Psychotherapy and Neuroimaging
Abstract
Technological advances in neuroimaging have enabled researchers to examine, in vivo, the relationship between psychotherapeutic interventions and markers of brain activity. This review focuses on two kinds of neuroimaging studies in psychotherapy: those that examine the patterns of brain activity associated with response to treatments and those that examine the changes that occur in brain activity during treatment. A general, hypothetical neural model of psychotherapy is presented, and support for the model is evaluated across anxiety disorders and major depression. Neuroimaging studies are broadly consistent in observing associations between response to psychotherapy and baseline activity in several key regions within the prefrontal cortex and limbic areas. These regions are involved in the generation and regulation of emotion, fear responding, and response to reward. Prepost examinations of change following psychotherapy also typically observe that psychological treatments for anxiety and depression can affect neural activity in these regions. Despite a general consensus that activity in these regions is associated with psychotherapy, substantial discrepancy persists regarding the precise direction of the observed relationships. Methodological challenges of the existing literature are considered, and future directions are discussed.
Psychotherapy’s efficacy is well established for a wide range of emotional disorders. Despite this, no psychotherapeutic intervention works equally well for all patients, and the mechanisms through which psychotherapy reduces symptoms and enhances functioning remain difficult to specify. With the advent of neuroimaging technologies, researchers have new tools with which to identify clinically meaningful markers of brain function that are associated with treatment response. Two kinds of associations have been examined. Treatment outcome prediction studies seek to identify those patterns of brain function that confer a higher likelihood that a treatment will work. Treatment mechanism studies examine changes in the brain as a result of the intervention in question, in order to help understand how treatments are exerting their effects. Both types of study hold promise for developing a more complete understanding of the neural mechanisms involved in successful therapy, and both may guide future treatment refinement, novel mechanistic treatment development, and personalized treatment prescriptions tailored to individual patients.
Below, a brief overview of the general neural architecture believed to be relevant for anxiety, depression, and psychotherapeutic interventions for these disorders is provided. Anxiety and depression are the focus of this review because these are the two disorders for which the most evidence has accrued, and also because there is good reason to believe that anxiety and depression share at least some underlying neural mechanisms. Functional neuroimaging studies of psychotherapy are reviewed, and recent advances toward improving the methodology and clinical relevance of research in this area are highlighted. Ideally, work will continue to progress toward greater relevance and import for the practicing clinician.
Neural Circuitry of Anxiety and Depression
Contemporary neurobiological models of anxiety and depression include both distinct (anxiety and depression specific) and overlapping networks of brain regions. As displayed in Figure 1, the neural circuitry of general emotion dysregulation and high negative affect, which is implicated in both types of disorders, includes an interconnected set of brain regions involved in the generation and regulation of emotion (1–3). Limbic structures (such as the amygdala, hippocampus, and insula) react to emotional information. Activity from these regions feeds forward through the anterior cingulate cortex ([ACC] involved in the appraisal and encoding of emotion), orbitofrontal cortex ([OFC] involved in the integration of affective and sensory information and reward processing), and finally to the dorsomedial and ventromedial prefrontal cortices ([DMPFC, VMPFC] involved in self-referential processing and in moderating emotional reactions). The initial activity in the limbic regions can be regulated by regions located within the prefrontal cortex (PFC). Lateral prefrontal regions, including the dorsolateral and ventrolateral prefrontal cortex ([DLPFC, VLPFC] both of which subserve higher-order cognitive functions), interact with the other frontal systems noted above, including the DMPFC, VMPFC, and ACC. These frontal systems are functionally interconnected with the amygdala and other limbic regions (3) and can modulate limbic activity during controlled processing of emotional stimuli (4).
Figure 1. Neural Circuitry of General Emotion Dysregulation and High Negative Affect as Seen From Medial (center of brain [top]) and Lateral (outside of brain [bottom]) Views
Limbic regions (white) such as the amygdala, hippocampus, and insula react to emotional information. Emotional information feeds forward from limbic regions (white) to cortical regions (gray), including the anterior cingulate cortex (ACC), orbitofrontal cortex (OFC), and finally to the dorsomedial and ventromedial prefrontal cortices (DMPFC, VMPFC). Prefrontal cortical regions (gray) including the ACC, DMPFC, VMPFC, as well as dorsolateral and ventrolateral prefrontal cortices (DLPFC, VLPFC), can provide top-down regulation over limbic regions (white arrows), modulating activity upward or downward depending on the context and goals.
The functioning of disorder-specific networks is also key to understanding the relationship between psychotherapy and brain function. In addition to the general emotion processing and emotion regulation networks described above, there is a partially overlapping set of regions that shows increased activation to fear-related stimuli. This system forms a “fear network,” and is particularly relevant to the anxious arousal and exaggerated fear responses that characterize anxiety disorders. This fear-responsive circuitry includes limbic regions such as the amygdala, hippocampus, and parahippocampal gyrus, as well as the insula, periaqueductal gray, and medial portions of the PFC (mPFC) including the VMPFC, OFC, and ACC (for a more detailed review, see de Carvalho et al. [5]). Finally, the functioning of an additional network, the reward circuit, is particularly relevant in the treatment of major depression (2), as it may play a role in anhedonia. This network of regions includes the ventral striatum, portions of the thalamus, amygdala, OFC, and mPFC (for a more detailed review, see Eshel and Roiser [6]).
Hypothesized Neural Circuitry of Psychotherapy
One potential cause for many of the core symptoms of depression and anxiety, particularly those associated with negative emotional experiences, could be an inefficiency of top-down cortical control over regions that respond to emotional stimuli (e.g., limbic and fear-network-related regions). Psychotherapy has broadly been hypothesized to remediate these neural abnormalities and reduce symptoms via a strengthening of the cortical, top-down emotion regulatory processes (Figure 1; see [7, 8]). Under this theory, improved PFC and cortical function would lead to enhanced regulation over limbic regions, thereby attenuating emotional reactions to negative inputs. Basic research in healthy control subjects suggests that psychotherapy skills, such as problem solving, cognitive reappraisal, extinction learning (the critical learning mechanism in exposure therapy for anxiety), and modification of clients’ self-representations each rely on the function of PFC structures (e.g., DLPFC, VLPFC, VMPFC, ACC; reviewed in [9]). This model therefore suggests that psychotherapy’s effects should involve improved regulation in PFC regions and corresponding modulation of regions responding to negative or threatening stimuli.
Whereas the neural models described above permit hypotheses regarding which regions of the brain are likely to be relevant in psychotherapy research, it is considerably more difficult to generate hypotheses regarding the expected direction of effects, e.g., increases versus decreases in activity. Following a review of the existing research findings, a discussion of the complexity in making directional hypotheses is provided.
Neuroimaging and Treatment Outcome Prediction
Anxiety Disorders
The most consistent finding in treatment outcome prediction of anxiety disorders is that a better response to therapy is associated with increased baseline hyper-reactivity in limbic or visual processing regions during provocation with threat-related visual stimuli. This pattern may indicate that a willingness or ability to adequately process threatening information is necessary to engage effectively with key psychotherapy strategies (e.g., exposure and habituation to feared stimuli). For instance, in a clinically anxious pediatric sample, better treatment response to either cognitive-behavioral therapy or fluoxetine was associated with greater amygdalar activation while attending to threatening social stimuli (10). In a relatively large sample of social anxiety disorder patients who completed cognitive-behavioral therapy, better response was predicted by increased activation to fearful faces in higher-order visual processing areas (occipitotemporal cortices [11]). An important strength of this last study was the use of a sophisticated nested cross-validation analysis to ensure that prediction estimates based on brain data were not inflated, as this is a common pitfall in neuroimaging research (12). Using this approach, occipitotemporal activations, in conjunction with behavioral variables, accounted for 41% of the treatment outcome variance. This represented a threefold increase in comparison to the variance explained by behavioral variables alone, suggesting the added value of neuroimaging techniques from a personalized medicine perspective. Similar findings were reported in a smaller sample of patients with social anxiety disorder, in which increased activation to threatening faces in higher-order visual processing areas (e.g., angular gyrus) predicted better response to cognitive-behavioral therapy (13).
With regard to PFC activation, one plausible hypothesis is that successful psychotherapy capitalizes on a patient’s existing strengths. Therefore, if psychotherapy indeed relies on PFC regulation of limbic regions, increased PFC function at baseline may confer an improved chance of response. Consistent with this hypothesis, an increased response to threatening faces in the DMPFC, ACC, and OFC has been found to predict better response to cognitive-behavioral therapy in patients with social anxiety disorder (13), and increased prefrontal (VMPFC/OFC) metabolism at rest has been found to predict better treatment response to behavior therapy in obsessive-compulsive disorder patients (14). Similarly, in a combined sample of adults with generalized anxiety disorder and panic disorder, a better response to cognitive-behavioral therapy was associated with increased activation of frontal regions during cognitive reappraisal, including DLPFC and precentral gyrus, as well as with increased limbic activations in the insula and a parahippocampal region (15). This latter study represents an important advance toward clinical applicability of neuroimaging findings, as the authors employed a random forest classification procedure to identify neuroimaging variables (i.e., regional brain activations) that would allow individual patients to be robustly classified as therapy responders or nonresponders. The combination of regions identified through this analysis yielded accurate classification for 79% of patients.
Of course, the opposing hypothesis is also plausible—that patients with the largest impairments in PFC function (i.e., decreased PFC function at baseline) may be in greatest need of the top-down regulatory skills provided during psychotherapy, and therefore may benefit most from this approach. Support for this contrasting hypothesis was found in two small studies, one examining obsessive-compulsive disorder (16), and the other focusing on posttraumatic stress disorder (17). Across these studies, a better response to cognitive-behavioral therapy was predicted by decreased activity in regulatory regions including the DLPFC (16), VMPFC (16), and ACC (16, 17).
Major Depression
The most common finding in the prediction of treatment outcome for depression involves activity in portions of the ventral ACC, including the pregenual/rostral ACC (portions of Brodmann’s area [BA] 24 and 32) and the subgenual ACC (BA 25). Over 20 studies have converged to indicate that activity in the ventral ACC is associated with response to medications, psychotherapy, and their combination (see [18–21]). However, the precise direction and the exact location of predictive effects within the ventral ACC remain controversial. Several studies have observed that increased activity in the rostral ACC is associated with a positive response to antidepressant medications (18). By contrast, Siegle and colleagues (19, 20) have demonstrated in three separate samples that lower sustained activity in the subgenual region of the ACC predicts positive response to cognitive-behavioral therapy for depression. Similar patterns in the ACC following cognitive-behavioral therapy have been observed by two independent research groups using different task designs (22, 23).
The picture that emerges from the studies reviewed above seems to indicate that increased activity in the ventral ACC should predict a superior response to medications whereas decreased activity should predict a superior response to cognitive-behavioral therapy. There are, however, findings inconsistent with this account. In two separate samples, Mayberg and colleagues (24, 21) examined predictors of general nonresponse to cognitive-behavioral therapy and medications. In both studies, pretreatment hypermetabolism in the ventral ACC was associated with a poor response across treatments. Moreover, using some of the same data, Mayberg and colleagues failed to find that activity in the ventral ACC was associated with a response to cognitive-behavioral therapy (21). Part of the reason for the discrepant findings could stem from differences in imaging technology (i.e., measures of metabolism versus blood flow) and methods (i.e., task-related activity versus resting activity). In addition, the ACC has been implicated both in the bottom-up generation of emotional experience and in the top-down regulation of limbic areas. As such, directional hypotheses about the nature of activity in this region are particularly difficult (20). It may be that relative increases or decreases in activity during a task are dependent on the resting activity prior to the start of the task (7). Moreover, the components of the ventral ACC, and the rostral and subgenual ACC, are distinct subregions, with different cellular properties and separate patterns of connectivity to other regions (see [18]). Specifying the precise role of each of these regions in emotion generation and regulation, as well as understanding how the functioning of these regions impacts different types of treatment, is an area of ongoing and active research.
To date, direct comparisons between brain-based predictors of psychotherapy and other active treatments have been rare. In an early, nonrandomized study, the connectivity between nodes of a fronto-limbic-thalamic circuit was examined (25). The authors observed that differences in the connectivity between the hippocampus and lateral PFC and between OFC and medial PFC differentiated responders to cognitive-behavioral therapy versus medications. More recently, and in the first study of its kind, McGrath and colleagues (26) examined prescriptive brain-based predictors of differential response to cognitive-behavioral therapy versus medication in the context of a randomized clinical trial. Focusing only on those subjects with the clearest clinical outcomes, i.e., those who met full remission criteria versus those who completed the trial without demonstrating much change, the authors identified six regions in which baseline metabolic activity predicted differential response: the right anterior insula, the right inferior temporal cortex, the left amygdala, the left premotor cortex, the right motor cortex, and the precuneus. Of these, the authors note that activity in the anterior insula constituted the strongest finding. Individuals with reduced metabolism in the anterior insula at baseline were more likely to remit following cognitive-behavioral therapy (and less likely following medications), whereas those with increased metabolism were more likely to remit following medications (and less likely following cognitive-behavioral therapy).
Other regions have been shown to predict response to psychotherapy for depression in single treatment-modality studies; however, the following relationships have not yet been replicated and should be treated as preliminary. In addition to the ACC findings reviewed above, Siegle and colleagues found evidence that activity in the right amygdala (19), and DLPFC (20) was associated with response to cognitive-behavioral therapy. The paracingulate gyrus (27) and vmPFC (28) may also be related to response to cognitive-behavioral therapy. Finally, striatal reactivity to rewarding outcomes may predict favorable response to cognitive-behavioral therapy among depressed adolescents (29).
Neuroimaging and Treatment Mechanisms
Several comprehensive reviews have examined brain-related changes following psychotherapy for anxiety and depression (see, e.g., [8, 30, 31]). The review below focuses only on those findings that speak directly to the hypothesized model of the effects of psychotherapy, and, where possible, only on those findings that have been replicated.
Anxiety Disorders
In studies of psychotherapy for anxiety disorders, changes in activity in portions of the PFC and limbic systems are typically observed; however, the direction of those changes varies based on the nature of the anxiety disorder and/or the study design. In the PFC, psychotherapy for posttraumatic stress disorder, for example, is associated with increased activation across multiple different task conditions, following both eye-movement desensitization and reprocessing therapy (e.g., [32, 33]) and cognitive-behavioral therapy (e.g., [34, 35]). Within the PFC, ACC increases following cognitive-behavioral therapy are emerging as a particularly well-replicated finding linked to posttraumatic stress disorder symptom decreases (34).
In obsessive-compulsive disorder, by contrast, several studies have reported decreased activity in multiple regions of the PFC, including the OFC (e.g., [36]), DLPFC (36, 37), mPFC (37), and ACC (e.g., 36) across different tasks. It is important to keep in mind that in both obsessive-compulsive disorder and posttraumatic stress disorder, single studies have reported contradictory findings (see [8, 30, 31]).
Less consistent evidence of change in PFC activity has been observed in social anxiety disorder. In one study of cognitive-behavioral therapy completers (13), decreases in VMPFC and DMPFC during symptom provocation were observed following treatment. However, in a large and well-controlled psychotherapy trial, patients with social anxiety disorder engaged in a cognitive reappraisal task following exposure to personalized negative self-beliefs (38). Cognitive-behavioral therapy was associated with increased DLPFC and DMPFC activity, earlier temporal onset of DMPFC activity, and increased DMPFC-amygdala functional connectivity, consistent with the hypothesized top-down cortical substrates of this specific therapy skill. This study is particularly notable in that it utilized a more ecologically valid, therapy-relevant cognitive task, allowing for a specific test of hypothesized neural mechanisms during the application of a particular skill.
Prefrontal effects following psychotherapy for panic disorder and specific phobia tend to involve those regions involved in the “fear network.” Specifically, two small, uncontrolled positron emission tomography studies of panic disorder found increased mPFC and decreased ACC metabolism at rest following cognitive-behavioral therapy (39, 40). A larger functional magnetic resonance imaging study in panic disorder found reductions following cognitive-behavioral therapy in the left DLPFC during fear acquisition, which was correlated with a reduction in agoraphobic symptoms (41). For specific phobia during symptom provocation, three small studies were consistent in finding decreased medial PFC/ACC activity following cognitive-behavioral therapy (42–44), whereas increased activation in the OFC was found in a distinct sample (e.g., 45).
Changes in limbic activity also appear to depend on methodology and the nature of the anxiety disorder in question. Reductions in amygdala activity following psychotherapy have been replicated only in studies of posttraumatic stress disorder (34, 35), whereas reductions in insula activity have been observed more generally. Following cognitive-behavioral therapy, decreased insula activity has been reported during presentation of social threat cues in social anxiety disorder (13), during fear learning/acquisition in panic disorder (41), and during symptom provocation in specific phobia (e.g., 43, 44). In the hippocampus, the direction of prepost changes appears to differ depending on the disorder. Posttraumatic stress disorder has been associated with increases in hippocampal activity following therapy (e.g., 34, 35), whereas reductions in this region have been observed following cognitive-behavioral therapy for social anxiety disorder (46) and following psychodynamic psychotherapy for panic disorder (47). Changes in limbic and other relevant regions following the treatment of obsessive-compulsive disorder have been relatively inconsistent across trials (see 8, 30, 31).
Major Depression
To date, at least two studies have been conducted to examine brain-related changes in response to each of the major schools of psychotherapy for depression: interpersonal psychotherapy, cognitive-behavioral therapy, behavioral activation, and psychodynamic psychotherapy. The majority of these studies have examined change following psychotherapy on its own, or in comparison to changes observed over the same period of time in healthy controls. Very few studies have compared changes following psychotherapy with changes following a different active treatment, and to date, only one study (48) has done so in the context of a fully randomized clinical trial.
Across psychotherapies for depression, reductions in activity or metabolism have been observed in several PFC regions, including the DLPFC, VLPFC, as well as medial prefrontal regions (e.g., dorsal and ventral mPFC, subgenual ACC, and OFC). Although the precise regions in the PFC differ somewhat across studies and treatments, reductions in one or more of these regions have been observed following interpersonal therapy (49), cognitive-behavioral therapy (48, 50), behavioral activation (27), and psychodynamic psychotherapy (51). One of the most well-replicated findings from this literature is that treatment with cognitive-behavioral therapy is associated with a reduction in emotional biases toward negative stimuli in the PFC, as evidenced by relative reductions in activity to negative stimuli and increases to positive stimuli (22, 28).
Regarding changes in other neural regions, results to date are more preliminary and mixed. For example, both decreases and increases in activity in the amygdala and hippocampus have been observed in response to cognitive-behavioral therapy (23, 28, 50). Furthermore, whereas studies have reported increases in activity in dorsal (23, 50) and ventral (48) areas of the cingulate cortex following cognitive-behavioral therapy, Siegle and colleagues (19) observed that cognitive-behavioral therapy responders did not show an increase in ventral ACC activity. Rather, the majority of those who responded to treatment had lower sustained activity in the ACC compared with healthy controls both before and after treatment. The precise location within the ACC of these sets of findings differs, and additional work will no doubt be needed to identify sources of these discrepancies.
Preliminary evidence from a single study suggests that treatment with behavioral activation can also affect reward systems in the brain. Behavioral activation was associated with increased activity in the caudate (a component of the striatal reward system) during reward anticipation and with changes to the paracingulate gyrus (in mPFC) during reward selection and reward feedback (52). Additional preliminary findings suggest molecular effects of psychodynamic therapy. In two small, preliminary studies, the authors observed an increase in serotonin transporter availability in the midbrain (53) and increases in serotonin receptor density in several frontal regions including the OFC, ventral ACC, mPFC, and DLPFC (54) following psychodynamic psychotherapy.
General Discussion
The findings reviewed above are broadly consistent with predictions regarding the neural substrates of psychotherapy. Activity in regions associated with negative emotion, emotion regulation, fear, and reward are associated with a response to psychotherapy, and psychotherapy appears to alter the functioning of these regions. Beyond understanding which regions are involved, however, the state of the field has not yet evolved sufficiently to make many specific conclusions regarding the direction of these effects. Conflicting directional observations may, for example, be due to differences in task states (imaging during a resting state versus symptom provocation versus application of a specific therapy skill). Increased PFC function at rest may, in fact, contribute to decreased capacity for activation of the PFC in response to symptom provocation or skill application, leading to findings in opposing directions depending on the task state that is examined (7). Furthermore, regional increases and decreases observed in neuroimaging are currently subject to multiple interpretations. Increased activation in a given region might be interpreted as reflecting an improvement in the strength of the region’s function, or as an impairment in the region’s efficiency, reflecting a need for greater activity in order to accomplish the same effect. One goal of future research will be to further clarify the precise nature of associations and to resolve the inconsistencies that have been observed.
The fact that brain function measured pretreatment is associated with the likelihood of response to psychotherapies is important. It suggests that future refinements regarding the precise direction of these effects across multiple kinds of tasks may enable treatments to be selected or individually tailored to the unique needs of the individual. Currently, several studies of psychotherapeutic treatment for anxiety have consistently implicated increased hyperresponsivity of regions reacting to threatening stimuli (e.g., the limbic and visual processing areas) as a marker of a better response to therapy. Such hyperreactivity may represent a clinically useful biomarker conferring a higher chance of a positive outcome from therapy, perhaps due to increased engagement with anxiety-provoking stimuli at baseline.
The pattern of predictive findings from studies of depression suggests that the functional state of the emotion regulation system, and potentially the reward system, prior to treatment has important consequences for the efficacy of cognitive-behavioral therapy. This may not be surprising. Cognitive-behavioral therapy is believed to engage and strengthen the patient’s ability to regulate and alter his or her emotional states. Little doubt remains that the ventral portions of the ACC play a key role in determining the likelihood of treatment response for depression. However, more work is needed to specify which specific subregions are critical, and what patterns of activity (to which tasks) are predictive of good or bad outcomes.
In order for brain-based predictive findings to become applicable in the clinic, greater clarity is needed regarding the precise task and imaging parameters that will lead to reproducible results at the single-patient level. Future work should aim to build on the strengths of recent studies, which used larger samples (11, 15), randomization of participants to different treatments (26), and more sophisticated analytic approaches that allow researchers to estimate the added benefit of neuroimaging data (11), to examine patterns of communication between brain regions (25), and to draw inferences that are valid at the individual patient level (15). These advances will help to address questions such as which treatment option is best for which patient. Additionally, when multiple brain regions are observed to predict response, as in (26), effort needs to be made to combine these multiple predictors in order to make a single treatment recommendation for the patient (see [55] for a new method for doing so.)
Examining change in brain function over the course of treatment can be a valuable tool with which to understand the mechanisms through which a given treatment is operating. As others have noted (see 30, 56), the best designs for such studies would have at least three groups: the group receiving the treatment in question, a group of similar psychiatric patients who either receive a different treatment or no treatment, and a group of relatively well-matched healthy control subjects. With such a design, researchers would be able to determine whether or not the mechanisms observed are unique to the treatment under investigation, and whether the treatment involves normalization of function or the recruitment of compensatory systems. Additionally, researchers must examine and demonstrate that the psychometric properties of neuroimaging-derived markers, e.g., test-retest reliability, are sound (57).
Currently, pre-post psychotherapy neuroimaging studies in anxiety disorders reveal some evidence consistent with hypothesized psychotherapy substrates (e.g., PFC increases, limbic decreases, and/or increased PFC-limbic functional connectivity). Evidence for increased activation in top-down PFC regions has been found most consistently in studies of posttraumatic stress disorder, whereas the findings in other anxiety disorders tend to suggest limbic (e.g., insular) decreases, either accompanied by concomitant PFC decreases (e.g., obsessive-compulsive disorder), or without consistent evidence of increases or decreases in regulatory regions.
Regarding changes during psychotherapy for depression, the most consistent findings are decreased activity across several regions of the PFC following multiple forms of psychotherapy, and reductions in negatively biased information processing in the PFC following cognitive-behavioral therapy. The fact that psychotherapy is broadly associated with decreased activity in several PFC regions may be surprising, and it appears to run counter to the hypothesis that psychotherapy should strengthen the ability of top-down control regions to modify the processing in down-stream systems. There is simply not yet enough information to resolve this issue, which is likely related to the methodological factors discussed above. It is also possible that the hypothesized mechanism of action of psychotherapy for depression is incorrect. More work will be needed to test these possibilities. The most informative studies will be those that test the core hypothesis that psychotherapy exerts effects via a top-down cortical route. If this hypothesis is true, psychotherapies should involve increasing communication between PFC and limbic areas, as evidenced by increased “functional connectivity” across these regions. To date, only one treatment study in depression has examined this issue (25), and many more are needed to determine under what conditions psychotherapy can and cannot affect the functioning of the key neural circuits relevant to depression.
Methodological Challenges and Future Research
Although neuroimaging studies have tremendous potential to advance our understanding of psychotherapy process and outcomes, this area of research is relatively new. As such, the clinical relevance and applicability of findings to date are limited. Treatment outcome prediction studies have rarely included more than one treatment condition, and thus cannot differentiate between prescriptive effects (which would allow for the selection of the specific treatment most likely to work for a given patient, relative to other treatments) and prognostic effects (which indicate more generally whether treatment itself is likely to be effective). Similarly, treatment mechanism studies have often used methods that cannot adequately distinguish between mechanisms of change through which the therapy is exerting its specific effects, practice effects associated with repeated testing, and nonspecific neural correlates of symptom improvement.
A handful of more recent studies offer a glimpse of the cutting edge within this area of research, through the inclusion of larger, well-controlled samples (26, 38, 41, 58), tasks designed to probe specific therapy skills (38, 58), and explicit analysis of the relationship between neural change and treatment efficacy (e.g., symptom reduction [41] or adherence [58]). Additional research is also needed to examine the effects of specific components of treatment. One notable example is a study utilizing EEG measures of cortical activity to examine the association between brain function and cortical activity before and after a brief, 30-minute, cognitive-behavioral therapy analog training session (59). Approaches like this may aid in identifying the neural mechanisms associated with specific therapeutic interventions. Finally, newer technologies, such as fNIRS, may provide researchers with additional tools with which to mitigate some of the methodological challenges noted above (60). During fNIRS, participants wear skullcaps containing devices that emit and detect near-infrared light, with which changes in cortical blood flow can be examined. fNIRS frees participants from the confines of a magnetic scanner, and as such, allows participants to engage in tasks not possible during more standard neuroimaging protocols.
Neuroimaging in psychotherapy is an active and growing area of research. It holds tremendous promise for helping clinicians tailor specific interventions to the needs of individual patients, and it may help scientists determine how psychotherapeutic treatments work. By continuing to build on recent methodological advances, research in this area has the potential to fulfill its “bench-to-bedside” promise and improve clinical care across the full range of psychological suffering.
Acknowledgments
Dr. Fournier is supported by NIMH award K23MH097889. Dr. Price is supported by NIMH award K23MH100259.
Footnotes
Jay C. Fournier, Ph.D., Department of Psychiatry, University of Pittsburgh School of Medicine
Rebecca B. Price, Ph.D., Department of Psychiatry, University of Pittsburgh School of Medicine
The authors report no financial relationships with commercial interests.
References
1.
Phillips ML, Ladouceur CD, Drevets WC: A neural model of voluntary and automatic emotion regulation: implications for understanding the pathophysiology and neurodevelopment of bipolar disorder. Mol Psychiatry 2008; 13:829-, 833–857
2.
Kupfer DJ, Frank E, Phillips ML: Major depressive disorder: new clinical, neurobiological, and treatment perspectives. Lancet 2012; 379:1045–1055
3.
Davidson RJ: Affective style, psychopathology, and resilience: brain mechanisms and plasticity. Am Psychol 2000; 55:1196–1214
4.
Beauregard M, Lévesque J, Bourgouin P: Neural correlates of conscious self-regulation of emotion. J Neurosci 2001; 21:RC165
5.
de Carvalho MR, Dias GP, Cosci F, de-Melo-Neto VL, Bevilaqua MC, Gardino PF, Nardi AE: Current findings of fMRI in panic disorder: contributions for the fear neurocircuitry and CBT effects. Expert Rev Neurother 2010; 10:291–303
6.
Eshel N, Roiser JP: Reward and punishment processing in depression. Biol Psychiatry 2010; 68:118–124
7.
DeRubeis RJ, Siegle GJ, Hollon SD: Cognitive therapy versus medication for depression: treatment outcomes and neural mechanisms. Nat Rev Neurosci 2008; 9:788–796
8.
Weingarten CP, Strauman TJ: Neuroimaging for psychotherapy research: Current trends. Psychother Res In press
9.
Frewen PA, Dozois DJ, Lanius RA: Neuroimaging studies of psychological interventions for mood and anxiety disorders: empirical and methodological review. Clin Psychol Rev 2008; 28:228–246
10.
McClure EB, Adler A, Monk CS, Cameron J, Smith S, Nelson EE, Leibenluft E, Ernst M, Pine DS: fMRI predictors of treatment outcome in pediatric anxiety disorders. Psychopharmacology (Berl) 2007; 191:97–105
11.
Doehrmann O, Ghosh SS, Polli FE, Reynolds GO, Horn F, Keshavan A, Triantafyllou C, Saygin ZM, Whitfield-Gabrieli S, Hofmann SG, Pollack M, Gabrieli JD: Predicting treatment response in social anxiety disorder from functional magnetic resonance imaging. JAMA Psychiatry 2013; 70:87–97
12.
Whelan R, Garavan H: When optimism hurts: inflated predictions in psychiatric neuroimaging. Biol Psychiatry 2014 75:746–748
13.
Klumpp H, Fitzgerald DA, Phan KL: Neural predictors and mechanisms of cognitive behavioral therapy on threat processing in social anxiety disorder. Prog Neuropsychopharmacol Biol Psychiatry 2013; 45:83–91
14.
Brody AL, Saxena S, Schwartz JM, Stoessel PW, Maidment K, Phelps ME, Baxter LR: FDG-PET predictors of response to behavioral therapy and pharmacotherapy in obsessive compulsive disorder. Psychiatry Res 1998; 84:1–6
15.
Ball TM, Stein MB, Ramsawh HJ, Campbell-Sills L, Paulus MP: Single-subject anxiety treatment outcome prediction using functional neuroimaging. Neuropsychopharmacology 2014; 39:1254–1261
16.
Olatunji BO, Ferreira-Garcia R, Caseras X, Fullana MA, Wooderson S, Speckens A, Lawrence N, Giampietro V, Brammer MJ, Phillips ML, Fontenelle LF, Mataix-Cols D: Predicting response to cognitive behavioral therapy in contamination-based obsessive-compulsive disorder from functional magnetic resonance imaging. Psychol Med 2013; 1–13
17.
Bryant RA, Felmingham K, Kemp A, Das P, Hughes G, Peduto A, Williams L: Amygdala and ventral anterior cingulate activation predicts treatment response to cognitive behaviour therapy for post-traumatic stress disorder. Psychol Med 2008; 38:555–561
18.
Pizzagalli DA: Frontocingulate dysfunction in depression: toward biomarkers of treatment response. Neuropsychopharmacology 2011; 36:183–206
19.
Siegle GJ, Carter CS, Thase ME: Use of FMRI to predict recovery from unipolar depression with cognitive behavior therapy. Am J Psychiatry 2006; 163:735–738
20.
Siegle GJ, Thompson WK, Collier A, Berman SR, Feldmiller J, Thase ME, Friedman ES: Toward clinically useful neuroimaging in depression treatment: prognostic utility of subgenual cingulate activity for determining depression outcome in cognitive therapy across studies, scanners, and patient characteristics. Arch Gen Psychiatry 2012; 69:913–924
21.
McGrath CL, Kelley ME, Dunlop BW, Holtzheimer Iii PE, Craighead WE, Mayberg HS: Pretreatment brain states identify likely nonresponse to standard treatments for depression. Biol Psychiatry In press
22.
Yoshimura S, Okamoto Y, Onoda K, Matsunaga M, Okada G, Kunisato Y, Yoshino A, Ueda K, Suzuki S, Yamawaki S: Cognitive behavioral therapy for depression changes medial prefrontal and ventral anterior cingulate cortex activity associated with self-referential processing. Soc Cogn Affect Neurosci 2014; 9:487–493
23.
Fu CHY, Williams SCR, Cleare AJ, Scott J, Mitterschiffthaler MT, Walsh ND, Donaldson C, Suckling J, Andrew C, Steiner H, Murray RM: Neural responses to sad facial expressions in major depression following cognitive behavioral therapy. Biol Psychiatry 2008; 64:505–512
24.
Konarski JZ, Kennedy SH, Segal ZV, Lau MA, Bieling PJ, McIntyre RS, Mayberg HS: Predictors of nonresponse to cognitive behavioural therapy or venlafaxine using glucose metabolism in major depressive disorder. J Psychiatry Neurosci 2009; 34:175–180
25.
Seminowicz DA, Mayberg HS, McIntosh AR, Goldapple K, Kennedy S, Segal Z, Rafi-Tari S: Limbic-frontal circuitry in major depression: a path modeling metanalysis. Neuroimage 2004; 22:409–418
26.
McGrath CL, Kelley ME, Holtzheimer PE, Dunlop BW, Craighead WE, Franco AR, Craddock RC, Mayberg HS: Toward a neuroimaging treatment selection biomarker for major depressive disorder. JAMA Psychiatry 2013; 70:821–829
27.
Dichter GS, Felder JN, Smoski MJ: The effects of Brief Behavioral Activation Therapy for Depression on cognitive control in affective contexts: an fMRI investigation. J Affect Disord 2010; 126:236–244
28.
Ritchey M, Dolcos F, Eddington KM, Strauman TJ, Cabeza R: Neural correlates of emotional processing in depression: changes with cognitive behavioral therapy and predictors of treatment response. J Psychiatr Res 2011; 45:577–587
29.
Forbes EE, Olino TM, Ryan ND, Birmaher B, Axelson D, Moyles DL, Dahl RE: Reward-related brain function as a predictor of treatment response in adolescents with major depressive disorder. Cogn Affect Behav Neurosci 2010; 10:107–118
30.
Barsaglini A, Sartori G, Benetti S, Pettersson-Yeo W, Mechelli A: The effects of psychotherapy on brain function: a systematic and critical review. Prog Neurobiol 2014; 114:1–14
31.
Quidé Y, Witteveen AB, El-Hage W, Veltman DJ, Olff M: Differences between effects of psychological versus pharmacological treatments on functional and morphological brain alterations in anxiety disorders and major depressive disorder: a systematic review. Neurosci Biobehav Rev 2012; 36:626–644
32.
Lansing K, Amen DG, Hanks C, Rudy L: High-resolution brain SPECT imaging and eye movement desensitization and reprocessing in police officers with PTSD. J Neuropsychiatry Clin Neurosci 2005; 17:526–532
33.
Pagani M, Högberg G, Salmaso D, Nardo D, Sundin O, Jonsson C, Soares J, Aberg-Wistedt A, Jacobsson H, Larsson SA, Hällström T: Effects of EMDR psychotherapy on 99mTc-HMPAO distribution in occupation-related post-traumatic stress disorder. Nucl Med Commun 2007; 28:757–765
34.
Felmingham K, Kemp A, Williams L, Das P, Hughes G, Peduto A, Bryant R: Changes in anterior cingulate and amygdala after cognitive behavior therapy of posttraumatic stress disorder. Psychol Sci 2007; 18:127–129
35.
Peres JF, Newberg AB, Mercante JP, Simão M, Albuquerque VE, Peres MJ, Nasello AG: Cerebral blood flow changes during retrieval of traumatic memories before and after psychotherapy: a SPECT study. Psychol Med 2007; 37:1481–1491
36.
Nakao T, Nakagawa A, Yoshiura T, Nakatani E, Nabeyama M, Yoshizato C, Kudoh A, Tada K, Yoshioka K, Kawamoto M, Togao O, Kanba S: Brain activation of patients with obsessive-compulsive disorder during neuropsychological and symptom provocation tasks before and after symptom improvement: a functional magnetic resonance imaging study. Biol Psychiatry 2005; 57:901–910
37.
Yamanishi T, Nakaaki S, Omori IM, Hashimoto N, Shinagawa Y, Hongo J, Horikoshi M, Tohyama J, Akechi T, Soma T, Iidaka T, Furukawa TA: Changes after behavior therapy among responsive and nonresponsive patients with obsessive-compulsive disorder. Psychiatry Res 2009; 172:242–250
38.
Goldin PR, Ziv M, Jazaieri H, Hahn K, Heimberg R, Gross JJ: Impact of cognitive behavioral therapy for social anxiety disorder on the neural dynamics of cognitive reappraisal of negative self-beliefs: randomized clinical trial. JAMA Psychiatry 2013; 70:1048–1056
39.
Prasko J, Horácek J, Záleský R, Kopecek M, Novák T, Pasková B, Skrdlantová L, Belohlávek O, Höschl C: The change of regional brain metabolism (18FDG PET) in panic disorder during the treatment with cognitive behavioral therapy or antidepressants. Neuroendocrinol Lett 2004; 25:340–348
40.
Sakai Y, Kumano H, Nishikawa M, Sakano Y, Kaiya H, Imabayashi E, Ohnishi T, Matsuda H, Yasuda A, Sato A, Diksic M, Kuboki T: Changes in cerebral glucose utilization in patients with panic disorder treated with cognitive-behavioral therapy. Neuroimage 2006; 33:218–226
41.
Kircher T, Arolt V, Jansen A, Pyka M, Reinhardt I, Kellermann T, Konrad C, Lueken U, Gloster AT, Gerlach AL, Ströhle A, Wittmann A, Pfleiderer B, Wittchen HU, Straube B: Effect of cognitive-behavioral therapy on neural correlates of fear conditioning in panic disorder. Biol Psychiatry 2013; 73:93–101
42.
Paquette V, Lévesque J, Mensour B, Leroux J-M, Beaudoin G, Bourgouin P, Beauregard M: “Change the mind and you change the brain”: effects of cognitive-behavioral therapy on the neural correlates of spider phobia. Neuroimage 2003; 18:401–409
43.
Straube T, Glauer M, Dilger S, Mentzel H-J, Miltner WHR: Effects of cognitive-behavioral therapy on brain activation in specific phobia. Neuroimage 2006; 29:125–135
44.
Hauner KK, Mineka S, Voss JL, Paller KA: Exposure therapy triggers lasting reorganization of neural fear processing. Proc Natl Acad Sci U S A 2012; 109(23):9203–9208
45.
Schienle A, Schäfer A, Hermann A, Rohrmann S, Vaitl D: Symptom provocation and reduction in patients suffering from spider phobia: an fMRI study on exposure therapy. Eur Arch Psychiatry Clin Neurosci 2007; 257:486–493
46.
Furmark T, Tillfors M, Marteinsdottir I, Fischer H, Pissiota A, Långström B, Fredrikson M: Common changes in cerebral blood flow in patients with social phobia treated with citalopram or cognitive-behavioral therapy. Arch Gen Psychiatry 2002; 59:425–433
47.
Beutel ME, Stark R, Pan H, Silbersweig D, Dietrich S: Changes of brain activation pre- post short-term psychodynamic inpatient psychotherapy: an fMRI study of panic disorder patients. Psychiatry Res 2010; 184:96–104
48.
Kennedy SH, Konarski JZ, Segal ZV, Lau MA, Bieling PJ, McIntyre RS, Mayberg HS: Differences in brain glucose metabolism between responders to CBT and venlafaxine in a 16-week randomized controlled trial. Am J Psychiatry 2007; 164:778–788
49.
Brody AL, Saxena S, Stoessel P, Gillies LA, Fairbanks LA, Alborzian S, Phelps ME, Huang S-C, Wu H-M, Ho ML, Ho MK, Au SC, Maidment K, Baxter LR: Regional brain metabolic changes in patients with major depression treated with either paroxetine or interpersonal therapy: preliminary findings. Arch Gen Psychiatry 2001; 58:631–640
50.
Goldapple K, Segal Z, Garson C, Lau M, Bieling P, Kennedy S, Mayberg H: Modulation of cortical-limbic pathways in major depression: treatment-specific effects of cognitive behavior therapy. Arch Gen Psychiatry 2004; 61:34–41
51.
Buchheim A, Viviani R, Kessler H, Kächele H, Cierpka M, Roth G, George C, Kernberg OF, Bruns G, Taubner S: Changes in prefrontal-limbic function in major depression after 15 months of long-term psychotherapy. PLoS ONE 2012; 7:e33745
52.
Dichter GS, Felder JN, Petty C, Bizzell J, Ernst M, Smoski MJ: The effects of psychotherapy on neural responses to rewards in major depression. Biol Psychiatry 2009; 66:886–897
53.
Lehto SM, Tolmunen T, Joensuu M, Saarinen PI, Valkonen-Korhonen M, Vanninen R, Ahola P, Tiihonen J, Kuikka J, Lehtonen J. Changes in midbrain serotonin transporter availability in atypically depressed subjects after one year of psychotherapy. Prog Neuropsychopharmacol Biol Psychiatry 2008; 32(1):229–237
54.
Karlsson H, Hirvonen J, Salminen J, Hietala J: Increased serotonin receptor 1A binding in major depressive disorder after psychotherapy, but not after SSRI pharmacotherapy, is related to improved social functioning capacity. Psychother Psychosom 2013; 82:260–261
55.
DeRubeis RJ, Cohen ZD, Forand NR, Fournier JC, Gelfand LA, Lorenzo-Luaces L: The Personalized Advantage Index: translating research on prediction into individualized treatment recommendations. A demonstration. PLoS ONE 2014; 9:e83875
56.
Linden DEJ: How psychotherapy changes the brain—the contribution of functional neuroimaging. Mol Psychiatry 2006; 11:528–538
57.
Barch DM, Mathalon DH: Using brain imaging measures in studies of procognitive pharmacologic agents in schizophrenia: psychometric and quality assurance considerations. Biol Psychiatry 2011; 70:13–18
58.
Goldin P, Ziv M, Jazaieri H, Hahn K, Gross JJ: MBSR vs aerobic exercise in social anxiety: fMRI of emotion regulation of negative self-beliefs. Soc Cogn Affect Neurosci 2013; 8:65–72
59.
Deldin PJ, Chiu P: Cognitive restructuring and EEG in major depression. Biol Psychol 2005; 70:141–151
60.
Roffman JL, Marci CD, Glick DM, Dougherty DD, Rauch SL: Neuroimaging and the functional neuroanatomy of psychotherapy. Psychol Med 2005; 35:1385–1398
Information & Authors
Information
Published In
History
Published online: 1 July 2014
Published in print: Summer 2014
Authors
Funding Information
Author Information and Disclosure:
Metrics & Citations
Metrics
Citations
Export Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
View Options
View options
PDF/EPUB
View PDF/EPUBGet Access
Login options
Already a subscriber? Access your subscription through your login credentials or your institution for full access to this article.
Personal login Institutional Login Open Athens loginNot a subscriber?
PsychiatryOnline subscription options offer access to the DSM-5-TR® library, books, journals, CME, and patient resources. This all-in-one virtual library provides psychiatrists and mental health professionals with key resources for diagnosis, treatment, research, and professional development.
Need more help? PsychiatryOnline Customer Service may be reached by emailing [email protected] or by calling 800-368-5777 (in the U.S.) or 703-907-7322 (outside the U.S.).