Entries in Paulus (1)



The disciplines I am most interested in are neuroscience and psychiatry, which are still distinct, though scientists hope to bridge the gap between these two fields. There are a number of neuroscience-based investigations conducted on psychiatric disease, but as Dr. Paulus mentions in a recent JAMA Psychiatry opinion piece, there is yet to exist a definitive biological explanation for any particular psychiatric illness. Yes, it is true that psychiatry deals with complex conditions, which are influenced by genetic and environmental factors exerting effects at the neuronal to systems levels.  Understandably, this makes it difficult to determine exactly what happens in different psychiatric illnesses. For medical conditions pertaining to other body systems (e.g., cardiovascular and infectious diseases), the biological determinants for disease manifestation and treatment efficacy are better understood. It would be great if psychiatric status could be examined in ways similar to monitoring for abnormal blood test results (e.g., CRP, HDL and LDL cholesterol, triglycerides, etc.) to minimize the chances of developing heart disease or MRI imaging to view a brightly-enhanced area likely to represent a tumor), but this approach to psychiatric diagnosis is too simplistic. Dr. Paulus believes that many scientific investigations focus on “mechanisms” and “mechanistic explanations” (admittedly vague expressions wildly thrown around in scientific publications) to study disease and that this way of conducting research is contributing to the rarity of biological breakthroughs in psychiatry.


In her opinion piece, Dr. Paulus also brings up the problem of reverse inference in neuroscience research. Yes, when discussing brain structure and function, you will frequently hear of a certain brain area being implicated in a particular behavior, emotion, affect, etc. Unfortunately, almost always secondary is the elaboration on the conditions in which the association was found. For example, if during resting state patients with a disease show increased activity in a brain region previously found to be active during tasks aimed at assessing reward-related behaviors, would you say there is heightened reward-associated activity in a disease? No, this is not necessarily true since brain areas are involved in multiple processes.


Dr. Paulus advocates for a prediction-based approach to solve clinical problems. However, is there sufficient information available on predictors to make estimates of diagnoses and prognoses for psychiatric illnesses, as she proposes to the neuropsychiatric research community?  I believe that there is need for computational predictive models to help determine the likelihood of developing a particular condition based on different factors, but we must also remain cautionary of an absolute utility of such models. How can a computational predictive model explain why a patient develops a disease despite expressing a neurobiological or genetic propensity for that disease?  What reasoning do you provide to the patient? Another vital piece of this puzzle is appropriate avenues for effective patient treatment.  How do you address the question of why something like CBT works for one patient but not another? What are the processes taking place in the brain leading to these outcomes? “Best practice” models for effective treatment are still subject to each patient’s unique biological makeup and environmental circumstances.

Furthermore, psychiatry is a field in which prescribed medications often have a variety of side effects, which can vary from person to person. Reliance on prediction models is unlikely to bring about novel understanding of disease or major improvements in therapeutic treatments. Neuroscience investigations should continue studying what exactly is altered in the brain of patients with a particular disease, and to what extent these changes are present. Research on!




Paulus, M.P. Pragmatism Instead of Mechanism: A Call for Impactful Biological Psychiatry. JAMA Psychiatry.2015; 72(7): 631-632.