False Positive Head-up Tilt:
Hemodynamic and Neurohumoral Profile
Fabio M. Leonelli, MD, FACC,* Ke Wang, MD,* Joyce M. Evans, MS,† Abhijit R. Patwardhan, PHD,† Michael G. Ziegler, MD,§ Andrea Natale, MD,* Charles S. Kim, BSE,† Kathleen Rajikovich, RN,* Charles F. Knapp, PHD†
Lexington, Kentucky, and San Diego, California
OBJECTIVES This study examined differences in mechanisms of head-up tilt (HUT)-induced syncope between normal controls and patients with neurocardiogenic syncope.
BACKGROUND A variable proportion of normal individuals experience syncope during HUT. Differences in the mechanisms of HUT-mediated syncope between this group and patients with neurocar-diogenic syncope have not been elucidated.
METHODS A 30-min 80° HUT was performed in eight HUT-negative volunteers (Group I), eight HUT-positive volunteers (Group II) and 15 patients with neurocardiogenic syncope. Heart rate and blood pressure (BP) were monitored continuously. Epinephrine and norepinephrine plasma levels, as well as left ventricular dimensions and contractility determined by echocardiography, were measured at baseline and at regular intervals during the test. RESULTS The main findings of this study were the following: 1) All parameters were similar at baseline
in the three groups; and 2) During tilt: a) the time to syncope was shorter in Group III than in group II (9.5⫾ 3 vs. 17 ⫾ 3 min p ⬍ 0.05); b) there was an immediate, persisting drop in mean BP in Group III; c) the decrease rate of left ventricular end-diastolic dimensions was greater in Group III than in Group II or Group I (⫺1.76 ⫾ 0.42 vs. ⫺0.87 ⫾ 0.35 and ⫺0.67 ⫾ 0.29 mm/min, respectively, p ⬍ 0.05); d) the left ventricular shortening fraction was greater in Group III than in the other two groups (39 ⫾ 1 vs. 34 ⫾ 1 and 32 ⫾ 1%, respectively, p⬍ 0.05); and e) although the norepinephrine level remained comparable among the groups, there was a significantly higher peak epinephrine level in Group III than in Group II and Group I (112.3⫾ 34 vs. 77.6 ⫾ 10 and 65 ⫾ 12 pg/ml, p ⬍ 0.05).
CONCLUSIONS Mechanisms of syncope during HUT appeared to be different in normal volunteers and patients with neurocardiogenic syncope. In the latter, there was evidence of an impaired vascular resistance response from the beginning of the orthostatic challenge. Furthermore, in the patients there was more rapid peripheral blood pooling, as indicated by the echocardio-graphic measurements of left ventricular end-diastolic changes, leading to more precocious symptoms. In syncopal patients, the higher level of plasma epinephrine probably mediated the increased cardiac contractility and possibly contributed to the impaired vasoconstrictive response. (J Am Coll Cardiol 2000;35:188 –93) © 1999 by the American College of Cardiology
Head-up tilt (HUT) is an accepted diagnostic test in the evaluation of unexplained syncope (1). Its use has become widespread, especially in the management of neurocardio-genic syncope, despite the lack of an accepted diagnostic “gold standard” for this condition, rendering the sensitivity of this test uncertain (2– 6). Furthermore, a variable number of normal individuals without any prior history of syncope will have a positive HUT table test (3,7–9). The percentage
of false positive studies seems to be related to the age of the subjects (10), duration and angle of tilt (11,12), and use of provocative agents (13). In fact, fainting may be considered one of the expected responses to this test as normal individuals left standing and dependent for a prolonged period of time will develop hypotension, cerebral hypoper-fusion and finally syncope (14). The time course of these events can be accelerated by exercise (15), elevated body temperature (16,17), and prolonged bed rest (18). Because fainting does not occur in these individuals during normal circumstances, it is likely that a positive HUT in normal controls is due to the attendant gravitational stress overcom-ing some of the essential adjustments to orthostasis.
In contrast, patients with neurocardiogenic syncope, who are symptomatic during normal daily activities, exhibit more
From the *Department of Cardiology, the †Center for Biomedical Engineering, and the ‡Department of Medicine, University of Kentucky, Lexington, Kentucky, and §University of San Diego, San Diego, California. Supported in part by NASA grant EPSCoR WKU 522611, NIH grant MO1RR02602, and an Industrial Grant from St. Jude Medical Co.
Manuscript received July 24, 1998; revised manuscript received July 8, 1999, accepted September 10, 1999.
Journal of the American College of Cardiology Vol. 35, No. 1, 2000
© 1999 by the American College of Cardiology ISSN 0735-1097/00/$20.00
profound abnormalities in the adaptive responses to ortho-stasis. These abnormalities can be easily uncovered by HUT testing. Comparing hemodynamic and humoral responses during positive HUT between normal volunteers and pa-tients with neurocardiogenic syncope could define more clearly the abnormalities in the reflex regulation of blood pressure. Should quantification of this response be possible with noninvasive methods, the specificity and sensitivity of the HUT could also be improved, rendering it a more accurate and useful clinical tool in the diagnosis of this elusive condition. With this purpose in mind, we compared hemodynamic and humoral responses during HUT-mediated syncope in a group of patients with neurocardio-genic syncope and a group of normal volunteers.
METHODS
Normal subjects (groups I and II). Sixteen volunteers
recruited from the staff or students of the University of Kentucky, gender- and age-matched to the patients (Group III), were included in this group. They had no previous history of syncope or presyncope and showed normal cardiovascular function by clinical exam, 12-lead electrocar-diogram (ECG), and standard echocardiographic evalua-tion. This group was further subdivided, according to the response to HUT; Group I with negative HUT response (8 subjects, 4 men and 4 women, mean age 34⫾ 2 years) and Group II, experiencing syncope during HUT (8 subjects, 4 men and 4 women, mean age 31⫾ 2 years).
Patients with unexplained syncope (group III). Patients
with unexplained syncope were included in the study if they met the following criteria:
1) At least two syncopal episodes in the last six months that remained unexplained despite a careful history, compre-hensive physical examination, full neurological evalua-tion, 12-lead ECG, 24-h ambulatory Holter monitor-ing, and echocardiography.
2) No history of hypertension or usage of drugs known to cause orthostatic hypotension.
3) Technically optimal echocardiographic images.
4) Normal heart structure and function by echocardio-graphic criteria.
5) Positive HUT that reproduced the clinical symptoms with evidence of hypotension (systolic blood pressure
[BP] less than 90 mm Hg) and/or bradycardia (heart rate [HR] less than 60 beats/min).
Fifteen consecutive patients, six men and nine women, mean age 30 ⫾ 4 years, referred to the University of Kentucky between June 1996 and August 1997, met these criteria and were included in the present study.
HUT protocol. After obtaining written permission from
all participants in the study, the test was performed in a quiet, temperature-controlled room in the morning follow-ing an overnight fast. The protocol included 20 min of baseline rest in the supine state followed by 30 min of HUT at 80° with a foot board for weight bearing and three waistbands around the body to fold the subject in case of syncope. The test was terminated prematurely if the subject experienced hypotension (systolic BP less than 90 mm Hg) and/or bradycardia (HR less than 60 beats/min). Blood pressure and ECG leads II, III, and aVF were continuously monitored using SpaceLabs ECG recording and Finapres TM BP Monitor (Ohmeda 2300). Blood pressure values recorded with the finger sphygmomanometer were com-pared every 4 min with values obtained from a standard cuff sphygmomanometer to confirm their accuracy. If more than a 10% difference between these two values was detected during the resting phase, the position of the finger sphyg-momanometer was changed. If such a discrepancy was found during tilt, the Finapres was allowed to undergo one or more cycles of automatic recalibration. In the situation where this maneuver did not normalize the readings, the entire data set was discarded.
Electrocardiographic and BP data were continuously displayed on a monitor, recorded on a strip chart recorder (Astromed 9000, Astromed, West Warwick, Rhode Island), digitized on line at the rate of 250 samples per second using a commercial system (DATA Q) and stored for subsequent analysis. To correlate BP and HR with left ventricular dimensions, a 30-s segment of these parameters, recorded at the time of echocardiography, was later retrieved and analyzed.
Echocardiographic analysis. Two-dimensional
echocardi-ography was performed using a Hewlett-Packard 77020 A sonos 1000 ultrasonograph and a 2.5-MHz phased-array transducer. A standard parasternal short-axis view at the level of the papillary muscles was recorded during the supine resting stage, 2 min after initiation of HUT and every 4 min thereafter until the end of the test or at the onset of symptoms. Left ventricular end-diastolic dimension (LVEDD) and left ventricular end-systolic dimension (LVESD) were determined using M-mode echocardiogra-phy. Percent of left ventricular fractional shortening (SF) was calculated: [(end diastolic dimension ⫺ end systolic dimensions)/end diastolic dimensions)⫻ 100].
All echocardiographic data were analyzed from stopped-frame videotape and strip-chart recordings by the same investigator so as to minimize interobserver variations. Each
Abbreviations and Acronyms
BP ⫽ blood pressure ECG ⫽ electrocardiogram HR ⫽ heart rate HUT ⫽ head-up tilt
LVEDD ⫽ left ventricular end-diastolic dimension LVESD ⫽ left ventricular end-systolic dimension SF ⫽ shortening fraction
value was obtained by averaging at least five consecutive heart beats.
Hormonal assay. An antecubital-indwelling catheter was
placed 1 h prior to the beginning of the study. Ten milliliters of blood was drawn at the end of the supine rest and at 8-min intervals during HUT. In subjects with positive HUT, a sample was drawn at the appearance of syncopal symptoms. All samples were immediately spun, plasma was extracted and frozen, and samples subsequently delivered to the laboratory where radioenzymatic assays for plasma norepinephrine and epinephrine were performed (19).
Statistical analysis. Group data are expressed as mean
value ⫾ SE. Two-factor analysis of variance (time and group) with multiple comparisons on the time factor was used to determine responses to tilt for the three groups of subjects. When significant effects were found, appropriate t tests were made using Bonferroni correction. A p value of 0.05 was considered to be statistically significant. A second investigator, blinded to the stage of the study, reviewed all the echocardiographic measurements of left ventricular end-diastolic and -systolic volume to allow an estimation of the interobserver variability. For interobserver variability, the correlation coefficients were 0.92 (p⬍ 0.001) and 0.94 (p ⬍ 0.001), respectively. Changes in left ventricular di-mension and SF over time were assessed for each group. Regression was used to compare the rate of change of these parameters among these three groups. The estimated beta coefficient obtained from regression slopes in the three groups was compared using the statistical analysis previously described.
RESULTS
Cardiovascular responses. Baseline mean BPs and HRs
were comparable in the three groups. During HUT, the mean BP in Group I did not change from baseline, while the mean HR increased from 62⫾ 5 to 95 ⫾ 10 beats/min (p⬍ 0.05). All subjects in this group completed 30 min of HUT. The duration of HUT was different in Groups II and III as symptoms of presyncope appeared at 17⫾ 3 min in the former and 9.5⫾ 3 min in the latter (p ⬍ 0.05) (see Table 1).
There was an equal distribution of vasodepressor and mixed hemodynamic responses (14,15) to HUT in both groups. In Group II there were four patients with vasode-pressor and four with mixed response, whereas in Group III, symptoms were associated with a vasodepressor response in nine and a mixed response in six patients. The BP response in volunteers with a positive HUT response (Group II) remained unchanged from baseline until the development of presyncope, while there was a progressive increase in mean HR throughout the test. This change was similar to that observed in Group I.
In the group of patients with neurocardiogenic syncope,
the HR response was comparable to the response in the other two groups. However, this group demonstrated an appreciable decrease in mean BP from the supine value within the first 2 min of tilt. This value decreased from 92⫾ 3 mm Hg to 83 ⫾ 4 mm Hg at 2 min (p ⬍ 0.05) and continued to decrease to 79⫾ 4 mm Hg at 2 min before symptoms became manifest. These values were different both from baseline and from the values recorded at the same stage in Group I.
Left ventricular dimensions. Left ventricular
end-diastolic dimension and LVESD, as well as shortening fraction (SF) values, were similar in the three groups at baseline (Fig. 1). Changes in these parameters were not significant in Group I throughout the entire upright tilt part of the protocol. In Group II, SF remains unchanged, while LVESD and LVEDD measurements decreased during the test, becoming significantly less than control 2 min before the end of the test, although they remained comparable to the values from Group I. In Group III (patients with neurocardiogenic syncope), all three parameters of left ventricular function were different from baseline at 2 min into the test, and LVESD and LVEDD measurements became different from Group I 2 min before the end of HUT. Moreover, SF significantly increased throughout the test to become, at 2 min before the end of HUT, statistically different from the other two groups (39⫾ 1% in Group III, 34⫾ 1 and 32 ⫾ 1% in Groups II and I, respectively; p ⬍ 0.05). The rate of decrease in LVESD and LVEDD was greater in Group III than in Group II or Group I (⫺1.76 ⫾ 0.42 vs.⫺0.87 ⫾ 0.35 vs. ⫺0.17 ⫾ 0.03 mm/min for the former and⫺1.68 ⫾ 0.4 vs. ⫺0.67 ⫾ 0.29 vs. ⫺0.11 ⫾ 0.03 mm/min for the latter, p⬍ 0.05).
Catecholamine changes. Epinephrine and norepinephrine
values were similar in the three groups at baseline (Fig. 2). Norepinephrine levels increased during HUT in a compa-rable manner in all three groups. Epinephrine levels in-creased during the test in every subject. Patients with neurocardiogenic syncope had a sixfold increase from the Table 1. Average Values of Mean Blood Pressure and Heart Rate⫾ SE During 80° HUT
Supine 2-min HUT Before end-HUT
Mean BP (mm Hg) Group 1 96⫾ 3 100⫾ 6 98⫾ 7 96⫾ 5 Group 2 91⫾ 5 90⫾ 3 83⫾ 2 51⫾ 3*† Group 3 92⫾ 3 83⫾ 4*† 79⫾ 4*† 50 ⫾ 2*† HR (beats/min) Group 1 62⫾ 5 86⫾ 7† 95⫾ 9† 95⫾ 10† Group 2 64⫾ 3 93⫾ 5† 105⫾ 5† 80⫾ 8 Group 3 68⫾ 3 94⫾ 4† 110⫾ 6† 82⫾ 4 †p⬍ 0.05 vs. supine. *p ⬍ 0.05 vs. Group I.
BP⫽ blood pressure; HR ⫽ heart rate; HUT ⫽ head-up tilt; supine ⫽ during supine position; 2-min HUT⫽ 2 min after starting HUT; Before ⫽ 3 min before end of HUT; end-HUT⫽ end of HUT.
baseline value (18⫾ 3 to 112 ⫾ 34 pg/ml, p ⬍ 0.05) at the onset of bradycardia and hypotension. This value was statistically different from the other two groups.
DISCUSSION
Earlier studies. Cardiovascular adjustments to
gravita-tional stress have been an object of study for well over a century. The HUT has long (20) been considered a conve-nient model to study reflex responses to gravitational stress. Both hypotension and bradycardia leading to syncope dur-ing HUT are not infrequent events in normal subjects (3,7–9). These responses were considered to be part of a reflex response triggered by a sympathetic-induced hyper-contraction of an almost empty left ventricular chamber (21). More recently (1), HUT has became an accepted diagnostic tool in the study of a poorly understood clinical entity often referred to as “neurocardiogenic syncope.” However, the frequency of false positive responses was immediately appreciated as a problem inherent in this test (6,13), leading some investigators to suggest that HUT should be used as a confirmatory test to support a clinically based diagnosis (13). Although HUT-mediated syncope can occur in normal subjects, hemodynamic and humoral responses leading to this event may be qualitatively different from the responses of HUT-positive individuals who
expe-rience syncope during normal daily activities. Comparison of the pattern of responses in these two groups could lead to a better definition of the abnormal regulatory mechanisms leading to pathological syncope and consequently improve the specificity of HUT.
Differences between groups. Our study shows that a
number of differences and similarities exist between these two groups. First, the time to syncope was about twice as long in the controls as in the patients. During the test, the HR and BP responses of the HUT-positive groups differed. The patients (Group III) exhibited a drop in BP as early as 2 min into the test accompanied, at the same time, by tachycardia. In syncopal volunteers (Group II), these changes were not present until 2 min before the onset of symptoms and were absent in the normal controls (Group I). The pattern of peripheral blood pooling in the three groups was indirectly estimated by analyzing changes in LVEDD. During the orthostatic challenge, both HUT-positive groups exhibited a progressive decrease in LVEDD, statistically different from baseline value, while the HUT-negative subjects exhibited an initial reduction that quickly stabilized, an effect that has been previously reported (22). Although the dimensions themselves are not different be-tween the two HUT-positive groups, the different rates of Figure 1. Changes in left ventricular end-systolic dimensions (LVESD), left ventricular end-diastolic dimensions (LVEDD), shortening fraction (SF), and rate of change of the same parameters during HUT in the three groups. Throughout the test, end-diastolic, end-systolic dimension, and SF did not change significantly in Group I. In Group II the LVEDD and the SF changed significantly from the baseline values but remained similar to Group I. In Group III all of these parameters changed from baseline; furthermore, the LVEDDs were different from Group I, and the SF was different from both groups. Although LVEDDs were comparable in Group II and Group III, the rate of change in Group III was different from the other two. *p⬍ 0.05 versus Group I. †p ⬍ 0.05 from baseline. ‡p ⬍ 0.05 versus baseline, Group I and II. mm⫽ millimeters; min ⫽ minutes; HUT ⫽ head-up tilt; before ⫽ 1 or 2 min before the end of HUT.
dimension reduction are suggestive of a more rapid periph-eral fluid shift in patients with neurocardiogenic syncope.
Decreased venous return: Possible etiology. The BP
response and the rapid reduction in venous return during orthostatic challenge are suggestive of an abnormality in vascular control.
This hypothesis is in agreement with previous work suggesting an impaired vasoconstrictor response in patients with neurocardiogenic syncope during tilt test (23) and during dynamic leg exercise (24,25). The mechanisms of this apparent failure of reflex vasoconstriction are unclear. Multiple reflexes are involved in the maintenance of BP during orthostatic stress (26). Although failure of any of these responses has never been conclusively demonstrated, intrinsic abnormality of cardiopulmonary mechanoreceptor function has been postulated to exist in these individuals (27). By contrast, epinephrine might contribute to inappro-priate vasodilation. Epinephrine peaked in patients with neurocardiogenic syncope just before the occurrence of symptoms. From our data it is unclear whether the signif-icant increase in epinephrine in Group III at the onset of symptoms represents a cause for the syncope or merely the failing attempt of the sympathetic system to maintain the cardiac output. This hormone dilates skeletal muscle and splanchnic resistance vessels at concentrations measured in humans under stress (28,29). The HUT-positive volunteers had normal SF and normal epinephrine levels, suggesting that this hormone plays only a marginal role in their syncopes. The reflex inducing syncope seems to differ
between patients and normal volunteers. The patients had left ventricular hypercontractility, possibly secondary to high plasma levels of epinephrine. The left ventricular SF of the HUT-positive volunteers was comparable to the nonfaint-ing controls. This findnonfaint-ing casts doubt on the importance of the Bezold-Jarish reflex (30) triggering bradycardia and hypotension during tilt, at least, in the normal controls.
SUMMARY
In summary, we have shown that the hemodynamic and humoral responses to HUT were different in syncopal patients and in controls. In patients, this test unmasked postural hypotension that could have been worsened by excessive circulating epinephrine. In HUT-positive normal volunteers, the progression to syncope was slower and accompanied by epinephrine responses similar to nonsyn-copal volunteers. Noninvasive quantification of BP, rate of change of left ventricular end-diastolic volume, SF of this chamber, and epinephrine plasma levels discriminated syn-copal responses of patients with neurocardiogenic syncope from that of positive normal volunteers. It may be possible to define an algorithm combining these parameters that defines an abnormal response to this orthostatic challenge.
Study limitations. In our study, we evaluated changes in
BP and left ventricular dimensions during HUT using finger plethysmography and echocardiography. Invasive he-modynamic monitoring is more accurate but sometimes induces syncope (30 –32). The rate of false positive response to HUT was higher in our subjects than any previously published rate in controls (1). Nevertheless, our data are strongly supported by a recent report (33) showing, in a large group of normal volunteers, a linear correlation be-tween the incidence of syncope and the duration of a 50° HUT. As in our group, at 30 min into the test, 50% of their subjects (33) had experienced hypotension and bradycardia, forcing termination of the tilt. This suggests that the specificity of this test is duration dependent and may be lower than previously reported.
Our conclusions were based on a relatively small prese-lected sample of subjects, which prevents us from general-izing our observations to the population at large. Therefore, the proposed monitoring of selected parameters to increase the specificity and sensitivity of HUT in the diagnosis of neurocardiogenic syncope awaits trials with larger numbers of patients. Although our observation of different rates of venous return in the three groups suggests an abnormality in vasoconstrictive response, we have performed no direct measurements to support our contention. Demonstration of this speculation should be the focus of future research. We hypothesized that the higher level of epinephrine could explain the different hemodynamic profile of the patients. Others have suggested (34) that patients with this condition have a beta-receptor hypersensitivity, which could exacer-bate responses to epinephrine. Finally, we cannot exclude Figure 2. Changes in epinephrine and norepinephrine plasma
levels during head-up tilt. The three groups exhibited similar increases in norepinephrine levels during the test. The increase in epinephrine level during HUT was higher in Group III than in the other groups. ‡p⬍ 0.05 versus baseline, Group I and II. HUT ⫽ head-up tilt.
that other reflexes (35) or circulating hormones (36,37) play a contributory role in patients’ abnormal response to HUT.
Reprint requests and correspondence: Dr. Fabio Leonelli, 740
South Limestone Street, Room L543, KY Clinic, University of Kentucky, Lexington, Kentucky 40536-0084.
REFERENCES
1. Benditt DG, Ferguson DW, Grubb BP, et al. Tilt table testing for assessing syncope. J Am Coll Cardiol 1996;28:263–75.
2. Fitzpatrick A, Theokorakis G, Vardas P, et al. The incidence of malignant vasovagal syndrome in patients with recurrent syncope. Eur Heart J 1991;12:389 –94.
3. Kenny RA, Bayliss J, Ingram A, Sutton R. Head-up tilt: a useful test for investigating unexplained syncope. Lancet 1986;1:1352– 4. 4. Sra JS, Anderson AJ, Sheikh SH, et al. Unexplained syncope evaluated
by electrophysiologic studies and head-up tilt testing. Ann Intern Med 1991;114:1013–9.
5. Raviele A, Gasparini G, DiPede F, Delise P, Bonso A, Piccolo E. Usefulness of head-up tilt test in evaluating patients with syncope of unknown origin and negative electrophysiologic study. Am J Cardiol 1990;65:1322–7.
6. Kapoor WN, Smith MA, Miller NL. Upright tilt testing in evaluating syncope: comprehensive literature review. Am J Med 1994;97:78 – 88. 7. Smith JJ. Circulatory response to the upright posture. Boca Raton, FL:
CRC Press, 1990:30 –3.
8. Graybiel A, McFarland RA. The use of the tilt table in aviation medicine. Aviat Med 1941;11:194 –211.
9. Wagner HN. Orthostatic hypotension. Bull Johns Hopkins Hosp 1959;105:322–59.
10. De Jong-de Vos van Steenwijk CCE, Wieling W, Johannes J, Harms MP, Kuis W, Wessling KH. Incidence and hemodynamic character-istics of near-fainting in healthy 6- to 16-year old subjects. J Am Coll Cardiol 1995;25:1615–21.
11. Natale A, Aktar M, Jazayeri M, et al. Provocation of hypotension during head-up tilt testing in subjects with no history of syncope or pre-syncope. Circulation 1995;92:54 – 8.
12. Lewis DA, Zlotocha J, Henke L, Dhala A. Specificity of head-up tilt testing in adolescents: effect of various degrees of tilt challenge in normal control subjects. J Am Coll Cardiol 1997;30:1057– 60. 13. Kapoor WN, Brant N. Evaluation of syncope by upright tilt testing
with isoproterenol. Ann Intern Med 1992;116:358 – 63.
14. Sander-Jensen K, Secher NH, Astrup A, et al. Hypotension induced by passive head-up tilt: endocrine and circulatory mechanisms. Am J Physiol 1986;251:R742– 8.
15. Bjurstedt H, Rosenhamer G, Balldin U, Katkov V. Orthostatic reactions during recovery from exhaustive exercise of short duration. Acta Physiol Scand 1983;119:25–31.
16. Rowell LB, Brengelmann GL, Blackmon JR, Murray JA. Redistribu-tion of blood flow during sustained high skin temperature in resting man. J Appl Physiol 1970;28:415–20.
17. Lind AR, Leithead CS, McNicol GW. Cardiovascular changes during syncope induced by tilting men in the heat. J Appl Physiol 1968;25: 268 –76.
18. Blomqvist CG, Stone HL. Cardiovascular adjustments to gravitational
stress. In Shepherd JT, Abboud FM, Geiger SR, editors. Handbook of Physiology. The Cardiovascular System: Peripheral Circulation and Organ Blood Flow, Sec. 2, Vol. III, Pt. 2. Bethesda, MD: American Physiological Society, 1983:1025– 63.
19. Kennedy B, Fisher MG. A more sensitive and specific radioenzymatic assay for catecholamines. Life Sci 1990;47:2143–53.
20. Allen SC, Taylor CL, Hall VE. A study of orthostatic insufficiency by the tiltboard method. Clin Sci Lond 1950;9:79 –90.
21. Oberg B, Thorne P. Increased activity in the left ventricular receptors during hemorrhage or occlusion of caval veins in the cat: a possible cause of the vasovagal reaction. Acta Physiol Scand 1972;85:164 –73. 22. Shalev Y, Gal R, Tchou PJ, et al. Echocardiographic demonstration of decreased left ventricular dimensions and vigorous myocardial contrac-tion during syncope induced by head-up tilt. J Am Coll Cardiol 1991;18:746 –51.
23. Sneddon JF, Counihan PJ, Bashir Y, Haywood GA, Ward DE, Camm AJ. Impaired immediate vasoconstrictor responses in patients with recurrent neurally mediated syncope. Am J Cardiol 1993;71: 72– 6.
24. Thomson HL, Lele SS, Atherton JJ, Wright KN, Stafford W, Frenneaux MP. Abnormal forearm vascular responses during dynamic leg exercise in patients with vasovagal syncope. Circulation 1995;92: 2204 –9.
25. Thomson HL, Atherton JJ, Khafagi FA, Frenneaux MP. Failure of reflex venoconstriction during exercise in patients with vasovagal syncope. Circulation 1996;93:953–9.
26. Abboud FM, Heistad DD, Mark AL, Schmid PG. Reflect control of the peripheral circulation. Prog Cardiovasc Dis 1976;17:371– 403. 27. Sneddon JF, Counihan PJ, Bashir Y, Haywood GA, Ward DE,
Camm AJ. Assessment of autonomic function in patients with neurally mediated syncope: augmented cardiopulmonary baroreceptor re-sponses to graded orthostatic stress. J Am Coll Cardiol 1993;21: 1193– 8.
28. Rowell LB, Seals DR. Sympathetic activity during graded central hypovolemia and hypoxemic humans. Am J Physiol 1990:H1197– 1206.
29. Rowell LB, Blackmon JR. Hypotension induced by central hypovol-aemia and hypoxhypovol-aemia. Clin Physiol 1989;9:269 –77.
30. Linden RJ. Reflexes from the heart. Prog Cardiovasc Dis 1975;18: 201–21.
31. McIntosh SJ, Lawson J, Kenny RA. Intravenous cannulation alters the specificity of head-up tilt testing for vasovagal syncope in elderly patients. Age Ageing 1994;23:317–9.
32. Imholz BPM, Wielding W, Langewouters GJ, Van Monfrants GA. Continuous finger arterial pressure utility in the cardiovascular labo-ratory. Clin Auton Res 1991;1:45–53.
33. Madsen P, Svendsen L, Jorgensen L, Matzen S, Jansen E, Secher N. Tolerance to head-up tilt and suspension with elevated legs. Aviat Space Environ Med 1998;69:781– 4.
34. Perry JC, Garson A. The child with recurrent syncope: autonomic function testing and beta-adrenergic hypersensitivity. J Am Coll Cardiol 1991;17:1168 –71.
35. Dickinson CJ. Fainting precipitated by collapse-firing of venous baroreceptors. Lancet 1993;342:970 –2.
36. Kaufmann H, Oribe E, Oliver JA. Plasma endothelin during upright tilt: relevance for orthostatic hypotension? Lancet 1991;338:1542–5. 37. Shen WK, Hammill SC, Munger TM, et al. Adenosine: potential
