The mechanisms leading to vasovagal syncope are still unclear. A simple discriminating test for the identification of syncopeprone subjects is not presently available. Fifty-two subjects had a stepwise orthostatic test with 60 ~ tilt and -20 and -40 mm Hg lower-body negative pressure before the appearance of impending syncope symptoms. Spectral and cross-spectral analyses of heart period and systolic pressure time series were performed to estimate the power of the high-frequency (= 0.25 Hz) and low-frequency (= 0.1 Hz) oscillations, the coherence between heart period and systolic pressure, and the mean low-frequency and high-frequency central frequency, phase shift, and transfer fimction at maximal coherence. According to time to presyncope, the 52 subjects were divided into two groups: 25 with normal orthostatic tolerance, and 27 with poor orthostatic tolerance. In the supine positions, the mean central low-frequency was significantly lower in poortolerance group than in normal-tolerance group, discriminating poor from normal orthostatic tolerance with 80% specificity and 83% sensitivity, and was significantly correlated to time to presyncope. In the 2 to 3 minutes preceding syncope, subjects with poor orthostatic tolerance had less tachycardia, lower low-frequency power of systolic pressure, higher respiratory frequency, and a less negative phase shift in highfrequency range. In presyncope, sympathetic activation is reduced in subjects with poor orthostatic tolerance. In addition, the higher breathing frequency and the smaller negativity of phase shift in high-frequency range, which may indicate an inadequate engagement of the baroreflex, suggest a causal role of respiration in the development of syncope. Supine central values of low frequency may be proposed as a valuable clinical index of orthostatic intolerance.
Received April 20, 2000; accepted as revised February 22, 2001Vasovagal, or neurocardiogenic, syncope is a sudden onset ofhypotension often associated with bradycardia and sometimes asystole, and leads to a short period of loss of consciousness that is usually followed by a rapid and complete recovery. Although the actual trigger mechanisms for the attack are still debated, it is now well established that there is an inappropriate change in autonomic nervous activity consisting of sympathetic withdrawal and often an increase in vagal efferent discharge [1,2]. Vasodilatation is the dominant syncope-initiating effect because prevention of bradycardia by administration of atropine or by cardiac pacing does not usually prevent the hypotension [3,4].Vasovagal attacks are associated with decreased cardiac filling and can be reproduced in a laboratory by applying a sufficiently severe orthostatic stress. Sometimes, a period of passive head-up tilting alone can induce the attacks [5], but often it is necessary to increase the stress by combining it with a vasodilator, such as isoproterenol [6] or nitrate [7], or by combining head-up tilting with suction applied to the lower body [8]. However, it may not alw...