weak for systolic pressure but more significant for diastolic pressure. Our data indicate that excessive lowering of diastolic pressure in hypertensive patients with CAD should be avoided.
For many years, there has been concern that lowering blood pressure excessively, especially diastolic pressure, leads to a paradoxical increase in cardiovascular morbidity and mortality, following a J-curve pattern. This concern stems from the fact that coronary perfusion occurs mainly during diastole and, hence, excessive lowering of diastolic blood pressure may be detrimental. In fact, 1 meta-analysis showed a J-shaped association between cardiac events and diastolic blood pressure, but this relationship was not shown for stroke and diastolic blood pressure.1 However, these findings are inconsistent with the views presented in the Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC VI), which asserts that the relationship between blood pressure and the risk of cardiovascular events is “strong, , and independent (italics added).”2
If, indeed, there is a J-shaped relationship between excessively low diastolic blood pressure and cardiovascular morbidity, based on the theoretical concerns of coronary hypoperfusion, it should be obvious in patients with already impaired coronary circulation, such as those with established coronary artery disease (CAD). To evaluate this possibility, we performed an analysis using the International Verapamil-Trandolapril Study (INVEST) cohort, which included patients with hypertension and CAD.
Subjects and methods
A total of 22,576 subjects with hypertension and CAD were included in the analysis. Subjects were considered to have CAD if they had 1 of the following: ischemia on 2 different stress testing modalities (radionuclide scans, electrocardiography, or echocardiography); a diagnosis of classic angina pectoris; more than 50% narrowing of at least 1 major coronary artery as shown on coronary angiography; or confirmed myocardial infarction (MI) ≥ 3 months before enrollment in the study.
Subjects were randomly assigned to receive treatment with atenolol (Tenormin) or sustained-release verapamil (Calan, Covera, Isoptin) using a PROBE (Prospective, Randomized, Open treatment with Blinded end point Evaluation) design. The first incidence of all-cause mortality, nonfatal MI, or nonfatal stroke as shown by intention-to-treat analysis was the primary end point of the study. The main secondary end points were the 3 elements considered separately. The results of the primary study showed that the 2 treatment strategies were equivalent with respect to the main end point, main secondary outcomes, and on-treatment systolic and diastolic pressures.3 For this analysis, the average post-baseline blood pressure (systolic and diastolic) was categorized in 10 mm Hg increments, and its relationship with the primary and individual end points were determined.
The primary end point occurred in 10.1% of subjects (n = 2269) after a follow-up of 61,835 patient-years. The relationship between systolic and diastolic blood pressure and the risk of the primary outcome followed a J-curve pattern. The J-curve relationship was comparatively weak for systolic pressure (
) but was more significant for diastolic pressure (
). In fact, the incidence of the primary end point doubled (17.4%) when the diastolic blood pressure was < 70 mm Hg and tripled (31.8%) when it was < 60 mm Hg (
For the end point of all-cause mortality, a J-shaped relationship with diastolic pressure was found (data not shown). As illustrated in
, a J-shaped relationship was also shown for fatal and nonfatal MI and, to a lesser degree, for fatal and nonfatal stroke. The relationship was present even after controlling for baseline clinical variables.
As the diastolic blood pressure decreased, a significant interaction between revascularization and the risk of the primary end point was noted, suggesting that patients who underwent revascularization before enrollment tolerated lower diastolic pressure relatively better than those who did not undergo revascularization.
The results of the present study show that a J-shaped association between the risk of the primary end point and systolic and diastolic blood pressures exists, with the relationship being more evident for diastolic than for systolic blood pressure. There is also a comparable association between diastolic blood pressure and the risk of each of the individual end points of all-cause mortality, fatal and nonfatal MI, and, to a lesser extent, stroke.
This J-curve relationship may be explained by 3 possible pathophysiologic mechanisms:
1. The blood flow to target organs, including to the heart, may be reduced by low diastolic blood pressure, which may cause cardiac ischemia. In 1 study, ischemic events were associated with diastolic rather than systolic hypotension in 13 of 14 occurrences.4 The researchers deduced that the low diastolic blood pressure had a temporally causal association with silent and symptomatic ischemia. In the current study, we noted that patients who were revascularized were able to tolerate a lower diastolic blood pressure, lending further credence to the hypothesis that myocardial perfusion may be compromised at low diastolic pressures.
2. A rise in pulse pressure, which is an indication of large artery stiffness and, hence, could be a sign of serious vascular disease, may result with low diastolic pressure. However, as has been published elsewhere,5 the drop in diastolic blood pressure paralleled the decrease in systolic blood pressure, although not to the same extent, making this mechanism less likely.
3. Low diastolic blood pressure may be related to an underlying chronic disease, which may be responsible for increased morbidity and mortality.
Compared with subjects with high systolic blood pressure, our subjects with low systolic blood pressure were more inclined to be male and to be leaner. They also were more likely to have heart failure, cancer, and MI. Compared with subjects with high diastolic pressure, those with low diastolic pressure were more inclined to be white, female, leaner, and older. They also were more likely to have diabetes, heart failure, cancer, and MI. However, even after controlling for these baseline variables, the adjusted model showed a J-shaped relationship between blood pressure variables and cardiovascular outcomes.
We showed that diastolic blood pressure < 70 to 80 mm Hg is associated with an increased risk of all-cause mortality and MI among patients with CAD and hypertension. Patients with increased systolic pressure are frequently undertreated, even though high systolic pressure is a strong risk factor for MI and stroke. However, results of our study indicate that caution should be used to avoid lowering diastolic blood pressure excessively in patients with hypertension and CAD.5 These results provide support for the assertion in JNC 7 guidelines that if diastolic blood pressure is low, patients with occlusive CAD have an increased risk of coronary events.6