Objective: To evaluate pulsatile components of the blood pressure as risk markers for carotid stenosis in isolated systolic hypertension.
Design: Duplex scans with Doppler measures of the blood flow velocity were used to diagnose carotid stenosis in 187 participants in the Systolic Hypertension in the Elderly Program and in 187 normotensive and mildly hypertensive control subjects.
Methods: The systolic blood pressure (SBP), diastolic blood pressure (DBP), pulse pressure, and mean arterial pressure (MAP) were selected as independent variables. A logistic regression model for carotid stenosis was used to adjust for potentially confounding risk factors. Serial models, each containing single or double blood pressure variables, were run to compare risk markers for carotid stenosis. Receiver operating characteristic curves were compared to assess the predictive value of each model.
Results: In the multivariate analysis, both the SBP (P = 0.005) and the pulse pressure (P < 0.001) were predictive of carotid stenosis, but the DBP and MAP were not. However, when either the SBP or the pulse pressure was included in the model, the DBP was associated negatively with carotid stenosis (P < 0.001 and P = 0.023, respectively). An increased pulse pressure and a decreased DBP were independent risk markers for carotid stenosis. Comparison of receiver operating characteristic curves indicated that the pulse pressure had superior predictive value to the SBP (P = 0.034).
Conclusions: The pulse pressure is the single best predictor of carotid stenosis. There is a negative correlation between the DBP and carotid stenosis for subjects with isolated systolic hypertension, but this can be demonstrated only after one has stratified for the SBP or for the pulse pressure. Thus, the pulsatile components of the blood pressure, increased pulse pressure and decreased DBP, are the most sensitive risk markers for the diagnoses of carotid stenosis.
1Hypertension Center, Veterans Affairs Medical Center, Long Beach, and the University of California, Irvine
2The Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh
3The Department of Medicine, Brookdale University Hospital and Medical Center, State University of New York, Brooklyn, USA.
4Requests for reprints to Dr Stanley S. Franklin, Hypertension Center (E109), VA Medical Center, 5901 East Seventh Street, Long Beach, CA 90822, USA.
Sponsorship: This study was supported by National Institutes of Health grants HL-39871 and HL-50439. This work was done while K.S.T. held an Established Investigatorship from the American Heart Association.
Received 19 February 1997 Revised 11 June 1997 Accepted 25 June 1997