1a
Joseph E. Beeney
One of my favorite graduate school professors liked to ask a provocative question about the increasing emphasis placed on neuroscience in Psychiatry and Clinical Psychology: “What from all of this work has influenced our clinical practice at all?” A nice variation of this question was, “What has clinical neuroscience produced that a therapist in the community could use?” These were meant to be rhetorical questions. I think the answer was supposed to be “nothing.” Still, I took the question as a serious, if humbling, challenge. Neuroscience is a fascinating, exciting field in its own right. Yet, I would guess for most clinical researchers, when we wrap up our careers and look back on our contributions, we want to see some work that is unambiguously relevant to clinical practice. So, “How does this inform clinical practice at all?” is a refrain to which I return. The chapter by Chan, Vaccaro, Rose, Kessler, and Hazlett (this volume) on “Neuroimaging in Personality Disorders” provides an opportunity for each of us to use this lens to interpret what the field has produced so far in the clinical neuroscience of personality disorders (PDs). At the same time, it is a chance to note the limitations of our approaches as an opening to find methods that might bring us closer to unambiguous relevance for clinical practice.
The National Institute of Mental Health (NIMH) declared an historic shift prior to the new millennium by calling the 1990s the “decade of the brain” – a development my graduate professor found irksome. However, in the 1990s, researchers began to develop an understanding of the brain that should matter a great deal to clinical researchers. During the decade, researchers began to understand that the brain is much more plastic and malleable than previously thought. Our experiences, even in adulthood, shape our neurocircuitry (Holtmaat & Svoboda, 2009) and these neural alterations are partly the result of epigenetic changes in gene expression (Houston et al., 2013). For researchers focused on identifying risk factors, prevention, or treatment approaches, these findings seem to place the human brain as the critical mediator between experience and psychopathology or health. As a consequence, there has been an enormous increase in structural and functional MRI studies. And at the same time, it can be difficult to see the clinical relevance of all of this work. Some even refer to these efforts as phrenology with better pictures (e.g., Uttal, 2002).
So, what might a neuroscience of PDs that informs clinical practice look like? It might tell us something about the mechanisms of the disorder, help us to carve out the boundaries between and commonalities across disorders, or maybe tell us how medication or psychotherapy work. The biggest obstacle for a clinically relevant neuroscience study is less-than-ideal research design. Study design is critical in how clearly we can interpret results. Poor design is likely much more frequently a function of practical issues (e.g., recruitment, limited funding), rather than researcher skill. Still, clinically relevant findings can be derived from studies that are less than perfect, if they are placed in the context of a network of findings. In addition, studies using network-based approaches and performance-based tasks can help with the interpretation that leads to clinical relevance.
Can Studies of Structural Morphology Be Clinically Relevant?
Identifying the neural mechanisms specific to a disorder that are distinct from other disorders is a major goal of clinical neuroscience. Yet, few studies provide unambiguous information on mechanisms specific to PDs. Researchers commonly employ a design in which people diagnosed with a PD are compared to people with no mental health difficulties. With multiple studies, this design can help generate a picture of differences in brain structure. But these are differences between people with no mental health difficulties and a group that is distinct in endless ways, in addition to having a specific clinical diagnosis. Drawing conclusions regarding specific abnormalities from such studies is problematic.
For instance, as Chan et al. (this volume) note, abnormal temporal lobe volumes may play a role in schizotypal PD (SPD) and psychotic symptoms. At the same time, it is unclear whether these differences are actually due to abnormalities specific to SPD, or simply having any mental health diagnosis or experience common to many mental health disorders (e.g., trauma). On the other hand, some of the studies the authors review that compare diagnostic near neighbors provide more compelling information, though these studies also have limitations. Several reports have compared people with SPD and schizophrenia, generally finding, on average, that there are differences between these two groups in terms of temporal lobe brain volumes, and that specific symptoms are related to temporal lobe abnormalities. In terms of brain structure, SPD resembles a less severe version of schizophrenia. However, it is possible other disorders may resemble a less severe version of schizophrenia when compared in this way. Temporal lobe abnormalities have been found in numerous disorders including autism, attention deficit/hyperactivity disorder, and psychopathy (Calhoun, Maciejewski, Pearlson, & Kiehl, 2008; Kobel et al., 2010; Lombardo et al., 2010).
Providing more context to these findings seems to provide a more compelling picture. We may know little that’s definitive, but the network of studies focusing on different levels of inquiry (e.g., genes, brain structure and function, phenomenology) does seem to be coherent. SPD and schizophrenia appear to share genetic risk. Similar brain pathology is observed in both disorders. Symptoms are similar in both disorders and the degree of temporal lobe abnormality is commonly associated with severity of symptoms (e.g., odd speech; Rosell, Futterman, McMaster, & Siever, 2014). In other words, combining multiple sources of information seems to support the idea of a schizophrenia spectrum in which the temporal lobe is critically involved. Even better evidence for the role of the temporal lobe in schizophrenia-spectrum symptoms could be garnered from studies that compare SPD and schizophrenia with other severe disorders (e.g., bipolar disorder, borderline personality disorder). Studies comparing severe disorders could rule out the possibility that severity of mental health diagnosis is driving the results (some studies already show differences between schizophrenia and bipolar disorder; e.g., Calhoun et al., 2008).
A network of results based in genetics, brain structure, and phenomenology provides clinically relevant information about diagnosis and prognosis. These results indicate that though a categorical boundary between SPD and schizophrenia may be useful for communicating severity, these are disorders that may be differentiated only by degrees of severity. In addition, as a starting point, these results suggest that the pharmacological and psychosocial interventions that are effective for treating other schizophrenia-spectrum disorders may be relevant for SPD. This is particularly important because there are is little research on treatment for SPD (Rosell et al., 2014).
Network Approaches to Functional Neuroimaging
Up until the past decade, studies of functional brain activation tended to focus on differences in neural activation for specific brain regions. Recently, as Chan et al. (this volume) note, researchers have emphasized neural networks over regions (Bressler & Menon, 2010; Bullmore, Bullmore, Sporns, & Sporns, 2009). Research suggests the idea that a brain region performs a specific cognitive function in isolation is untenable. Rather, the brain is organized into distinct but interacting networks (Sporns, 2012). A single region may support different aspects of cognition or emotion depending on the network with which it is coactive. Ideally, such a shift in perspective and analysis could aid in the ongoing problem of interpreting functional brain activation. Activation of the amygdala may be related to any of the multitude of psychological functions the amygdala has been found to participate in. At the same time, if evidence generated from neural connectivity models indicates that amygdala activation co-occurs with that of the anterior insula, dorsal anterior cingulate, ventral striatum, and medial temporal lobe (e.g., hippocampus), this coactivation constrains our interpretation somewhat. These regions are part of the salience network (Menon, 2015), a neural network that filters and amplifies stimuli that has biological or learned value. Thus, amygdala activation in this instance might plausibly be interpreted as contributing to this process.
As indicated by Chan et al. (this volume), most fMRI studies of PDs have used a regional approach; yet, studies that did not necessarily focus on neural networks, or use a connectivity-based approach, have provided results highly suggestive of the activation of a neural network. The evidence is imperfect, but the combination of regional activations, task, and design can decrease the ambiguity of our interpretations. Again, the integration of multiple sources of data is helpful for this goal. For instance, Silbersweig and colleagues (Silbersweig et al., 2007) found that individuals diagnosed with BPD evidenced less activation in the ventromedial PFC (vmPFC) and increased activation in the limbic and reward systems when trying to inhibit behavior, particularly in the presence of negative emotional stimuli. The vmPFC and its connections to limbic and reward systems supports social and emotional behavior. Other studies have found this circuit is central to tasks in which participants need to inhibit a proponent response in the face of emotional stimuli (e.g., Kanske & Kotz, 2011). Among non-clinical participants, inhibiting such responses appears to be due to increased connectivity between the vmPFC and amygdala and down-regulation of amygdala activity. Individuals diagnosed with BPD appear to struggle with this process, made clear because the researchers used a performance-based task in which the people with BPD do not perform as well. Mixing these behavioral results with the neuroimaging data, the Silbersweig study suggests that negative emotion is detrimental to response inhibition because individuals with BPD fail to regulate limbic and reward systems in the face of emotional stimuli. Again, this finding could be used to inform treatment. The context that is provided both by the activation of a relevant neural network, along with the context provided by well-designed performance-based tasks, aids in the interpretation of these findings and increases their clinical relevance.
New Nosology, New Neuroscience?
The diagnostic definition of personality disorders is also currently undergoing revision (Bender, Morey, & Skodol, 2011). The promise of any new diagnostic system is greater diagnostic clarity that can aid in clinical relevance. Researchers have often pointed to the limitations of our classification models as a major impediment to translating advances in neuroscience to the clinical realm. There is currently extensive empirical work being done to develop a nosology of personality disorders (and mental health disorders, generally) that do not have the same problems of within-diagnosis heterogeneity, comorbidity across disorders, low reliability and artificial categorical structure of the current system (Hopwood et al., 2011; Insel et al., 2010; Kotov et al., 2017). The promise of such efforts is that we may have better ability to understand the neural mechanisms of symptoms that are present in multiple disorders in the current system (e.g., impulsivity), but are not always present within two people with the same disorder.
Chan et al. (this volume) focus their chapter on categorical diagnosis, likely because the field has yet to shift in a substantial way towards a more dimensional approach to personality disorders. At the same time, perhaps we should begin to envision what neuroscience research might look with a more dimensional approach to PDs. This could include questioning how a dimensional view affects interpretability of results to focus on transdiagnostic symptoms and the neural circuits from which these symptoms manifest, as suggested by the Research Domain Criteria (RDoC; Insel et al., 2010) approach promoted by NIMH. Moreover, this raises important questions about how a review of relevant research regarding symptom domains and brain circuits from an RDoC point of view might provide different understanding of more basic phenomena and/or clinical implications compared to that provided from a categorical perspective.
The alternative DSM-5 model for personality disorders is also an effort to move to a more dimensional diagnostic approach. The alternative model prescribes a dimensional assessment of self (identity and self-direction) and interpersonal disturbance (empathy and intimacy), followed by evaluation of pathological personality traits (negative affectivity, detachment, antagonism, disinhibition, and psychoticism). Afterward, a categorical diagnosis is determined (antisocial, avoidant, borderline, narcissistic, obsessive-compulsive, schizotypal, or personality disorder-trait specified). This model combines a determination of severity of dysfunction (self and interpersonal dysfunction) with assessment of extreme elevations on the five-factor model of personality. Each of these constructs and diagnoses has at least a small body of neuroscience literature. It is important to evaluate whether our current literature suggests such an approach will bring greater clarity and clinical utility to our structural and functional neuroscience of PDs. At the same time, there seems to be an opportunity to consider how the existing neuroscience of PDs can inform this model.
Conclusion
The field of clinical neuroscience continues to advance in ways that may yield exciting new understanding of the character of personality disorders and effective approaches to treatment. Statisticians are rapidly developing new methods and tools for conducting connectivity analyses (e.g., Gates, Molenaar, Hillary, & Slobounov, 2011; Wang, Zuo, & He, 2010). Some researchers are using longitudinal designs to map neural and symptom changes over time, while others are identifying neural changes following successful psychotherapy. At the same time, these and other advances offer the possibility of improved interpretability of our findings, which generally translates into increased clinical relevance. We may not win over my skeptical professor, but we increase the possibility he may find something in the neuroscience literature that is directly applicable to his psychotherapy practice.
References
Bender, D. S., Morey, L. C., & Skodol, A. E. (2011). Toward a model for assessing level of personality functioning in DSM-5, part I: A review of theory and methods. Journal of Personality Assessment, 93(4), 332–346.
Bressler, S. L., & Menon, V. (2010). Large-scale brain networks in cognition: Emerging methods and principles. Trends in Cognitive Sciences, 14(6), 277–290.
Bullmore, E., Bullmore, E., Sporns, O., & Sporns, O. (2009). Complex brain networks: Graph theoretical analysis of structural and functional systems. Nature Reviews Neuroscience, 10(3), 186–198.
Calhoun, V. D., Maciejewski, P. K., Pearlson, G. D., & Kiehl, K. A. (2008). Temporal lobe and “default” hemodynamic brain modes discriminate between schizophrenia and bipolar disorder. Human Brain Mapping, 29(11), 1265–1275.
Gates, K. M., Molenaar, P. C. M., Hillary, F. G., & Slobounov, S. (2011). Extended unified SEM approach for modeling event-related fMRI data. NeuroImage, 54(2), 1151–1158.
Holtmaat, A., & Svoboda, K. (2009). Experience-dependent structural synaptic plasticity in the mammalian brain. Nature Reviews Neuroscience, 10(9), 647–658.
Hopwood, C. J., Malone, J. C., Ansell, E. B., Sanislow, C. A., Grilo, C. M., McGlashan, T. H., … Morey, L. C. (2011). Personality assessment in DSM-5: Empirical support for rating severity, style, and traits. Journal of Personality Disorders, 25(3), 305–320.
Houston, I., Peter, C. J., Mitchell, A., Straubhaar, J., Rogaev, E., & Akbarian, S. (2013). Epigenetics in the human brain. Neuropsychopharmacology, 38(1), 183–197.
Insel, T., Cuthbert, B., Garvey, M., Heinssen, R., Pine, D. S., Quinn, K., … Wang, P. (2010). Research Domain Criteria (RDoC): Toward a new classification framework for research on mental disorders. American Journal of Psychiatry, 167(7), 748–751.
Kanske, P., & Kotz, S. A. (2011). Conflict processing is modulated by positive emotion: ERP data from a flanker task. Behavioural Brain Research, 219(2), 382–386.
Kobel, M., Bechtel, N., Specht, K., Klarhöfer, M., Weber, P., Scheffler, K., … Penner, I. K. (2010). Structural and functional imaging approaches in attention deficit/hyperactivity disorder: Does the temporal lobe play a key role? Psychiatry Research: Neuroimaging, 183(3), 230–236.
Kotov, R., Krueger, R. F., Watson, D., Achenbach, T. M., Althoff, R. R., Bagby, R. M., … Zimmerman, M. (2017). The Hierarchical Taxonomy of Psychopathology (HiTOP): A dimensional alternative to traditional nosologies. Journal of Abnormal Psychology, 126(4), 454–477.
Lombardo, M. V., Chakrabarti, B., Bullmore, E. T., Sadek, S. A., Pasco, G., Wheelwright, S. J., … Baron-Cohen, S. (2010). Atypical neural self-representation in autism. Brain, 133, 611–624.
Menon, V. (2015). Salience network. In A. Toga (ed.), Brain Mapping: An Encyclopedic Reference (Vol. 2, pp. 597–611). New York: Academic Press.
Rosell, D. R., Futterman, S. E., McMaster, A., & Siever, L. J. (2014). Schizotypal personality disorder: A current review. Current Psychiatry Reports, 16: 452.
Silbersweig, D., Clarkin, J. F., Goldstein, M., Kernberg, O. F., Tuescher, O., Levy, K. N., … Stern, E. (2007). Failure of frontolimbic inhibitory function in the context of negative emotion in borderline personality disorder. American Journal of Psychiatry, 164(12), 1832–1841.
Sporns, O. (2012). From simple graphs to the connectome: networks in neuroimaging. NeuroImage, 62(2), 881–886.
Uttal, W. R. (2002). Précis of the new phrenology: The limits of localizing cognitive processes in the brain. Brain and Mind, 3(2), 221–228.
Wang, J., Zuo, X., & He, Y. (2010). Graph-based network analysis of resting-state functional MRI. Frontiers in Systems Neuroscience, 4, 16.