Skip to main content

Transforming the understanding
and treatment of mental illnesses.

Celebrating 75 Years! Learn More >>

NAMHC Minutes of the 248th Meeting

February 9, 2017

Department of Health and Human Services
Public Health Service
National Institutes of Health
National Institute of Mental Health


The National Advisory Mental Health Council (NAMHC) convened its 248th meeting in closed session at 8:00 a.m. on February 9, 2017, at the Neuroscience Center in Rockville, Maryland, to review the report from the Board of Scientific Counselors (BSC), the group that provides scientific oversight for the NIMH Division of Intramural Research Programs (DIRP). This portion of the meeting recessed at approximately 8:50 a.m. The NAMHC then reconvened in open policy session at 9:00 a.m. and adjourned at approximately 12:45 p.m. In accordance with Public Law 92-463, the policy session was open to the public. The NAMHC then reconvened for a closed session to review grant applications at 2:00 p.m. until adjournment at approximately 5:00 p.m. (Appendix A: Review of Applications). Joshua Gordon, M.D., Ph.D., Director, National Institute of Mental Health (NIMH) presided.

Council Members Present at the Grant Review and/or Open Sessions

(Appendix B, Council Roster)


Joshua Gordon, M.D., Ph.D.

Executive Secretary

Jean Noronha, Ph.D.

Council Members

  • Patricia A. Areán, Ph.D.
  • Deanna M. Barch, Ph.D.
  • David A. Brent, M.D.
  • Benjamin G. Druss, M.D., M.P.H.
  • Hakon Heimer, M.S.
  • Michael F. Hogan, Ph.D.
  • John H. Krystal, M.D.
  • Marsha M. Linehan, Ph.D.
  • Maria A. Oquendo, M.D., Ph.D.
  • Rhonda Robinson-Beale, M.D.
  • Mary Jane Rotheram, Ph.D.
  • Hyong Un, M.D.
  • Christopher A. Walsh, M.D.

Ad Hoc Members

  • Tami D. Benton, M.D.
  • Randy D. Blakely, Ph.D.
  • Ian H. Gotlib, Ph.D.
  • Alan E. Greenberg, M.D., M.P.H.
  • David C. Henderson, M.D.
  • Lisa H. Jaycox, Ph.D.
  • Gregory A. Miller, Ph.D.
  • Elyn R. Saks, J.D., Ph.D.

Ex Officio Members

  • John W. Davison, M.B.A, Ph.D. Department of Defense

Liaison Representatives

  • Anne Mathews-Younes, Ed.D. Center for Mental Health Services, Substance Abuse and Mental Health Services Administration (SAMSHA)

NIMH Board of Scientific Counselors Present at the Open Policy Session

  • Edwin (Ted) Abel, Ph.D., Iowa Neuroscience Institute, University of Iowa Carver College of Medicine

Others Present at the Open Policy Session

  • A. David Redish Ph.D., University of Minnesota
  • Bryan Roth, M.D., Ph.D., University of North Carolina Chapel Hill
  • Cheryl King, Ph.D., University of Michigan

Open Policy Session Call to Order and Opening Remarks

Dr. Gordon called the open policy session to order and welcomed all in attendance. He noted that the policy session is open to the general public and the press, and is being videocast on the web to the NIH community and the public. After reviewing the agenda, Dr. Gordon said that eight council nominees for fiscal year 2017 are joining the Council for the first time and are serving as ad hoc members until they are officially on board. He also thanked the recent Council graduates who have kindly accepted NIMH’s request to stay on for this meeting to ensure appropriate representation. Members of the Council briefly introduced themselves and described their backgrounds.

Approval of Minutes of the Previous Council Meeting

Turning to the minutes of the September 2016 Council meeting, Dr. Gordon asked whether Council members had any comments, revisions, or questions. Receiving none, the Council unanimously passed the motion to approve the minutes.

NIMH Director’s Report

Joshua Gordon, M.D., Ph.D. Director, NIMH

Dr. Gordon reported NIMH news and said that after his update, Council members would hear from three extramural scientists who would provide their perspectives on opportunities and challenges in three areas that he has identified as priorities to address in his first few years as Director of NIMH. Dr. Gordon first reported on transition activities and said that NIMH looks forward to working with the new administration to ensure that the Institute can fulfill its mission of improving, through research, the plight of patients who are suffering from mental illness. NIMH has contributed to reports to the transition teams, and some of the Institute’s major priorities were highlighted as NIH-wide priorities to bring to the transition team. Highlighted priorities include the BRAIN Initiative  in basic neuroscience as well as suicide prevention as a clinical goal. Dr. Francis Collins has been invited to continue as NIH Director for an undetermined time and is guiding NIH through the transition.

Dr. Gordon commented that this is also a time for transition within NIMH, and thanked Bruce Cuthbert, Ph.D., who served as NIMH Acting Director for the time between Dr. Thomas Insel’s departure and Dr. Gordon’s arrival. Dr. Cuthbert retired from NIH in November 2016, but is continuing to serve as a part-time consultant directing the Research Domain Criteria (RDoC) project. Regarding changes in NIMH Extramural leadership, three different divisions appointed Deputy Directors: Meena Hiremath, Ph.D., P.M.P., in the Division of Extramural Activities; Julie Bronder Mason, Ph.D., in the Office of Science Policy, Planning, and Communications; and Beverly Pringle, Ph.D., in the Office of Research on Disparities in Global Mental Health. In the Intramural Research Program, Karen Berman, M.D., was elected to the National Academy of Medicine; Francis McMahon, M.D., received the 2016 Colvin Prize; and Miles Herkenham, Ph.D., received the 2017 Winter Conference on Brain Research Pioneer award.

Regarding legislation, Dr. Gordon reported that the non-partisan 21st Century Cures Act  was signed in December 2016 and passed with overwhelming majorities in the Senate and the House of Representatives, demonstrating a national commitment to mental health research. Many components of the Act involve NIMH. Examples include the BRAIN Initiative, and more generally prevention, treatment, and recovery for a number of different disorders, including opioid addiction. The Act appoints the Directors of NIMH, NIDA, and NIAAA to the advisory councils of the Substance Abuse and Mental Health Services Administration (SAMSHA), to further the use of evidence-based care practices in SAMSHA fund states. The Act also encourages collaboration, strengthens existing mental health and behavioral care parity legislation, and includes broad provisions to assist special populations. For NIH overall, the Act stimulates rigor and reproducibility, provides stronger data sharing provisions, and requires Certificates of Confidentiality for clinical research. It also exempts some NIH research from the Paperwork Reduction Act of 1995 and exempts genomic data from the Freedom of Information Act, enhancing NIH’s ability to conduct clinical research. The Cures Act provides funds for research in several areas, including the Precision Medicine Initiative , the Cancer Moonshot , the NIH Center for Regenerative Medicine , and provides about a billion and a half dollars over the next 10 years for the BRAIN Initiative. Dr. Gordon said that in this context the legislative and policy landscape appears to be good for NIMH and NIH.

Dr. Gordon next provided updates for some major initiatives. NIMH and NINDS co-lead the BRAIN Initiative , a collaborative effort among NIH Institutes that carry out neuroscience research. In October 2016, the third round of BRAIN Initiative grants was announced, with more than 100 new awards to 170 investigators from 60 institutions, totaling more than $70 million. This nearly doubled the fiscal year 2016 investment in BRAIN, and further increases are planned as established in law by the 21st Century Cures Act. For NIMH programs, he explained that the applications success rates for fiscal year 2016 was through the 25th percentile of scores, and he then reviewed the NIMH budget. He noted that the number of research projects is holding steady, even though the average project cost is increasing. Dr. Gordon pointed out that in FY16, the success rate for NIMH was approximately 20 percent which is consistent with past success rates for NIMH. Dr. Gordon said that the Institute tries to fund the top 10 percent of grants, but occasionally an individual grant scores within the 10th percentile but reflects science that is not central to the NIMH mission. Such grants are typically brought to the Council for input and approval. He emphasized that NIMH has been successful in funding many grants in the teens and that the Institute will review grants from new investigators, in particular, up to the 25th percentile.

Dr. Gordon also explained why some good science in the 25th percentile is not being funded. He noted that beginning in the mid-2000’s, the congressional funding effort that had doubled the NIH budget ended, and a period of overall budget stagnation resulted in an actual real dollar decrease in the ability to fund research. Award rates for research project grants varied from 19 percent in fiscal year 2013, to 20 percent in fiscal year 2015 and 22 percent in fiscal year 2012. In fiscal year 2016, although NIMH award rates stayed essentially at 20 percent, there was an uptick of award rates to 23 percent, which enabled the Institute to provide more funding of investigator-initiated research projects. He added that NIMH is hopeful that modest budget increases for NIH will continue. He said that NIH is funded through a continuing resolution through April 28, 2017.

Dr. Gordon next highlighted a number of scientific achievements and recent papers supported by NIMH funding. Following the work of the Human Genome Project , intense efforts have been made to examine the genetics and genomics of psychiatric disorders through the identification of numerous genes that contribute to the development of schizophrenia and psychosis. In particular, he referenced copy number variant genes, which confer a very high propensity for developing a psychotic disorder. Over the past decade, NIMH has been invested in looking at some of these high penetrance alleles as a way of examining the neurobiology of schizophrenia. NIMH now funds research identifying low penetrance alleles, which each contribute a very small piece of risk but together account for a much greater portion of schizophrenia. In discussing a paper published in Nature Medicine, entitled Thalamic miR-338-3p mediates auditory thalamocortical disruption and its late onset in models of 22q11.2 microdeletion  by Chun, et al., Dr. Gordon explained that this is an attempt to examine the neurological basis of psychosis by focusing on a high penetrance allele for schizophrenia known as the 22q11.2 microdeletion - also known as DiGeorge syndrome or velocardiofacial syndrome - which arises when an individual lacks a copy of about 20 genes on chromosome 22. Because of the high penetrance for psychosis, if an individual is missing one copy of these genes, he or she has about a 30 percent chance in late adolescence or early adulthood of developing a psychotic disorder that is otherwise indistinguishable from idiopathic schizophrenia. The syndrome accounts for about 1 percent of patients with schizophrenia.

Dr. Gordon stated that Dr. Zakharenko’s research is promising because it assessed a mouse model designed with a corresponding missing segment of chromosome 16 and showed that these mice had a deficit in the ability of information to travel from the medial geniculate nucleus of the thalamus through connections into the auditory cortex. Study results showed that electrically stimulating neurons in the normal mouse thalamus elicits strong responses, but in the mouse model, there is a weaker response in the auditory cortex. Dr. Gordon explained that these researchers studied the auditory cortex because human studies have repeatedly shown that individuals with schizophrenia show inappropriate activation of the auditory cortex during hallucinations, which could account for auditory hallucinations. Dr. Gordon also described Dr. Zakharenko’s earlier research showing that by giving haloperidol to these mice, the activation can be reversed.

Investigators are beginning to assess the links from genes to behavior, with steps along the way that are suggested mediators, and researchers need to know how these genes contribute to the phenotypes seen in patients. This requires not just knowing the gene and the behavior, but understanding all the steps involved. Dr. Zakharenko’s study, for example, did not definitively demonstrate why patients hallucinate because there is no behavioral model for psychosis in mice. However, Dr. Gordon said, it is tantalizing to think that researchers are at least beginning to build a chain of events on this gene to show how individuals might develop psychosis.

Efforts are also underway to understand the heterogeneity of psychiatric disorders, and there is a major effort in the development of large imaging datasets, particularly using assessments such as resting functional connectivity. A group of investigators including Dr. Conor Liston, who has been supported by NIMH, is trying to assess the relationship between resting state functional connectivity and heterogeneity in depression to potentially identify depression subtypes. Dr. Gordon outlined the issues highlighted in Resting-state connectivity biomarkers define neurophysiological subtypes of depression  by Drysdale, et al., published in Nature Medicine. Dr. Gordon described how resting state functional connectivity is determined and noted that this study included resting state and clinical data collected from several hundred individuals. The authors posed two questions.  First, based on functional connectivity and clinical data, would individuals cluster into different groups? Second, if researchers looked at a different group of several hundred patients, would the same clustering appear?

Various depression scores segregated differently in the four different subtypes identified in Dubin and Liston’s research, and these groups differ clinically. Dr. Gordon showed the subtypes and their likelihood of responding to transcranial magnetic stimulation (TMS) applied to the dorsolateral prefrontal cortex (dlPFC) - a treatment under investigation for depression that is also being applied in the clinic. He said that this study demonstrates the importance of large sample sizes and computational modeling.  In this case, data from a large group of people with depression were analyzed using a clustering system. Clustering based on analysis of a large group was applied to sort another group of patients. Thus, a subset of patients identified as likely to respond to dlPFC TMS treatment was not from the group used to construct the model—an important technique for validation. NIMH has funded other efforts in this area, but this study using clustering algorithms is starting to bear fruit. Moving forward, it will be important to employ this method cross-diagnostically.

In describing Behavioral economic incentives to improve adherence to antiretrovirals: early evidence from a randomized controlled trial in Uganda  by Linnemayr, et al., published in AIDS, Dr. Gordon said that this work highlights some of the exciting work emerging from NIMH’s significant investments in HIV and AIDS, specifically involving behavior. The experiment assessed the utility of paying small amounts or giving small gifts to individuals infected with the virus to get these individual to come to the clinic for treatment with anti-retroviral therapy (ART).  Dosing events were monitored by automated methods to ensure that participants were taking the right amount of medication. Many individuals in the study had initial adherence rates of less than 80 percent, putting them at risk of an increase in viral load and disease progression, but researchers found that either of two different economic incentive interventions improved adherence, and adherence rates were boosted to 90 percent or higher. As such, this was a low-cost and cost-saving intervention.

Describing another scientific highlight, Dr. Gordon indicated that Dr. Bryan Roth and his group analyzed the crystal structure of LSD when it is bound to a human serotonin receptor and reported this in a paper published in Cell, entitled Crystal Structure of an LSD-bound Human Serotonin Receptor  by Wacker, et al. Although it is known that two-way receptor binding agonists are not all hallucinogenic, the mechanisms are not well understood. To move forward, it is important to understand how hallucinogenic and non-hallucinogenic agonists interact with the receptors. This paper starts on that path, and Dr. Roth and his group have begun to define physically what happens to the receptor that differs between compounds that are hallucinogenic and those that are not. This work also furthers understanding of the chemical structure of these compounds, and is an example of how research on a molecular structure can inform clinically relevant phenomena.

Dr. Gordon highlighted a paper published in the American Journal of Preventive Medicine, entitled Improving Suicide Risk Screening and Detection in the Emergency Room Department  by Edwin Boudreaux, et al. The authors reported on enhancing suicide risk screening and detection in the emergency department (ED). Dr. Gordon said that to develop effective treatments to prevent suicide and improve the lives of patients who are suffering, we need to identify those individuals with relatively high fidelity. He noted the remarkable effectiveness of simply asking a person if he or she is thinking about suicide along with a few other related questions such as, “how often are you thinking about suicide?” and “how are you feeling?” The work by Dr. Boudreaux and colleagues demonstrates this concept by making it standard practice to ask ED patients three simple questions about suicide, a process that can be automated and that can dramatically increase the rate of positive screening for suicidal ideation. On average, the rate of detection for suicidal ideation doubles when people are questioned about it, as compared to when they are not, which suggests that many people who arrive at the ED thinking about suicide will talk about it if asked.

Dr. Gordon said that he had intended to discuss the Nature Neuroscience paper, entitled Gene expression elucidates functional impact of polygenic risk for schizophrenia  by Fromer et al. from the Common Mind Consortium - but in the interest of time he had to move on.

Dr. Gordon provided an overview of current NIMH priorities, emphasizing that therapeutic and clinical trials to acquire target validation continue to be important. Target validation can be just as useful and important for psychosocial interventions as pharmacological or device interventions. NIMH has issued separate clinical trial funding opportunity announcements to ensure that the scientists who are studying psychosocial interventions have a clear funding path going forward. Dr. Gordon said that work continues on the RDoC project and that updates will be forthcoming during future meetings. Dr. Gordon remarked that NIMH has increased its emphasis on genomics and that a Council working group has been asked to provide a report on genomics. In behavioral neuroscience, the NIMH is beginning to think about how knowledge gained from rodent models is translated to lay the foundation for future therapeutics. He noted the importance of working to understand how animal models are useful and how to ensure that NIMH’s portfolio reflects this usefulness.

Dr. Gordon concluded by discussing his priorities and said that he wants to ensure that the NIMH portfolio is balanced with short-, medium-, and long-term investments. An important short-term investment is suicide prevention. Suicide rates have been rising consistently over the past 15 years, despite what is known and perhaps because of what is not known. It is an issue that needs to be tackled both in terms of the research agenda, and in terms of the 21st Cures Act that encouraged NIMH to build relationships with SAMSHA and other provider organizations. Medium-term goals include determining whether the neurocircuit technology - which is very advanced in rodent models - can be brought to humans and what this would entail. Dr. Gordon noted that Dr. Roth will discuss this later in the meeting. A long-term goal is to bring more computational and theoretical approaches into psychiatry, which will be discussed by Dr. Redish. Dr. Gordon noted that, although this goal is already scattered across the NIMH portfolio, it will be important to consider how to encourage more data mining and biophysical modeling, and to use such modeling to establish phenotypes.


Deana Barch, Ph.D. noted that suicidal thoughts occur much earlier in development than was believed—as early as preschool—and that these thoughts are predictive of later suicide attempts. She said there is some fear about asking young children about suicidal thoughts, but if the data establish that the thoughts are predictive of more active thoughts later, it may be that earlier intervention could help prevent children from engaging in impulsive behaviors when they become teens. Dr. Gordon said that one of the most frightening things in the CDC data is the increase in suicide in the preteen age range and that although it represents a very small fraction of overall suicides, it indicates the need for early intervention.

Marsha Linehan, Ph.D. said that a significant problem in this field is that the average university faculty does not train people in treating suicidal behavior, due to a fear of treating them. Often suicidal people are sent to hospitals, even though no data support this approach. She pointed to the need for an increased focus on training and reducing the fear of litigation.

Dr. Gordon emphasized the need for training in effective, evidence-based treatments, such as Dialectal Behavioral Therapy (DBT), and Dr. Linehan replied that there are proven treatments that are not being used. Tami Benton, M.D. echoed Dr. Linehan’s comments and mentioned that addressing this concept is very exciting because her primary clinical work involves suicidal youth. She said that children are presenting with suicidal thoughts at six, seven, and eight years of age—and most of the interventions that are available for adults do not necessarily have demonstrated efficacy in youth. Dr. Benton said that this could be an opportunity to begin intervention at much earlier ages. Dr. Linehan remarked that the adult interventions do work in adolescents, which was something she did not expect. Dr. Barch indicated that some of the adult interventions have been shown to work in school-age children.

Computational Psychiatry

A. David Redish, Ph.D., Distinguished McKnight University Professor Department of Neuroscience, University of Minnesota

Dr. David Redish began his presentation on computational psychiatry by explaining that he would first define terms, then talk about computation in fundamental (or basic) neuroscience, and finally discuss computation in clinical science, or computational psychiatry. He said that when people hear the term “computational neuroscience,” they think of models—typically compartmental models. He described a computational model from a Gerstner textbook and noted that a classic example of such a model is Eric de Schutter’s famous 3,000-compartment Purkinje cell. Dr. Redish remarked that although these examples involve replicating observed experimental results, computational neuroscience is more than replicating results, which is only a small piece of the computational neuroscience puzzle. A second piece of the puzzle involves mathematical techniques, and he said that by using these techniques researchers have been able to see what could not be seen before. In his lab, for example, the research team has been able to: use mathematical technologies to extract representations of future possibilities; show that rats deliberate over choices in the same way that humans do; and assess what those mechanisms of deliberation are and how to get to them. Finally, computational neuroscience is also more than big data, or the analysis of large datasets.

To set the stage for his discussion on computation in fundamental (or basic) neuroscience, Dr. Redish outlined the importance of understanding the differences among conceptual frameworks, theories, and models in computational neuroscience. The conceptual framework defines the way to think about a problem, creating a family of theories. A theory is a specific hypothesis that makes predictions. A model, which is an instantiation of the theory, can be used to test the viability of a hypothesis. Computation informs conceptual frameworks, theories, and models. This led to his discussion about computation in fundamental neuroscience in terms of explaining interactions and making connections across conceptual, spatial, and temporal levels in neural systems. He added that computation can play a big role in moving from genetics, to circuits, to behavior.

Dr. Redish offered one of his favorite quotations, from Frank Herbert’s novel Whipping Star: “the most common use of abstraction is to conceal contradictions” and said that his corollary to this statement is that abstraction is an important part of understanding. If we are looking for computational answers, we need to build at the right level and focus on connections between levels. He emphasized that many analysis techniques are available at different temporal and spatial scales, but also pointed out that most reviews of the multiple analysis techniques ignored behavior, which also provides information about neural function. He said that computation allows us to connect those interactions across levels even when they interact in complex ways. Analyzing behavior in these systems would involve answering questions such as how behavior arises from neural firing patterns. This would require a computational understanding of that connection. Similarly, computation will be needed to ascertain complex connections in the analysis of how behavior changes the physical nature of synaptic circuits and how those changes in circuits alter neural patterns. This loop includes temporal dynamics, and this adds even more complexity and requires computation to make the connections.

As an example of the usefulness of computation in basic neuroscience, Dr. Redish described his work in hippocampal place cells, noting that the place cells provide a spatial map in the brain of the location of the animal. He noted that a potential conceptual framework would be that the hippocampus has information about space and memory; concepts that need to be brought together. The hypothesis that the spatial information in hippocampal place cells derives from another type of cell in the entorhinal cortex that also represents space (but in a different way) could be supported by a model. Such a model could analyze grid cells, model connectivity, and determine whether place cells would appear. He noted that the analysis of this hypothesis is a computational one, which is about connecting levels. Dr. Redish also emphasized that the use of computation takes considerable time and effort and that it is important to recognize that like many theoretical sciences, computational neuroscience requires its own effort.

Discussing the importance of computational psychiatry, Dr. Redish referred to an Ernst Strüngmann Forum that he and Dr. Gordon organized in 2015. The meeting brought together 22 experts in computation and 22 experts in psychiatry to discuss how to advance psychiatric research through the use of computational techniques. Based on the assumption that behavior is derived from a complex emergent circuit that can be regarded as the putative cause of the behavior. The four working groups at the meeting independently arrived at the consensus that the Diagnostic and Statistical Manual of Mental Disorders (DSM) and psychiatric diagnoses should be thought of as observations instead of as theoretical syndromes or definitions. However, these observations are useful and mathematical tools can connect computational hypotheses with these observations. Dr. Redish noted that the latent constructs are the computational hypotheses about how the putative causes drive the behavior. He suggested that RDoC is essentially a computational hypothesis about latent constructs and itself is computational. He said that although the idea of computational hypotheses and latent constructs is not new, being similar to ideas of endophenotypes and trans-disease processes, but that it will be important now to build these theories based on computational hypotheses from fundamental neuroscience.

Returning to the idea that RDoC is fundamentally computational, Dr. Redish suggested that it should include the concept of the failure mode—an engineering term indicating a vulnerability inherent in the machinery that can be identified if one knows where to look. In discussing neurophysiological failure modes, he gave the example of epilepsy as an imbalance between excitatory and inhibitory connection strengths. Computationally, typical brains function at a balance between excitation and inhibition partly because it becomes easier to react quickly. He pointed out that the many genetic abnormalities that lead to epilepsy create that computational imbalance. He said that researchers could use a computational approach to understand how excitatory and inhibitory imbalance leads to epileptic seizures, and genetic or direct treatments might be identified that could restore that balance. He said we could also use computational analyses to identify oncoming seizures. Thus by knowing the location of the failure modes, we can apply the appropriate treatment. He further remarked that interactions among underlying dysfunction in genetics, pharmacology, and the environment lead to observed dysfunction in behavior, and behavior is the fundamental issue. These components are computationally translated through the brain, and computational techniques are needed to make that connection. He also argued for the use of functional translation by modeling computational dysfunction instead of causal dysfunction, and emphasized that the key to modeling is to do it at the right level.

Dr. Redish concluded his presentation by discussing how computation can play a role in future research. He remarked that many new experimental techniques depend on extensive computational processing. He emphasized that obtaining this kind of analysis is extremely complicated and that new training is needed not only in the proper use of statistics, but also in the ability to derive one’s own statistics to fit the experiment. Assessment techniques such as functional MRI and diffusion tensor imaging require a series of extremely complex processes, and it is important to ask if the assumptions underlying those processes are valid for the data being used. He emphasized that computational modeling takes a great deal of time and personnel. Further, to translate findings to the clinic, multidisciplinary research teams are needed that combine computational expertise, clinical expertise, and fundamental science expertise, as was demonstrated beautifully in the Strüngmann Forum. Having all three of these groups working together provides a huge advantage that will be key to clinical translation.


Randy Blakely, Ph.D. commented that one of the problems involved in modeling neuropsychiatric disorders in animals to design new therapies has been that the intermediate phenotype has been so gross that many look-alike phenocopies have been allowed to drive research and investment. He said that it is important that researchers working in genetics and genetic modeling interact with researchers involved at the behavioral level, because animal models are flawed and failure occurs in ways that do not match real-world situations. He also remarked that it is important to work together as teams to avoid the counterproductive noise in the modeling. He mentioned ADHD and hyperactivity as an example.

Dr. Redish replied that although the two ends are critical, the problem is that we have been ignoring the intermediate level. We need to, for example, select a hyperactivity model and determine the intermediate level that is driving the hyperactivity. We need to ask if it is similar in any way to the human condition. Thus, instead of going directly from the genetic level to the hyperactivity behavior model, we need to determine if the intermediate level is matching. He said that we are now starting to talk about taking specific dysfunctions in specific intermediate structures. We need to ask, for example, if we can create a model of hippocampal dysfunction and connect that to specific behaviors. However, behavior is not enough because the same behavior can be generated by many different brain structures, and it takes great care to get a behavior for which the failure model can be separated out in the brain structure. He said that the intermediate level is important here as well.

Dr. Blakely emphasized the importance of using the rules derived from this intermediate analysis in the most productive ways to address idiopathic disorders. He noted that letting the investigation of rare genetic models guide research toward translation may not be the best way.

Dr. Linehan remarked that although she fully supports the approach, she would also like to see the same model applied to behavioral interventions because we know that behavioral interventions work, and they work noninvasively. Dr. Redish commented that behavior changes the brain and that we need to learn how this happens. A behavioral manipulation is, in fact, a neural manipulation. Dr. Linehan emphasized that we need to know what behavioral changes are required to change a disorder, but this does not require that we know where it is changing the brain.

Dr. Redish said that he would argue that researchers need to know how behavior changes the brain and that the intermediate level failure mode concept is the key. To identify appropriate behavioral intervention, we need two pieces of data, both of which are just starting to become available. Specifically, we need to know the brain dysfunction that is generating this behavioral change and what the behavioral manipulation is doing. He emphasized that this double step that involves taking the behavior to get to the intermediate point - and bringing it back - is where the breakthroughs will occur because researchers are beginning to understand how behavior is generated by neural circuits at various kinds of mesoscale levels.

Dr. Barch remarked that she would like to see Drs. Redish and Linehan hammer out the issue of neural primacy, because she believes it is a mistake because environmental or behavioral factors can lead to outcomes that are changing the brain. As such, there may not be a need to spend time determining the neural instantiation or mechanisms in order to go directly to change. Also, a problem in computational psychiatry is that most success has been in cognition and it has not penetrated many of the complex behavioral domains that drive disorders such as emotional dysregulation, mood, social relationships, and interpersonal relationships. She said that we do not have a good conceptual framework in much of computational psychiatry for thinking about how these disorders arise and that we need to start to push on this. She also remarked that many scientists do not believe that computational psychiatry applies to their work because of the lack of models. As well, a challenge in computational psychiatry is that too often researchers get stuck at the conceptual framework. However, she said, not knowing how this model moves beyond the conceptual level means that it will not influence psychiatrists and psychologists. She noted that we need to think about how experiments should be conducted differently, how we should interpret our data differently, and how to think about our patients differently.

Dr. Redish said that in the computational world most of this work has been at the model level and only in the last 10 or 15 years have conceptual frameworks started connecting to experimental validity. He disagreed that the biggest successes in computation have been in cognition, and argued that the greatest successes have been in areas that have closer neurophysiology, such as epilepsy. Computation has not hit the psychiatric connection yet but the time is right due to concrete fundamental science breakthroughs over the past five years. Because we are beginning to achieve an understanding of mechanisms driving behavior, we can make accurate second-, third-, and fourth-order predictions involving the behavior of cell activity or a structure in a complex experiment. He agreed with Dr. Barch’s point about behavior and said that the diagram he highlighted in his slide presentation needs to include behavior and a social dynamic section. Models of social dynamics are available and it would be valuable to include computational scientists who are involved in social dynamics. Suggestions have emerged from social dynamics conceptual frameworks, modeling, and computational analysis indicating that contingency management is working differently than currently believed. If correct, this could change treatment.

John Krystal, M.D. suggested that it is premature to try to determine at what levels computational psychiatry will have its greatest impact because it is, in part, a strategy for developing formal theories that can be tested in translation across any level. He commented that our successes are illustrations of how one might cross levels, but do not necessarily indicate where the greatest impact will be. As well, he noted that many domains are not being addressed, such as developmental levels. Another issue involves whether computational psychiatry will embrace analytics and modeling or whether these are two divergent domains of the work. He emphasized that the application of machine learning techniques to cluster individuals will probably become standard as new scientists receive better training and become more computational, but modeling will always be critical. He also referred to the huge amount of data and the different kinds of experiments needed to validate a model.

Dr. Redish replied that although computational techniques are being used in data analysis, in modeling, and in clustering, the biggest breakthroughs will be in using a computational understanding that connects fundamental science with clinical practice and clinical observations.

Patricia Areán, Ph.D. said that these models would also be applicable to the problems that are occurring in service delivery and that the discussion around failure analysis is a wonderful way to start thinking about service gap problems and utilization issues. In addition, using these computational models would be helpful in moving toward the implementation of current services at a greater scale, particularly because large datasets are now available.

Neural Circuit Technologies

Bryan Roth, M.D., Ph.D. Professor, Department of Pharmacology, School of Medicine, University of North Carolina - Chapel Hill

Dr. Roth began his presentation with a historical perspective of early neuroscience research. He shared that from the 1940s through the 1960s, neuroscience research was focused on electrical stimulation of the brain. He referenced 1961 paper published in Science, entitled Distinct ‘feeding’ and ‘hunger motivating’ systems in the lateral hypothalamus of the rat , in which P.J. Morgane examined rat feeding behavior. In this study, investigators used electrodes to stimulate parts of the hypothalamus and mapped the areas that reliably elicited feeding behavior. An important finding was that electrical stimulation of the hypothalamus not only activated neurons, but also axons in passage. However, some significant questions arose from this work. If researchers can elicit a particular behavior, is this because the behavior is encoded by the neurons or circuits in this area or is it externally produced? Is this top-down control? He noted that the assumption, from the 1960s to the present, has been that these behaviors are not encoded discretely in the brain, but are distributed networks. In this case, it could be assumed that a network is not being activated. However, data from Dr. Roth’s work suggest that this may not be the case.

Currently, his lab examines neural circuits from a drug discovery perspective. Dr. Roth said that as a psychiatrist, he specialized in schizophrenia treatment and ran an inpatient ward for 10 years solely devoted to schizophrenia. He remarked that during this time it became clear that the drugs patients were prescribed were having diverse effects on many circuits and receptors in the brain. Dr. Roth wanted to deconstruct this to understand more about circuits and receptors in order to help spur drug development and improve medications. His lab was interested in developing a synthetic receptor that could assess whether specific neurons of interest could mediate a drug’s effect. He noted, that if a drug were specifically designed to target only the synthetic receptor, those receptors could potentially be activated in a neuronal population, and might lead to further understanding about behavior and therapeutics.

He shared that rather than starting with a receptor, the lab began working with an inert metabolite of the antipsychotic clozapine (clozapine N-oxide). This drug was selected because of its properties of excellent oral bioavailability and CNS penetrability, in addition to its long record of safety in humans. Led by staff scientist Dr. Blaine Armbruster, the lab decided to create a receptor that could be activated by this relatively inert drug, place the receptor in various brain regions, and turn the various brain regions on and off. If the approach worked, the lab could assess where chemical actuators of these synthetic receptors or modified receptors could be used – with the ultimate goal of being used for a drug approved in humans. He added that directed molecular evolution was used to develop receptors in yeast. These receptors can increase firing neurons, inhibit firing neurons, and modulate the signaling of neurons, and all of them are freely available to the scientific community through Addgene. Dr. Roth acknowledged NIMH support for the post-doctoral researcher who conducted the studies in model organisms.

Next, Dr. Roth reviewed some of the key differences between optogenetics and chemogenetics. He explained that rather than inducing synchronized firing or synchronized unfiring of neurons, chemogenetics raises or decreases the probability that neurons will fire. And, the addition of the drug increases the probability of spontaneous electrical activity. He referenced three of the few thousand papers that have been published since his lab began this research. In a paper published in Science, entitled Stellate cells drive maturation of the entorhinal-hippocampal circuit  Flavio Donato, et al. examined the development of a circuit that regulated place fields and were able to silence the circuits. Dr. Roth described this as a “big effect.” In a recently published paper in Cell, entitled AgRP Neurons Control Systemic Insulin Sensitivity via Myostatin Expression in Brown Adipose Tissue  Jens Brüning, et al. addressed the question of how neurons might control functions outside of the brain and showed how particular cells in the hypothalamus control insulin sensitivity in brown adipose tissue. Finally, in a paper published in Nature, entitled A shared neural ensemble links distinct contextual memories encoded close in time , Alicino Silva et al. reported on a study that involved the use of chemogenetic technology to look at how memories are encoded in the brain.

Dr. Roth also discussed his research in feeding. He said that since the earliest experiments with electrical stimulation of the brain, scientists have attempted to understand which parts of the brain might be essential to feeding behavior. He added new technologies such as optogenetics and chemogenetics enable the continued exploration of feeding. Focus on the agouti-related peptide (AgRP) neuron, (which has been deconstructed in great detail), has revealed that these neurons are located in the arcuate nucleus of the hypothalamus, which is regulated by leptin, a hormone that may regulate feeding behaviors. Dr. Roth said that to achieve cell-type specificity in his mouse studies, he combined two technological platforms, Cre driver line mice from the mouse genome consortium and viral delivery of chemogenetic or optogenetic actuators. He explained that mice are available that allow for the expression of these constructs only in particular and defined neurons - AgRP neurons). With the activation of a chemogenetic platform, a very strong increase occurs in neuronal firing and these neurons can be reliably turned on and off.

In one experiment, when these neurons were turned on in mice, they ate voraciously at a time of day when mice are not normally hungry and do not normally eat. When120 neurons were activated in their brains, the mice ate continuously for five hours. Within two hours, an 18-gram mouse ate 1 gram of food, which is equivalent to a 190-pound human eating 20 to 30 pounds of pasta in one sitting, when not hungry. Dr. Roth asserted that when those neurons are activated in the brain, a behavior can be reliably induced. In response to those that believe feeding is not a complex behavior and that there is no indication that mice are motivated to eat or seek food, Dr. Roth said that he could point to an experiment conducted by Dr. Michael Krashes, now with the NIH. Dr. Krashes placed a mouse whose neurons had been activated into a cage without food and it crawled around the cage eagerly looking for food, while the control mouse did not.

Dr. Roth remarked that through this non-interventional technology, the drug to turn on the receptor can be delivered through the mouse’s drinking water and that its ensuing behavior can be observed under relatively normal circumstances. He said that about 100 papers have assessed variations of this experiment, and the neurons and circuits are being mapped in exquisite detail. However, although these technologies are very useful for the basic science of circuit deconstruction, we do not know if they will be useful for treating diseases. He described a depression experiment from his lab, to determine whether directly turning serotonin neurons on could act as an antidepressant. The research team focused on one small population of dorsal raphe serotonin neurons and activated them in mice. Dr. Roth pointed out that all effective antidepressant drugs increase the vigorousness and duration of swimming behavior in mice. In this study, a very strong antidepressant drug-like response was elicited by activating about 100 neurons in the mouse brain. He noted that although this finding does not mean that the activation of these neurons would have an antidepressant effect in humans, it is something that ultimately could be tested.

Regarding the BRAIN Initiative, Dr. Roth hopes that ultimately we will gain an understanding of circuitry alterations in neuropsychiatric diseases - opening the door for circuit-based therapeutics such as chemogenetics. He summarized some of the studies his lab conducts with opiate receptors and said that once more is known about the structure of these receptors and how they function, new chemogenetic tools could be constructed. He described a new DREADD (designer receptors exclusively activated by designer drugs) that his lab created using a structure to design the ligand and the receptor. Noting that the experiment assesses feeding because it is a behavior that is easy to measure, he said that chemical actuators were used with rapid onset and relatively rapid loss of effect. In these experiments, mice groom themselves before receiving the actuator. After a mouse receives the actuator, it eats the food even though it is not hungry. He said that one of these actuators is a non-generic antipsychotic drug, that DREADD activators can be FDA-approved medications, and that DREADDs have been reported to work in human cells and in non-human primates.

In discussing opportunities for enhancing drug delivery systems in larger brains, Dr. Roth mentioned convection-enhanced adenovirus vector virus delivery, which is approved by FDA as a delivery vehicle for human CNS clinical trials. He suggested that it would be helpful to be able to administer the viruses peripherally, perhaps even by nasal insufflations, to target them directly to brain regions of interest. Enhancing cell-type specific targeting can be done through enhancer-based and two virus-based specifications. Looking forward, Dr. Roth said that to use these and related technologies, such as optogenetics, in humans, it is important to ensure that they are expressed and functional in humans before conducting a clinical trial. He concluded by emphasizing that PET/SPECT ligand development to visualize and confirm DREADD delivery and expression will be key, along with medicinal chemistry to optimize DREADD ligands and structure-based design of new DREADDs.


Alan Greenberg, M.D., M.P.H. asked if they could get the mice to do the opposite behavior and not eat. Dr. Roth replied that they can and that there are a few ways to accomplish this. One is by silencing the neurons, and another is by activating inhibitory neurons that are feeding in. He said that they can make mice thin and that there are models of obesity in mice in which different neurons in the circuit are silenced or activated, and diet-induced obesity can be decreased.

Dr. Blakely asked, what some of the approaches may be that would allow researchers to gain higher temporal control over these receptors, as it relates to the temporal aspects of stimulation with the systemic delivery of drugs that are long lasting and continually activated. Dr. Roth said that the new DREADD, based on the kappa-opioid receptor with Salvinorin B as a ligand, has effects in a minute or two and that the effect can be seen in many behaviors. The pharmacokinetics are rapid on and rapid off with this ligand. He commented that this is a solvable medicinal chemistry issue and that we have made some of these prodrugs. However, if millisecond control is needed, a light-based actuating system would be required.

Ian Gotlib, Ph.D. remarked that many researchers are doing work in neuro-modulation and neuro-feedback and thus are more interested in the spatial issue than in the temporal issue. He wondered if neural modulation can be used in those areas to affect drive and asked if being able to localize a signal in the smallest region in which those neurons reside would affect those neurons. He also asked whether this would affect eating behavior, motivation, and appetite. Dr. Roth replied that if the neurons’ genetic signature is known, there is no question that we will be able to accomplish this in the future. No new technology would need to be developed.

Dr. Gordon said that work is needed to bring the technology into humans and that some of the questions involve the extent to which this can be accomplished using the existing model systems such as mice and rats and to what extent we need to think about expanding our circuit repertoire to other model systems.

Dr. Roth explained that this will be the golden era of the mouse neurobiology, if it has not begun already. Ultimately, we will have a catalog of every mouse neuron, know their expression repertoire, and be able to target every “sub-sub-subtype” of mouse neuron.

Dr. Gordon remarked that this is one of the stated goals of the BRAIN Initiative and that funding opportunity announcements will finalize this effort. He said that the catalog will not be 100 percent complete the first time around, and multiple revisions are expected.

Dr. Roth said that this work is progressing, but because mice are not humans and are not even small primates, non-trivial technology problems and challenges are involved in moving from the mouse brain to the non-human primate brain. He said that we do not need to invent a new technology, but the technologies that we have need to be optimized for larger brains.

Dr. Redish asked if a catalog of neurons can be compiled, given that some of those neuronal functions depend on developmental and environmental issues. Dr. Roth said the catalog is being compiled and that data in the ultimate datasets will be profoundly useful, although not perfect.

Opportunities to Reduce Suicide

Cheryl King, Ph.D. Professor, Department of Psychiatry University of Michigan

Dr. King began by remarking that the discussion would be more applied and solidly based in clinical and public health models, with the goal of considering options to move the dial in reducing the rate of suicide in the United States. She said that we know that suicidal individuals are highly heterogeneous across levels of analysis and that many possible prevention strategies can be pursued. She reviewed the work of the National Action Alliance for Suicide Prevention’s Research Prioritization Task Force in A Prioritized Research Agenda for Suicide Prevention: An Action Plan to Save Lives , noting that NIMH was heavily involved in the public/private partnership that developed this comprehensive document. The action plan prioritized six areas for suicide research: (1) why individuals are suicidal; (2) how to better detect/predict risk; (3) effective interventions; (4) effective services; (5) preventive interventions; and (6) research infrastructure. Various models are discussed in the document about the impact of different approaches, and the Emergency Department (ED) was identified as a particularly fruitful area for suicide prevention.

She shared that the American Foundation for Suicide Prevention (AFSP) Project 2025  is another prevention effort that aims to reduce the suicide rate in our nation by 20 percent by 2025. Modeling was used to identify the programs, policies, and interventions with the greatest potential for impact, including those involving the ED and large healthcare systems interventions and firearm education. Dr. King said it is important to decide how to invest resources through decisions that are guided by modeling and to think about the implications of the public health model and the opportunity to impact more people with preventive efforts. Also, we know from developmental psychopathology that the earlier we intervene, the more chance we have to alter developmental trajectories.

Dr. King said that we know suicide is a leading cause of death and that many of the people at risk for suicide go unrecognized. She said that we know from psychological autopsy studies that the first attempt is fatal much of the time; for nearly half of adolescent suicide victims the first attempt is fatal. We also know that males have a much higher suicide rate than females; the male suicide rate is 3.5 to 4 times higher than the female rate. We also know that men and boys are less likely to seek mental healthcare services. However, we are not treating males at a commensurate level in mental health clinics and psychiatric hospitals across the country, and we are psychiatrically hospitalizing more teen girls than teen boys for suicide risk. Dr. King stressed the importance of focusing on reaching males as well as females as we try to move the dial. She also said that reducing the suicide rate substantially will require interventions and preventive strategies that occur outside and inside the mental health system.

Dr. King said that through CDC’s National Violent Death Reporting System (NVDRS)  we are learning much more about violent deaths in the United States than we used to know from national mortality statistics, which provided socio-demographic data. More data are being collected, and the current effort covers at least 40 states. Initial data from the first 17 states show that males are more likely to die on their first attempt, again emphasizing the importance of early identification and screening. She said that one of the advantages of screening in the ED is that both males and females are well represented across the lifespan. Because teen boys do not access primary care services as frequently as teen girls, the ED is a useful place to find them. A fair number of adults and children visit EDs each year, therefore the ED sample is slightly enriched in terms of suicide risk. In addition, the rate of self-harm seen in EDs has increased. Some studies show that the clinical recognition of a mental disorder in the ED is associated with the patient being less likely to return with self-harm. This provides a hint about the potential for recognition and brief intervention when some of the concerns and problems of these patients are acknowledged and validated.

Dr. King, along with David Brandt and Jackie Grupp-Phelan, is leading a large-scale NIMH-funded collaborative study called Emergency Department Screening for Teens at Risk for Suicide (ED-STARS), which will develop and validate a computerized adaptive suicide risk screen for youths aged 12 to 17 who present to pediatric EDs for any complaint. Dr. King said that the goal is to develop a short, adaptive screen that is feasible to implement in EDs broadly. This multisite study is being conducted with 13 EDs from the Pediatric Emergency Care Applied Research Network (PECARN)  a nationwide network of pediatric EDs with infrastructure support for research, and an Indian Health Service hospital serving the White Mountain Apache tribal nation. She explained that the study involves screening youth for a wide range of known suicide risk factors, and that working with Robert Gibbons of the University of Chicago, a computerized adaptive screen will be developed and evaluated for its psychometric properties and compared to a commonly used brief screen. A behavioral test will also be evaluated to see if it adds incremental validity in terms of identifying those teens who do not share suicidal thoughts but who might be suicidal. The study team will also develop a risk stratification algorithm with three levels of risk that can guide the ED providers in terms of the level of intervention needed, with some room for the ED to decide where it wants to set those levels in terms of sensitivity and specificity. The first level may be acute risk, suggesting the need for a mental health. The second level may be patients at risk, who could be provided with resource information. Dr. King discussed studies that have shown that providing individuals many pamphlets or online information does not have significant impact, but that behavioral rehearsal with patients as they seek other services (e.g., call crisis line) is a promising area. She would like to have an opportunity to see what impact behavioral rehearsal might have in this middle risk group. The third level is a no risk group, with no further action needed.

Dr. King said that a great deal of work remains to be done following the recognition of suicidal ideation or attempts. We know from two studies that used large national Medicaid databases that many patients, both youths and adults, who present to the ED with episodes of deliberate self-harm receive no mental health services. We also know that the risk of suicide increases substantially during the 12 months following ED treatment for deliberate self-harm. Dr. King said that although many different types of brief interventions are available, more research is needed on effectiveness. Work is also needed in online confidential services to help link individuals to treatment, because many people are more comfortable with this arrangement. She remarked that we do not leave our chairs for much of what we do now; we go online, and this is moving into the world of care. Some ideas for moving forward include developing adaptive interventions and designing treatment trials so that they inform their development. Dr. King also noted that the approach of conducting online screening and confidential counseling to inform individuals about available resources and then linking them to those resources is nearly ready, if not ready, to be used for college students and the workforce, and it could be rolled out in different workplace settings. Dr. King said that the challenge is to implement this kind of screening along with screening in the ED and primary care across the lifespan. She concluded by outlining the learning healthcare system approach  for empowering and changing systems, which involves gathering solid data, analyzing it, interpreting it, using feedback to make modifications, and determining how to disseminate it in the field.


Lisa Jaycox, Ph.D. asked if there would be opportunities for discussing means prevention with families in the ED. Dr. King responded that they routinely discuss firearm safety in many of their clinical settings with those at elevated risk, and they chart the discussion. More and more work is ongoing in this area, one in which we can make a difference because firearms account for about half of the suicides in the United States.

Dr. Linehan discussed an NIMH-funded study for alcoholics at high risk for suicide that involves asking individuals if they prefer to talk to a computer or to a person if they felt suicidal, and then providing online treatment. It was found that men who were alcoholics preferred to talk to a computer and not to a person, and the online treatment was just as effective as in-person treatment, both for the alcoholics and for people with extreme depression. They also conducted outreach in grocery stores and other places men frequent, using signage to ask if anyone felt suicidal and wanted to talk to someone about it and received a big response from men (but not women). Dr. Linehan said she believes that the next world is one of computerized interventions and that it is important to focus on the development of these interventions and do the necessary research.

Rhonda Robinson-Beale, M.D. asked about the effectiveness of suicide hotlines. Dr. Linehan replied that their patients generally are not interested in accessing suicide hotlines. Michael Hogan, Ph.D. said that Madelyn Gould has done studies on suicide prevention hotlines showing they are fairly effective, and it has been found that follow-up calls are also effective. He said he appreciates the work and how it is being carried forward, but huge gaps in knowledge and catastrophic gaps in applying what we already know persist. Measures, accountability, and standards of care are lacking, and Project 2025 is probably one of the most significant efforts to galvanize interest. He also said that the modeling of possible impacts is impressive, but might be more guesswork related to what might happen than actual modeling. He suggested that conducting some actual modeling or measurement in this area would reinforce focus and attention.

Dr. Gordon said that NIMH met with the AFSP and reviewed the modeling done, and realizing it was superficial, decided to develop the modeling within the Institute, with outside help if needed. He said that the Institute wants to focus on not just where the opportunities are, but also on the evidence for efficaciousness including identifying gaps where more work is needed. He also commented that to meet Project 2025  goals, steps that go beyond research need to be taken now. Dr. Gordon said that he was encouraged by Council members’ comments about efforts that could be undertaken at this time and that NIMH looks forward to receiving additional feedback from members in this area, as well as on the extramural and intramural programs.

Dr. Robinson-Beale had a question about the data and the consistency of the data, noting that some cardiac health research involves linking mortality and morbidity data in a way that provides a clear idea of an intervention’s effect. In mental health, the inability to link this kind of information back to systems that may have been in contact with individuals makes it difficult to get attention and develop solutions. She said that in mental health, there has been a tremendous movement to get primary care to conduct depression screening in end stage renal disease, an area in which researchers could begin to bring the data together.

John Davison, M.B.A, Ph.D. said that he was energized by this discussion as it pertains to the military population, in which a high degree of stigma in this area persists. Relating to the idea that an alcoholic male is more likely to talk to a computer than to a person, he wondered if Edwin Boudreaux is working on the computer-adapted testing for adults. Dr. Gordon said that he was not sure whether Edwin Boudreaux was involved or not but that an iPad version of the screening is available. Dr. King commented that the adult ED screen did not involve computerized adaptive testing but that Robert Gibbons at the University of Chicago has worked with other groups to develop different adaptive screens for depression, and that a number of them are available and more are under development.

Dr. Benton remarked that a primary care setting is an ideal place to implement the screens for young people and their parents.

Dr. Gordon said that these areas will be revisited over time as NIMH comes to Council with ideas for RFAs and asks for members’ expertise to help guide the research agenda.

Concept Clearances

Dr. Dianne Rausch, Director of the Division of AIDS Research (DAR), NIMH, presented three concept clearances in important gap areas for which applications will be requested. DAR focuses on the neurologic and neuropsychiatric consequences of AIDS infection as well as behavioral interventions and the consequences of the disease.

Mobile and Connected Health Interventions to Improve Care Continuum and Health Outcomes Among Youth Living with HIV

The goal of this initiative is to develop and test the next generation of interventions delivered through mobile health (mHealth) technology to improve testing, care linkage and retention, and antiretroviral therapy (ART) medication adherence among youth living with HIV (YLWH), domestically and internationally. In 2015, 670,000 young people between the ages of 15 to 24 were newly infected with HIV globally, and AIDS-related deaths among adolescents have increased over the past decade, despite decreased death rates in all other age groups. In addition, adults are diagnosed more readily than adolescents, with age-related disparities evident at each step in the HIV care continuum. Recent data have also shown that the earlier treatment is initiated after HIV infection, the better the long-term health outcomes. In fact, because virally suppressed individuals are no longer infectious, it is critical that infected individuals take and continue taking ART, and better ways are needed to increase youth testing and entry into care.

Interventions delivered through mHealth offer a promising approach for improving outcomes among YLWH, using social support and technology that incorporates the use of wireless devices and gamification, to improve testing, linkage to care, retention in care, and medication adherence. Because youth are avid adaptors and heavy users of smart phones and digital technologies, these technologies offer opportunities for developmentally appropriate personalized interventions that can be delivered in real time in private settings. They also can be adopted on demand and remotely, thus addressing stigma and issues involving time constraints. The development and testing of the next generation of remote treatment services includes technologies that can be used to: increase the uptake of rapid home HIV self-testing; rapidly link youth to HIV care after they test positive; help re-engage youth who have fallen out of care; help youth in the transition from pediatric to adult care; and provide novel remote treatment services to improve care retention (including phone counseling sessions and other opportunities to receive care without going to a clinic).


Dr. Davison commented on the importance of reaching youth and delivering testing and treatment to them and asked if this concept includes trying to improve access to diagnosis.

Dr. Rausch responded that testing is the first key step in this concept, because individuals cannot be linked to care if they do not know they are infected. She said that a goal is to use new strategies to target youth to get tested and get the test kits to them. She also emphasized that one of the most important strategies is home testing, which can extend to partners and also become a way to screen potential partners.

HIV Healthcare Systems Approaches to Improve Viral Suppression (HH-SAIVS)

Dr. Rausch noted that although many past interventions have been focused on targeting patients and getting them tested and linked to care. Though there are some interventions available for care teams, more research is needed in structural interventions, which have the potential for broader impact. The purpose of this initiative is to identify, develop, and test healthcare system interventions (healthcare delivery changes) that would improve the connection to care and treatment and thus improve HIV viral suppression, which is key to treatment and prevention. The goal is to develop structural interventions in healthcare systems that can improve population health and that have the potential to reach larger numbers of HIV-infected individuals.

Dr. Rausch said that creative new ways are being sought to promote adherence. She reviewed a model of a four-level healthcare system  that was developed by the Institute of Medicine and discussed a cardiovascular care study published in Health Affairs, entitled Eliminating Medication Copayments Reduces Disparities in Cardiovascular Care , by Choudhry et al. that involved measuring the incidence of the first major event or revascularization among non-white patients by altering prescription coverage. In this study, the elimination of the co-payment improved medication refills by 6 percent, cardiovascular events were reduced by 35 percent, and healthcare spending was reduced by 70 percent. She commented that the development of this kind of intervention in HIV healthcare may more effectively increase uptake and adherence to ART medications and would include prescription-based strategies, such as synchronizing prescription refills, extending refill intervals, or improving packaging formats or labeling. Efforts could also entail testing reduced medication co-payments or value-based insurance designs to see if such coverage-based approaches can improve rates of ART medication refills and viral suppression. Researchers can also investigate “test and start” HIV care models designed to expedite ART treatment initiation and the achievement of viral suppression: (1) test the impact of specialized HIV care clinics or differentiated HIV care models designed to meet the needs of underserved patient populations; (2) enhance and optimize comprehensive HIV care team composition to improve rates of viral suppression; and (3) test electronic health informatic “dashboards” and quality improvement approaches to strengthen the HIV care continuum at the clinic, municipal, or state level.


Dr. Benton commended the approach as worthwhile to pursue in health systems, because the infrastructure is generally sufficient to allow it to propagate. She suggested also keeping in mind other underlying factors found in systems such as Accountable Care Organizations, patient-centered homes, and the use of data and patient registries. She also stressed the importance of pursuing questions involving whether or not the resources designated within a system are sufficient for reaching out to people who have fallen out of the mainstream. She said she hopes these factors are identified and explored because they can make a difference in the ultimate outcome.

Dr. Rausch said that the focus on benefit design, and particularly co-payments, is critical because it often constitutes a significant barrier to treatment, and she emphasized the importance of removing barriers, incentivizing behaviors, and using behavioral economic principals to promote adherence. She also commented that there is great enthusiasm for applying behavioral economic principles to improve adherence.

Dr. Krystal asked how this strategy is integrated with other NIH Institutes or efforts. Dr. Rausch explained that the Office of AIDs Research (OAR) is the overarching coordinating body for AIDS research at NIH and has coordinating teams across the different disciplines to develop initiatives and focus on priorities. She said that all of the Institutes will have an opportunity to participate in these initiatives, and that she believes that this initiative will be cosponsored by a number of other Institutes.

Altered Neural Pathways, Receptors and Networks in HIV-Induced CNS Dysfunction

Dr. Rausch began by noting that most people who are currently infected are on ART. She went on to say that the neuropathology of HIV-associated neurocognitive disease under therapy is no longer massive encephalitis and neuro-degeneration, but is subtler and in many cases subclinical and difficult to assess. In fact, symptoms specific to HIV-1 induced CNS pathology are seldom seen in ART-suppressed individuals, who exhibit mild to moderate forms of cognitive dysfunction. The goal of this initiative is to foster change through a focused Funding Opportunity Announcement (FOA). The FOA would encourage researchers to decipher, in the context of ART and through the use of novel research models and diagnostic tools, the causal role played by altered neuronal networks, neural receptors, and pathways in the mild to moderate HIV-1 induced CNS dysfunction that is observed in clinical studies. Dr. Rausch said that there is a need to learn what kind of pathology is being seen in HIV-infected people on treatment for long periods who may be aging and who may have been on medication for a long time and also have comorbidities and polypharmacy. She explained that there is also a need to develop/adapt quantitative biomedical diagnostic tools that can provide the true clinical picture of the nature and extent of neurocognitive impairment and that such tools would diminish reliance on neuropsychiatric batteries. The initiative encourages research in the following areas: neuronal dysfunction using novel in vitro and in vivo neuro-electrophysiological techniques pursuing outcomes related to mild to moderate HIV-1 induced CNS dysfunction; in vitro and in vivo models to investigate the abnormalities in neuronal receptors, neural activity, and neural networks that represent current clinical outcomes specific to the context of ART; clinical studies to adapt existing diagnostic neuroimaging- and electrophysiology-based methods to measure and monitor the activity of neuron ensembles and pathways; and novel technologies, such as virtual reality, 3D organoids, magnetoencephalography (MEG), neural probes, ultrasonic neural modulation, and photonics to decipher the underlying pathophysiology of mild to moderate forms of HIV-1 induced CNS dysfunction.


Dr. Greenberg commented that these are three significant high-priority issues, and he emphasized the importance of including mobile health interventions for youth in the continuum of care; providing structural interventions in healthcare systems; and exploring neural AIDS in the era of treatment. He also commended NIMH for pursuing these initiatives.

The Council unanimously passed the motion to approve the three concepts.

Public Comment

Dr. Maria Kovacs from the University of Pittsburgh commented on the need for two different general pathways, or approaches, to suicide prevention that take into account the differences between people who succeed on their first suicide attempt and those who repeatedly show suicidal ideation with or without a plan and have previous attempts. She noted that the window for intervention is very narrow for the first group and emphasized the importance of developmental psychopathology. Dr. Kovacs said that effective long-term prevention requires that the developmental precursors of the psychopathological behaviors are acknowledged and that interventions are conducted earlier in life.

Dr. Gordon thanked Dr. Kovacs for her remarks. Hearing no further comments from the public, Dr. Gordon adjourned the meeting at 12:45 p.m.

Joshua A. Gordon, M.D., Ph.D.

Appendix A

Summary of Primary MH Applications Reviewed

Council: January 2017

IRG Recommendation
Category Scored # Scored Direct Cost $ Not Scored (NRFC) # Not Scored (NRFC)
Direct Cost $
Other # Other Direct Cost $ Total # Total Direct Cost $
Research 594 $802,655,115 450 $525,776,614 0 $0.00 1044 $1,328,431,729

Research Training

25 $40,266,780 9 $12,142,169 0 $0.00 34 $52,408,949


67 $49,411.660 36 $26,984,046 0 $0.00 103 $76,395,706


0 $0.00 1 $1,678,320.00 0 $0.00 0 $1,678,320.00


686 $892,333,555 496 $566,581,149 0 $0.00 1182 $1,458,914,704

Appendix B

Department of Health and Human Services
National Institutes of Health
National Institutes of Health
National Advisory Mental Health Council
(Terms end 9/30 of designated year)


  • Joshua A. Gordon, M.D., Ph.D.
    National Institute of Mental Health
    Bethesda, MD

Executive Secretary

  • Jean Noronha, Ph.D.
    Division of Extramural Activities
    National Institute of Mental Health
    Bethesda, MD


  • Patricia A. Areán, Ph.D. (16)
    Director of Targeted Treatment Development
    Department of Psychiatry and Behavioral Sciences
    University of Washington
    Seattle, WA
  • Deanna M. Barch, Ph.D. (16)
    Gregory B. Couch Professor of Psychiatry
    Department of Psychology, Psychiatry and Radiology
    Washington University
    St. Louis, MO
  • Tami D. Benton, M.D. Ad Hoc (Pending)
    Department of Child and Adolescent Psychiatry
    And Behavioral Sciences
    Children’s Hospital of Philadelphia
    Philadelphia, PA
  • David A. Brent, M.D. (17)
    Academic Chief
    Child & Adolescent Psychiatry
    Endowed Chair in Suicide Studies
    Professor of Psychiatry, Pediatrics and Epidemiology
    Director, Services for Teens at Risk
    University of Pittsburgh School of Medicine
    Pittsburgh, PA
  • Randy D. Blakely, Ph.D. Ad Hoc (Pending)
    Department of Biomedical Sciences
    Charles E. Schmidt College of Medicine
    Florida Atlantic University
    Jupiter, FL
  • Benjamin G. Druss, M.D., M.P.H. (18)
    Rosalynn Carter Chair in Mental Health and Professor
    Department of Health Policy and Management
    Rollins School of Public Health
    Emory University
    Atlanta, GA
  • Ian H. Gotlib, Ph.D. Ad Hoc (Pending)
    David Starr Jordan Professor and Chair
    Department of Psychology
    Stanford University
    Stanford, CA
  • Alan E. Greenberg, M.D., M.P.H. Ad Hoc (Pending)
    Professor and Chair
    Department of Epidemiology and Biostatistics
    School of Public Health
    George Washington University
    Washington, DC
  • Hakon Heimer, M.S. (16)
    Founding Editor
    Schizophrenia Research Forum
    Brain and Behavior Research Foundation
    Providence, RI
  • David C. Henderson, M.D. Ad Hoc (Pending)
    Department of Psychiatry
    Boston University School of Medicine
    Boston, MA
  • Michael F. Hogan, Ph.D. (18)
    Consultant and Advisor
    Hogan Health Solutions LLC
    Delmar, NY
  • Richard L. Huganir, Ph.D. (17)
    Professor and Director
    Department of Neuroscience
    Investigator, Howard Hughes Medical Institute
    Co-Director, Brain Science Institute
    The Johns Hopkins University School of Medicine
    Baltimore, MD
  • Lisa H. Jaycox, Ph.D. Ad Hoc (Pending)
    Senior Behavioral Scientist
    Health Program
    Rand Corporation
    Arlington, VA
  • John H. Krystal, M.D. (19)
    Robert L. McNeil, Jr. Professor of
    Translational Research
    Chair, Professor of Neurobiology
    Chief of Psychiatry, Yale-New Haven Hospital
    Department of Psychiatry
    Yale University School of Medicine
    New Haven, CT
  • Marsha M. Linehan, Ph.D. (17)
    Professor and Director
    Behavioral Research and Therapy Clinics
    Department of Psychology
    University of Washington
    Seattle, WA
  • Gregory A. Miller, Ph.D. Ad Hoc (Pending)
    Professor and Chair
    Department of Psychology
    University of California, Los Angeles
    Los Angeles, CA
  • Maria A. Oquendo, M.D. (17)
    Ruth Meltzer Professor of Psychiatry & Chairman
    Department of Psychiatry
    Perelman School of Medicine
    University of Pennsylvania
    Philadelphia, PA
  • Rhonda Robinson Beale, M.D. (19)
    Senior Vice President and Chief Medical Officer
    Blue Cross of Idaho
    Meridian, ID
  • Mary Jane Rotheram, Ph.D. (16)
    Bat-Yaacov Professor of Child Psychiatry
    And Behavioral Sciences
    Director, Global Center for Children and Families
    Director, Center for HIV Identification Prevention
    And Treatment Services (CHIPTS)
    Semel Institute and the Department of Psychiatry, University of California, Los Angeles
    Los Angeles, CA
  • Elyn R. Saks, J.D., Ph.D. Ad Hoc (Pending)
    Orrin B. Evans Professor of Law
    Gould School of Law
    University of Southern California
    Los Angeles, CA
  • J. David Sweatt, Ph.D. (16)
    Professor and Chairman
    Department of Pharmacology
    Vanderbilt University
    Nashville, TN
  • Hyong Un, M.D. (17)
    Head of EAP & Chief Psychiatric Officer
    Blue Bell, PA
  • Christopher A. Walsh, M.D. (19)
    Chief, Division of Genetics and Genomics
    Boston Children’s Hospital
    Bullard Professor of Pediatrics and Neurology
    Harvard Medical School
    Boston, MA 

Ex Officio Members

Office of the Secretary, DHHS
Norris W. Cochran
Acting Secretary
Department of Health and Human Services
Washington, DC

National Institutes of Health
Francis Collins, M.D., Ph.D.
National Institutes of Health
Bethesda, MD

Department of Veterans Affairs
Theresa Gleason, Ph.D.
Acting Deputy, Chief Research & Development Officer
Office of Research & Development
Department of Veterans Affairs
Washington DC

Department of Defense
John W. Davison, M.B.A., Ph.D.
Chief, Condition-Based Specialty Care Section
Clinical Support Division, Defense Health Agency
Department of Defense
TRICARE Management Activity, OASD (HA)
Falls Church, VA

Liaison Representative

Paolo del Vecchio, M.S.W.
Center for Mental Health Services
Rockville, MD