The Electroencephalogram in the Management of Psychiatric Conditions

Psychiatric TimesPsychiatric Times Vol 30 No 5
Volume 30
Issue 5

The electroencephalogram (EEG) has a limited but definitive role in understanding and managing psychiatric conditions. When the presentation is unusual, a neurological workup that includes an EEG is essential.

Table: Indications for clinical EEGs in patients with psychiatric disorders

Table: Indications for clinical EEGs in patients with psychiatric disorders

Figure: Summary of the types of abnormalities an EEG can detect

Figure: Summary of the types of abnormalities an EEG can detect

CME credit for this article is now expired. It appears here for your reference.

The electroencephalogram (EEG) has a limited but definitive role in understanding and managing psychiatric conditions. When the presentation is unusual, a neurological workup that includes an EEG is essential. In conditions in which EEG abnormalities are demonstrably common, an EEG should be considered part of the essential workup. A simplified EEG can have substantial diagnostic usefulness, particularly in emergency department settings. The abbreviated EEG can be done in a matter of minutes using a 10- or 12-channel recording instrument.

The EEG as a diagnostic tool

The EEG is an extremely sensitive voltmeter. Typical signals range from approximately 30 to 80 μV, but can be as low as 10 μV in some tracings or as high as 150 or 200 μV in some high-voltage “spike” epileptic discharges. Electrical potentials measured between any 2 EEG electrodes fluctuate or oscillate rapidly, usually many times per second. These EEG signals are the result of summated field potentials generated by excitatory and inhibitory postsynaptic potentials in vertically oriented pyramidal cells of the cortex.

The simultaneous recording of brain waves from many scalp locations is important because it:

• Allows direct comparisons between homologous cortical regions

• Permits recording arrays to locate focal or regional abnormal features more clearly

• Increases the ability to detect various artifacts (ie, waveforms of non-brain origin) that can contaminate the recording

The EEG of an awake and relaxed healthy adult is usually dominated by frequencies between 8 and 13 cycles/second (8 to 13 Hz alpha activity). Once the person focuses his or her attention or becomes stressed, the frequency increases to the beta range (above 13 Hz). When an adult begins to get drowsy, EEG rhythms slow to the theta (4 to 8 Hz) range and finally to the delta (below 4 Hz) range with sleep. The appearance of excessive theta or any delta during wakefulness is a definite abnormality. While excessive beta during wakefulness has been linked with anxiety disorders, it is not considered an EEG abnormality (by today’s standards) and cannot be used to diagnose anxiety states.

The prevalence of EEG abnormalities in psychiatric patients is significantly elevated and ranges from 20% to 68% higher than in healthy controls.1-4 EEG findings in psychiatric populations include generalized or focal slowing of cortical activity and a variety of focal or generalized paroxysmal EEG discharges. Part of the reason that these patients have such a wide range of EEG abnormalities may be because the EEG signal is quite sensitive to many variables (eg, metabolic, vascular, endocrinological) that affect CNS function.

Although EEG abnormalities may not have direct primary psychiatric diagnostic specificity, they suggest the presence of other, organic/medical/physiological variables that could contribute to the psychiatric presentation. Moreover, EEG discharges without overt seizures (ie, isolated epileptic discharges) may have behavioral consequences such as emotional lability, irritability, or temper dyscontrol that cut across broad diagnostic labels.5,6


Indications for clinical EEGs in patients with psychiatric disorders

The EEG may reveal 1 of 2 classes of deviations: the slowing of normal rhythms and the appearance of abnormal paroxysmal (episodic) electrical activity suggestive of an epileptic process. Both kinds of abnormalities can be focal and suggestive of a localized pathological process, or generalized and more diffused, suggestive of a degenerative or metabolic process. If slowing is suspected, the patient must be awake while the EEG is recorded; if paroxysmal activity is suspected, the EEG needs to be done during both sleep and wake periods. The Table summarizes indications for which EEGs might prove useful for patients with psychiatric disorders.

The need for an EEG

There are few psychiatric conditions in which the prevalence of epileptic activity has been documented to be elevated, and which warrant consideration of an EEG on the basis of presenting symptoms. In general psychiatric practice, the leading indication for initiating some form of organic or neurological workup is an atypical presentation (Figure).


Summary of the types of abnormalities an EEG can detect

A major question is whether resistance to standard treatment interventions is an indication to obtain an EEG. For example, a number of case reports describe patients who received a diagnosis of borderline personality disorder (BPD) and who were subsequently found to have isolated epileptic discharges over one or both temporal regions.7-9 Snyder and Pitts8 showed that patients with BPD have a significantly higher rate of both definitive and less definitive EEG abnormalities than persons with dysthymic disorder. Abnormalities (mainly cortical slowing) were most frequently bilateral and of frontal, temporal, or frontotemporal distribution.

BPD patients were found to exhibit a much higher prevalence of symptoms commonly seen in complex partial seizures or episodic dyscontrol than patients with unipolar depression.8 Similarly, in patients with BPD, there was a much higher prevalence of paroxysmal EEG activity, particularly in posterior sharp waves (suggestive of an epileptic-like profile).

Archer and colleagues10 found that 6.3% of adolescents with BPD and 2.6% of patients with dysthymic disorder had bilateral spike and wave discharges. None of the subjects in a comparison group of other personality disorders had similar discharges. Thus, an EEG may enable identification of a brain abnormality that is responsible for treatment resistance.

Exploring paroxysmal activity

Autism. Epilepsy is common in children with autistic spectrum disorders (ASD). A significant proportion of children with ASD have abnormal EEGs, even those who have never had a seizure. EEG abnormalities can range from mild slow wave abnormalities to frank epileptiform discharges (epileptiform discharges may only be detected during sleep and at times may require prolonged monitoring).

Tuchman and Rapin11 examined the EEGs (including sleep) of children with ASD. Epileptiform discharges were seen in 59% of epileptic children but in only 8% of non-epileptic children. The authors concluded that EEGs are not indicated in the absence of seizures. When historical clinical deterioration is evident, the rate of epileptiform EEG abnormalities is significantly higher (14%) than in children without obvious deterioration (6%). The average child who participated in this study was first evaluated neurologically 4 years after the onset of deterioration. This point highlights the need for studies conducted at a much earlier stage of the illness.

Whether treatments to eliminate EEG spikes have a therapeutic effect on the behavioral abnormalities in children with pervasive developmental disorders and autism remains an open and important question.12 Nevertheless, it may be most crucial to attempt treatment at the earliest indication of the illness rather than after many years of damaging spikes before the onset of the first seizure.

Panic disorder. Young and colleagues13 described 5 patients with brief simple partial seizures that mimicked panic attacks. They concluded that the most common psychiatric disorder that must be differentiated from temporal lobe epilepsy is panic disorder. In their sample, seizures were briefer and more stereotyped than panic attacks. In addition, aphasia and dysmnesia accompanied seizure activity in some patients. This differentiation could be diagnostically challenging because patients with documented complex partial seizures of temporal lobe origin may have concomitant nonictal episodic emotional symptoms, including phobia, true panic attacks, and anxiety.14,15

Toni and colleagues16 compared the symptoms of panic disorder–agoraphobia with symptoms of complex partial seizures. Because of the similarities, they concluded that the two may have a common neurophysiological substrate. Whether the source and frequency of the isolated epileptic discharges may influence the responses to antiepileptic drugs and make treatment response difficult to predict also remain open and important questions.

Repeated violence and aggression. The prevalence of abnormal EEGs ranges widely, from 6.6% in patients with rage attacks and episodic violent behavior to 53% in patients with antisocial personality disorder.17,18 Monroe19 showed that anticonvulsants can block EEG epileptiform discharges and can lead to dramatic clinical improvement in individuals who exhibit aggressive behavior. Similarly, Neppe20 provided evidence on the clinical usefulness of adding carbamazepine to the treatment of violent schizophrenia patients with repeated aggressive episodes who also exhibit temporal lobe abnormalities on the EEG but who have no history of a seizure disorder.

In an earlier study, a reduction of aggressive episodes was seen when carbamazepine was added to the neuroleptic regimen of women with schizophrenia who also had EEG abnormalities.21 Other studies suggest that anticonvulsant therapy may have a beneficial effect on aggressive tendencies irrespective of the presence or absence of EEG abnormalities.22

Until large, multicenter, double-blind efficacy studies are done, patients should be given the benefit of the doubt, and an anticonvulsant should be used when an EEG demonstrates isolated epileptic discharges.

Recalcitrant ADHD. The majority of patients with ADHD respond favorably to psychostimulants. Therefore, an EEG may be useful only when there is no response to treatment. A number of studies found variable rates of EEG abnormalities in children with ADHD. Phillips and colleagues23 reported on EEG screening in children hospitalized over an 18-month period for behavioral problems; 9% had abnormalities with background electrocortical slowing or paroxysmal discharges. The authors concluded that EEG screening may be of limited value in children with behavioral problems who had no clinical evidence of a neurological disorder.

Frank24 reported significantly higher rates of abnormalities. In a sample of children with ADHD, 31% had abnormal routine EEGs. Of the children with abnormal EEGs, 84% had spikes or spike-wave discharges; the others had excessive background cortical slowing.

Millichap25 reported an incidence of 7% of definite abnormalities suggestive of seizure disorder and an additional 19% moderately abnormal dysrhythmias not diagnostic of seizure disorder in children with ADHD. On the basis of their findings, they suggested these 6 indications for an ECG in a child presenting with ADHD:

• Personal or family history of seizures

• Inattentive episodes characterized by excessive daydreaming and/or periodic confused states

• Comorbid episodic, unprovoked temper or rage attacks

• Frequently recurring headaches

• A history of head trauma, encephalitis, or meningitis preceding the onset of ADHD

• Abnormalities on neurological examination

Atypical bipolar/rapid cycling. Patients with rapid cycling bipolar disorder may exhibit isolated epileptic discharges on an EEG. Levy and associates26 compared the EEGs of patients with rapid cycling with those of non-rapid cycling patients. The EEGs of 3 of the 5 rapid cycling patients had isolated epileptic discharges, compared with none of the controls.

Subictal mood disorders. The characteristics of what has been termed “subictal mood disorders” include brief euphorias, mixed bipolar episodes, brief severe depressive dips with impulsive suicide attempts, compulsive symptoms, irritability and hostile outbursts, and marked premenstrual worsening.27 Patients with subictal mood disorders may also have paradoxical reactions to mood-active drugs (lithium and antidepressants), with a better response to anticonvulsants.

Note that the EEG does not contribute to the diagnosis of schizophrenia or bipolar disorders except that it helps the clinician rule out a neurological cause for the symptoms when a patient presents with an atypical picture (eg, unusual age of onset).

Cortical slowing in psychiatric conditions

Dementia versus pseudodementia. Because patients with advanced dementia rarely have a normal EEG, a normal EEG can play an important role in diagnosing cases of pseudodementia (dementia secondary to depression or psychosis). When dementia and depression coexist, it becomes important to have some idea about the relative contribution of each disorder to the overall clinical presentation. Brenner and associates28 compared the EEGs of patients with depression, dementia, pseudodementia, and dementia plus depression with the EEGs of age-matched healthy controls. Their findings indicate that the more abnormal the EEG is, the less likely it is that a patient will respond favorably to an antidepressant. This is particularly important because aging increases the likelihood of experiencing adverse effects from antidepressants.


A 56-year-old woman with Alzheimer disease who has been in a nursing home for 4 years stopped eating and was admitted to the neuropsychiatry ward of a psychiatric hospital. In reviewing her history, a complete workup for possible reversible causes of dementia had never been done. A comprehensive evaluation was started that included blood work, a CT scan, and an EEG. There was nothing significantly abnormal on any test; the EEG was normal.

A normal EEG was incompatible with advanced stage Alzheimer disease. A trial of antidepressant therapy was therefore initiated. Nothing happened for 2 weeks. During the third week, the patient started speaking. Her speech was completely disorganized with severe psychotic content. Haloperidol was added, and the patient was ambulatory and able to leave the hospital within 2 weeks.

Possible focal problem. Patients with a history of head injury or an early brain insult during infancy or childhood may have developed behavior that helps them compensate for the deficit resulting from the injury. Such behavior could be deemed aberrant. An EEG may detect focal slowing (most commonly frontal or temporal), which indicates a disordered brain region. Knowledge of the focal deficit followed by detailed neuropsychological evaluation could shed much light on the particular case and help guide treatment and rehabilitation efforts.

Delirium. The differential diagnosis of acutely disorganized and disoriented patients often includes delirium. In acutely agitated delirious patients, the EEG often helps determine the cause of altered consciousness: a diffuse encephalopathic process, a focal brain lesion, or continuous epileptic activity without motor manifestations (ambulatory nonconvulsive status epilepticus). Most often, patients with delirium have a toxic-metabolic encephalopathy.

In general, as encephalopathy progresses, there is diffuse slowing of the background rhythms, including alpha (8 to 13 Hz) to theta (4 to 7.5 Hz) activity. Delta (less than 3.5 Hz) activity usually does not become prominent until the patient approaches nonresponsiveness. The major exception to the above rule is seen during withdrawal from alcohol and during delirium tremens (DTs). Excessive fast activity (rather than slowing) dominates the EEG (beta activity: 13 to 30 Hz) in patients with alcohol withdrawal delirium.29


A 44-year-old man with a long history of alcohol dependence presented to the emergency department with a blood alcohol level of 0.42%. He was admitted for detoxification. As his alcohol level started coming down, he became progressively more confused. The benzodiazepine dosage was increased. He progressively became agitated and was placed in 4-point restraints. A neurology consultation was called. The consultant diagnosed DTs and suggested increasing the benzodiazepine dose. Symptoms continued to escalate despite increasing doses of the benzodiazepine. Later that day, the patient had a medical consultation. It was suggested that the patient was in iatrogenic delirium and that all medications should be stopped. An abbreviated (8 leads) EEG was obtained and revealed a moderate degree of diffuse slowing, consistent with encephalopathy and not the fast low-voltage picture expected with DTs. Medications were rapidly tapered and the patient recovered.


Electroencephalography is burdened by several constraints that place limitations on the information provided. These constraints have been extensively discussed elsewhere30 and include:

• The scalp: recorded activity reflects mainly cortical activity with minimal to no reflection of subcortical or deeper neural processes

• Skin, skull, dura, and brain tissue impose varying degrees of impedance between the source of the electrical discharge and the scalp electrodes

• Time sampling (many important EEG abnormalities are paroxysmal, or “episodic,” and may require prolonged recording time)

• Non-specificity of results (the brain’s electrical activity can only respond to stimuli or insult by becoming faster or slower in frequency or higher or lower in voltage, or a combination of these two responses)

A number of problems make detection of isolated epileptic discharges in nonepileptic psychiatric patients difficult and likely an expensive process. Standard neurological training in EEG interpretation strongly emphasizes under-interpretation to avoid stigmatizing the patient with the diagnosis of seizure disorder. In psychiatric EEGs, abnormalities, when present, tend to be infrequent in the record and require full vigilance on the part of the interpreter. The chance of a false negative can be lessened by repeating the EEG or by doing the EEG testing for extended periods (eg, 1 to 3 days), which increases the chances of recording the EEG signal when the patient is actually in a panic attack or a rage episode.

Additional problems may arise because the patient does not understand the special needs associated with obtaining an accurate recording-eg, staying still, activating procedures such as hyperventilating or staying up all night before the procedure (sleep deprivation improves the detection of isolated epileptic discharges), or relaxing and falling asleep during the test. Furthermore, a patient may behave inappropriately during the test. These factors necessitate that the laboratory be placed in a psychiatrically oriented space with psychiatrically trained technicians who can help the patient (particularly children) relax and cooperate.

The contents presented herein are largely derived from the textbook Standard Electroencephalography in Clinical Psychiatry: A Practical Handbook.31

Note: This article was originally published as a CME in the May 2013 issue of Psychiatric Times. Portions of it may have since been updated. 


Nash Boutros, MD, has no disclosures to report.

Natalia Jaworska, PhD, MSc (peer/content reviewer), has no disclosures to report.


1. Tucker GJ, Detre T, Harrow M, Glaser GH. Behavior and symptoms of psychiatric patients and the electroencephalogram. Arch Gen Psychiatry. 1965;12:278-286.

2. Struve FA. The necessity and value of securing routine electroencephalograms in psychiatric patients: a preliminary report on the issue of referrals. Clin Electroencephalogr. 1976;7:115-130.

3. Struve FA. EEG findings detected in routine screening of psychiatric patients: relationship to prior expectation of positive results. Clin Electroencephalogr. 1977;8:47-50.

4. Fenton GW, Standage K. Clinical electroencephalography in a psychiatric service. Rev Can Psychiatrie. 1993;38:333-338.

5. Boutros NN. Epileptiform discharges in psychiatric patients: a controversy in need of resurrection. Clin EEG Neurosci. 2009;40:239-244.

6. Shelley BP, Trimble MR, Boutros NN. Electroencephalographic cerebral dysrhythmic abnormalities in the trinity of nonepileptic general population, neuropsychiatric, and neurobehavioral disorders. J Neuropsychiatry Clin Neurosci. 2008;20:7-22.

7. Cowdry RW, Pickar D, Davies R. Symptoms and EEG findings in the borderline syndrome. Int J Psychiatry Med. 1985-1986;15:201-211.

8. Snyder S, Pitts WM Jr. Electroencephalography of DSM-III borderline personality disorder. Acta Psychiatr Scand. 1984;69:129-134.

9. Messner E. Covert complex partial seizures in psychotherapy. Am J Orthopsychiatry. 1986;56:323-326.

10. Archer RP, Struve FA, Ball JD, Gordon RA. EEG in borderline personality disorder. Biol Psychiatry. 1988;24:731-732.

11. Tuchman RF, Rapin I. Regression in pervasive developmental disorders: seizures and epileptiform electroencephalogram correlates. Pediatrics. 1997;99:560-566.

12. Tharp BR. Epileptic encephalopathies and their relationship to developmental disorders: do spikes cause autism? Ment Retard Dev Disabil Res Rev. 2004;10:132-134.

13. Young GB, Chandarana PC, Blume WT, et al. Mesial temporal lobe seizures presenting as anxiety disorders. J Neuropsychiatry Clin Neurosci. 1995;7:352-357.

14. Signer SF. Seizure disorder or panic disorder? Am J Psychiatry. 1988;145:275-276.

15. Spitz MC. Panic disorder in seizure patients: a diagnostic pitfall. Epilepsia. 1991;32:33-38.

16. Toni C, Cassano GB, Perugi G, et al. Psychosensorial and related phenomena in panic disorder and in temporal lobe epilepsy. Compr Psychiatry. 1996;37:125-133.

17. Blake PY, Pincus JH, Buckner C. Neurologic abnormalities in murderers. Neurology. 1995;45:1641-1647.

18. Stafford-Clark D, Taylor FH. Clinical and electro-encephalographic studies of prisoners charged with murder. J Neurol Neurosurg Psychiatry. 1949;12:325-330.

19. Monroe RR. Anticonvulsants in the treatment of aggression. J Nerv Ment Dis. 1975;160:119-126.

20. Neppe VM. Carbamazepine as adjunctive treatment in nonepileptic chronic inpatients with EEG temporal lobe abnormalities. J Clin Psychiatry. 1983;44:326-331.

21. Hakola HP, Laulumaa VA. Carbamazepine in the treatment of violent schizophrenics. Lancet. 1982;1:1358.

22. Yassa R, Dupont B. Carbamazepine in the treatment of aggressive behavior in schizophrenic patients: a case report. Can J Psychiatry. 1983;28:566-568.

23. Phillips BB, Drake ME Jr, Hietter SA, et al. Electroencephalography in childhood conduct and behavior disorders. Clin Electroencephalogr. 1993;24:25-30.

24. Frank Y. Visual event-related potentials after methylphenidate and sodium valproate in children with attention deficit hyperactivity disorder. Clin Electroencephalogr. 1993;24:19-24.

25. Millichap JG. Attention deficit-hyperactivity disorder and the electroencephalogram. Epilepsy Behav. 2000;1:453-454.

26. Levy AB, Drake ME, Shy KE. EEG evidence of epileptiform paroxysms in rapid cycling bipolar patients. J Clin Psychiatry. 1988;49:232-234.

27. Himmelhoch JM. Cerebral dvsrhythmia, substance abuse, and the nature of secondary affective illness. Psychiatr Ann. 1987;17:710-727.

28. Brenner RP, Reynolds CF 3rd, Ulrich RF. EEG findings in depressive pseudodementia and dementia with secondary depression. Electroencephalogr Clin Neurophysiol. 1989;72:298-304.

29. Reilly EL, Glass G, Faillace LA. EEGs in an alcohol detoxification and treatment center. Clin Electroencephalogr. 1979;10:69-71.

30. Struve FA. Clinical electroencephalography as an assessment method in psychiatric practice. In: Hall RCW, Beresford TP, eds. Handbook of Psychiatric Diagnostic Procedures. Woodstock, GA: Spectrum Publications; 1985.

31. Boutros NN, Galderisi S, Pogarell O, Riggio S, eds. Standard Electroencephalography in Clinical Psychiatry: A Practical Handbook. Chichester, UK: John Wiley & Sons; 2011.

Related Videos
© 2024 MJH Life Sciences

All rights reserved.