Vascular Mechanics and Stroke: Study the Arteries But Do Not Forget the Flow

To the Editor: We have read with interest the article “Usefulness of Vascular Wall Deformation for Assessment of Carotid Arterial Stiffness and Asso ciation With Previous Stroke in Elderly”1 by Tsai et al. Since the first publication from Oishi et al.2 in 2008, a number of studies have been performed to validate vascular circumferential strain and strain rate in the ascending, descending thoracic aorta, the abdominal aorta, or the carotid arteries. Locally assessed carotid stiffness can influence carotid strain (CS) and carotid strain rate (CSR), but it is also conceivable that both vascular deformation indexes could be influenced by blood flow. In fact, we have recently demonstrated in a group of consecutive patients with moderate to severe aortic stenosis that the most important determinant of circumferential ascending aortic strain was the stroke volume index.3 It remained the strongest predictor of aortic circumferential strain when adjusted for covariables such as the stiffness index, systemic resistance, and vascular compliance and impedance. This means that circumferential vascular strain was highly dependent on the aortic flow, in a similar way as the left atrial and the left ventricular (LV) longitudinal strain are influenced by phasic volume changes.3 In the study of Tsai et al.,1 CS and CSR were significantly correlated with local vascular stiffness, but the influence of blood flow was not taken into account. We believe that noninvasive, echo-Doppler–derived carotid blood flow should have been performed because blood flow could probably influence CS and CSR. Moreover, a number of echocardiographic variables such as LV ejection fraction, LV diastolic function and filling pressures, and LV mass index could have been assessed because they influence LV myocardial performance, stroke volume, carotid blood flow, and, consequently, CS and CSR. This information would have been especially important in elderly hypertensive patients with a history of stroke. In fact, approximately 25% of all ischemic strokes are due to cardiac embolism, with heart failure and hypertension being two major risk factors for stroke.4 We therefore believe that adjustment for those covariables should have been done, to validate and improve the quality of the multivariable model for the prediction of stroke. The authors declared no conflict of interests.