Psychological Medicine, 2005, 35, 1385–1398. f 2005 Cambridge University Press doi:10.1017/S0033291705005064 Printed in the United Kingdom REVIEW ARTICLE Neuroimaging and the functional neuroanatomy of psychotherapy JOSHUA L. ROFFMAN*, CARL D. MARCI, DEBRA M. GLICK, DARIN D. DOUGHERTY AND SCOTT L. RAUCH Department of Psychiatry, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA ABSTRACT Background. Studies measuring the effects of psychotherapy on brain function are under-rep- resented relative to analogous studies of medications, possibly reflecting historical biases. However, psychological constructs relevant to several modalities of psychotherapy have demonstrable neuro- biological correlates, as indicated by functional neuroimaging studies in healthy subjects. This review examines initial attempts to measure directly the effects of psychotherapy on brain function in patients with depression or anxiety disorders. Method. Fourteen published, peer-reviewed functional neuroimaging investigations of psycho- therapy were identified through a MEDLINE search and critically reviewed. Studies were compared for consistency of findings both within specific diagnostic categories, and between specific mod- alities of psychotherapy. Results were also compared to predicted neural models of psychother- apeutic interventions. Results. Behavioral therapy for anxiety disorders was consistently associated with attenuation of brain-imaging abnormalities in regions linked to the pathophysiology of anxiety, and with acti- vation in regions related to positive reappraisal of anxiogenic stimuli. In studies of major depressive disorder, cognitive behavioral therapy and interpersonal therapy were associated with markedly similar changes in cortical–subcortical circuitry, but in unexpected directions. For any given psy- chiatric disorder, there was only partial overlap between the brain-imaging changes associated with pharmacotherapy and those associated with psychotherapy. Conclusions. Despite methodological limitations, initial neuroimaging studies have revealed con- vergent and mechanistically sensible effects of psychotherapy on brain function across a range of psychiatric disorders. Further research in this area may take advantage of emerging neuroimaging techniques to explore a broader range of psychotherapies, with the ultimate goal of improving clinical decision-making and treatment. INTRODUCTION more effort towards understanding the neural mechanisms of medication than of psycho- Functional neuroimaging studies provide a therapy (Fig. 1). This disparity has persisted means to observe and characterize changes in brain function related to psychiatric interven- despite the similar costs and clinical efficacy of medication and psychotherapy for common tions. Since the introduction of this research psychiatric disorders (Antonuccio et al. 1995; tool, investigators have devoted considerably Goldman et al. 1998; Satcher, 1999). Historical bias toward medications as being a clearly * Address for correspondence: Joshua L. Roffman, M.D., defined ‘biological’ intervention, compared with Department of Psychiatry, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114, USA. the more complex psychosocial intervention of (Email: jroff[email protected]) psychotherapy (Westen et al. 2004), has likely 1385 1386 J. L. Roffman et al. 40 35 30 25 20 15 Number of studies 10 5 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Year FIG. 1. Imaging/medication (%) and imaging/psychotherapy (&) studies by year. Method: An Ovid MEDLINE search was com- pleted using key words related to neuroimaging (e.g. PET, fMRI, SPECT), and medication (e.g. psychotropic) to find studies including both neuroimaging and medication between the years 1966 and 2003. Based on the abstracts generated from this search, we selected studies based on four criteria: included studies were published in English between 1990 and 2003, used human subjects, and investigated psychiatric (e.g. depression) rather than neurological (e.g. Parkinson’s Disease) disorders. A similar search was conducted using key words related to neuroimaging and psychotherapy (e.g. psychotherapy, intrapersonal therapy, cognitive behavioral therapy, and psychodynamic therapy). contributed to this imbalance. However, as first discuss how future research efforts may refine suggested well before the era of functional neuro- our physiological understanding of how psycho- imaging, the changes in affect, behavior, and therapy works, and why this knowledge may be cognition that are mediated by psychotherapies clinically useful. undoubtedly have biological underpinnings (Freud, 1895). In recent years, the number of studies using neuroimaging to explicitly evalu- ate neural correlates of psychotherapy has PUTATIVE NEURONAL MECHANISMS steadily increased. These studies answer the call OF PSYCHOTHERAPY for more biologically rigorous approaches to Many potentially unhealthy cognitive and psychotherapy research (Kandel, 1998). emotional patterns targeted by psychotherapy In this review, we will address how neuro- appear to have measurable biological analogs imaging research is beginning to reveal the (Beutel et al. 2003). One salient example in- relationship between psychotherapy and brain volves repression, or the unconscious ‘forget- function. We shall first give examples of how ting’ of threatening ideas or experiences. In a psychological constructs relevant to psycho- recent investigation by Anderson and associates therapy, such as extinction, cognitive restruc- (2004), healthy subjects were instructed either to turing, and repression, have been associated remember or ‘forget’ target words. In a recall with discrete brain activity. While long con- task, presentation of ‘forget’ words was asso- sidered the ‘building blocks’ of psychotherapy ciated not only with poor recall of those words, on a theoretical level, these constructs appear to but also with increased activation of the pre- have parallel meaning on the level of neuro- frontal cortex (PFC) and decreased hippo- anatomy. Second, we will evaluate the emerging campal activation. These results replicated a literature on psychotherapy-related changes in previous study conducted by Bunge and col- brain activity profiles, drawing some preliminary leagues (2001), who also found that the an- comparisons among different psychotherapies, terior cingulate gyrus directs attention away and between psychotherapeutic and psycho- from unwanted memories. Thus, active ‘for- pharmacological approaches. Finally, we will getting’ modulated by the PFC and anterior The functional neuroanatomy of psychotherapy 1387 cingulate may have relevance to the psycho- studies implicates the ventral PFC and amyg- dynamic concept of repression. dala in this process (Milad et al. in press). The process of psychotherapy may also Targeted lesions or specific pharmacological engage dedicated neural circuitries that are par- inhibition of the amygdala interfere with fear ticularly responsive to a discrete mode of treat- conditioning in rodents (see Maren & Quirk, ment. We find examples of this in studies related 2004 for a review), while functional neuro- to psychodynamic, cognitive, and behavioral imaging studies in humans have consistently therapy; such studies have used healthy subjects associated amygdalar activation with con- to investigate analogs of specific therapy pro- ditioned fear responses (Buchel et al. 1999; cesses. When a psychodynamically oriented Fischer et al. 2000; Charney, 2003; Cheng et al. therapist ‘takes a history’, this elicits episodic 2003). Analogous studies in rats (Lebron et al. memories from the patient in a focused way. 2004) and humans (Gottfried & Dolan, 2004; However, when the same therapist observes the Phelps et al. 2004) suggest that the ventral PFC patient ‘free associate’, episodic recall occurs in mediates the retention and recall of extinction a less organized, more random way. Andreasen for conditioned fear responses by keeping and co-workers (1995) observed that while the amygdala in check. One might, therefore, random memories engaged assocation cortex in expect extinction-based behavioral therapies in frontal, parietal, and temporal regions, focused humans to work either by potentiating ventral memories selectively activated verbal areas PFC activity or attenuating the amygdala (or (including Broca’s area and the left frontal both). operculum). Thus, the less ‘censored’ process of Collectively, the growing number of studies free association may engage wider networks related to the psychotherapy process point to a of association cortex, facilitating exploration of plausible neuronal substrate for psychother- latent aspects of the patient’s symptomatology apeutic interventions. In this setting, we have or personality. also seen emerge a critical mass of studies that In cognitive behavioral therapy (CBT) for directly evaluate the neurobiological effects of depression, patients are sometimes asked to psychotherapy in patients with mood and anxi- revisit bad or painful memories and explicitly ety disorders. re-evaluate their negativity toward the memory. Using a related paradigm in healthy subjects, Ochsner and colleagues (2002) observed a re- lationship among reappraisal of negative stim- THE PROBLEM OF VARIABLE uli, improvement in mood, and brain activity METHODOLOGY patterns. Subjects rated their mood before and Before examining the literature on psycho- after being asked to ‘re-interpret’ highly nega- therapy and neuroimaging in