While psychiatrists generally believe that biological disturbances have a significant influence on psychiatric disorders, case formulations typically do not include a discussion or exploration of brain abnormalities. For most of its history, psychiatry has not had the tools to explore brain function. Although neuroanatomical and physiological approaches to understanding mental illness are burgeoning, they remain the province of psychiatric research. An essential question that the psychiatric community faces is how to best incorporate these new tools into clinical practice.
Residency education can play an important role in answering this question. Psychiatric training programs can facilitate the assimilation of neuroanatomical, physiological, and psychological testing into clinical practice by promoting an understanding of and familiarity with these emerging technologies (1). The psychiatry residency training program at Columbia University has addressed this goal with a pilot project in which psychiatric residents perform extensive neurobiological evaluations of patients. This program pairs one or more second year residents with a faculty mentor and together they select a patient with an Axis I diagnosis to undergo neuroanatomical, neurophysiological, and neuropsychological tests. The residents then propose a neurobiological formulation that integrates data collected about the patient with the existing literature available on the neurophysiology of the Axis I condition present in the patient.
In this article, we provide an approach and format for constructing neurobiological formulations. Thus far five such formulations have been presented by residents to staff at Columbia University. Based on these presentations we gathered data investigating the impact these presentations had on psychiatry residents. Following the paper, we include an appendix with an example of a full neurobiological formulation presented by a resident.
Residents meet weekly with faculty mentors for 4 months to complete a neurobiological case formulation. When finished, each resident presents his case to the department faculty for discussion, allowing for feedback both on the clinical material and on the neurobiological formulation.
Before beginning the formulation, residents and faculty mentors meet to identify a patient who is on an inpatient unit, has an Axis I diagnosis, no known neurological conditions, and is able to sign an informed consent to take part in this IRB approved project. The absence of a neurological condition was assessed through the patient’s history and physical examination. We tried to exclude patients with known neurological condition because the project aimed to explore neurobiological models of psychiatric disorders, rather than exploring psychiatric manifestations of neurological conditions. If however over the course of developing the neurobiological formulation, a neurological condition was uncovered, we would not exclude the patient from the project.
The residents then review the literature on the neurobiology of the patient’s Axis I diagnosis. Once the residents have found salient articles, they review them with the faculty mentor, assessing the quality and relevance of the studies. Residents attempt to incorporate some of the techniques of evidence-based medicine (2, 3). For example, together with faculty mentor, residents can assess the reliability of the outcome measures used in the study, the relevance of the studied population to a clinical sample, and determine if the study’s findings have been replicated. By reviewing the literature, the resident and mentor compile a battery of tests that may reveal underlying brain abnormalities in the patient.
A neurobiological formulation consists of 1) a Clinical Summary, which reviews the patient’s history, symptoms, and diagnoses; 2) Neurobiological Data, which presents the relevant findings from neuroanatomical, neurophysiological, and cognitive testing; and 3) a Case Formulation that integrates the patient’s clinical presentation and history with the neurobiological data.
The case formulation suggests possible relationships between phenomenology and brain abnormalities. Similar to a psychodynamic formulation where a developmental history helps one appreciate a patient’s current symptoms, the neurobiological formulation uses neurobiological data to help clinicians conceptualize symptoms. We stress to residents that a neurobiological approach neither establishes the etiology of symptoms nor confirms diagnoses. It uses a cross-sectional approach that cannot test etiological hypotheses in a controlled fashion; it cannot establish the etiology of observed or historical signs and symptoms and therefore cannot confirm diagnostic impressions. The value of case formulations is that they require residents to integrate complex findings, to generate hypotheses regarding a patient’s condition, and provide a framework for predicting treatment outcomes (4). A neurobiological formulation can also yield greater appreciation for the role of neuroscience in clinical psychiatry. In addition to its value in residency education, it establishes a new model for presenting cases in the literature by emphasizing the integration of neurobiological approaches to understanding and conceptualizing clinical phenomena.
When developing this project, we were concerned that patients might have no evidence of brain abnormalities in spite of a detailed examination. Until neurobiological formulations are systematically assessed and reported, the prevalence of detectable neurobiological abnormalities must be considered unknown. While all of the cases we have examined have evidenced neurobiological abnormalities, a neurobiological formulation could be constructed in the absence of neurobiological findings. In this case, residents would present the prevailing neurobiological model of the Axis I condition diagnosed in the patient and then generate hypotheses explaining why data could not be found to support this model. This could include a discussion of the limited sensitivity of the diagnostic tests used, shortcomings in the neurobiological models proposed in the literature, or a revised diagnostic formulation given the normal neurobiological assessment.
Five neurobiological formulations have been presented to faculty and residents at Columbia University. After the presentations, a survey was conducted among residents attending (all of the residents surveyed had attended at least two of the presentations) to assess their impressions of this pilot program. The 15 residents surveyed were asked if listening to and discussing the neurobiological formulations increased their knowledge of available neuroanatomical, neuropsychological and /or neurophysiological instruments. On a Likert scale (in which 1 = not at all and 7 = greatly), the residents’ mean response was 5.87 (SD +/− 0.92). Residents were then asked if the formulations increased their sense of the relevance of neuroanatomical, neuropsychological, and/or neurophysiological information to patient assessment and treatment. The mean response was 5.60 (SD +/− 0.99). Lastly residents were asked if the project convinced them that an etiological formulation of psychiatric illness is inadequate if reference to brain function is not included. The mean response was 5.27 (SD +/− 1.28).
Residents were further asked if they felt that the project should be continued and if so, how often they wanted to participate in neurobiological evaluations. Residents overwhelmingly (100%) were in favor of maintaining neurobiological formulations in the curriculum with most residents indicating that it should be done at least once yearly.
Comments from residents include: “[the neurobiological formulations] were a very applicable and educational experience,” “very good at stimulating interest” and “helped me to understand how a neurobiological viewpoint can help in understanding a patient’s illness.” Also the presentations provided “a structure in which to frame and conceptualize [neurobiological] material and make it possible to begin to utilize it.”
In sum, the neurobiological formulations have been well received by residents as an educational experience. Moreover, we suspect that they contribute to residents’ knowledge of neuroscience, but follow-up studies are needed to assess this.
Over the course of the next decade, neuroanatomical and neurophysiological testing will likely be incorporated into psychiatric practice. Current psychiatric residents will benefit from developing an understanding of these modalities and may help facilitate making these tools, currently used almost exclusively in research, more applicable to clinical practice. We presented an approach and format for constructing neurobiological formulations. To clarify our concept of a neurobiological formulation, we provide an appendix with a complete neurobiological case formulation presented by a resident.
A total of five neurobiological formulations have been presented to staff at Columbia University and residents were surveyed to assess their interest in these cases and their impressions of the relevance of neurobiological assessments to clinical practice and resident education. The results suggest that residents found the neurobiological assessments interesting, helpful in expanding their knowledge of neurobiological testing, and worth incorporating into residency education.
“B” is a 55-year-old male with a long history of depression and anxiety (the patient signed a consent form approved by the Institutional Review Board of the New York State Psychiatric Institute regarding all procedures included in this study; some details of the patient’s history have been changed to protect confidentiality). He has attempted suicide twice and was transferred to the New York State Psychiatric Institute (NYSPI) in the winter of 2005 from the intensive care unit (ICU) of a local community hospital after being medically treated for an overdose with a tricyclic antidepressant. While in the ICU, the patient had been comatose for nearly 48 hours, suffered aspiration pneumonia, and cardiac conduction abnormalities. Ten days prior to the overdose, B had been discharged from NYSPI, where he had been hospitalized for depression and suicidal ideation.
B was born and raised in a large metropolitan area. He was enrolled in a gifted program in elementary school and described himself as popular among classmates, athletic, and a good student. He reported obtaining a combined score of 1300 on his SAT examination and was offered an academic scholarship to a prestigious university. He had a stable career, working throughout his adult life at a community not-for-profit organization.
B has a family history significant for depression in his mother, who was depressed most of her adult life and committed suicide when B was in his early twenties. Soon after her death, B developed symptoms of dysthymia. He was treated successfully with fluoxetine at age 40 and remained asymptomatic for over 12 years. B developed a major depressive episode in his early fifties in the setting of financial difficulties and was hospitalized briefly. He was treated with venlafaxine, responded well, and was symptom free until he developed his second major depressive episode shortly after divorcing his wife at age 54. He became actively suicidal and was hospitalized after overdosing on lorazepam (his first hospitalization to NYSPI). During this hospitalization, he responded well to nortriptyline; however, 10 days following discharge, he attempted suicide by overdosing on this medication.
Admitting Mental Status Examination
On the initial interview, B endorsed prominent anxiety, anhedonia, poor concentration, feelings of hopelessness, insomnia and active suicidal ideation to overdose on his medications. He described his anxiety as an “unrelenting sense of dread” that seemed to come and go throughout the day in reaction to minor stressors. He denied symptoms of panic disorder, obsessive compulsive disorder, posttraumatic stress disorder, or other anxiety disorders.
His affect was constricted and dysphoric, but he was articulate, cooperative, and forth-coming. His Folstein Mini-Mental State Exam (MMSE) score was 29 (of 30). He appeared to be of above-average intelligence.
The patient’s second hospitalization at NYSPI lasted nearly 3 months. His hospital course was marked with brief periods of improvement where he appeared euthymic with nearly all symptoms remitting. These periods were followed by precipitous declines in mood following relatively mild social stressors. For example, prior to a holiday, the patient scored less than 8 on Hamilton Depression and Anxiety Scales; however, the patient felt that his family had not spent sufficient time with him during the holiday and soon thereafter, he relapsed with severe anxiety and thoughts of suicide. This pattern recurred several times over the course of his hospitalization. He was eventually discharged to a day treatment program on a regimen of escitalopram, bupropion, clonazepam, and quetiapine. At the time of discharge, his thoughts of suicide had remitted but he continued to have intermittent bouts of severe anxiety.
Two head MRI studies were obtained. The first was obtained during the patient’s first hospitalization at NYSPI (prior to the overdose of nortriptyline) and the second was obtained during his second hospitalization at NYSPI. Both MRI studies revealed bilateral atrophy of the dorsal prefrontal cortex. The atrophy was well circumscribed; there was little to no evidence of atrophy in the other cerebral lobes and his ventricular volumes were normal. (See Figure 1 and Figure 2)
B’s overall IQ score (WAIS–Third Edition (WAIS-III)) was in the normal range; however, his verbal and performance scores differed with a verbal score in the 98th percentile and a performance score in the 5th percentile. (See Figure 3).
Translating these findings into anatomical terms, we suspected that B’s performance IQ deficits were related to his right cerebral hemisphere. This hypothesis was based on abundant data suggesting that in right-handed individuals visual-spatial abilities are subserved by the right hemisphere, whereas verbal processes are subserved by the left hemisphere. We gathered further support for this hypothesis by assessing B’s memory with the Wechsler memory test and the Benton Visual retention test as shown in Figure 4.
Similar to his performance on the IQ test, B demonstrated competency on the left hemispheric tasks measured on the Wechsler Verbal Memory Scale, but his performance was impaired on the right hemispheric visual-spatial skills tested on the Benton Visual Retention Test.
To further investigate the asymmetry in B’s left versus right hemispheric functioning, we employed dichotic listening tasks. Using dichotic tones, we found a deficiency in left tonal processing (consistent with a right hemispheric deficit) that was 3 standard deviations below normal. Similar to the IQ and memory assessments, this dichotic listening task suggested a marked deficit in right hemispheric functioning.
The two most prominent findings from our neurobiological assessment of B were: a) bilateral dorsal prefrontal atrophy on MRI; and b) impaired right hemispheric function on neuropsychological and physiological testing. We will first address the MRI findings and then discuss the neuropsychological testing.
LeDoux (5) suggested that fear is subserved by two distinct neural pathways. The first pathway, anchored in the amygdala, provides the CNS a quick, but imprecise, means of responding to dangerous stimuli. The second pathway, centered in the primary sensory cortices, association cortices, and prefrontal cortex, is slower to react, but more accurate in assessing stimuli. When these two pathways work together, the prefrontal cortex can inhibit activation in the amygdala in response to benign stimuli (5). Thus, the prefrontal cortex is thought to catch false alarms activated by the amygdala, restoring affective equilibrium by reducing fear and anxiety.
B’s head MRI demonstrated substantial prefrontal cortical atrophy. Based on the models of fear described above, impairments in prefrontal function can alter and inhibit the brain’s capacity to regulate fear and anxiety. This loss of prefrontal function was evident in B’s clinical presentation; he seemed to have a marked inability to regulate his anxiety when faced with environmental stressors. In neurological terms, we suspect that B’s amygdala-based fear pathway was functioning normally, but his damaged prefrontal cortices impaired his ability to regulate anxiety. In theory, cognitive therapy provides patients with a greater capacity to modulate subcortical affective systems through prefrontal inhibition (6). In the short-term psychotherapy done with B during his hospitalization, he demonstrated little capacity to modulate his mood through cognitive interventions. We suspect that B’s damaged prefrontal cortex impaired this process, resulting in the amygdala-based fear system operating without inhibition.
An association has long been noted between cerebral asymmetry and mood disorders (7–9). This association was first described in stroke victims. Patients with left frontal strokes were noted to be at a much greater risk for depressive reactions than patients with right frontal strokes (8, 10). EEG studies have similarly noted that patients with mood disorders have greater right versus left frontal activation both at baseline and during stressful events (7). These findings seem at odds with our patient’s presentation where his functional deficits were localized to the right hemisphere. However, investigators have suggested that the apparent association between left frontal deficits and mood disorders may be more complicated than has been presumed (11). Indeed, lesions of the left frontal lobe may lead to disinhibition of subcortical structures or even disinhibition of contralateral structures. Available data on the relationship between cerebral asymmetry and mood disorders suggest that marked differences in the functioning of the two cerebral hemispheres places individuals at risk for mood disturbances (8) regardless of the direction of the asymmetry. The mechanisms by which asymmetry contributes to depression are unknown; however, investigators suggest that it may be similar to the loss of frontal inhibition implicated in anxiety, as described above (12, 13). Frontal asymmetry leads to disinhibition of subcortical affective structure, and in patients predisposed to depression, such as B, this can lead to an exacerbation of depressive symptoms (14, 15).
We do not know the etiology or date of onset of B’s abnormal neuropsychological and neuroanatomical findings, but given his history of strong academic performances as a child and stable employment as an adult, it seems unlikely that his cognitive impairments were present throughout his life. More likely, B suffered an insult to his frontal lobes during adulthood. With ample cognitive reserves in his left hemisphere he was able to compensate for the damage to his left frontal lobe; however, lacking these reserves in his right hemisphere, right hemispheric deficits became apparent.
Without a biopsy, the etiology of the patient’s frontal lobe damage cannot be established. Head trauma is a possible cause; the patient denied any history of trauma, but given his cognitive impairments this cannot be fully excluded. Another possible etiology is an idiopathic fronto-temporal lobe dementia (FTD). Early presentations of FTD can include prominent mood disturbances and frontal lobe atrophy. (The temporal lobes can remain spared until a later course in the disease.) If this were the case, follow up neuroimaging should demonstrate temporal lobe involvement. A third possibility is that B was demonstrating long-term manifestations of his mood disorder. Data exist on the short-term effects of mood disorders on brain structures, but the neurobiological effects of a mood disorder over a 30-year period have not been explored in controlled, longitudinal studies. Nevertheless, several investigators have suggested that mood disorders accelerate cerebral atrophy, possibly secondary to increased stress hormones, diminished neuroprotective cytokines, and impaired neurogenesis (16, 17).