Originally found on

In a remarkable study published in the American Journal of Psychiatry, Kymberly Young and colleagues report that depressive symptoms can be substantially diminished by training a region of the brain known as the amygdala to respond more strongly to positive memories. The amygdala is heavily involved in regulating emotions. Earlier research has demonstrated that activity in the amygdala linked to the retrieval of positive memories is blunted in depressed persons as compared to non-depressed individuals.

Over the last decade, functional neuroimaging techniques have advanced to the point that it is possible to measure increases or decreases in brain activity in real time, even in regions as small as the amygdala. In earlier work, Young and associates used a technique called real-time functional MRI neurofeedback to show that individuals can train their amygdalae to increase hemodynamic responses (a surrogate for “activity”) if provided immediate feedback of amygdala activity in real time during a positive memory retrieval task.

In this recent study, 33 adults with moderately severe major depressive disorder completed the research protocol. None was being treated with an antidepressant. The study consisted of four visits: During the first visit, participants completed several clinical and self-report assessments of depression and anxiety, as well as an autobiographical memory test. This set of assessments was repeated at each subsequent visit. Five to seven days after the first visit, study participants completed the first neurofeedback training session, and one week later, they completed a second session. During the fourth visit 5 to 7 days later, participants completed the assessments one last time.

At the start of the study, participants were divided into two groups that had almost identical scores on the various depression scales. One group (18 individuals) received neurofeedback from the amygdala; the other (15 individuals) received neurofeedback from a region of parietal cortex that is not known to be involved in emotional regulation.

One week after the first neurofeedback session, participants in the amygdala training group exhibited decreases in depression scores compared to baseline scores. By the end of the study, decreases in depressive symptoms were even greater in this group and were associated with increased amygdala activity during positive memory retrieval. Twelve of 18 participants experienced a greater than 50 percent decrease in depression scores (defined as a “response”), and 6 met criteria for remission (defined as reporting very few depressive symptoms). Only 2 depressed participants who received real-time feedback from the parietal cortex demonstrated 50 percent decreases in depressive scores, and only 1 remitted.

Many important questions remain: Can these results be replicated? Will more sessions lead to a higher percentage of people who respond and who achieve remission? How long do the antidepressant effects last? Can occasional booster training sessions prevent relapse? Can people practice the method in the absence of imaging feedback to sustain improvements?

Currently, this technique requires an individual to be trained for several hours utilizing expensive functional imaging equipment. However, if additional research indicates long-term effectiveness of this approach, it is likely that more affordable methods will be developed.

In recent posts, we have noted several new pharmacologic approaches that might provide rapid treatment of severe depressive symptoms. Here, we describe work showing that a person can utilize biofeedback to train the amygdala and decrease depressive symptoms in the absence of pharmacological treatment.

These preliminary findings again support the idea that basic science research examining specific brain regions may provide a way to develop novel clinical treatments.

Written by Eugene Rubin, MD, Ph.D.