Borderline personality disorder: no demonstrated additional clinical benefit from amygdala neurofeedback

An add-on tested during intensive psychotherapy

The trial included 36 people who still had clinically significant symptoms halfway through their dialectical behavior therapy. Random allocation assigned 23 participants to the group also receiving amygdala neurofeedback using real-time functional MRI, and 13 to the group continuing psychotherapy alone.

Here, neurofeedback uses a measure of brain activity to allow participants to practice modulating it. The hypothesis was that learning to reduce amygdala activity would lead to further improvements in emotion regulation. This biological hypothesis still needed to translate into a benefit in daily life.

The primary outcome was emotional instability, assessed through repeated measurements in daily life. The researchers also collected questionnaire responses and brain imaging measurements before and after the intervention, then at 3- and 6-month follow-ups.

Brain changes, without a consistent clinical advantage

Participants managed to reduce amygdala activity during neurofeedback. Some imaging measurements also showed changes after the intervention, including temporary reductions in activity in this region during an emotion regulation task.

But daily-life measurements did not show a consistent advantage from adding neurofeedback to psychotherapy. Negative emotions, inner tension and emotional instability improved over time, without a clearly established additional advantage for this group.

The questionnaires did not reveal any additional benefit either. Borderline symptom scores measured using the BSL-23 increased after discharge in both groups. The brain findings therefore cannot be presented as evidence of clinical improvement.

A finding limited by sample size and loss to follow-up

Post-treatment assessments were completed by 15 participants in the neurofeedback group and 9 in the control group. At the 6-month follow-up, the numbers were 12 and 7, respectively. These losses to follow-up, combined with the small initial sample, reduce the precision of the conclusions.

This trial does not demonstrate that neurofeedback is ineffective in all settings. It does not establish an additional clinical benefit when added to this intensive inpatient psychotherapy. Nor does it allow conclusions about its use alone or in other populations.

What we can take away

Neurofeedback enabled participants to modulate amygdala activity, but this pilot trial does not justify promising additional symptom benefits when it is added to intensive dialectical behavior therapy.