Jury Out on EEG to Predict Depression Treatment Response

Megan Brooks

January 10, 2019

At present, there is not enough evidence to support the use of quantitative electroencephalographic (QEEG) biomarkers to predict how a patient with major depressive illness will respond to treatment, conclude the authors of a meta-analytic review.

"QEEG is already being promoted and commercialized as a means for selecting [depression] treatments and predicting responses," Alik Widge, MD, PhD, of the Department of Psychiatry and Behavioral Sciences, University of Minnesota, Minneapolis, told Medscape Medical News.

But this is "very premature," said Widge. "Our review shows that the science is evolving, but that the studies haven't been done yet to prove that QEEG methods are reliable enough for use in routine clinical practice."

"The main message for the general psychiatrist is that QEEG methods are not ready for prime time. They have not shown the kind of reliability we need in a diagnostic or prognostic test, and it would not be a wise use of healthcare dollars to order such testing outside of a formal research protocol," Widge added.

The study was published online January 1 in The American Journal of Psychiatry.

Publication Bias

The meta-analysis included 76 articles covering 81 distinct QEEG biomarkers.  For all biomarkers taken together, the meta-analysis suggested predictive power "above chance," hinting that QEEG may have predictive power for treatment response in depressive illness, the researchers say.

The meta-analytic estimate of sensitivity was 0.72 (95% confidence interval [CI], 0.67 - 0.76), specificity was 0.68 (95% CI, 0.63 - 0.73), and logarithm of the diagnostic odds ratio was 1.89 (95% CI, 1.56 - 2.21). These correspond to an area under the curve of 0.76 (95% CI, 0.71 - 0.80).

However, funnel-plot analysis suggested "substantial publication bias," with QEEG's apparent predictive power driven by small studies with strong positive results, with a lack of publication of small studies reporting null effects. "Most studies did not use ideally rigorous statistical practices," the authors note.

They conclude that QEEG "may become clinically useful, but only with further and more rigorous study." For now, "it is not ready for routine clinical use." 

The results of this meta-analysis, they add, do not mean QEEG research should be stopped or slowed, as many popular QEEG markers have meaningful biological rationales. "If future studies can be conducted with an emphasis on rigorous methods and reporting, and with specific attempts to replicate prior results, QEEG still has much potential," the authors conclude.

Currently Not a Useful Biomarker

Commenting on the study in a podcast, Daniel Pine, MD, deputy editor of The American Journal of Psychiatry, said, "Both clinicians and researchers are searching very hard to find biomarkers that are useful for predicting treatment response. QEEG has a long history of being used in this way." This new meta-analysis "synthesizes all the data that we have on the relationship between QEEG profiles and response to antidepressant medication."

"The bottom line to come out of the article," said Pine, "is to suggest that at the current time, QEEG is not a clinically useful biomarker to predict treatment outcome, and clinicians should attend to this before they think about recommending [it] for their patients who are initiating antidepressant treatment."

The study was supported in part by the Brain and Behavior Research Foundation, the Harvard Brain Science Initiative, and the National Institutes of Health. Widge has pending patent applications related to the use of electrographic markers to characterize patients and select neuromodulation therapies, and has received device donations and consulting income from Medtronic. A complete list of author disclosures is available with the original article.

Am J Psychiatry. Published online January 1, 2019. Full text

For more Medscape Psychiatry news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.