Pathogenesis and etiology of syncope

Brian Olshansky, MD Christopher S Cadman, MD

Feb 3, 2000

Syncope is the abrupt loss of consciousness associated with the absence of postural tone; it is followed by a rapid and usually complete recovery. This symptom is alarming for the individual, witnesses, family, and physicians. Although syncope can be a harbinger of a multitude of disease processes and can mimic the appearance of a cardiac arrest, it is most often benign and self-limited. Nevertheless, injuries associated with syncopal attacks occur in 35 percent of patients, and recurrent episodes can be psychologically devastating. In addition, syncope can be a premonitory sign of sudden cardiac death (SCD), especially in patients with organic heart disease.

The cause for syncope is not often obvious, and the individual at highest risk for sudden death can be difficult to detect (show figure 1A-1B). It is important to recognize that syncope and SCD are two different entities that must be distinguished from each other in order to accurately assess future outcome and prognosis. Patients in whom cardiopulmonary resuscitation, or electric or pharmacologic cardioversion have been required should be labeled as having SCD and not as having syncope. These two diagnostic entities are inextricably intertwined, however, since patients with cardiac syncope have a very high incidence of subsequent SCD (approximately 24 percent) [1].

This card will review the epidemiology, pathogenesis, and etiology of syncope. The evaluation and management of this disorder are discussed separately. (See "Evaluation of the patient with syncope" and see "Management of the patient with syncope").

The pathogenesis and etiology of syncope are the same as that for presyncope, which is the prodromal symptom of fainting. Such patients usually present with symptoms of dizziness. (See "Approach to the patient with dizziness").

EPIDEMIOLOGY — Syncope is a relatively common clinical problem. Data from the Framingham study, for example, suggest that syncope occurs annually in about 3 percent of men and 3.5 percent of women in the general population [2]. However, the causes of syncope differ between men and women and men are more likely than women to have a cardiac cause for syncope (show figure 2) [3]. Other studies of relatively healthy individuals, however, have reported an incidence as high as 40 percent [4]. There appears to be a moderate female predominance (60 percent in one report) [1].

The variability in incidence is due to differences in study populations, definitions of syncope, methods, and criteria. Nevertheless, approximately one-third of individuals is likely to have a syncopal episode during their lifetime [1,5,6]. In over 75 percent of cases in nonelderly subjects, syncope occurs as an isolated event (ie, not due to known neurologic or cardiovascular disease) [1].

Syncope is also a common clinical complaint of patients treated in the emergency room and is the source for a significant number of hospital admissions. Data from several reports showed that 3 to 5 percent of all emergency room admissions and 1 percent of all hospital admissions were for syncope [1,6,7,8].

Incidence of syncope in the elderly — The elderly patient is more likely to have syncope, is more commonly injured, and is more likely to seek medical advice than the younger patient [9]. The Framingham study, for example, noted an annual incidence of six percent for at least one episode of syncope in subjects over the age of 75; in comparison, the annual incidence was only 0.7 percent in men aged 35 to 44 (show figure 3) [2].

The increased risk of syncope in the elderly is due to age- and disease-related abnormalities that impair the ability to respond to physiologic stresses that would ordinarily not cause syncope [9].

CAUSES OF SYNCOPE — Determining the cause of syncope is important for both prognostic and therapeutic reasons. Several large studies have assessed the causes for syncope. One study, for example, evaluated 510 patients [10]. This study, however, was flawed because it was retrospective and data of 500 additional patients, who were initially evaluated, was not included due to incomplete information. Nevertheless, the findings continue to remain useful by providing insight into the differential diagnosis and the frequency of the underlying causes of syncope. Other studies have also provided useful information [1,7,8].

The organization of the underlying causes of syncope into broad based categories related to the presence or absence of cardiovascular disease (CVD) permits the clinician to accurately assess a patient's prognosis and provides a framework for evaluation and treatment [11].

  Cardiovascular causes — Patients with an underlying cardiovascular cause of syncope have higher rates of SCD rate and all-cause mortality than those with a noncardiovascular cause of syncope. The mortality rate in patients with CVD after five years of follow-up has been reported to approach 50 percent, with a 30 percent incidence of death in the first year (show figure 4) [1,7,8]. Mortality in these patients is in large part due to the severity of the underlying CVD.

  Noncardiovascular causes — Patients with a noncardiovascular cause of syncope have a substantially lower mortality rate (about 6 percent or less in the first year) [10].

  Unknown causes — The mortality for those patients in whom the cause is unknown (syncope of unknown etiology) is similar to that in patients with noncardiovascular etiologies (show figure 4).

There is one caveat concerning this classification: the noncardiovascular group includes some CVD. Vasovagal syncope, for example, which is often considered a noncardiovascular cause, is a cardiovascular reflex. However, patients with vasodepressor syncope have a favorable prognosis. As a result, inclusion of these patients in the noncardiovascular group maintains the prognostic significance of having a CVD cause for syncope.

CARDIOVASCULAR DISEASE — CVD causes of syncope result from arrhythmias or from cardiac diseases which severely obstruct blood flow, such as aortic stenosis and hypertrophic obstructive cardiomyopathy (show figure 1A-1B). In one study, for example, less than 50 percent of those with a CVD cause for syncope had ventricular tachycardia (VT) [7].

Treatment of the underlying cause of syncope does not necessarily improve prognosis. As examples, the mortality from dissecting aneurysm or aortic valvular disease remains high despite treatment. In addition, it remains unclear whether treatment of syncope in patients with hypertrophic cardiomyopathy alters the high risk of death from this disease [12]. On the other hand, effective treatment of syncope in patients who have the long QT syndrome can alter the risk of SCD. (See "Prognosis and management of the long QT syndrome and torsade de pointes"),

Syncope has been found to be an important prognostic predictor among certain subgroups of patients with cardiac disease and impaired left ventricular function [13]. These data, however, have not been corroborated in all reports; furthermore, treatment of the presumed cause for syncope in these patients (tachyarrhythmias) does not necessarily improve prognosis. One possible explanation for these discordant findings is that the cause of syncope and perhaps death in these patients was not a tachyarrhythmia but a bradycardia with a malignant form of vagal reflex.

Arrhythmias — One study of 510 patients with syncope reported that "paroxysmal tachycardia" and "Stokes-Adams attacks" were thought to be the underlying cause of symptoms in only 8 and 17 patients, respectively [10]. Arrhythmic causes of syncope may have been grossly underrepresented in this series, however, since sophisticated techniques to detect cardiac arrhythmias were not available at the time. Primary cardiac rhythm disturbances (bradycardias and tachycardias) are currently recognized as common causes of syncope when other causes cannot be found [14]. This change in recognized etiology is largely due to the use of more sophisticated diagnostic tools such as prolonged monitoring techniques and electrophysiology testing. With such diagnostic aids, a primary arrhythmic etiology is often discovered.

Although an arrhythmic etiology for syncope is often suspected clinically; the culprit arrhythmia frequently cannot be diagnosed since most are paroxysmal and infrequent. An additional factor which confounds the accurate diagnosis of an arrhythmia is that rhythm disturbances are a common finding among patients, including those in whom syncope has not occurred. As a result, the presence of an arrhythmia (such as asymptomatic sinus bradycardia or nonsustained VT on an ambulatory monitor recording) does not necessarily mean that the arrhythmia is the cause for syncope. Furthermore, if the true etiology of syncope results in a secondary arrhythmia, treatment of the rhythm disturbance will not prevent recurrent syncope although the arrhythmia is coincident with symptoms. As an example, a "relative" bradycardia is often seen during a vasovagal episode and although it is associated with syncope, treatment of the bradycardia rarely prevents recurrent symptoms.

Tachycardias are usually more hemodynamically unstable and are less well tolerated than bradycardias, especially at the onset of the arrhythmia. Both forms of rhythm disturbances are more likely to cause a syncopal response if the rhythm begins abruptly. This is illustrated by the following observations:

  •  Chronic sinus bradycardia is often well tolerated, while sinus rhythm with abrupt sinus arrest and a long sinus pause frequently causes symptoms.

  •  Persistent atrial flutter with a ventricular response rate of 150 beats per minute will often be asymptomatic, while paroxysmal VT at the same rate may cause syncope.

  •  The initiation of a tachycardia is frequently associated with hypotension and syncope; however, if the rhythm persists, the blood pressure can rise and the patient becomes asymptomatic.

The hemodynamic stability resulting from any arrhythmia is also influenced by other factors including rate, ventricular function, body position, medications, and baroreceptor sensitivity [15,16].

VT and bradyarrhythmias secondary to ischemia were considered to be the cause for syncope in only seven patients (1.4 percent) in the original series of causes of syncope [10]. These data are similar to others in the literature. As a result, although patients who are admitted to the hospital with syncope are commonly tested for a recent myocardial infarction (MI), this procedure is rarely (if ever) necessary since syncope is infrequently caused by MI.

The more common arrhythmic causes of syncope include sinus bradycardia, atrioventricular (AV) nodal block, sustained VT, and supraventricular tachycardia (SVT).

  Sinus bradycardia — Syncope resulting from this bradyarrhythmia may be caused by intrinsic sinus node disease (sick sinus syndrome, tachy-brady syndrome), drugs (beta blockers, including ophthalmologic application), or autonomic imbalance (high vagal or decreased sympathetic tone). (See "Sinus bradycardia").

  Atrioventricular nodal block — The forms of AV nodal block which can result in syncope are generally second or third degree AV blocks. In general, type I (Wenckebach) second degree AV block is benign and not progressive. If the QRS is narrow, the majority of cases are due to block at the level of the AV node. Type II second degree AV block is associated with significant conduction disease, such as a bundle branch block, and the site of block is usually infranodal. Type II block is often progressive, is more commonly associated with syncope, and warrants a permanent pacemaker. (See "Second degree atrioventricular block: Mobitz type II").

  Ventricular tachycardia — Sustained VT (monomorphic or polymorphic) was associated with syncope in only 11 percent of patients in one series. Prompt treatment is indicated, however, if syncope is present and a serious arrhythmia is diagnosed. It is likely that recurrent syncope will be prevented if the arrhythmia is successfully treated [1].

Syncope resulting from sustained VT is most commonly due to structural heart disease, particularly coronary heart disease. In these patients, other factors, such as baroreceptor sensitivity, may be important in determining hemodynamic tolerability of VT. As an example, one study of 24 patients with sustained monomorphic VT after an MI reported that patients with syncope or presyncope baroreceptor sensitivity (as assessed with an infusion of phenylephrine) tolerated VT significantly less well than those with normal baroreceptor responses [16]. Age, left ventricular function, VT rate or duration, and heart rate variability did not correlate with the sequelae of the arrhythmia.

Patients with dilated cardiomyopathy and right ventricular dysplasia are also prone to have VT presenting with syncope. In addition, an unusual form of VT, torsade de pointes, may cause syncope in patients with either the congenital or acquired forms of the long QT syndrome.

Ventricular fibrillation does not cause syncope. Instead, this rhythm disturbance causes cardiac arrest and death.

  Ventricular bigeminy — Ventricular bigeminy may occasionally be associated with hypotension, bradycardia, and syncope. This is seen primarily in patients with an underlying sinus bradycardia or those with advanced heart disease and significant left ventricular dysfunction.

  Supraventricular tachyarrhythmias — Supraventricular tachyarrhythmias have rarely been associated with syncope [17]. When syncope occurs with these rhythm disturbances it may be neurally mediated (neurocardiogenic) rather than resulting directly from the tachycardia [18].

Organized and regular SVTs cause syncope more frequently than do disorganized supraventricular rhythm disturbances. Such regular SVTs include AV nodal reentry or AV reentry tachycardia (in which action potentials are conducted via the AV node), or the Wolff-Parkinson-White Syndrome (in which action potentials are conducted via both the AV node and an accessory pathway). By comparison, disorganized supraventricular rhythms, such as atrial fibrillation, rarely cause syncope, which primarily occurs the ventricular response rate is rapid.

Cardiovascular disease with blood flow obstruction — Syncope may be caused by obstruction to blood flow due to cardiovascular abnormalities, the most frequent being aortic stenosis and hypertrophic cardiomyopathy (show figure 1A-1B). Less common conditions include pulmonic stenosis, primary pulmonary hypertension, atrial myxomas, and pulmonary embolism.

  Aortic stenosis — Although the patient with a stenotic aortic valve rarely presents with syncope, recognition of this association is exceedingly important since it is associated with a high mortality if untreated. (See "Natural history of aortic stenosis").

Syncope in patients with aortic stenosis is often associated with exertion. In most such cases, syncope results from an inability to produce a compensatory increase in cardiac output (due to the obstruction) which normally occurs in response to exercise-induced peripheral vasodilation [19]. In others, the most probable cause of syncope is an exaggerated, more malignant form of a vasovagal response. This may have the same mechanism as neurocardiogenic syncope, which is due to the stimulation of ventricular mechanoreceptors (see below). A bradyarrhythmia or ventricular tachyarrhythmia may also be a cause for syncope in some of these patients [20].

  Hypertrophic cardiomyopathy — Hypertrophic cardiomyopathy is associated with syncope, which occurs in up to 30 percent of patients who have dynamic outflow obstruction exist [21]. The obstruction may intensify with postural changes, hypovolemia, or drugs. VT has also been reported to occur in approximately 25 percent of patients [22].

  Other causes of outflow obstruction — Severe pulmonic stenosis, primary pulmonary hypertension, atrial myxomas, and pulmonary embolism can all cause syncope due to obstruction.

NONCARDIOVASCULAR CAUSES OF SYNCOPE — Syncope due to noncardiovascular abnormalities is associated with a relatively benign prognosis. The most prominent causes include neurally mediated syncope (vasovagal syncope, orthostatic hypotension, carotid hypersensitivity and autonomic reflexes) (show figure 5) and seizures.

Vasovagal syncope — Vasovagal syncope is also known as vasodepressor or neurocardiogenic syncope. It is common and was the second most frequent cause of syncope in the Wayne study [10].

Patients with vasovagal syncope are usually young and have a prodrome of diaphoresis, nausea, fatigue, and pallor. Often there is an increase in heart rate prior to an episode, suggesting sympathetic nervous system activation and an increase in epinephrine [23]. Syncopal episodes are frequently induced by prolonged standing, venipuncture (experienced or witnessed), heat exposure, painful or noxious stimuli, and fear of bodily injury. Patients may occasionally have recurrent episodes without identifiable causes.

Understanding the physiology involved in vasovagal syncope is essential to understanding the clinical manifestations (show figure 5). Both neural and endogenous chemical pathways may be involved:

  Carotid sinus reflex — The blood pressure and heart rate are normally controlled by several factors, including input from baroreceptors located within the carotid sinus and aortic arch. An increase in blood pressure or pressure applied to the carotid sinus enhances the baroreceptor firing rate and activates vagal efferents, thereby slowing the heart rate and reducing the blood pressure. This process is known as the carotid sinus reflex.

  Adenosine receptor activation — An animal model that replicated upright tilt testing was associated with increased secretion of adenosine, activation of the adenosine A1 receptor, and paradoxical bradycardia [24]. Dipyridamole, which blocks the uptake of adenosine and raises its extracellular concentration, enhanced the bradycardic response while theophylline, an adenosine receptor antagonist, prevented the bradycardia.

  Bezold-Jarisch reflex — Autonomic nervous system dysfunction plays an important role in the occurrence of neurocardiogenic syncope. It has been hypothesized that regional sympathetic denervation with subsequent denervation supersensitivity and hypercontractility contribute to the pathogenesis of this autonomic dysfunction; however, sympathetic innervation of the left ventricular myocardium appears to be normal in most patients with neurocardiogenic syncope, and cardiac denervation does not appear to play a significant role in its pathogenesis [25]. Receptors in the atria, great veins, and left ventricle also exist whose activation results in the Bezold-Jarisch reflex. Activation of the mechanoreceptors in the left ventricle and stretch receptors in the great vessels with pressure or volume loading (as may occur with sympathetic stimulation) stimulate C fibers; such stimulation may result in activation of the vagal efferents and the same type of blood pressure and heart rate response as observed with the carotid sinus reflux.

The Bezold-Jarisch and carotid sinus reflexes have two components: a cardioinhibitory response; and a vasodepressor response [26,27]:

  •  The cardioinhibitory response results from increased parasympathetic tone and may be manifested by any or all of the following: sinus bradycardia, PR prolongation, and advanced atrioventricular block.

  •  The vasodepressor response is due to decreased sympathetic activity and can lead to hypotension. As an example, one study of 53 patients with vasovagal syncope reported that the final trigger for symptomatic hypotension appeared to be the abrupt disappearance of muscle sympathetic nerve activity [28].

The cardiac autonomic changes associated with these reflexes can be established with the use of heart rate variability measured at the onset of vasovagal syncope. In one study of 22 healthy patients who first experienced syncope during a tilt table study, two different patterns of autonomic changes were noted: a progressive increase in cardiac sympathetic modulation up to sudden onset of bradycardia; and gradual sympathetic inhibition with a concomitant enhancement of vagal modulation [29]. (See "Heart rate variability: Technical aspects").

  Alpha-adrenergic response — Vascular tone is maintained in part by alpha-adrenergic tone. Although these reflexes are physiologic, some individuals have an exaggerated or even a pathologic response. As an example, one study used subhypotensive lower body negative pressure to reduce central blood volume and blood pressure in patients with vasovagal syncope [30]. Although this technique should normally inactivate the cardiopulmonary mechanoreceptors and increase vascular resistance and blood pressure, these patients responded with a decrease in peripheral vascular resistance, suggesting a paradoxical activation of these receptors. This finding may result from reduced cardiopulmonary baroreceptor sensitivity or from a pathologic response to an increase in left ventricular inotropy due to reduced venous return and left ventricular volume.

In addition, pure cardioinhibitory or pure vasodepressor responses can occur, but most patients have a mixed response.

Transcranial Doppler monitoring has shown that there is sudden intense cerebral vasoconstriction that occurs concomitant with or prior to the loss of consciousness in patients with neurocardiogenic syncope [31]. This response is usually due to hypotension but approximately 1 percent of patients with neurocardiogenic syncope have cerebral vasoconstriction in the absence of systemic hypotension, suggesting a derangement of cerebral autoregulation [32].

  Central serotonergic mechanism — Several studies have shown that central serotonergic pathways participate in the regulation of blood pressure. Drugs that enhance central serotonergic activity, such as fenfluramine and clomipramine, increase the plasma levels of prolactin and cortisol and have been used to evaluate the activity of the serotonergic system. One study administered clomipramine to 20 subjects, eight of whom had a history of neurocardiogenic syncope; those with a history of syncope had higher levels of prolactin and cortisol compared to controls without syncope [33]. Serum concentrations of prolactin and cortisol obtained during tilt table testing (without clomipramine) were significantly elevated only in those subjects with a positive test.

Orthostatic hypotension — Orthostatic hypotension is defined as a postural decrease in systolic blood pressure of at least 20 mmHg, in diastolic blood pressure of at least 10 mmHg or symptoms. The major causes of orthostatic hypotension associated with syncope include:

  •  Decreased intravascular volume, as may occur with diuretics
  •  Drug effects, especially antidepressants (tricyclics, phenothiazine) and antihypertensive agents (beta and alpha blockers, hydralazine, ACE inhibitors, ganglionic blockers), particularly  vasodilators, including calcium channel blockers and nitrates (more commonly observed in the elderly). Other drugs include opiates and bromocriptine
  •  Primary autonomic insufficiency or failure (including Shy-Drager syndrome, Parkinson's disease, and others)
  •  Secondary autonomic insufficiency (eg, diabetes mellitus and amyloidosis)
  •  Alcohol consumption which impairs vasoconstriction [34]

(See "Mechanisms and causes of orthostatic and postprandial hypotension").

Carotid sinus hypersensitivity — Hypersensitivity of the carotid sinus results in vagal activation which in turn leads to similar physiologic manifestations to those seen with vasovagal syncope (show figure 5). Three different types of hypersensitivity are therefore described: cardioinhibitory, vasodepressor, and mixed. A patient is considered to have carotid sinus hypersensitivity if the application of pressure to the carotid sinus leads to a pause of three seconds or more. (See "Evaluation of the patient with syncope", section on Carotid sinus massage for a discussion on how to perform the test).

Cerebrovascular disease — Atherosclerotic disease of the cerebral arteries had previously been considered a common cause for syncope. However, it rarely causes true syncopal symptoms. Instead, stroke and transient ischemia attacks cause focal neurologic deficits that do not recover rapidly or completely. As examples:

  •  If the posterior cerebral circulation is impaired (vertebrobasilar artery insufficiency), symptoms such as dizziness are more apt to occur than is syncope.

  •  If the anterior circulation is compromised, a focal neurologic deficit and not a global decrease in consciousness will occur.

The rare exception in which syncope can occur is with severe obstructive four vessel cerebrovascular disease; however, other neurologic findings are likely to occur prior to the loss of consciousness in these patients.

Metabolic or toxic abnormalities — Metabolic or toxic abnormalities are rarely associated with an abrupt onset or complete brisk recovery; syncope is therefore rare in this setting. Hypoglycemia and encephalitis can cause coma, stupor, and confusion, but rarely syncope. Nonetheless, if a patient does not recall the history surrounding the event or if the event was unwitnessed, distinguishing coma from syncope can be difficult.

Micturition syncope — Micturition syncope is a form of situational vasovagal syncope, seen in 3.3 percent of patients in one study [10]. The cause for this type of syncope is probably related to an abrupt change in position combined with a strong vagal stimulus (show figure 5). Initial studies suggested a clear male predominance in this condition; however, one report found a higher incidence of syncope after evening micturition in women [35].

Autonomic reflexes — Several other autonomic reflex mechanisms for situational vasovagal syncope can occur including those causing deglutition syncope, posttussive syncope, and syncope associated with defecation. While the mechanisms for these entities are not identical, the autonomic nervous system appears critically involved in the initiation of the episode. Syncopal symptoms in these settings generally involve a poorly tolerated hemodynamic response to specific autonomic cardiovascular reflexes.

Seizures — Seizures can mimic syncope, especially when:

  •  The seizure is atypical and not associated with tonic-clonic movements
  •  A complete history cannot be obtained
  •  A grand mal seizure occurs (see below)

Seizures are the likely cause for 10 to 15 percent of apparent syncopal episodes. Patients with seizures, however, rarely have an abrupt and complete recovery. Instead, the postictal state is characterized by a slow and complete recovery.

A potentially confounding factor is that loss of cerebral blood flow due to any cause of syncope can result in a seizure-like state. As an example, the initiation of a rapid VT may be associated with impaired cerebral blood flow, followed within seconds by tonic-clonic movements. This apparent seizure activity is associated with brain wave slowing, not epileptiform spikes, on the EEG.

Seizures are frequently mistaken for syncope. In the original syncope series, for example, seizures were mistaken for syncope in 26 of 510 patients [10]. This problem was more recently addressed in a study in which only 31 percent of physicians evaluating patients with syncope agreed on whether or not seizure was the cause for the observed episode [36]. This distinction can be especially difficult if the seizures are atypical or if there is no witness to the event. One clue that may be helpful, if present, is evidence of soft tissue injury at multiple sites due to tonic-clonic movements during the seizure.

SYNCOPE OF UNKNOWN ORIGIN — Patients in whom an etiology for the symptoms of syncope is not found are categorized as having syncope of unknown origin. This definition, however, is strongly dependent upon the investigators and the modalities used for diagnosis. Thus, a patient with an unknown diagnosis in one study may be recognized as having a clear diagnosis in another based upon different methods of diagnosis and analysis. As an example, a patient in whom electrophysiologic testing is performed for the analysis of syncope may be considered to have syncope of unknown origin if the study is negative, even though an arrhythmic cause for syncope was suspected. The prognosis would be considered to be excellent in this setting if other studies were noncontributory. The same patient, however, might have a significantly different diagnosis and a poorer prognosis if one or more episodes of asymptomatic nonsustained VT were noted on an ambulatory monitor.

Such differences in the degree of evaluation complicates interpretation of different reports in the literature.

Incidence — The reported incidence of syncope of unknown origin has varied over time. In the original syncope study, for example, syncope of unknown origin was present in only 23 of the 510 patients (4.5 percent) [10]. More recent data suggests approximately one-half of patients who are admitted for syncope are ultimately categorized as having syncope of unknown etiology [8]. This apparent increase in incidence has occurred despite improved diagnostic technology because the degree of certainty tolerated for diagnosis has become more stringent with time. In essence, it is only possible to know the cause for syncope if the episode is witnessed, and if the electrocardiogram, the blood pressure, and the electroencephalogram are monitored. Even then, the diagnosis may not be clear. As a result, the causes thought to be responsible for syncope are often based upon flawed methods and incorrect assumptions.