Skip to content

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.

PLOS BLOGS The Official PLOS Blog

How the depressed brain changes after psychotherapy, by Lauren Sakowski

By Lauren Sakowski

The struggle with depression involves more than long-term bouts of sadness, helplessness, and hopelessness.  Just ask anyone that has lived through it.  Other emotionally-based symptoms such as dysphoric affect and anxiety have been reported by individuals with depression, and studied by neuroscientists.  Positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) have been used to examine regional activation in the brain, implicating the prefrontal cortex and subgenual cingulate cortex as the most commonly found areas of dysregulation in mood disorders.  The ability to measure activation levels in various regions of the brain makes PET and fMRI suitable to study outcome measures following treatments for depression such as psychotherapy or medication.

In their 2009 Frontiers in Human Neuroscience paper, Emma J. Thomas and Rebecca Elliott reviewed several studies of emotional and non-emotional cognitive impairments in subjects with depression and illustrated different findings regarding brain activity among different research groups.  The amygdala has been shown to play a role in tasks involving facial emotion processing and exposure to emotional words and pictures, demonstrating greater responses to emotional stimuli in depressed subjects.  Other areas of interest based on the outcome of these task-based studies are the anterior cingulate and prefrontal cortices.  Although the neuronal basis of mood disorders has been explored to some extent, the correlation between these abnormalities and clinical symptoms remains unknown.

More recently, Buchheim et al., 2012 utilized fMRI to examine levels of activation in regions of the brain before and after a 15-month course of psychotherapy, free of medication.  This study was the first of its kind to combine long-term treatment for recurrent major depression with psychodynamic therapy.  Patients with a DSM-IV diagnosis of depression were recruited from mental health facilities and age-, sex-, and education- matched controls were recruited from the community.  Subjects were exposed to stimuli from the Adult Attachment Projective Picture System (AAP), as shown in Figure 1, and asked to describe the scene in each of seven pictures in the context of the accompanying sentences.  The examiner alternated personally relevant and neutral cue sentences.  Participants’ brains were scanned during the task and then they filled out questionnaires with respect to emotional arousal and depression.  Individuals with a diagnosis of depression received 15 months of psychodynamic psychotherapy following a baseline assessment.  Treatment involved between two and four hours of weekly therapy from a certified psychotherapist.  Control subjects did not undergo any type of therapy.  At the end of the 15 months, all subjects underwent follow-up testing.

Figure 1. Example stimuli used in the personal attachment task (Buccheim et al. 2012, Fig. 1).
Figure 1. Example stimuli used in the personal attachment task (Buccheim et al. 2012, Fig. 1).

Over the course of treatment, the average scores for the Beck Depression Inventory (BDI) and the Global Severity Index (GSI) decreased significantly between baseline and endpoint.  These data suggest changes in degree of depression from severe to mild-moderate.  No changes in either score were observed in control subjects.

The major focus of this study was to test whether there was a change in brain activity for patients following treatment that was not observed in controls when exposed to personally relevant cue sentences.  This interaction is displayed in Figure 2, illustrating the changes observed in patients only.  The red circle in panel A highlights an effect in the left portion of the medial temporal lobe.  Only the left side was affected, confirming the left-lateral pattern of activation associated with the AAP interview task.  The ventral anterior cingulate cortex (vACC) and medial prefrontal cortex (mPFC) are shown in the blue and yellow circles, respectively, in panel B.  In the medial temporal lobe and mPFC, individuals with depression showed higher levels of activation at baseline than control subjects and a decrease to control levels following treatment.  Depressed subjects showed more activity in the vACC than controls at baseline when exposed to personally relevant cue sentences, but activity below control levels at endpoint.   Brain activity in the vACC correlated with BDI scores, and there was a trend toward correlation between BDI scores and activity in the medial temporal lobe.  However, both the vACC and mPFC activity correlated with GSI scores.

Figure 2. Maps of the interaction between personally-relevant and neutral cues in patients and controls over time (Buccheim et al., 2012, Fig 2).
Figure 2. Maps of the interaction between personally-relevant and neutral cues in patients and controls over time (Buccheim et al., 2012, Fig 2).

Previous studies focused on the effects of cognitive behavioral and interpersonal therapies, with brain activity measured at rest.  The use of fMRI and AAP by Dr. Buccheim and colleagues in this context was novel in several ways: the long-term duration of treatment, the study of subjects with recurrent depression, and the psychodynamic nature of therapy. One potential limitation of this particular study is a lack of a no treatment control group of depressed subjects; however, it would be unethical to keep depressed individuals waitlisted for the 15-month duration of the study.  Additionally, the personalization of the AAP from the interview process could present a confounding variable due to differences among subjects.  To address this, depressed subjects and controls rated their level of emotional arousal.

Since this article has been published, Dr. Buchheim and colleagues have continued to use AAP coupled with fMRI in a case study involving a female dysthymic patient with narcissistic tendencies.  Therapy hours were rated either “difficult” or “easy” based on the quality of the patient’s day.  A significant interaction was found between the personally relevant statements and goodness of therapy hours in posterior cingulate/precuneal region, indicating increased activation when looking at personally relevant images during therapy that was rated poorly.  This region of the brain has been shown to be associated with discriminative avoidance, which correlates to the defense mechanisms exhibited by the patient.  This study presents only one example of many potential uses of neuroimaging-based correlation of therapeutic mechanisms.

Taken together, the data gathered in Dr. Buchheim’s 2012 study show an association between a long-term course of psychotherapy and changes in activation in regions of the brain that are involved in emotional regulation and processing, as observed in depression.  These regions are responsive to treatments such as antidepressant medication and deep-brain stimulation.  Studies correlating clinical and neuroimaging data can be used in the future to compare different therapeutic approaches and monitor the progress of any given therapy in individuals.  This method could also be used to analyze a combined approach of medication and therapy, which is often the course of treatment used for major depressive disorder.

I would like to leave the PLOS Neuro community with a few thoughts/questions to be open for discussion here or on Twitter:

  • Cognitive behavioral therapy is another commonly used form of treatment for major unipolar depression.  The goal of cognitive behavioral therapy is to alleviate symptoms of depression and distress in a typically short timeline.  Psychodynamic therapy focuses on uncovering the cause of distress and is more long-term.  Based on these differences, I would expect to see a more subtle decrease in activity between baseline and endpoint measures in the medial temporal lobe, vACC, and mPFC.  I would love to hear your thoughts and encourage you to comment.
  • How could this correlation of neuroimaging, therapy, and emotional/cognitive assessment be applied to other disorders?

Any views expressed are those of the author, and do not not necessarily reflect those of PLOS.


Ressler, K.J. and H.S. Mayberg (2007) “Targeting abnormal neural circuits in mood and anxiety disorders: from the laboratory to the clinic,” Nature Neuroscience 10 (9): 1116-1124. doi:10.1038/nn1944

Thomas, E.J. and R. Elliott (2009) “Brain imaging correlates of cognitive impairment in depression,” Frontiers in Human Neuroscience 3 (30). doi: 10.3389/neuro.09.030.2009

Buchheim A. et al. (2012) “Changes in Prefrontal-Limbic Function in Major Depression after 15 Months of Long-Term Psychotherapy,” PLoS ONE 7 (3): e33745. DOI: 10.1371/journal.pone.003374

This article is part of the Disordered Cognition section of the PLOS Neuroscience Collection.

Buchheim, A. et al. (2013) “A clinical case study of a psychoanalytic psychotherapy monitored with functional neuroimaging,” Frontiers in Human Neuroscience 7 (677). doi: 10.3389/fnhum.2013.00677

SakowskitwitterpicLauren Sakowski is a molecular biologist/neuroscientist at the Nemours/A.I. duPont Hospital for Children studying the role of inflammation in neurodegeneration in a leukodystrophy mouse model.  @LaSaks87

  1. This neat left-lateralization again cannot possibly be explained by lateralization of “responses to the task” but only by lymphatic- and adventitia-borne inflammation mostly from mast cells – and what CBT provokes is a voluntary (striate muscle!) neuro-feedback on the inflammatory functional increase (reverting later to atrophy) – via the PPG. Psychodynamic therapy only really works as ISTDP (of Habib Davanloo) and this resolves the causes of ongoing shame reflex via PPG which unleashes BBB mast cells. Then one would NOT “expect to see a more subtle decrease in activity between baseline and endpoint.” The other schemes (Mayeberg etc.) are biologically adventurous.

  2. Thank you for your comments! I had not looked at the Mayberg study previously, but the idea of finding biomarkers to determine a suitable treatment is fascinating. The connection between inflammation and depression is somewhat new to me and I appreciate your explanation of the difference between CBT and psychodynamic therapy.

  3. Your definition of psychoanalytic psychotherapy is a bit outdated. For those of us who have studied neuroscience and infant mental health research the goal of treatment is first to attain mutual regulation with the patient. Seriously depressed patients are people who have perhaps never had another human being actually mutually regulate with them. This experience is healing for the patient in and of itself. Even psychoanalysts and psychoanalytically informed psychotherapists understand how important the attainment of regulation is to the formation of a “good enough” patient therapist dyad though other terms have been used through the decades to describe this experience. The work goes on from that point.

  4. Interesting. But, how or why would neurological correlations become the decision-maker for therapy? That is, only IF it shows-up on the fMRI would the therapy be of any use? That is quite a bit of a leap. I believe neurological correlates have been specially useful for some sort of validation in the psychotherapy field, as fMRI’s and other bio and neurological variables functioning as correlates result in reserachers and therapists increasing the “belief” in their interventions IF there appears to be a sign of biological change. What are your thoughts?

  5. “Even psychoanalysts and psychoanalytically informed psychotherapists understands how important the attainment of regulation of regulation is to the formation of a “good enough” patient therapist dyad though other terms have been used through the decades to describe this experience”???? I think we need to set the history straight here. Attachment was first discussed by Freud and later gradually developed to the most empirically prolific branch of psychoanalytic theory by John Bowlby, another psychoanalyst. Emotion regulation as an aspect of attachment was predominately championed by Allan Shore, another psychoanalyst, although many other psychoanalysts, such as Rene Spitz, suggested this connection already in the 30’s by studying separations between children and parents in hospitals. Attunement between infant and caretaker was championed by Daniel Stern, again a psychoanalyst. “The good enough” parent/child or patient/therapist dyad is a rip off from the theory of Donald Winnicott, another classic psychoanalyst. There is nothing wrong with Buchheim et al.’s definition of the treatment. However, your attempt to “cut out” an important part of the psychoanalytic heritage, pretend it has nothing to do with psychoanalytic theory and then arrogantly suggest that “even” psychoanalysts know about this is a severe distortion of history and reflects badly on your level of education and professionalism. Your grade: D-

  6. I am by no means an expert in this field, so if anyone else wants to chime in, I welcome it. I would not necessarily refer to correlations as a hard and fast “decision-maker” but more of a guide. Taking into account that this is still a fairly new concept and treating it as such, it could provide insight into whether or not treatments are effective.

  7. Dr R. Carl
    Sadly your information is quite deficient
    Many clinicians researched emotion regulation b4 Alan Schore. Just ask him He will tell you!
    Also Daniel Stern will also tell you who reserched attunement prior to him….
    Time to study up!

Leave a Reply

Your email address will not be published. Required fields are marked *

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Back to top