July 2020, Volume XXXIV, Number 4
The P Factor
A new framework for assessing mental health
inding ways for accurate and consistent diagnosis in mental health has been sought since the emergence of self-awareness. For millennia we have struggled to understand who we are and how our internal experience matches with the reality of the external world. We work to describe the complexity of this experience so others can relate, understand, and interact. At times, our internal state or interactions with the external world do not match our expectations, which becomes distressing. This can lead to dysphoria and to cycles and behaviors we consider abnormal. We seek out care in the hope of understanding and intervention, with the goal of improved feelings and a return to health. This is the practice of psychiatry and medicine in general.
The practice of medicine has evolved over thousands of years, driven by a well-defined process of description, measurement, diagnosis, and intervention. Over time, science has sought to establish systems to standardize the description of disease and identify potential solutions or treatments. This is especially true regarding the subjective nature of the dysphoria described above. Chinese historical records dating back to 1100 BCE show motivation for the classification of mental health problems. These records included detailed observations of behaviors and emotions, grouping them into categories for study and treatment. The ancient Greeks attempted to explain possible causes of notable afflictions with the goal of testing ways of healing them. Kraepelin, Bleuler, and Freud all crafted theories around symptom clusters and drives. These explanations were guided by our early, and sometimes erroneous, understanding of neurology and the brain. These pioneers were left feeling an incomplete understanding of the system and knew there were still many missing pieces of the puzzle.
The DSM and its limitations
In 1952, the American Psychiatric Association developed the Diagnostic and Statistical Manual of Mental Disorders (DSM). This groundbreaking manual harkened back to the work done in 1100 BCE to offer a taxonomy of symptoms grouped into diagnoses. The goal was to provide a common language for describing various mental health presentations in order to provide specificity for clinical care and future scientific discovery.
Half of individuals who meet diagnostic criteria for one disorder also meet criteria for a second disorder.
While highly successful in clinical care, it was well-known that, despite the DSM’s taxonomy and descriptions of multiple symptoms, the manual does not necessarily reflect groupings of underlying neuropathophysiology. A simple example: not all sadness is vegetative depression, just as not all chest pain is myocardial infarction. Studies of the DSM also reported significant overlap between different diagnoses, leading to ambiguity and potential misidentification of an illness, as well as to subsequent misapplication of a treatment. A study by Newmann et al. in the Journal of Abnormal Psychology proposed a “rule of 50%,” which states that half of individuals who meet diagnostic criteria for one disorder also meet criteria for a second disorder.
For example, anti-NMDA receptor encephalitis was classically diagnosed as schizophrenia, but now, due to advanced diagnostic techniques, is recognized as an auto-immune disorder that manifests as psychosis.
Another example is the childhood spectrum of disorders called PANDAS, which were often labeled as obsessive-compulsive disorder but are now are also thought to be related to strep infections and a possible auto-immune connection. While the symptoms of both of these disorders are consistent with the DSM’s descriptions, underlying etiology is developing clarity in some of the “OCD” disorders and not others, likely because they are different disorders, even though the symptoms are similar.
Providers clearly recognized a need for more precision in not only our language, but also our understanding of underlying physiology.
Comparisons to medicine
Historically, the successful practice of medicine has been dependent on the sophistication of the tools available. In some areas of medicine, problems can be described and measured easily due to their macroscopic nature. This allows for development of a reliable mechanistic understanding, which then leads to a clear rubric of steps required to address the problem. Other areas are based on microscopic findings. These topics are less visible initially, but as high-resolution microscopic tools have become available, medicine has once again developed understanding and suggested a rubric for these otherwise hidden problems.
While improvements in microstructure tools have provided immense discovery, scientists have also found disease to be more than just structural disruption. There are often processes functioning out of homeostasis. While these disruptions are still physical in nature, often chemical, the issue frequently arises from a change in the steady state of the system. These are measured via tests of levels, markers, and even genetic interactions.
Psychiatry and neurology are among a group of specialties where the substrate of study is both physical and functional, but also dependent on something else. Physical issues such as brain injuries are visible at the macroscopic level and variations in dopamine receptors related to psychosis are visible at the microscopic level. Sympathetic overdrive, as seen in post-traumatic stress disorder, is a disruption of functional homeostasis between sympathetic and parasympathetic nervous systems and can be seen at the homeostatic functional level, visible through changes in heart rate variability. These are all reliably measurable but also do not entirely explain the outcomes we observe. Something is still missing from our understanding.
The “g” and “p” factors
Modern science continues to search for definition of this missing piece. One area of study focuses on large-grained observational constructs. This is most notable in the study of general intelligence, or g factor. General measures of intelligence were developed to provide practitioners with a less-subjective measure of general cognitive brain function. This is used to predict when individuals may have a harder time navigating the world in areas such as education, work, and independence. This inspired a series of papers on a construct called the p factor. The intent of this factor was to provide a less-subjective assessment of the liability/sensitivity one may have to stressors over time. Caspi described this construct via conduct disorder in the Journal of Clinical Psychological Science, stating that “the propensity to persistent conduct disorder symptoms from adolescence to midlife is indicative of General Psychopathology rather than specific to an Externalizing style.” The authors suggest that, while individuals may have many other factors or behavioral patterns that determine one’s way of interacting with the world, there is a common factor that will determine whether they are at higher risk for future psychopathology and other diagnoses.
The difficult part of the p factor is that it doesn’t really diagnose.
The difficult part of the p factor is that it doesn’t really diagnose—more than anything else, it provides a likelihood for suffering. This lack of diagnosis is what makes it a bit provocative. It’s more a unifying factor for emotional sensitivity and for potential distress. On the positive side, it may predict when a normally effective set of “brain tools”—e.g., proclivities for managing situations in certain ways—prove to be ineffective under stress. In addition, the p factor suggests that all “disorders” are connected in a common way: if there is one, then there are often others.
Both the g factor and the p factor have been met with much skepticism. Many suggest that they are far too large-grained to be practical and lack predictive guidance. Others report that they are subject to tremendous biases in areas such as culture, education, and experience. The clear conclusion of many of the studies is that psychiatry needs additional tools to address the gap left by considering only the physical or functional aspects of the human nervous system. Assessment and understanding of how the nervous system functions at an informational level is vital to a more complete understanding of both the diagnosis and treatment of these complex and debilitating disorders.
Research Domain Criteria
Recognizing the need for translational study, The National Institute of Mental Health announced in 2009 a new research framework, called Research Domain Criteria (RDoC), intended to change the way mental health was studied and understood. RDoC integrates many levels of study (genomics, circuits, behavior, self-report) to ground the understanding of mental health across various dimensions. The goal is not to serve as a diagnostic guide or to replace current systems of diagnosis, but to provide a framework that can benefit diagnosis and intervention through a solid, translational, mechanistic footing.
While initially controversial, RDoC has now been in place for over 10 years and is beginning to provide recognizable benefits to the field. Many interesting data sets have been generated and findings produced, yielding insight into symptoms, diagnosis, and intervention. Yet we still lack a complete path from presentation to mechanism to intervention to health. RDoC is a complicated framework that has produced a tremendous amount of data. We continue to struggle with the sheer complexity of the data, our complex nervous system, its actions, and interactions between multiple nervous systems, which we call human behavior. Not only is this complexity difficult to navigate scientifically, but it also often produces far too much data to be practical for a clinical environment.
Navigating data and translation into understanding and practical clinical tooling for diagnosis and treatment is the next frontier. This is a new field called neuroinformatics. The raw computing power that has become recently available provides us with the capability to analyze and understand data at a level not previously available. Computational neuroscience, large data sets, and machine learning give us insights into how thought is produced and factors that relate, and even cause, large-scale changes in thought and behavior. There are now new ways to model and measure the information flow through complex nervous system components. It is through these models and new techniques that the next generation of diagnosis and intervention will be sought. The models become more human-like each day and provide us with easier ways to test theories and play out possible scenarios. We can produce probabilities of outcomes that increasingly predict future states accurately. We once again feel as if we have discovered a missing piece, if not the missing piece.
It is an exciting time for neuroscience research and mental health treatment in general, given the new understanding being gathered each day. It is important for us to remember the history of where we have been and how the mystery of the brain has driven us. Given the new computing power we can access, we are not only provided with new understanding but also profound responsibility, as groups such the AI for Good Foundation remind us. Ultimately it is our drive to understand ourselves and our relationship with the world that will provide us with answers and the ability to provide hope and relief from suffering.
© Minnesota Physician Publishing · All Rights Reserved. 2019