Home » Psych Central Professional » Using EEG to Predict Medication Response

Using EEG to Predict Medication Response

Using EEG to Predict Medication ResponseMany psychiatrists are already hearing from patients about “this new technique” of using EEG to predict whether a given antidepressant will work. How good is this technology? Is it even remotely ready for prime time?

First, let’s review a little background on EEG. First developed in the 1920s, EEG entails applying electrodes to the scalp’s surface in order to visually examine brain waves. Brain waves are labeled according to their frequency. The faster frequency bands are associated with wakefulness, the lower frequency bands with relaxation or sleep. A good way to memorize the confusing names is to use the mnemonic BAT-D (progressing from most alert to most asleep):

    Beta (16 hz): Alert, intellectual activity.
    Alpha (8-11 hz): Relaxed, daydreaming.
    Theta (4-8 hz): Deep relaxation, meditation.
    Delta (1-3 hz): Deep sleep.

The two most common clinical uses of EEG are in diagnosing the presence of epileptic activity, and diagnosing sleep disorders.

EEGs are typically read visually by a neurologist who looks for irregularities in the different bands. Quantitative EEG (QEEG) is a newer technique that relies on computer software to read the tracings, theoretically replacing the subjective human element with an objective printout that can be interpreted by physicians will less specialized training.

Using QEEG to predict response to antidepressants.

While there are many controversial claims regarding the utility of QEEG for psychiatric diagnosis (see, for example, Coburn KL et al., J Neuropsychiatry Clin Neurosci 2006;18(4):460-500), a series of reputable studies (the lead investigator is Andrew Leuchter, M.D., at UCLA) have examined QEEG for predicting response to antidepressants. The main company involved in such trials is Aspect Medical Systems (, and they recently announced results of their large study called BRITEMD (Biomarkers for Rapid Identification of Treatment Effectiveness in Major Depression). The full set of results has not yet been presented, but a summary of the findings are available on Aspect’s website (http://library.corporateir. net/library/73/737/73770/items/273305/ BRITEStudyResults1207.pdf.)

In this study, 375 patients with major depression were given a baseline EEG and were then started on Lexapro 10 mg QD. After one week, they were assessed using Aspect’s proprietary QEEG system called ATR indicator (antidepressant treatment response indicator). This device is essentially a laptop computer attached to 5 electrodes, which are placed on each ear lobe and on the forehead. The computer digitizes the EEG signals from the Theta and Alpha frequencies (or bands) coming from the prefrontal lobe, and provides a numerical measure of the amount of energy in these bands. This measure is made easy for clinicians; it varies from 0, indicating a low likelihood of eventual response, to 100, indicating the highest likelihood of response.

After the 1 week EEG, patients were randomly assigned to one of three groups: continuation of Lexapro, switch to Wellbutrin, or addition of Wellbutrin to Lexapro. The goal was to see if an ATR prediction of response would correlate with an actual response to Lexapro at 7 weeks. While we’ll have to wait for the May 2008 APA meeting for the full results, the Aspect website reports that the overall response rate among patients randomized to Lexapro was 51%; interestingly, the subset of patients whose ATR predicted a response had a 67% response rate, while patients with a low ATR score (predicting a nonresponse) had only a 28% response rate to Lexapro. Furthermore, looking specifically at patients who were randomized to Wellbutrin, those whose ATR predicted a poor response to Lexapro had a pretty robust 53% response to Wellbutrin. If these numbers hold up, Aspect’s QEEG technology is potentially useful in helping to inform our antidepressant decisions early in treatment.

There are a few important caveats, however. The company did not include a “best clinical judgment” arm in their study. It’s possible that clinical judgment alone would do as well, if not better, than QEEG at predicting response. For example, if a patient complains of worsening depression or of intolerable side effects after one week on Lexapro, many psychiatrists would switch to another agent. Depending on the accuracy of this judgment, these “clinician-predicted” nonresponders might have a high rate of response to a different medication, even higher than ATR’s 53% response to Wellbutrin.

In addition, there are only a few clinical situations in which this test would be useful. For example, it would not be useful for a patient who is feeling a little better after the first week (whether due to active medication or placebo). If the ATR reading predicted a response in such a patient, you’d obviously continue treating; if it predicted nonresponse, you’d probably also continue, because the overall estimated accuracy of the device is only about 75%. The machine would be primarily useful for those patients who are not responding at one week, and who you believe will eventually respond. In this case, an ATR prediction of response would supply you with objective “ammo” to convince your patient to stay the course.

At any rate, the discussion is still theoretical. The device does not have FDA approval, and the company has yet to begin a pivotal trial to win approval. It could take a few years before we see the results of this definitive trial.

TCPR VERDICT: QEEG: Promising early results, but replication is needed

Using EEG to Predict Medication Response

This article originally appeared in:

The Carlat Psychiatry Report
Click on the image to learn more or subscribe today!

This article was published in print 1/2008 in Volume:Issue 6:1.

The Carlat Psychiatry Report


APA Reference
Psychiatry Report, T. (2013). Using EEG to Predict Medication Response. Psych Central. Retrieved on December 14, 2018, from


Scientifically Reviewed
Last updated: 31 Aug 2013
Last reviewed: By John M. Grohol, Psy.D. on 31 Aug 2013
Published on All rights reserved.