LAUSR

After decades of psychiatric neuroscience research, we have not made enough progress in offering causal explanations for disturbed experience and behavior, diagnosing true-types of psychiatric disorders, or choosing rational treatments.1 As a result, our field often reverts to utilitarian approaches that propose identifying measurable neurobiological abnormalities among people with psychiatric disorders as a potential basis for a diagnosis and treatment, regardless of causation. But these approaches fare poorly because differences between patients and psychiatrically healthy individuals are quite modest, similar brain abnormalities are seen across disorders, and there is great heterogeneity within patient populations. There are exciting examples of brain abnormalities related to psychopathology (eg, in the amygdala in anxiety disorders and in the nucleus accumbens in substance use disorders), but these are not found among all individuals with similar presentations. They do not provide us with an adequate explanation of why certain individuals get sick and why they get their specific symptoms. More broadly, these research leads do not explain salient features of psychiatric disorders such as why most patients have a relapsing-remitting course of illness despite chronic underlying abnormalities in brain and why many patients present with mixed clinical pictures, which often include psychotic, mood, anxiety, and addiction symptoms. We need to approach psychiatric neuroscience differently to answer these questions. We will discuss recent advances in several fields outside of psychiatry that may inform our thinking. Two shared themes are that dynamic complex systems have properties that need to be understood at the system level—not at the level of individual elements—and that time is a critical dimension in understanding the behavior of complex systems. The brain is one such complex system; its interactions with the world are characterized by continually receiving and processing information and appropriately directing action. A healthy brain responds effectively to its environment and outputs.2 Thus, health is not a state but a process. Additionally, the brain has multiple feedback and compensatory mechanisms that maintain important features of the system within appropriate ranges, even in the face of disruptive internal or environmental factors. This ability to return important parameters to previously set ranges is a form of resilience. If these mechanisms are disrupted in an individual brain, as they are in psychiatric disorders, one would expect to see instability in multiple interconnected systemic features and not an abnormality in just 1 parameter.3 In biological psychiatry research, we often focus on measuring specific parameters, such as mRNA and protein levels; blood oxygen level–dependent signals in specific brain regions; or performance on cognitive tests. But we ignore the fact that the brain works hard to maintain important parameters within healthy ranges. If an important parameter has become abnormal, this signals a significant disruption that has overwhelmed compensatory mechanisms. In this case, the measured abnormality is more likely to be the result of the disease process than its cause. In fact, we know that the brain maintains critical parameters in a normal range by sacrificing other less important parameters in the process even in physiological extremes. Catastrophic collapse occurs when this is no longer possible.4 We might obtain deeper insight into brain abnormalities in psychiatric disorders by developing models of how the brain as a system evolves in a balanced and regulated manner, what forces drive it out of balance, and what the system does to compensate. What concepts and tools can investigators use to study complex system behavior in psychiatric neuroscience? Here, we introduce a series of Viewpoints to stimulate thought in addressing that question. The Viewpoint in this month’s issue of JAMA Psychiatry by McEwen5 concerns allostasis. Allostasis refers to the adaptive physiologic processes that are activated by novel, often stressful, experiences. When physiologic parameters are forced out of their usual ranges by environmental perturbations, the body responds with homeostatic mechanisms to restore them. As physiologic challenges mount, homeostasis becomes more and more difficult to maintain and the system may need to switch to an “allostatic” state: a state of dysregulated activity, indeed a disease state, that is nonetheless capable of stability in the face of the environmental demands. McEwen5 reviews the concept of allostasis as it applies to neuroscience and discusses biological mediators that play a role in the emergence of an allostatic overload in psychiatric disorders. In future issues, we plan to publish Viewpoints that provide a different but complementary perspective. One will concern mathematical models of complex systems and describe concepts such as metastable states and winnerless competition. Metastable states occur when a set of brain activity parameters different than the equilibrium state persist during information processing. These can contribute to winnerless competition, in which the characteristic of the complex system is not a stable state but rather a continual change based on internal dynamics and environmental inputs. Another Viewpoint will focus on the concept of “network medicine,” which characterizes disease states as perturbations in physiological networks as opposed to manifestations of a specific structural or molecular lesion. This approach highlights the nature of disease as a property of the entire system that can only be treated by modifying the abnormal system properties. This Opinion