Who Should Take Antidepressants?

A brain scan could help objectively identify who will benefit, and who won't.

Lexapro (Wikimedia)

Amid calls for more science in psychiatry, part of the $100 million the U.S. government is spending this year on the BRAIN Initiative will go toward "mapping the human brain" in the interest of more concrete diagnoses. Basing treatments on harder data could cut down on variations from doctor to doctor and get people healthier faster. For example, a recent study at Johns Hopkins found that more than 60 percent of adults who were diagnosed by their doctor as having depression actually did not meet the official diagnostic criteria for the disorder upon re-evaluation by Hopkins psychiatrists. Some of them may have been prescribed antidepressant medications when their real problem was something else entirely.

Even when diagnosis is accurate, only some depressed people respond well to antidepressants. For others, they don't really help. Dr. Thomas Insel, director of the National Institute of Mental Health, explained to me, "We have reasonably good treatments for depression; both medications and psychotherapies. But we don't know who will respond best to medication and who will respond best to psychotherapy."

We do know about certain interesting factors that predict responses. For example, depressed people with a personal history of child abuse seem to respond better to therapy than to medications. But an empiric test to decide who should receive what type of treatment, instead of relying on possible correlations and likelihoods from the patient's history, could save time and money in effectively getting people what best works for them.

As Insel told me, "We increasingly think of depression as a brain disorder and thus, turning to the brain to predict treatment response seems like a smart thing to do."

Neurologist Helen Mayberg and her team at Emory University published research this week in the Journal of the American Medical Association Psychiatry that used PET imaging of the brain, which lights up to indicate levels of metabolism, in patients with major depressive disorder. Mayberg's group specifically looked at a part of the brain called the insula. They found that patients who had high levels of metabolism in the insula responded well to the commonly prescribed antidepressant escitalopram (Lexapro). People with low levels of metabolism in the insula didn't do well on Lexapro, and were better off with cognitive behavioral therapy.

Screen Shot 2013-06-13 at 2.00.45 PM.png

(McGrath et al., JAMA Psychiatry)

Since SSRI antidepressants and psychotherapy take time to work, identifying who-needs-what as quickly as possible could save people months of experimenting with ineffective therapies.

This particular study was small, but the concept is promising. In practical terms, as Insel put it, "It's hard to imagine that this form of brain imaging will be of immediate use -- each year 6.7 percent of adults meet criteria for major depression in this country, about 2 percent of adults are disabled by this common disorder (which has the highest rate of disability for all medical disorders for Americans between 15 and 49). Only about half receive any treatment and only half of those receive even minimally adequate care. Against that public health challenge, a high-tech imaging study is probably not going to be widely adopted even if it is useful. That said, this is an exciting proof of concept."

And the concept is getting good attention. Larger studies at multiple medical centers, now well federally-funded, are underway. If they pan out, Insel says, it "could presage a new era when treating depression depends less on trial and error, less on whether you see a psychiatrist or a psychologist, less on your insurance coverage, and more on science. Wouldn't that be enlightened?"