Objectives: The aim of this study was to determine whether pulse pressure is a risk factor for coronary artery disease using data from the MRC trial of treatment of mild hypertension, and whether the effect of anti-hypertensive drug therapy on pulse pressure may be a determinant of outcome in treated patients.
Methods: Logistic regression and Cox regression analyses were used to compare systolic and diastolic blood pressure, pulse pressure and mean blood pressure as predictors of coronary events and stroke in the MRC Mild Hypertension Trial. The effects of anti-hypertensive drug treatment with bendrofluazide and propranolol on pulse pressure were assessed using 1-year follow-up data. Event rates in the placebo-treated group and responses to anti-hypertensive treatment were measured in quartiles of age-adjusted entry pulse pressure. A 'four-corners' analysis was performed, with subjects divided into the upper and lower halves of the distributions of systolic and diastolic blood pressure at entry.
Results: Pulse pressure was a stronger predictor of coronary events than systolic, diastolic or mean blood pressure in males by logistic regression. Pulse pressure was similar to systolic pressure as a coronary event predictor on Cox regression. Stroke was predicted most strongly by mean blood pressure. Fatal and non-fatal coronary event rates increased progressively in ascending quartiles of age-adjusted pulse pressure, but there was also a strong correlation with systolic blood pressure. The values of partial logistic regression coefficients in models containing both systolic and diastolic blood pressure also supported a role for pulse pressure in predicting coronary events and for mean blood pressure in predicting stroke. Coronary risk, but not stroke, was inversely related to diastolic blood pressure in the four-corners analysis. In a Cox model, regressions of coronary event probability on systolic blood pressure at entry were significantly and inversely related to diastolic blood pressure categorized in quartiles. Bendrofluazide but not propranolol decreased pulse pressure significantly and was associated with a reduction in cardiovascular events overall, but no definite relationship between the effect of drugs on pulse pressure and specific responses to treatment was seen.
Conclusion: Pulse pressure is a strong risk factor for coronary events in untreated hypertensive male subjects in the MRC Mild Hypertension Trial, whereas stroke is best predicted by mean blood pressure. Bendrofluazide and propranolol have different effects on pulse pressure which may be related to their relative efficacy in the treatment of hypertension, but this possibility requires further study in more suitable populations.