Because the brain is the central organ that regulates behavior, a change of behavior may come about with disturbances of the brain, either through alteration of its structure or via functional change, functional here being used in its original meaning to emphasize disturbance of brain function.⁹² Neurologists interested in behavioral disorders have mainly concerned themselves with patients with lesions that cause structural changes, whereas psychiatrists have dealt more with the consequences of disturbed function, where underlying structural lesions have been more difficult to discern. Epilepsy is one of a number of conditions in which there is often an underlying structural abnormality to be found if the appropriate technique is used (e.g., neuropathology, magnetic resonance imaging), but profound functional changes also occur, either as a consequence or independently. These may be reflected in the seizure, one manifestation of the epilepsy process, but may also be associated with some of the less dramatic but nonetheless clinically significant behavioral manifestations of epilepsy.

Psychosis, as used in the International Classification of Diseases (ICD)-10,²⁵ defines a disorder with the presence of “hallucinations, delusions, or a limited number of severe abnormalities of behavior, such as gross excitement and overactivity, marked psychomotor retardation, and catatonic behavior.” In the Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV), the term psychotic refers to delusions, any prominent hallucinations, disorganized speech, or disorganized or catatonic behavior.⁷

Hallucinations and delusions, the hallmark of psychosis, suggest some deviant neurologic processing, underlying which are usually structural but sometimes solely functional alterations of activity. Although in epilepsy hallucinations and delusions may be experienced in certain settings for which patients have clear insight, in the majority of cases insight is lacking, and the condition is truly psychotic.

In the middle of the 19th century, European psychiatrists noted the high incidence of psychotic episodes in institutionalized patients with epilepsy. Several authors described the specific psychopathology of psychiatric complications occurring in the context of epilepsy using such terms as “épilepsie lavvée,”⁵⁰ “grand mal intellectual,”⁹ “epileptoid states,”¹⁸ and “epileptic equivalents.”²⁴ Samt⁶² put forward the idea that the pathophysiology of certain psychoses occurring in the context of epilepsy, especially episodic twilight states, was identical to the pathophysiology of motor seizures. He suggested that in the absence of epileptic seizures, such epileptic equivalents could be sufficient for a diagnosis of epilepsy.

Some authors in the 19th century explicitly noted the rarity of chronic paranoia or true madness in patients with epilepsy. These observations resulted in intensive discussions on the nature of the relationship between epilepsy and schizophrenia, a subject frequently chosen in theoretical disputes on definitions of terms such as disease and symptom complex in psychiatry at the beginning of the 20th century.^20,30

Combined seizures and schizophrenialike symptoms have generally been interpreted either as symptomatic seizures secondary to cerebral sequelae of insanity—for example, brain edema in catatonia—or as symptomatic psychoses caused by seizures or the underlying epileptic process.^32,36 In cases with no obvious temporal relationship between epileptic seizures and psychotic symptoms, it was speculated that both were not directly linked but were caused by the same underlying brain pathology.^54,73 Ganter,¹³ Krapf,³² and Glaus¹⁶ published clinical case series with prevalence rates of combinations lower than expected. These studies, together with observations of alternating periods with seizures and seizure-free periods with psychosis in some patients and the improvement of psychotic symptoms after spontaneous seizures in others, led to the theory of functional dependency and biologic antagonism of schizophrenic and epileptic symptoms, a concept that influenced von Meduna⁴⁶ to introduce iatrogenic convulsions into the treatment of schizophrenia.

With progress in diagnosis and treatment in epilepsy, epileptology shifted conceptually to the realm of neurologists in many countries. Psychiatric aspects were neglected until they were “rediscovered” in the 1950s and 1960s.^14,38,82 American and English authors reported an excess of schizophrenialike psychoses in epilepsy patients, especially in those suffering from temporal lobe epilepsy.^15,55,70

Slater et al. published a detailed analysis of 69 patients from two London hospitals who suffered from epilepsy and interictal psychoses. On the basis of this case series, the authors challenged the antagonism theory and postulated a positive link between epilepsy and schizophrenia. Although Slater was criticized for drawing conclusions on the basis of insufficient statistics,⁷⁷ the temporal lobe hypothesis soon became broadly accepted and stimulated extensive research into the role of temporal lobe pathology in schizophrenia. The use of epileptic psychoses as a biologic model or “mockup” of schizophrenia⁵⁹ is largely based on Gibbs and Slater’s work.

The possible impact of research into epileptic psychosis on the understanding of the pathophysiology of endogenous psychoses explains the bias in the literature toward study of interictal schizophrenialike psychoses. The spectrum of psychotic syndromes in epilepsy is, however, much more complex, and psychotic complications are not restricted to patients with temporal lobe epilepsy.

Some of the earlier studies were reviewed in the first edition of this textbook (Chapter 197). The more recent studies have employed an improved methodology. Bredkjaer² in a record linkage study looked for associations between epilepsy from the national patient register of Denmark and the equivalent psychiatric register. The incidence of nonorganic, nonaffective psychoses, which included schizophrenia and schizophrenia spectrum disorders, was significantly increased in epilepsy, even when patients with learning disability or substance misuse were excluded.

Stefansson et al.⁷⁶ in a case-control study compared the prevalence of nonorganic psychiatric disorders in patients with epilepsy to those with other somatic diseases, the groups being taken from a disability register in Iceland. Although the difference in psychiatric diagnoses overall was not significant, there was a higher rate of psychoses, particularly schizophrenia and paranoid states, among males with epilepsy.

Qin et al.⁵⁶ in another study from Denmark have confirmed the increased risk of schizophrenia and schizophrenialike psychoses in epilepsy, and in this study a family history of psychoses and a family history of epilepsy were significant risk factors for psychosis.

Studies from Japan examining new referrals for epilepsy quote a 6% prevalence of psychoses in those with normal intelligence (in contrast to 24% in those with learning disability).⁴⁴

There are several studies of much more selected populations, such as hospital case series. Thus, Gureje,¹⁹ in patients attending a neurologic clinic, quoted that 37% of patients were psychiatric cases and that 29% of these were psychotic. Mendez et al.⁴⁸ in a retrospective investigation reported that interictal psychotic disorders were found in over 9% of a large cohort of patients with epilepsy in contrast to just over 1% in patients with migraine. The epilepsy sample had more complex partial seizures, more auras, and less generalized epilepsy.

None of the above studies has been able to examine issues related to epilepsy classification in any detail, cohorts being derived from case registers lacking detailed information from, for example, brain imaging. Certain risk factors have been defined, but not from these data, and are noted below.

There is no internationally accepted syndromic classification of psychoses in epilepsy. Most of the previously proposed classification systems for these psychoses^3,10,29,85 are based on a combination of psychopathologic, etiologic, longitudinal, and electroencephalographic (EEG) parameters. Unfortunately, because of a lack of taxonomic studies, our knowledge about regular syndromic associations is still limited.

“Atypical” syndromes are not unusual, and presentations such as those associated with forced normalization (see below) and postictal psychoses make simple divisions between what is ictal and what is interictal difficult to discern. In other words, the above two examples are of psychotic states closely tied to the biology of the ictus but which are interictal in their timing. A new multiaxial approach to the classification of psychoses in epilepsy can be found in the proposal by Krishnamoorthy.³³

It is suggested that patients with epilepsy and psychoses receive two separate diagnoses according to either ICD-10 or DSM-IV,⁷ but in addition, the relationships between onset of psychosis and seizure activity, antiepileptic therapy, and changes of EEG findings should be noted.

For pragmatic reasons, however, it remains convenient to group psychoses in epilepsy according to their temporal relationship to seizures.

The various syndromes are described in Table 1. The ictal psychoses are more likely to be linked to complex partial seizure status but have never been examined in any detail. In clinical practice they are not uncommon in seizures of temporal origin, but some of them are secondary to frontal lobe seizures. Simple focal status or aura continua may cause complex hallucinations, thought disorders, and affective symptoms. The continuous epileptic activity is restricted and may escape scalp EEG recordings. Insight usually is maintained, and true psychoses emerging from such a state have not been described. Nonconvulsive status epilepticus requires immediate treatment with intravenous antiepileptic drugs.