The Search for a Blood Test for Depression

Why don't we have biological tests for mental illness?

May 25 2015, 11:00am

Image: Hartwig HKD/Flickr

If you think you have diabetes, you go to a doctor and get a blood test. Heart disease? An electrocardiogram can make sense of symptoms. And with cancer, a biopsy can cement a diagnosis.

But when it comes to behavioral health, your diagnosis is based on, well, your behavior.

While mental illness is increasingly treated biologically and chemically—think Prozac, Adderall, Loxapine—illnesses are typically diagnosed by specialists ticking symptoms off checklists provided in the Diagnostic and Statistical Manual of Mental Disorders (DSM) or the International Classification of Diseases. In most cases, these are symptoms that are self-reported.

"This test brings mental health diagnosis into the 21st century."

But in the last year, there have been two promising studies examining efforts to create a biological or physical test to diagnose mental illness.

Blood Will Tell

At Northwestern University's School of Medicine, Eva Redei, a neuroscientist, led research that was published in September's Translational Psychiatry.

With her fellow researchers, she developed a blood test to diagnose depression in adults. Depression, one of the most common mental illnesses is also one of the most difficult to diagnose and to treat.

"This test brings mental health diagnosis into the 21st century," said Redei, who noted that diagnoses now are done the same way they were done a century ago.

The study, which Redei co-authored, found that people with depression have markers in their blood—levels of RNA molecules, which are the messengers that interpret DNA genetic codes and carry out instructions—that show the disorder. That same blood work could even identify people who are prone to depression before symptoms actually show up, and those same biomarkers can also indicate if a treatment is working.

This new finding addresses what Redei calls one of the biggest hindrances in treating and studying mental illness: when it comes to looking for variations in the brain or biomarkers, scientists often don't know what they are looking for.

"We know very little about the etiology [cause] of these illnesses and that doesn't help," said Redei.

While she has spent decades working on establishing a biological test for depression, other researchers are working on tests using biomarkers for schizophrenia, bipolar disorder, and Alzheimer's.

Pics Or It Didn't Happen

One of the handicaps in using biomarkers is the blood-brain divide—not all the genes present in the blood are present in the brain, leaving room for error.

Keeping that divide in mind, Marcel Just, a professor of psychology at Carnegie Mellon University is looking at an alternate way to diagnosis mental illness by looking at the brain.

Just created a brain-reading technique that uses neural representations of social thoughts to predict autism diagnoses with 97 percent accuracy. Normally, autism is diagnosed by behavior. Just used FMRi—a neuroimaging procedure that measures brain activity and takes pictures of what's happening in your brain—and his machine technique to monitor brain activation patterns that allowed him to figure out a person's thoughts or emotions.

Just and his fellow researchers used this to identify autism in those who had already been diagnosed by detecting changes triggered by words in the brains of those with the disorder. His research was published in the journal Plos-One in December.

"If this approach were wildly successful, then you could have someone be in the scanner and have them think of 500 different concepts and you can say 'Oh they are all in the normal range, or this subset is altered and that's associated with such and such brain alteration, and in conventional diagnoses that would mean you're a paranoid schizophrenic,'" said Just, whose research was funded by the National Institute of Mental Health.

Just says this work could fuel a significant change in the way autism is diagnosed. The DSM (often called the DSM-5, a reference to the current version), which is now the standard, can be improved upon.

"It's the best we have. And the people who did it, I'm sure, are extremely well intentioned and did the best possible job that could be done using this approach," he said. "But this isn't the way to understand autism. Maybe it's not intended for that."

He's now working on a pilot study using his technique on people with suicide ideation.

Next Steps

Just's autism study was done on just 17 people and Redei's was done on a few dozen. Redei's work is also ongoing, and neither test will be at your doctor's office anytime soon. There's still a lot of testing and independent verification that has to be done before it can go before The Food and Drug Administration for approval, and even then it wouldn't become the standard overnight.

Both techniques appear promising, however, and feed into the RDoc or Research Domain Criteria movement being led by The National Institute of Mental Health to change the way psychiatric illnesses are viewed and understood.

If these biological tests were to become commonplace, doctors would already have the capabilities to carry them out.

As Redei says, when you go to a doctor, you may already be getting tested for an array of ailments. If her technique proves effective as a physical test for behavioral conditions, "they'd draw blood into one more tube—and that's it."

And diagnosing mental illness, a historically gray area, may become more of an exact science.