Clinical manifestations

Presentation and course

Functional seizures are involuntary, intermittent events that resemble epileptic seizures but are associated with biopsychosocial stressors (24; 74). Functional seizures can have certain characteristic manifestations that help to differentiate them from epileptic convulsions: duration longer than 5 minutes, hip-thrusting, high seizure frequency, many seizure types, ictal hallucinations, closed eyes, and pre-ictal headache (63; 92). Although other descriptors of behavior have been described, the following occur less often: gradual onset, asymmetrical thrashing movements of the limbs and side-to-side head movements (rather than bilaterally symmetrical tonic-clonic movements), opisthotonic posturing, lack of stereotypic pattern, talking or screaming throughout the seizure, weeping, prolonged bilateral involvement without impairment of consciousness, sudden return to consciousness following a prolonged generalized seizure, resolution with whispering or rapid and non-stertorous breathing, and a waxing and waning of any of the behaviors (23; 138; 26; 120; 114).

Functional seizures can mimic epileptic ictal events exactly, including pupillary dilation, urinary and fecal incontinence, tongue-biting, Babinski responses, and even bodily injury due to accidents that occur during an episode (84; 63).

However, ictal events that appear not to respect conventional understanding of neuroanatomical correspondences or have highly unusual features, including those previously mentioned as typical of functional seizures, can occasionally be epileptic (118; 99). Patients and witnesses also may be no more accurate than chance at recalling details of the seizures (126; 127). Consequently, it can be extremely dangerous to make a diagnosis of functional seizures based on description or observation of the ictal event alone (11; 78). We find that patients report of a large list of comorbidities, especially migraines, chronic pain, and asthma; many medications for those conditions and other historical factors (eg, concussion, older age of onset, and major psychological trauma including sexual abuse) may be at least as helpful as descriptions of ictal behavior (65; 64). To assist clinicians in differentiation, multiple likelihood scores to differentiate functional seizures from epilepsy have been developed using a limited number of questions (66; 135; 140), including the functional seizures likelihood score.

Studies show that neurology trainees and internal medicine physicians have an area under the receiver operating curve (AUC) of around 75% when distinguishing between functional seizures and epileptic seizures using video recordings without EEG (121). Although videos of seizures were only available in less than 1% of patients queried, when they were available, they were helpful (130). Although the AUC of experienced epileptologists was 90%, there is significant morbidity inherent in misclassified cases in both cases of error. For example, 10% of cases of empirically treated “status epilepticus” in the ESETT trial were determined to have prolonged functional seizures, leading to known iatrogenic risks associated with large doses of sedating medication, intubation, and hospitalization in the intensive care unit (60). Similarly, there are anecdotal reports of sudden unexpected death in epilepsy (SUDEP) in patients thought to have functional seizures alone whose antiseizure medication was discontinued but instead had both epileptic and functional seizures.

Prognosis and complications

With targeted psychotherapy, around 30% of patients will be seizure-free and a total of 80% will have at least a 50% reduction in seizure frequency (85; 51). Prognosis is worse for patients whose delay from first seizure to diagnosis of functional seizures is prolonged; who have comorbid epileptic seizures; whose personalities have greater somatization or dissociation, or include positive phenomena; or for patients who receive disability benefit, have higher anxiety and depression scores, are at an older age at onset, are unemployed, have more functional symptoms, have lower educational qualification, and have a lesser belief in the diagnosis and in CBT as “logical” treatment (111; 50). There are two key barriers to treatment for functional seizures: lack of availability of trained psychological or psychiatric providers and nonadherence to psychotherapy. At the time of diagnosis, cessation of all antiseizure mediations in patients without epilepsy and motivational interviewing have been shown to improve attendance at subsequent psychotherapy appointments (133). Due to functional disorders existing at the intersection between mental health and neurology, many psychotherapists are uncomfortable with or unfamiliar with the treatment for functional seizures. Psychotherapists or psychiatrists who misinterpret the seizures as epileptic, despite more definitive video-EEG documentation of seizures, can confuse patients and thereby be counterproductive (113; 55). Additionally, there are fewer than 20 centers in the United States that provide specialized neurobehavioral treatment (15). To address this limitation in access to providers familiar with the condition, a randomized trial for neurobehavioral therapy for functional seizures in the United Kingdom showed that therapists could provide effective therapy with relatively limited additional training (51).

Similar to epileptic seizures, functional seizures impact patients’ lives beyond the times of the seizures; therefore, patient outcome should not be based solely on reductions in seizure frequency (101). Seizure impact spans multiple aspects of quality-of-life and may be greater than the impact of epileptic seizures (128; 61). Prognosis in children may be better than in adults (141). An Australian investigation suggests that patients with functional seizures have a 2.6 times higher standardized mortality ratio than the general population, not significantly different from the rate of SUDEP in patients with epilepsy (96; 94). Importantly, a quarter of those deaths were suicide (96). We suggest that neurologists continue to be involved in the care of patients with functional seizures because, while uncommon, patients can develop nonseizure functional neurologic disorders during the treatment course and cessation of antiseizure medications can reveal comorbid epileptic seizures.

Clinical vignette

A 25-year-old woman developed episodes of loss of consciousness at 15 years of age. The initial episodes were a sudden loss of consciousness with collapse and without associated movements. A cardiac evaluation identified baseline low blood pressure and an abnormal tilt table test, so treatment with midodrine and nadolol was begun. This treatment was effective for about 2 years. The episodes then returned during the patient’s first year of college but were different at their onset, which became a shaking of the right shoulder that spread across the entire right side. Around this time, the patient identified the development of a persisting right hemiparesis that continues to impair ambulation to the extent that a cane is necessary. Based on the development of shaking, epilepsy was diagnosed, and levetiracetam replaced the other medications. With this treatment, the episodes were controlled again. The episodes returned and increased in frequency despite continued treatment with levetiracetam. Their current manifestation does not include asymmetric shaking and instead is intermittent truncal contractions that sometimes are associated with oral injury or urinary incontinence. The duration of the episodes varies from 2 to 30 minutes and resolve with headache and fatigue.

To address the underlying diagnosis, video-EEG monitoring was performed. Two episodes were recorded and identified by the patient’s mother as similar to the habitual episodes. The episodes included intermittent truncal movements not typical of epileptic seizures, and the EEG was normal both during and between the episodes. Neurologic examination identified an inconsistent and variable right-sided weakness, and MRI of the brain was normal. The diagnosis was changed to functional seizures, and a psychiatric consultation did not identify a formal diagnosis of a psychiatric disorder. The psychiatrist agreed with the diagnosis of functional neurologic disorder with associated psychosocial stressors. No risk factors for functional seizures were identified. However, the patient’s interaction with her family demonstrated an abnormal dependence on them that was in conflict with her stated desires. Furthermore, she demonstrated no negative emotions when discussing the dependence or her seizures. With the patient’s understanding of the new diagnosis, levetiracetam was discontinued, and neurobehavioral therapy was recommended.

Biological basis

Etiology and pathogenesis

Traditionally, functional seizures are classified as within the umbrella functional neurologic disorders, which sometimes also has features of dissociative disorder, which is a different category in the DSM-V (04). Regardless of the classification, the causative etiology of functional seizures is not well established. The factors associated with functional seizures are heterogeneous and can include mild traumatic brain injury, epileptic seizures, various forms of psychological disorders, emotional conflicts, inappropriate coping mechanisms, and psychoses (46; 115; 91). Functional neurologic disorders have been attributed to impairment in connectivity between neuroanatomic areas involved in emotion regulation and areas involved in motor planning and conscious awareness. Patients with functional neurologic disorder often have difficulty consciously characterizing their emotions or feelings with words. This has been termed alexithymia, and individuals with functional seizures are more likely to have alexithymia than control individuals with epilepsy (59). Ineffective coping responses also may contribute because patients with functional seizures more often employ emotion-focused or avoidance-oriented strategies than the more effective task-oriented approach (93). Overall, the presence of trauma, abuse, or neglect during childhood is associated with the subsequent development of functional seizures (59; 93). In military veterans, there is significant comorbidity with posttraumatic stress (39). It is important to emphasize that 10% to 15% of patients with functional neurologic disorder do not acknowledge psychological stressors as contributing to their symptoms; therefore, some patients may resist etiological descriptions or delivery of the diagnosis that rely entirely upon psychological factors. In addition to these psychologically based associations, functional neurologic disorders are associated with mild traumatic brain injury and other reportedly mild biological stressors, such as vaccinations and mild viral illnesses (44; 54).

Identifying the primary and secondary factors that contribute to or perpetuate the functional seizures is important because identification of triggers and developing alternating coping strategies is one of the tenants of successful treatment (49; 77). Psychodynamics may differ depending on the type of functional ictal manifestations (104). Personality abnormality can be present with features that most often resemble borderline or avoidant personality disorder, or an overly controlled or somaticizing personality (110; 28; 22). Epileptic seizures can also progress into functional seizures, and functional behaviors can exaggerate or elaborate epileptic auras (32).

Functional seizure disorders are psychiatric conditions and, as such, are due to abnormality in the complex behavioral systems within the brain. Brain imaging of patients with functional seizures demonstrates subtle quantitative abnormalities in structure and resting state functional MRI when compared to controls (34; 75; 88; 52; 129; 70). One finding is greater functional connectivity between the insula and both the precentral sulcus and the parietal lobe (136). Moreover, this abnormality correlates with the dissociation score. Large-scale brain networks also differ from controls with increased local specialization and decreased global integration, which indicates a global network with less efficient information propagation (35). These regions have incomplete overlap with regions associated with impaired emotion regulation, as well as depression, bipolar disorder, schizophrenia, posttraumatic stress disorder, and mild traumatic brain injury (41). The clinical implications of this finding are rudimentary at this time.

Another perspective on the etiology of functional seizures and other functional neurologic disorders is based on a model comprising abnormal inferences within cognition and perception (40). Essentially, abnormal top-down modulation may result from prior beliefs and experiences and produce abnormal behavior. From extracranial and intracranial electroencephalographic recordings of functional seizures, this may involve the posterior parietal cortex, which is an area involved in self-referential cognition (06).

Furthermore, organic brain damage is not uncommon among patients with functional seizures (83; 10; 68), especially when it affects the nondominant hemisphere (33).

Epidemiology

There are few reliable statistics on the incidence and prevalence of functional seizure disorders, but the incidence has been reported as 4.9 per 100,000 per year and prevalence reports range from 2 to 33 per 100,000 (105; 18; 38). Functional seizures are three times more common in women than men (84; 34). Functional seizures rarely occur in children prior to puberty but must be distinguished from other nonepileptic paroxysmal events that are common in the younger age groups, especially in those children with intellectual impairment (05; 98).

Approximately 10% of patients with functional seizures also experience comorbid epileptic seizures that may or may not be similar in semiology (17; 122; 62). Some studies place this proportion much higher (47; 56), and this discrepancy can be explained by the fact that patients with both functional and epileptic seizures present extremely difficult differential diagnostic problems, and the overall greater rate at referral centers may be due to a selection bias toward the more complicated conditions. Additionally, this rate tends to be higher in patients with video-EEG-documented functional seizures and suspicion for comorbid epileptic seizures that were not observed during video-EEG monitoring.

Prevention

Patients with functional seizures commonly have experienced severe biopsychosocially traumatic events, such as sexual abuse in childhood (115). Patients with coexistent functional and epileptic seizures frequently have become dependent on their illness; when medical or surgical management greatly reduces or eliminates the habitual epileptic ictal events, they develop functional seizures to take the place of their epileptic seizures (48). Functional seizures also can occur in around 1% of intracranial surgeries for epilepsy, but this is rare (109). Unfortunately, there are no identified strategies to prospectively prevent or mitigate the risk of functional seizures in patients before the seizures start.

Differential diagnosis

Functional seizures need to be distinguished from epileptic seizures as well as intermittent nonepileptic events due to systemic, neurologic, or other psychiatric disturbances. Frontal lobe seizures are most commonly misdiagnosed as functional (118), but, in video-EEG populations, functional seizures are epidemiologically more common than confirmed frontal lobe epilepsy (72). Patients with both epileptic and functional seizures pose the most difficult diagnostic dilemmas. In addition, some patients with focal seizures with loss of awareness or focal to generalized seizures that are well controlled with medication may highlight or elaborate on residual auras (46). These apparent functional events, therefore, have an epileptic basis that, if diagnosed, might be treatable. On the other hand, epileptic seizures can occasionally be precipitated by psychological factors, a condition that has been referred to, confusingly, as psychogenic epileptic seizures.

History, careful description of typical ictal events, physical examination, and routine laboratory studies (including EEG) may be sufficient to permit identification of a specific epileptic condition or raise concern about the existence of a systemic, neurologic, or psychiatric illness associated with nonepileptic intermittent behavioral disturbances. However, an abnormal (even epileptiform) EEG can occasionally be seen in people with functional seizures who do not have epilepsy (108; 01).

Historically, a broad term for nonepileptic seizures was pseudoseizures, and the term “pseudoseizures” is sometimes still used as a synonym for functional seizures; however, pseudoseizures are any paroxysmal behavior that resembles an epileptic seizure, including other paroxysmal neurologic disorders. Therefore, the term functional, dissociative, or psychogenic seizure is more specific. Consequently, the terms functional seizure or “psychogenic nonepileptic seizure” are used more because they are more precise. Patients also find the term “pseudoseizures” pejorative because it suggests that their seizures are “fake.” (See terminology discussion above.)

Functional seizures are not a form of malingering, which would involve epileptic-like events that are consciously feigned; nor are they factitious seizures, which would involve a seizure history or actual events that are fabricated solely for the purpose of gaining patient status (Munchausen syndrome) (116). "Factitious seizures by proxy" refers to a situation where an individual, usually a relative of the patient, fabricates the seizure history or seizures in order to have the patient admitted to a hospital (36).

Diagnostic workup

In distinguishing between epileptic seizures and functional seizures, the physician should initiate an initial neurologic evaluation that is appropriate to the type of epilepsy suspected. If nonepileptic seizures due to medical, neurologic, or other psychiatric causes are considered, then additional testing needs to focus on these likely causes. To differentiate nonepileptic seizures from malingering, which is rare, it is important to document potential secondary gain.

The diagnosis of functional seizures should be considered based on the patient’s history and reported or observed semiology. There are multiple interview-based or questionnaire-based clinical scores to identify patients who were more likely to have functional seizures (65; 64; 69; 135; 140). Due to evidence that longer delay to diagnosis is associated with poor long-term outcome, functional seizures should not be viewed as a diagnosis of exclusion (139; 16; 57; 106; 100). The level of evidence needed for the diagnosis of functional seizures is summarized in table 1 (78).

Table 1. Diagnostic Levels of Certainty for Functional Seizures Reproduced from LaFrance and Colleagues with permission

As suggested by their prominent place in the diagnostic certainty criteria, videos of seizures can be particularly helpful. However, it can be difficult to obtain videos of sufficient quality. A high-quality patient- or caregiver-provided video should record the ictal period in its entirety; include the patient’s entire body within view; record as much of the postictal period as possible; and ensure proper lighting, sound, and focus while maintaining interaction with the patient to assess responsiveness (130). This high-quality video, however, may only be available for less than 1% of patients unless these videos were specifically requested (130). When a quality video was viewed by an epileptologist, the accuracy of the diagnosis of functional seizures was 95%, but this accuracy diminished substantially with level of expertise, including an area under the receiver operating characteristic curve (AUC) of 89% for neurologists, 71% for emergency physicians, 69% for nurses with experience in epilepsy, and 49% for other nurses (21).

Additional tests can influence the likelihood of the diagnosis, including but not limited to, psychological measures, psychiatric assessment, review-of-systems questionnaires, self-protective maneuvers during the event, as well as preserved memory during generalized convulsive episodes (125; 28; 29; 22; 112). To date, the sensitivity and specificity of serum markers taken around the event are limited because these also are unchanged in some epileptic seizures (23; 03; 27; 25; 29). Some patients with functional seizures also have a template of a friend or relative with epileptic seizures (137). Furthermore, demonstration of brain damage, or even epileptic seizures, does not necessarily prove that the patient does not also have functional seizures (83; 27; 71).

When clinical doubt persists following routine evaluations, it may be necessary to perform video-EEG monitoring to record the habitual seizures. Due to frequent emergency room visits and lost productivity from frequent seizures, the cost of untreated functional seizures was estimated to be 21,000 euros per patient per year (14; 86). Although inpatient video-EEG monitoring is costly, it has been found to produce a subsequent decrease in total healthcare costs, even with increased behavioral health costs (02). Inpatient video-EEG monitoring is most helpful when seizures involve impairment of consciousness because focal seizures with intact awareness commonly have no EEG correlates, and a negative ictal EEG recording cannot rule out epilepsy for these events. Moreover, movement artifact on the EEG may give the appearance of epileptic abnormality to less experienced EEG interpreters (16). In some patients, ictal or peri-ictal heart rate, or heart rate variability, may distinguish between epileptic and functional seizures (95; 103; 107), and ictal SPECT can be helpful (30).

Examination of the patient during the behavioral seizure can also be useful, particularly when conscious behavior is demonstrated in an apparently unconscious patient (130). This can lead to erroneous conclusions, however, when the examiner is not familiar with ictal signs and symptoms. For instance, certain myoclonic phenomena can be bilaterally synchronous and relatively long-lasting without loss of consciousness (eg, paroxysmal kinesigenic dyskinesia), and patients can often respond appropriately to complicated stimuli during focal impaired awareness seizures but have no memory of their actions afterward. Nevertheless, diagnostic information is often obtained by an experienced observer witnessing a typical seizure.

In lieu of inpatient video-EEG monitoring for this purpose, suggestion has been used as a provocative test for functional seizures (45), but this approach can lead to misdiagnosis when concurrent EEG is not obtained and has been challenged as unethical by some because of the potential impact on trust for the physician. Other provocative tests such as hyperventilation and photic stimulation (20) and hypnosis (12) are reported to be useful. In the context of video-EEG monitoring with provocative techniques, patients with epileptic seizures may experience functional seizures with atypical semiology compared to their typical events due to a conscious or unconscious pressure to perform. Consequently, definitive demonstration of an epileptic seizure or a functional seizure does not rule out the possibility that both seizure types exist unless all seizure types have been recorded. A careful history of each of the different seizure types experienced by the patient is essential because patients may have both functional seizures and epileptic seizures.

Management

Effective management of functional seizures requires that the physician, family, and patients recognize that the events are involuntary, legitimately disabling, not epilepsy, and treatable (53). Appropriate treatments for functional seizures include neurobehavioral therapy, group psychotherapy, behavioral modifications, stress reduction, reassurance, and, at times, antidepressant or antipsychotic medication to treat comorbid psychiatric disease (46; 115; 81; 49; 51; 90; 85). Neurobehavioral therapy also has been called cognitive behavioral-informed psychotherapy. Although it has been established that psychotherapy targeting the events is superior to treatment-as-usual, there isn’t clear evidence that one particular type of psychotherapy is superior to another. LaFrance and colleagues published a comprehensive review of management (82; 100). When patients have only one seizure type and it is functional, antiseizure medication can be safely tapered and discontinued, ideally during the diagnostic video-EEG admission.

When patients have several seizure types, and one or more may be epileptic, antiseizure medication therapy must be continued with counseling that the medication therapy only addresses the epileptic seizures. The patient, as well as family and friends, should be made aware of which ictal events are functional, which are definitely epileptic, and which may be epileptic so that seizure logs can record them independently. It is then necessary to see which seizures respond to antiseizure medication and which to psychotherapy. The specific neurobehavioral therapy for functional seizures was modified based on neurobehavioral therapy to address biopsychosocial factors that can worsen epileptic seizures, therefore response to medication or therapy does not perfectly indicate that the seizures were epileptic or functional. The question as to whether patients with functional seizures should be permitted to drive has not been adequately answered (19).

The interdisciplinary functional seizure clinics that have been established are a collaboration between psychiatry, psychology, social work, and neurology (15). Although all of these providers can be trained in psychotherapy for seizures, the bulk of the psychotherapy can be accomplished with psychologists, whereas psychiatry manages the psychiatric medications and neurology manages seizure medications, if indicated, and evaluates any new or changing characteristics of the seizures or other functional neurologic disorders (85). In absence of an interdisciplinary clinic, we highly encourage keeping the door open for follow-up with neurology, if only for continued reassurance, clarification of the diagnosis, evaluation of new possibly functional symptoms, and coordination of psychological services. When patients are discharged from neurologic practice, there are fears that the 80% of patients who are not seizure-free after neurobehavioral therapy will seek out neurologists or other healthcare providers for re-evaluation, leading to redundancies in healthcare utilization.

The following website has a list of psychotherapy providers for functional seizures: nonepilepticseizures.com.

Outcomes

Simply establishing and delivering the diagnosis in a sensitive manner can have a profound impact on health care utilization as well as the direct and indirect costs of illness (139; 87; 16; 02; 86). In particular, motivational interviewing may improve outcomes (133; 132). In addition to improvements in seizure frequency and rates of seizure freedom, these improvements are made through education regarding the utilization of emergency services, the reduction of costs associated with further diagnostic assessment, the costs of medications, and improvements in the quality of life of patients.

Of patients who complete treatment as described above, the prognosis for control of seizures is modest. Some patients do extremely well, but many continue to either have seizures or other forms of psychiatric disability despite appropriate therapy (89; 73; 42; 43; 119). Seizure freedom has occurred in 16% to 35% of study populations with follow-up periods ranging from 6 to 67 months (139; 43; 58; 97; 51).

However, as few as 60% of patients attend the first psychiatric follow-up visit, and 15% to 25% of patients complete the full recommended course of therapies described above. This adherence varies substantially across studies; therefore, identification of barriers to care and facilitation of follow-up is an active area of research and policy (55; 85; 133; 132). Telepsychiatry has been effective for improving adherence and delivery of therapy both prior to the COVID19 pandemic and during the pandemic (80; 131).

Special considerations

Pregnancy

Functional seizures can occur during pregnancy and may first develop during pregnancy. Prompt definitive diagnosis, ideally with video-EEG monitoring, should be performed because treatment with antiseizure medications may be teratogenic (31).

Anesthesia

Although there are limited peer-reviewed publications on the topic, our clinical experience is that functional seizures can be precipitated by anesthesia and are observed in the post-anesthesia recovery area. We have cared for multiple patients who have a history of seizures only after general anesthesia that were demonstrated to be functional with concurrent extracranial surgery and video-EEG monitoring that was started before surgery and continued in the post-anesthesia care unit. In these cases, we provided reassurance that these functional seizures were not physically dangerous and did not reflect acute neurologic injury from the surgery. We counseled that these seizures may occur after other episodes of general anesthesia and likely warranted time-sensitive neurologic reevaluation before aggressive treatment with benzodiazepines that have risk of iatrogenic harm. Functional seizures are not treated by antiseizure medications, so pre- or post-operative prophylactic antiseizure medications were not recommended. Anecdotally, we also have cared for a patient with medication resistant epilepsy who had functional seizures during the first few days of stereo-EEG monitoring but has not experienced functional seizures at other times.