Brain imaging can predict effectiveness of CBT

In the 1980's there was some attention interest in being able to figure out if a given case of depression was "psychological" or "biological" in origin.  The idea was, that if you could show that a person was depressed for psychological reasons, then you also would know that the patient should receive treatment with psychotherapy.  Likewise, a case of depression with a biological cause should be treated with medication.

Nothing ever came of that.  In fact, what happened was that practitioners came to think that there is no meaningful difference.   Even if we find some replicable way to distinguish between depression with a biological cause and depression with a psychological cause, there is no particular reason to think that  knowing the cause would enable us to predict which treatment would be more likely to be effective.

Still, there are some interesting ideas there to explore, even if the exploration cannot necessarily be expected to lead to improved treatment matching.  

Now we see a study that offers just such an exploration.  Some folks over at the University of Pittsburgh School of Medicine have done fMRI studies of depressed people, compared to control subjects, then looked to see if there were any findings that correlated with later response to cognitive-behavioral therapy.  They report that there were two findings that did tend to correlate with responsiveness to CBT.  The treatment responders were more likely to have decreased activity in the subgenual cingulate cortex.  They also were found to have "increased activity after reading negative words in a brain region called the amygdala."

This would be more exciting if they had a larger sample size.  There were only 14 depressed patients in the study, and 21 controls.  So at this point, it is fun to think about what the study means, but we certainly can't go around changing our clinical practice -- or our belief systems -- based upon the results.