Hearing voices in patients with functional seizures: Case reports and insights into mechanisms and treatments

Abstract

Introduction

Functional seizures (FS), paroxysmal alterations with motor, sensory, cognitive, and autonomic features without ictal epileptiform activity, are the most common subtype of functional neurological disorder (FND). FS often coexist with other symptoms, one of them being auditory hallucinations (AH); however, evidence examining this relationship is outdated and limited. The link between FND and AH has been explored, with mixed opinions on whether hallucinations should be considered part of the functional cognitive disorder. Importantly, the specific relationship between FS and AH has not been systematically examined.

Methods

We present 2 patients (one pediatric and one young adult) with AH and FS who underwent a 6-week multidisciplinary FND therapy program (weekly physiotherapy and psychotherapy) and experienced complete remission of AH.

Results

Our cases add to the limited literature on the overlap between AH and FS. Whereas prior reports have typically related AH in patients with FS to schizophrenia or other psychotic disorders, our patients did not meet the full DSM-5 diagnostic criteria for schizophrenia, and presented hallucinations without additional psychotic features. In both, hallucinations were temporally and/or contextually related to FS episodes and disappeared after a multidisciplinary FND therapy program.

Conclusions

Hallucinations are a nonspecific symptom of a wide range of psychiatric and nonpsychiatric conditions. The mechanisms underlying AH are multiple and remain under debate; however, AH and FND likely share common mechanisms involving disrupted self-agency and interoception, underscoring the importance of integrated therapeutic approaches over isolated psychiatric treatments.

Resumen

Propósito

Las crisis funcionales (CF) – alteraciones paroxísticas con síntomas motores, sensitivos, cognitivos o autonómicos, sin actividad epileptiforme asociada – son el subtipo más común de trastorno neurológico funcional (TNF). Las CF frecuentemente coexisten con otros síntomas, y uno de ellos son las alucinaciones auditivas (AA); sin embargo, la evidencia que evalúa este fenómeno está desactualizada y es limitada. La relación entre TNF y AA se ha explorado con puntos de vista mixtos acerca de si las AA pertenecen al trastorno cognitivo funcional. La relación específica entre CF y AA no se ha examinado de manera sistemática.

Metodología

Presentamos dos casos (uno pediátrico y un adulto joven) con AA y CF quienes participaron en un programa multidisciplinario de 6 semanas para TNF (fisioterapia semanal y psicoterapia) y presentaron remisión completa de las AA.

Resultados

Nuestros casos se suman a la escasa literatura de la relación entre AA y CF. A diferencia de artículos previos que evalúan AA en CF con esquizofrenia u otros trastornos psicóticos, nuestros pacientes no cumplían criterios DSM V para esquizofrenia y presentaban alucinaciones sin otros síntomas psicóticos. En ambos casos las alucinaciones estaban relacionadas temporal o contextualmente con las CF y desaparecieron tras el tratamiento multidisciplinario.

Conclusiones

Las alucinaciones son un síntoma inespecífico que se presenta en una amplia gama de trastornos psiquiátricos y no psiquiátricos. Los mecanismos fisiopatológicos de las AA son múltiples y todavía muy debatidos; Sin embargo, las AA y el TNF probablemente compartan mecanismos que afectan el sentido de agencia y la interocepción, subrayando la importancia de enfoques terapéuticos integrados sobre tratamientos psiquiátricos aislados.

Introduction

Functional (dissociative) seizures (FS), defined as paroxysmal alterations with motor, sensory, cognitive, and autonomic features without ictal epileptiform activity, are the most common subtype of functional neurological disorder (FND).¹ This highly disabling entity has an annual incidence of 1.5–6.17 cases per 100 000 population.² Recent advances in understanding of the underlying mechanisms have conceptualized FS as brain network disorders, understood from a biopsychosocial perspective.^3,4 Seventy-two percent of patients with FS present various medical comorbidities (migraines, sleep difficulties, irritable bowel syndrome, learning difficulties, chronic pain, and fatigue).⁵ The coexistence of FND and other specific symptoms has also been described. One such symptom is auditory hallucinations (AH), although evidence examining this relationship is outdated and limited.^6–8

AH are defined as sounds experienced by a waking individual in the absence of an external stimulus.⁹ Historically, a difference has been established between true AH and pseudohallucinations. The latter are defined as subjective internal (coming from within rather than from an external source) sensory perceptions and are not accompanied by the full sensory experience associated with true AH. Hallucinations in FND generally occur with preserved insight, without psychotic behavior, and often involve multiple sensory perceptual modalities. It is worth noting that AH are considered to be nonspecific, and may present in an array of diseases.¹⁰

One of the proposed contemporary models for AH involves disorders of the generation of inner speech (mental use of language that often occurs consciously and in the absence of overt articulation) and impaired self-agency and spatial externalization. However, this is only one component of AH and multiple factors are associated with them.¹¹ Neuroimaging studies have shown overactivation of areas involved in the production and perception of language during verbal HA.¹² This precept suggests that AH are generated by the patient's language production mechanisms without self-agency.¹³ Agency is directly related to interoception (the perception of one's physiological condition),¹⁴ as it allows stability of self-awareness by constantly portraying oneself as an active agent.¹⁵ Patients with schizophrenia and AH have been shown to present aberrant interoception.¹⁶ Similarly, patients with FND also have altered interoception, hinting at a common pathophysiological mechanism for FND and related hallucinations. Nevertheless, the link between FND and AH has been explored, with mixed opinions on whether hallucinations should be considered part of functional cognitive disorder. Importantly, the specific relationship between FS and AH has not been systematically examined; available reports are scarce and mostly describe patients with FS and AH in the context of comorbid schizophrenia. This gap highlights the need to further characterize the coexistence of FS and AH within the spectrum of FND.

Non-pharmacological interventions such as cognitive behavioral therapy (CBT) centered around reappraisal of the meaning of AH have been studied for the treatment of AH, with an average effect size of 0.44 in one study; however, the effect is inconclusive, given that the final objective of CBT is not the disappearance of hallucinations, but rather a decrease in the emotional interference associated with them.^12,17 Literature on the therapeutic management of AH and FND is lacking. Overall, treatments directly targeting hallucinations have shown limited efficacy, and the literature on therapeutic approaches addressing AH within FND is scarce.

If these symptoms indeed share underlying mechanisms, therapeutic interventions targeting FND would be expected to improve both. In line with this, we present 2 cases (one pediatric and one young adult) with AH and FS who underwent a 6-week multidisciplinary FND therapy program (weekly sessions of physiotherapy and psychotherapy) and experienced complete remission of AH.

Methods

Patients were evaluated at the FND unit of the Clinical Neuroscience Institute of Vithas Madrid University Hospitals between January 2023 and October 2024. Each patient underwent a 6-week therapy program with physiotherapy and psychotherapy, with follow-up neurology visits at 6 weeks and 3 months. Both patients (or their legal guardian, if needed) gave written consent to be included in the present study.

Case presentations

Patient 1

A 15-year-old boy was evaluated for convulsive episodes. He had a previous diagnosis of autistic spectrum disorder without intellectual disability, with obsessive traits (particularly orderliness and cleanliness) and hypersensitivity to lights and sounds, requiring the use of headphones over the past 2 years. He was under pediatric neurology follow-up, and treated with methylphenidate and clonazepam (2 mg TID) for attention-deficit/hyperactivity disorder and associated symptoms.

The patient's family sought medical care at our institution due to frequent paroxysmal episodes characterized by loss of consciousness and rigidity in all 4 limbs, typically lasting more than 5 min. Six months prior to the appearance of FS, he began experiencing AH, characterized by voices perceived as coming from the outside (and not being real), talking about negative experiences in the patient's life. AH usually presented in the afternoons and were not temporally related to FS. On one specific occasion, AH preceded an episode of loss of consciousness and generalized body rigidity lasting approximately 4 h. He was discharged without a definitive diagnosis, started on lorazepam (2.5 mg TID), and advised against returning to the emergency department for similar episodes. Additionally, treatment with venlafaxine and olanzapine was initiated, with no improvement in symptoms.

Upon initial assessment by a neurologist, his symptoms were attributed to anxiety. Lorazepam was subsequently replaced by clonazepam, and olanzapine, paliperidone, and sertraline were added, resulting in no improvement of AH. However, the frequency of episodes increased, with 2–10 episodes occurring daily. FS were suspected during the initial evaluation at our center, and confirmed following a 24-hour EEG study, which revealed no interictal or ictal epileptiform activity during a typical episode. Additionally, brain MRI findings were unremarkable.

Multidisciplinary management was initiated, comprising weekly psychotherapy and specialized physiotherapy for 6 weeks. Physiotherapy focused on identifying prodromal symptoms, developing sensorimotor control strategies to actively prevent seizures, promoting sensory processing and regulation strategies, and establishing a self-care routine. Psychotherapy aimed to reinterpret hallucinations as an internal narrative, and utilized a relational approach to explore the patient's interaction with his symptoms and their impact on interpersonal relationships. Concurrently, clonazepam dosage was gradually reduced, resulting in improved alertness. Upon completion of the multidisciplinary treatment program, AH resolved, and the patient achieved greater control over FS, evidenced by a reduction of greater than 50% in their frequency and a noticeable decrease in episode intensity.

Patient 2

A 21-year-old man with no relevant medical history presented for evaluation of seizures. The index event consisted of generalized rigidity with profuse salivation and jaw clenching, lasting 10–15 min. He received brivaracetam, with no benefit, and experienced a similar episode 6 weeks later that required intensive care unit admission and endotracheal intubation. On discharge, he was started on lacosamide 150 mg twice daily.

Despite treatment, events recurred monthly. Patient-supplied video of a typical spell showed pelvic thrusting and bicycling (“pedaling”) movements. He also developed sudden losses of muscle tone with daily falls. AH had begun approximately 2 years before the onset of FS, and consisted of male voices delivering derogatory comments about his self-worth. The AH occurred daily, were perceived as originating “inside his head,” and were recognized by the patient as unreal. A psychiatrist diagnosed schizophrenia and started lurasidone 74 mg daily; however, both AH and FS persisted without changes.

Findings from neurological examination, 3T brain MRI, and routine EEG were unremarkable. A 24-hour EEG captured a habitual episode without an ictal EEG correlate, but demonstrated low-persistence bilateral frontotemporal epileptiform discharges interictally. Based on the interictal EEG findings and the family's description of the semiology of the first event, he was diagnosed with FS coexisting with mild focal epilepsy. Lacosamide dose was reduced to 100 mg twice daily, and a multidisciplinary FS program, comprising physiotherapy and psychotherapy modeled on previously described protocols, was initiated.

Psychotherapy targeted an entrenched internal dialogue of inferiority and worthlessness linked to issues surrounding his adoptive family. He was encouraged to externalize emotions through writing and, with therapeutic guidance, to examine and reframe negative self-talk. After 6 weeks of multidisciplinary treatment, FS frequency declined markedly, with only one event reported over the subsequent 3-month period, and AH resolved completely. During follow-up, lurasidone dose was successfully tapered without recurrence of AH.

Discussion

In this study, we describe the cases of 2 patients with AH and FS who experienced symptom improvement after a multidisciplinary therapeutic approach designed for FND. Notably, AH resolved completely even though the intervention did not specifically target hallucinations.

Our findings suggest that AH associated with FND may share underlying mechanisms, such as impaired self-agency and altered interoception. Importantly, AH resolved completely following targeted FND treatment, without separate interventions specifically targeting hallucinations. The link between FND and AH has been explored, with mixed opinions on whether hallucinations should be considered part of functional cognitive disorder.¹⁸ The specific relationship between FS and AH has not been examined in detail; only a few case reports describe concurrent FS and AH, in patients who were also diagnosed with schizophrenia.¹⁹

First, the neural processes underlying FS remain an active area of research. Classical Freudian concepts of internalized conflict manifesting as sensory and motor symptoms have given way to a multifactorial biopsychosocial model. This contemporary framework integrates neurobiological, psychological, cognitive, and sociodemographic variables to identify predisposing, precipitating, and perpetuating factors for functional symptoms.²⁰ The biopsychosocial approach emphasizes cognitive and perceptual factors (top-down and bottom-up processes), including attentional allocation, predictive coding, interoception, and emotional and somatosensory processing.²¹ Top-down processes interact with physiological hyperarousal, involuntary conditioned responses, and local and central processing disturbances, leading to activation of a seizure scaffold based on prior experiences and perceptions.²²

AH are commonly defined as auditory perceptions experienced by a conscious individual in the absence of external stimuli.⁹ Prior research links AH to disturbances in inner-speech generation and self-agency, whereby internally generated speech is misattributed as originating externally.^11,13 Neuroimaging studies consistently report hyperactivation during AH in regions involved in language production and perception, including Broca's area, the insula, inferior frontal gyrus, superior temporal gyrus, and hippocampal regions.^12,13 The concept of agency (recognizing oneself as the initiator of actions or thoughts) is closely tied to interoception, the internal perception of bodily states.^14,15 Patients with schizophrenia who experience hallucinations show aberrant interoceptive processing.¹⁶ Similarly, patients with FND exhibit impaired interoception, suggesting a shared pathophysiological mechanism contributing to both FS and hallucinations.²³

It is worth noting that discussion of AH in FND often brings to mind post-traumatic stress disorder (PTSD). However, none of the patients met DSM-5 or ICD-10 criteria for PTSD (lack of intrusive flashbacks, no avoidance of specific triggers).²⁴ Moreover, hallucinations in PTSD tend to be discrete flashbacks to the traumatic event, cause distress, and have an internal perceived locus. With this in mind, patient 1 had AH about negative experiences in his own life, but they were clearly perceived as having an external locus; patient 2 had AH with an internal locus, however they were about self-worth and not a specific traumatic experience.²⁵ Differentiating hallucinations in FND, PTSD, and other neurological and psychiatric disorders is not only based on internal inconsistency; rather, a thorough investigation into the phenomenology is warranted. All hallucinations have internal inconsistency and are based on changes in neural function rather than a specific structural lesion.

A notable implication of our cases is therapeutic: standard treatment for FS within a multidisciplinary FND program was sufficient to resolve AH, without additional psychiatric interventions, which were previously proven ineffective. This supports the hypothesis raised in the introduction that, if FS and AH share such mechanisms as impaired self-agency and altered interoception, targeting the underlying FND can lead to improvement of both. Multidisciplinary therapies tailored to FND, including education, psychotherapy, and physiotherapy, aim to improve self-awareness, sensorimotor control, emotional regulation, and cognitive reinterpretation of symptoms. Psychotherapy emphasized the recognition of hallucinations as manifestations of internal narratives rather than external threats, thereby enhancing self-agency and interoceptive awareness. Recent reviews support the efficacy of cognitive-behavioral and psychodynamic therapies for positive symptoms of schizophrenia (e.g., hallucinations and delusions) and for FND symptoms, indicating common therapeutic pathways.^26,27

Our cases add to the limited literature on the overlap between AH and FS. Whereas prior reports have typically related AH in patients with FS to schizophrenia or other psychotic disorders,¹⁹ our patients did not meet full diagnostic criteria for schizophrenia according to the DSM-5,²⁸ and presented hallucinations without additional psychotic features. In both patients, hallucinations were temporally and contextually related to FS episodes. In one patient, the hallucinations closely mirrored his negative internal dialogue, reinforcing the view that AH in this context may arise from diminished interoception and self-agency, leading to misinterpretation of inner speech as being externally generated (Fig. 1), rather than reflecting a distinct psychotic phenomenon.

Figure 1.

Proposed physiopathology of auditory hallucinations. (A) Predictive brain modeling and processing of inner speech in a healthy person depends on agency and interoception. Brain areas associated with speech perception and production generate an internal dialogue that is processed as inner speech thanks to self-agency, which depends on interoception. Inner speech (bottom-up) is compared with brain-generated predictions (top-down), based on past experiences and beliefs, to create perception. (B) The lack of self-agency and introspection in FND results in failure to recognize the internal dialogue as having been produced by the patient, with the inner dialogue being perceived as having come from somebody else (auditory hallucination). Bottom-up information does not match top-down predictions, creating a perception error.

These findings are clinically significant: AH in patients with FND may not represent a separate psychiatric condition but rather an associated phenomenon sharing pathophysiological mechanisms with FS. Accordingly, assessing and addressing AH within the broader context of FND treatment may improve outcomes and help avoid unnecessary (and potentially counterproductive) polypharmacy, which may impair alertness and insight, hindering functional recovery. The small number of cases limits the generalizability of our results. Prospective studies with larger cohorts and control groups are needed to validate these observations.

In conclusion, hallucinations are a nonspecific symptom that presents across a wide range of psychiatric and nonpsychiatric conditions. The mechanisms underlying AH are multiple and remain under debate; however, AH and FND likely share mechanisms involving disrupted self-agency and interoception, underscoring the importance of integrated therapeutic approaches over isolated psychiatric treatments. Our cases illustrate that AH associated with FS may respond effectively to multidisciplinary FND treatments, even in the absence of comorbid psychotic disorder. To our knowledge, this represents one of the first reports documenting complete remission of AH in patients with FS after a standard FND intervention. Patients will require close follow-up to determine symptomatic relapse or appearance of new symptoms suggestive of a comorbid trauma-related response. A multidisciplinary therapeutic approach focused on retraining neural processes may therefore be a promising strategy not only for AH in FND, but also for other conditions where impaired self-agency and interoception play a role.

Conflict of interest

The authors declare no conflict of interest.