Skip to main content

Transforming the understanding
and treatment of mental illnesses.

Celebrating 75 Years! Learn More >>

 Archived Content

The National Institute of Mental Health archives materials that are over 4 years old and no longer being updated. The content on this page is provided for historical reference purposes only and may not reflect current knowledge or information.

Social Disconnection and Late-Life Suicide: Mechanisms, Treatment Targets, and Interventions - Day Two


Jovier Evans: Good morning everyone we're really glad that you made it back for day two. The emphasis on today is to try to take the information from yesterday around translation of targets to actual implementation of interventions. With that, I'm going to turn the agenda and the day over to our co-chairs Dr. Martha Bruce and Dr. Yeates Conwell.

Dr. Yeates Conwell: Thank you, Jovier. I'll kick things off. Let me just echo, thanks to our technical crew for making this work and all of you for participating as well. Today we really want a lot of creative thinking and input as we move along this translational research spectrum from looking yesterday at mechanisms and targets and today our mandate being translating that knowledge and those targets, those opportunities to implementation of interventions through encouraging mechanistic thinking. We had two sessions yesterday, just very quick overview of where we've come looking at mechanisms, things about the relationship between social disconnectedness and late-life suicide, social connectedness, and health and well-being clearly are intimately tied.

That's very compelling evidence and really important background understanding for why we think this is such an important effort we're undertaking here today that we then looked at interpersonal and environmental factors, issues of function, coping style, family, spiritual and cultural context, lots of important issues raised there by our speakers yesterday and then moving into targets and which overlaps of course with our understanding of mechanisms, looking across the spectrum from neurobiology to what big data can tell us about that and then cognitive and functional decline.

Dr. Martha Bruce: Yates? Yates? I just wanted to say that one of the most wonderful things about yesterday, besides, it being a wonderful day, was having Alex Dombrovski and Amy Buyers in the same session, talking together and people from such different perspectives and yet having such a meaningful conversation. That was just really cool.

Yeates: It was, there's a lot of cool things. Our discussion tried to focus on themes and threads, and I think we realized how challenging that is because of the richness of the information that we have to call on here and making this work. We know a lot about life suicide. I think we've learned over time. We know a lot about social connectedness, and I learned more yesterday was terrific.

We know a little bit about how the two interact, but really need to flesh out what that overlap is, how they come together, how they influence each other. One of the things that just struck me so much is the role that social connectedness and social factors play as moderators. It's like the air that we breathe as these other factors are operating at the interface between social connectedness or other risk factors rather and late-life suicide.

The moderating influences which poses real challenges for us to study and to understand, but I think tremendous opportunities as well and that's our job today is to look at what those opportunities are actually for translating that knowledge into implementable interventions that are going to save lives. Any other thoughts, Marty, by way of background and how we want to set this up?

Dr. Martha: No, we have a couple of minutes so if anybody from yesterday panel want to jump in and say anything, this is a moment to do it quickly or any of the participants, we probably have a second to have any comments that would move the conversation forward.

Yeates: I think we might mention also in case there are folks speaking today who weren't able to join us yesterday. That's a possibility that we'd ask you to introduce yourselves very briefly at the beginning of your presentations, then we will be operating a little differently today, in that, we will have the presentations a round table discussion but then we will be moving into breakout groups and then coming back together, that should be a productive part of the day as well, please, bear with us till the end of the day, which will be at three o'clock I think.

Dr. Martha: I think the other thing yesterday, there were some nuances in the schedule that had to get changed and that will likely happen today and we'll try to keep people abreast of those things, but we still view ourselves as sitting around a table someplace in the Washington Metro area and having that flexibility. I think we will do our best to keep people aware of what's changing, but we will also try to be flexible to accommodate the conversations and those needs.

Dr. Martha: I think we've already said this, but I'll say different than yesterday is that all the people who are on the participant list will have their mics available to open up so when we are in the discussion components of the talks. I will just join Yeates and saying welcome, we're really pleased you're here.

It's really a gift to be able to step back from our everyday lives, which for many of us are on Zoom and Zoom, Zoom, Zoom, and to be able to step back and hear our colleagues who we don't see often and think and it's a gift. I'm going again, thank Sylvia and Liz Neca for putting this together. All right. I think we can go, so we get going on the next session. You have a couple of minutes.

Yeates: Sure. We can always [crosstalk]--

Dr. Martha: Always use the time. This session as Yeates said is if the overall session is from talking from translation of targets to implementation of interventions and we're thinking of interventions using mechanistic thinking, thinking about what those interventions should be, but also can we think about interventions in a way that would make them something that actually could be implemented and disseminated and scaled in the real world?

Our first part of the session and we have two talks, and these are both addressing, improving measurement of social disconnection and enhancing behavioral engagement, and late like psychosocial interventions. We have two talkers, Carla and Dimitri and again, when you start your talking, you give a very brief introduction of who you are and then just get going. Carla, if you would share your slides and we'll start, thanks.

Carla Perissinott: Good morning, everyone so my name is Carla Perissinott I'm an associate professor of medicine at UC San Francisco. I'm a geriatrician and palliative medicine physician and I've spent a little bit over the last 10 years, really looking specifically at loneliness, actually in health outcomes, and looking at it through the lens of being a geriatrician and how this impacts actually my own patients. I am also the medical director of care at home, which is a practice that sees homebound older adults in San Francisco.

This is something that's very near and dear to me and very relevant against my clinical practice. What I want to start with today is really again, do a brief review of the definitions and how this impacts our measurement tools. I'm going to use two examples from my work, one being worked with something called the Curry Senior Center here in San Francisco and then second is a study that we have ongoing on isolation during COVID to really talk about some of the challenges we've had as researchers doing a real-time study during the pandemic, and how this relates to measurements and interventions, and then I'll spend a little bit of time talking about the role of health care and then really focus on the gaps.

We spent a lot of time talking about this yesterday, and I thought Julian did a really great job obviously, really talking about how-- when we talk about loneliness and social isolation, often they're really looked at almost as-- they often are talked about simultaneously or is the same thing. We're not often discussing the different factors of all this including the emotional, functional, and structural aspects of this.

As we talked about yesterday, really, when we think about this concept, the term social disconnection can really encompass all of this. The challenge of this when we get into measurement as we'll see is many of the measurement tools that we're currently using, either in research or in practice, don't actually encompass the entire term. What I want to use as a framework and what helps me think about this is that when I am either in my own research or in practice trying to think about what tool do I want to use or how will I be measuring these concepts? The first thing I have to do is really think about what is the issue that I'm trying to define and what am I looking at? Am I looking at isolation? Am I looking at loneliness? Or am I looking at exactly both or the entire congregate term of social disconnection?

The second part is thinking about who is the population that I'm interested in? For my work, as I said, I focus on older adults which means that while I may pick a tool specifically for isolation and loneliness, if I'm not also assessing some other co-variables that can impact loneliness and isolation, then I'm also potentially impacting my results and the interventions. What I mean by that, for example, in older adults, some of the things that we see that impact these experiences are, for example, sensory impairments. The third part is understanding the mechanism.

Again, in this population, once you've decided, how do I think, or what are the factors that could be making that person isolated, or lonely? Driving from my own work with my homebound, older adults, I have many older adults who are homebound and isolated because of stairs. That's kind of the structural components and if I'm not thinking about, again, those mechanisms. Then when we think about interventions, we're potentially failing by not looking at the whole picture.

Then getting into the interventions, again, one of the challenges with interventions in the work of interventions right now that was highlighted in the National Academy of Sciences publication is that there is not a lot of great data and the data that is one of the challenges is really that there are interventions that are applied to a broad area of circumstances or a broad group of people without truly understanding what is the theory of that intervention? How do I think it's working and what am I targeting? My hope is that through these conversations today, we get into those nitty-gritty and help the greater community be at the research community, or those of us working in the community understand what we're trying to do and be better targeted.

Then, of course, the end result is measuring our impact and then how do we implement and scale this, which really relates to all the topics that I just went through. As I know, some of this was discussed yesterday, but just to review, from a medical perspective, or from a healthcare perspective for me as a clinician, when I think about wanting to screen or identify loneliness or isolation, we do not currently have a gold standard for that.

That makes it challenging to know and to think about policy in terms of how to encourage health system to do this. There are many different measurement tools as we know, all of them have benefits and drawbacks, most of them have been developed for research purposes, and not truly for clinical use, so that creates some challenges with validation. Some of the most commonly used skills are here, you're familiar with what many of them I'm sure and one of the newer ones is called the ESIR is which is Extreme Social Isolation Risk, and which was supported by AARP and the idea behind this screeners that it really is looking at the multi-dimensions of social disconnection and thinking about us, putting people into risk categories.

This is a new scale this year and so I think we're still-- though it's been tested in quite a few different communities I think the next steps here to really think about can these types of new skills be applied more broadly both in research and clinical settings? But of course, other than the ESIR, to my knowledge, we do not really have a great congregate social disconnection measurement tool that is widely used clinically. That creates again some challenges for us and opportunities as researchers and as clinicians.

Going through the two examples, I wanted to start with the study that I've been working on this project with the Curry Senior Center in San Francisco, and this is from my own work. The Curry Senior Center is a senior center that is a nonprofit in partnership with the Department of Public Health, it is in the Tenderloin of San Francisco, which is known as one of the most diverse and underserved neighborhoods in the city. There tends to be a lot of old adults with premature aging because of mental health issues and substance abuse problems. I was approached by Curry Senior Center several years ago to help them think about how to evaluate their program which was a peer mentorship program.

What started with this and what was interesting about this is that for me, as a researcher as a geriatrician was that there were many assumptions that were being made with this program, and using this as an example, but of money. For example, the term isolated was being used broadly without really understanding what are we talking about? Are we talking about truly being physically isolated or are we or are we talking about loneliness?

That was some of the first things that we needed to get through. The reason I bring that up is that this is an important thing-- that important point for me, as a researcher to learn about when I'm working with community-based organizations, is to we often speak different languages, and how do you come to the middle ground to make sure that we're talking about the same thing? We adhere to research methodology, but while also respecting the rights and the needs of the community that we're working with.

Once we got through some of these terms, it became really important for us to think about what are we actually trying to do with this program, and are we really thinking about this, peer mentorship program as a tool for isolation? Or are we targeting loneliness? The second part, which is also a really difficult thing with a combining the role of a community-based health organization and a researcher is that there's sometimes this fear or this hesitancy of working with academicians and researchers that we're going to ruin their work by getting involved which is probably sometimes true, but there's a way to how do we align and make things tactful and say, it is actually okay for us to ask these things in a structured way, but how do we build that trust to be able to do that? That was a lot of my work upfront to be able to figure that part out.

The last part, in this project, and I'll go in and more detail is understanding the unique experiences of this population and how to capture that. What I mean by that is saying, as I mentioned that the Tenderloin is one of the most diverse neighborhoods in the city which meant that we were going to have to understand the linguistic differences of our clients, we would have to understand the mental health challenges of some of the clients there and substance abuse and other stigmatization that can happen.

These were all things that we really needed to sort out as Curry Senior Center develop their program and intervention, but for me, as a researcher, as I thought about what are the research questions, and how can I make sure I'm fully capturing the experience of these adults? With that in mind, I was so grateful to work with a fantastic group at Curry Senior Center. They have such experience working in the community that they were so helpful for me to understand exactly this part of who's the community that I'm working with? Because they have, they're so deeply rooted in the community, they had a list of items that they knew were affecting their clients ability to engage and in the outside world.

We use that list to really come up with a tool to really think about what are the barriers that this community has to going in socializing or engaging with others? This is just a screenshot from our redcap survey and as you can see, some of the things that we ask are really in layman's terms, or in general terms for our clients to understand which is do you just do you ever decide not to go out and engage because of your mood? How about your physical ability? Safety of the neighborhood is a very big a concern amongst the participants because of the location, substance abuse here and then there's things that are often very hard to talk about such as incontinence, which can also affect people wanting to engage. What we learned from this, again, is that if we did not ask these questions, we would be missing on a lot of the reasons why these clients were isolated and lonely, as we found out. Physical ability and safety were very high reasons for not going out and engaging, as were financial insecurities.

Again, I think if we looked solely at one of the dimensions of either loneliness or isolation, we missed a whole opportunity to intervene. What we ended up learning from this study, and our results are still waiting to be published, although we're excited to say that we actually did see dramatic decreases in loneliness scores, as measured by the UCLA loneliness score, as well as decreases in depression scores, as measured by the PHQ. What we know is that if we don't ask these questions, we don't know how we can help these individuals.

We did also find that by really delving into the intricacies of what this program is aimed to do, which is really an implementation science framework, we recognized that the intervention was really that the use of peers is more really targeting loneliness, specifically, as we find out, but in the process, affecting isolation as well. The other finding we have from here, which is a challenging thing, is what you find may not be what your funders want.

What I mean by this is that this program was funded initially by the Mental Health Services Act because the presumption was that these adults living in the Tenderloin needed to be connected to services. That was the goal of the program from the funders' point of view. The thought was that if the peer connects with the individual, they will connect them to a social worker, they will connect them to whatever.

What we actually found is that that is not at all what these individuals wanted and that the peers served a much different role of actually being not an employee and not someone associated with the city and county, theoretically, but really a peer and a friend who could accompany someone. It was very peer and participant oriented and goal-directed in terms of what their goals were, and not what the city and county wanted, which is fantastic in many ways because, again, we were able to uncover unmet needs but very different from what was desired by the funders.

The next thing I would be remiss without saying is that the effects of the pandemic. This intervention is very much reliant on in-person interaction. We were at the end of the study period, though, the project or the program is ongoing, and suddenly, the shelter-in-place orders in San Francisco wreaks havoc on this program, for both the peers themselves and the clients, in terms of the peers becoming worried about their clients and not being able to see them, and having to shift to telephone-based work, and not knowing if that was going to be as effective.

Again, that to me highlights the importance of in-person personal contact for some of the interventions that we're using, but not really knowing the extent of what's needed. I believe, as Julian mentioned yesterday, that's one of the things that's missing in the interventions. Literature is knowing, "Is there a difference in effect sizes and effects in general when you look at in-person versus virtual interventions?" I will run through this quickly.

What was again very interesting about the Curry Senior Center project was really understanding what were the peers doing in this program, and why I think it was actually targeting loneliness, more specifically, is that the peers with the clients were really focused on many things. They were absolutely focused on enhancing social support. They ended up increasing opportunities for social interactions, which was part of, I think, one of the goals of the MHSA, and probably doing some on improving social skills but less on maladaptive social cognition.

I bring these up because I find that paper by Masi et al from 2011 really fantastic in terms of thinking about a framework for how we think about interventions, specifically on loneliness, but I believe you could think about this more broadly. Turning now to my second example, which is thinking about loneliness and isolation during COVID-19, so my colleagues Ashwin Kotwal, Julianne Holt-Lunstad, and myself embarked on a study early in the shelter-in-place orders here in the Bay Area of San Francisco Bay Area, surrounding San Francisco.

What we wanted to do is really understand, "How were adults going to be impacted by shelter-in-place orders?" We had some questions to start with, us really thinking about, were we concerned about loneliness or isolation or both, and more broadly, the social disconnection, but knowing we didn't have a good measurement tool for that. The other questions we had, given that this was a new situation with the pandemic, is not being entirely clear as to how often we needed to measure and ask our participants in our study about loneliness and isolation.

The third question that came up and more towards the middle and now is ongoing is, are our current measurement tools that we have sufficient during shelter-in-place orders, given that many of the measurement tools and the one that we used had elements of the Berkman-Syme that really are more reliant, I believe, on in-person interactions and don't distinguish between in-person or telephone interactions? Are we really capturing things well?

Lastly, again, what's interesting with some of these questions that are very nuanced on loneliness, that we know that there are differences in how people respond, depending on if they're by phone in person, or mail, and knowing how would that change things given that everything that we're doing is certainly by phone or mail. Our study, we recruited from three different sites across the Bay Area. We did use some senior centers and we also used some of our geriatric clinics at UCSF.

We used volunteers to interview. We actually took advantage of being in the pandemic and having medical students and other students who did not have to go to class and so could help us and wanted to get some research experience. In terms of the timing, we did decide to do phone surveys primarily. We were conducting these surveys every two weeks. Again, thinking that as shelter-in-place started, things were progressing so quickly that if we didn't ask these questions so frequently, we might miss big changes, given that we weren't really sure how the pandemic was evolving.

The study is ongoing, and I'll talk a little bit about that in a second. This slide has since been updated, as we have more participants, but a large proportion of our study, participants are female, over 60% lived alone, a lot of them have financial insecurities, pretty significant comorbid depression, anxiety, and pretty significant functional impairments as well, which is relevant to the work that we talked about yesterday, in terms of the role of functional impairments in social disconnection.

We had a pretty diverse population, with multiethnic participants, and about 9% of our respondents were Black, which is higher than some of the other surveys I've seen. That's still lacking. As we also talked about yesterday, we know that there's a huge gap in understanding some of these risks in minority populations. Our key findings from this are that what we have seen so far in the pandemic, and specifically in the shelter-in-place orders in San Francisco in the Bay Area, is that many older adults experience worse in loneliness and depression, and anxiety during the pandemic, but interestingly, we also see some people that, over time, have been coping much better.

We also found, which is very important as we think about interventions and how we think about interventions during pandemics, is that the majority of older adults that we surveyed, and though our population was certainly geared towards more in low-income populations that still certainly highlights a gap and a divide in our population, is that most were not using technology for social connections. That is so important because we've had to move everything so much to technology-based, and it's showing us that there's a big gap, and we're missing a lot of people.

We did ask questions, both on loneliness and isolation, and I've seen effects on that. As I mentioned, what's been challenging to understand with these results is that people were obviously forced into a shelter-in-place. How does that impact how we think about this? The key questions that we still have at this point and we're trying to sort through are, how quickly does isolation or loneliness change during a pandemic? Again, we struggled with this in knowing the time interval for our questionnaires and how we define isolation was very challenging as well, again, because we have people that are having to shelter-in-place, but within that, there's clearly some differences in people who are more isolated than others because there are some people that were still having some contacts, with some people in person or by phone, and some that had zero.

That, again, led us to this question, is what is the right tool to use during shelter-in-place orders at this type of situations, which is automatically limiting social interactions, and do we need to think about new tools, for example, either pandemic-like situations or even populations who are homebound, how do we distinguish, again, between the interactions, as I said.

Then, the last parts that we're struggling with is thinking about-- We started the bulk of our recruitment at the beginning of shelter-in-place in March, yet we have this desire to continue this study and think about what about having a new cohort, but clearly things are different. As we've seen from the data in our own study, if we are seeing adults that are adapting over time, what will that mean for this cohort now and how do we do analytics on that? These are some of the ongoing research questions we have.

Then, finally, when to end the study. I think we were a little bit too optimistic in February and March when we started this and thought, "Six months, we'll be done." Our consent really informed our participants that we would be doing surveys every two weeks and then six months post-pandemic. Well, this post-pandemic can now be 10 years from now. What does that mean? We're really struggling with what are the correct time points to use. I look forward to discussion and guidance from some of you as we move forward on this.

Moving on to what does this all mean for health care, knowing that we have community-based organizations such as Curry that have great interventions that are potential for some of our patients, or my patients, or some of the information that we were learning from COVID-19. In the role of health care, I think it's important to think about social disconnection from a population health framework and thinking about-- As a clinician, or as a public health advocate, I think about primary prevention and secondary prevention.

This is a good framework for us to use. When we're thinking about evaluating or identifying social disconnection, the first part is primary prevention, so identifying those people who are at risk so that we can really try to prevent things from setting in. There's a lot of different ways to look at this, but we know that people of lower socioeconomic status, women, LGBTQ, older adults, and people with recent losses are probably some of the people amongst the highest risk.

When we think about secondary prevention, this is thinking about people who are already lonely, or isolated, or socially disconnected, and how do we mitigate the effects of this. This will require knowing which interventions work, which is why the topics today we're discussing are so important. Another way to think about this, and this is some work from my colleagues, EAKC at AARP and Jeremy Nobel at Harvard are thinking about a pyramid of vulnerability, is who is at risk, which is going to be the biggest group of people, who's already starting to disconnect, and who is already incredibly isolated that's probably going to need more intensive care.

This pyramid really follows a population health framework in terms of having work that is really at the screening and evaluation level first and then building up to more intensive interventions, the more at-risk or the more isolated and disconnected someone is. The other part-- Yes.

Dr. Martha: I think we'd have to have to wrap up.

Carla: Yes, [crosstalk].

Dr. Martha: [crosstalk] to wrap up because it's so interesting, but--

Carla: No problem, yes.

Dr. Martha: Okay, thank you.

Carla: I will say that the last part of this, we know that health care is really focused on social determinants of health right now, and social disconnection probably feeds into this. What we know is that clinicians are actually interested in understanding social disconnection, but they don't know how to implement and how to measure. I will skip the slides on what the IOM has recommended, but what I will say, and this is probably a good place to stop, is that what we know from the National Academy of Science's report from February is that there's a recommendation of identifying those at highest risk, which is what I started with saying, thinking about periodic assessment, but we don't know what that periodic means and which tool to say, although we know that there are validated tools.

Then, lastly, it's forming those partnerships between clinicians and researchers to really think about how do we bring the science, and the clinical worlds, and the community worlds together towards interventions. I will stop there. I apologize for going over time, I get too excited talking about this. [chuckles]

Dr. Martha: Thank you. This is really great. As Dimitris gets ready to present, I just want to thank you for helping us all really frame ourselves in thinking about who are we really targeting and what is the focus, isolation, loneliness, both, how do we measure those things, what is the context of all of that, and what do we want to do about it, and how do we go about it, even from computers to telephone to in person. I think keeping that in mind is going to be so important for us going forward. Thank you. This was great. Thank you.

Carla: Absolutely. Sure.

Dr. Martha: Our next speaker is Dimitri Kiosses, a long-term colleague from Cornell. Dimitris, if you'd put up your screen and introduce yourself, that would be

Dimitris: Okay, great. Thank you so much. I'm Dimitris Kiosses, and I'm an Associate Professor of Psychology and Clinical Psychiatry at Weill Cornell Medicine and an Investigator at the ALACRITY Center of Cornell, which is directed by George Alexopoulos. My research focuses on psychotherapeutic interventions, and that's what I'm going to talk about today. I want to thank Liz Neca and Jovier Evans for organizing and for inviting me to this excellent workshop. Also, I want to thank Martha Bruce and Yeates Conwell who are chairing this workshop. It's a great pleasure and honor to be here.

This is my research support. Today, I'm going to speak a little bit-- I'm going to give some data from the R61/R33 grant from NIMH. Co-investigators are James Gross, who's an expert in emotional regulation, and Greg Hajcak who's an expert on EG. My presentation is going to be divided in four small parts. In the beginning, I'm going to talk a little bit about some challenges, and some thoughts, and questions about suicide prevention research in general.

Then, I'm going to describe where we stand in terms of psychosocial interventions for late-life suicide prevention and some considerations. When I'm talking about psychosocial interventions in this presentation, I'm going to focus mostly in therapy like psychotherapeutic interventions. Then, I'm going to discuss a little bit about social disconnection and late-life suicide. Finally, I'm going to present the work that we're doing here in our lab about the mechanistic approach to suicide prevention.

There are a lot of questions and a lot of challenges about suicide prevention research. There are some that I want to put out there that we may discuss also later. Of course, there is difference between suicidal behavior and suicidal ideation. Then, there is the questions, do the risk factors for suicidal behavior are the same for suicidal ideation, and it seems that they're not.

I have another question. In terms of the wide range of suicidal ideation, does active suicidal ideation with intent or plan, is it closer to suicidal behavior or to passive suicidal ideation? Understanding a little bit more about whether active suicidal ideation with intent may be closer to what we're talking about, suicidal behavior. Another question is could social isolation be a suicide risk factor for a specific person in one situation and not in another situation, how do different risk doctors, and this is what we have been discussing, how do different risk factors interact with social isolation.

Most importantly, for me, is that how do risk factors change as we approach suicidal behavior, so what happens a few days before the suicidal behavior? What happens a few hours before? What happens right before suicidal behavior? Our lab focuses on a period close to suicidal behavior. It's very difficult to have just suicidal behavior but appeared close to suicidal behavior. We focus on the personal situation, specific approach that I'm going to discuss. Most psychosocial interventions for late-life suicide prevention tested in RCTs. More psychosocial interventions with suicide-related outcomes, were initially designed for specific disorders, either depression or borderline personality disorder, which implies in a way that we think that by reducing depression, we may reduce suicide risk.

The mechanism of action for suicide, it seems to be different from the mechanism of action for depression. As a result of these psychosocial interventions that were mostly designed for specific disorders. Many studies have secondary analyses of a sample with patients with relatively mild forms of suicidal ideation. We need more studies in older adults with more severe forms of suicidal ideation.

We also need to consider and to address comorbidities such as cognitive impairment, disability, and pain that are prevalent in older adults. One of the thinking that we have been having a discussion in our lab is like, do we need different interventions for different severity of suicidal ideation or suicidal behavior? Or does the dosage of the intervention vary or depend on the intensity of suicidal ideation or the severity of previous suicidal behavior?

These are some questions that we struggle, about the psychosocial interventions. Of course, we do have underrepresented groups. That's true for late-life psychosocial interventions in general, racial and ethnic minorities, patients with low education level and low socioeconomic status that we definitely need to address. Social disconnection, this is a summary of the slide that Dr. Conwell put yesterday.

He describes different aspects of social disconnection and social connection, social support, social isolation, loneliness, and I add the interpersonal conflict that Dr. Conwell has been talking about in his peer intervention study. One of the things that I just want to mention here is that I'm a researcher that focuses on emotion regulation. My question is, what is the role the negative emotions play in suicide prevention related to social disconnection? The way that we have addressed this issue so far, and I'm repeating what we have come as a conclusion, is social disconnection may directly lead to suicidal ideation or behavior or create the environment through which suicidal ideation or suicidal behavior develops. The question for me is that do negative emotions play a role in that connection between social connectedness and late-life suicide?

Again, the role of cognitive impairment that I think it's much more prevalent than we think, in late-life suicide. In the studies that I'm doing for people who have been hospitalized for suicidality, 50 years old and older, about 30%-- I will talk about the study about 50 years old to 84. 30% have dementia rating scores comparable to MCI and 10% have scores comparable to dementia.

What I'm thinking is that as people become more emotionally charged or more closer to suicidal behavior, it may mean that their cognitive performance is not the same, so that their negative emotions may affect how they react to the situation. We definitely need to address this in psychosocial interventions, but also, I believe in the connection between social disconnection and in the late-life suicide. I'm going to talk a little bit about the approach that we have in our lab. What we're trying to do is to identify factors related to suicidal behavior.

What we're doing is we tried to-- We said, "What is the best approach that we can do?" We decided to focus on those patients who are hospitalized for suicidality. These are patients who either have had a suicide attempt, or they have had active suicidal ideation, in the Columbia Suicide Severity rating scale above or equal three, which means not a specific active suicidal ideation.

The assumption is that maybe these two groups are similar. We don't know but this is the assumption. We say, "Okay, let's identify this group. Let's see what happened in this group before they were hospitalized, and how we can help them in the three months after hospital discharge which is a period of high suicide rate." What I'm going to describe is the theory and then the personal situation-specific approach that we have. I hope that this may give us some understanding or some help in how to go forward in developing psychosocial interventions for social connectedness and suicide.

The assumption that we have is that hospitalization for suicidality is preceded by an emotional crisis, the perfect store, that leads to suicidal ideation and suicide attempt. Now, the second assumption is that this crisis is patient and situation specific. These triggers are patient and situation specific. For this particular person, in this particular situation before the suicidal behavior. This combination contributed to negative emotions that contributed to suicidal ideation or suicidal behavior. The idea is, if we identify the emotional crisis right before the active suicidal ideation before the hospitalization, before the suicide attempt.

Identify these personalized triggers and the negative emotions associated with them, and provide the patients with emotional regulation strategies, then we might be able to help them after they are discharged. That's the theory that we have. What we did is we developed an emotion regulation intervention for suicide prevention. We selected a particular type of emotional regulation because we wanted the many techniques about emotion regulation.

We wanted to see, is there a particular type of emotion regulation, and specifically cognitive reappraisal that may help reduce, that we may help improve in these patients after hospital discharge and they may reduce in suicide risk? When we talk about cognitive reappraisal, we talk about changing perspectives, modify the appraisal of the situation, to alter its emotional significance. What we did is we went through the neuroscience literature to our psychotherapeutic literature, and we collected the best cognitive reappraisal strategies and we incorporated it. The idea for the R-61 was to evaluate the target engagement and the target is cognitive reappraisal.

We have 35 subjects, the mean age 60 years old, and 50 to 84, as I said. 82% Caucasian, 60% African Americans, and 16% Hispanic. One of the things that I noticed after we did a study is that I looked and I said, "How many of these people may have some difficulty in the two items of getting along, which is dealing with people they don't know, or maintaining a friendship. I found that 70% of these patients have at least moderately difficulty, not mild, moderately difficulty in dealing with people they don't know, or in maintaining friendships.

There's an aspect of social disconnection there that I wasn't aware actually, before. You can tell within the personal conflicts that you see when you do the therapy, but this is a high percentage. These are people who have been hospitalized for suicidality. We developed this intervention. What else we did is that we also added, I'm sorry about the slide, we also added an application. A tablet that we gave to the patient, we incorporate the techniques that we wanted them to use when they were faced with a crisis. Then we asked them to use. Not all of the patients used them, about 14 or 15 subjects used them. Here are some of the results. We measured cognitive appraisal using two measures. One is a subscale of the emotion regulation questionnaire, which asks the patient whether they have changed the perspective they have used cognitively appraisal strategies to influence their emotion. We see here that there was an increase in cognitive reappraisal strategies and cognitive reappraisal scores, and higher score is more cognitively appraisal.

Another subscale that we also have in ERQ which is Emotion Regulation Questionnaire developed by James Cross is expressive suppression, which is suppressing the expression of the emotions. One we asked him about cognitive reappraisal, we also asked him about expressive suppression. We didn't see any significant difference here from admission to week 12. Now, the only thing that we did is we gave EEG electroencephalogram, before and after, right at the beginning of the intervention, and 12 weeks at the end of the intervention. Here, what we do is we show them a negative picture about zero, and then about 1200 milliseconds, we ask them to reappraise or react.

Here, what we see in the preliminary result is that the higher line is ATP when they reappraise negative at week 12. The lower line is when they appraise negative at baseline, like before the beginning of the treatment. After week 12, we see a lower emotional response. That's how I said it in [unintelligible 00:51:51]. We may have some signal that this may affect cognitive reappraisal.

Dr. Martha: Dimitris, you're going to have to start wrapping up. I know you have a lot to say, but--

Dimitris: No, no.

Dr. Martha: Thank you.

Dimitris: This is the intensity of suicidal ideation for the treatment. The last slide that I have is I also checked whether there was any significant improvement in the two items that we talked about. We saw a significant improvement in these two items. Also, we asked those patients that had high scores that had these difficulties in dealing with people to maintain a friendship, what were the negative emotions that they may have.

They were sad, angry, nervous, lonely, and negative emotions of themselves. Potential targets that we thought is increasing social connection, and [unintelligible 00:52:50] have great techniques that we can do, or maybe reducing the negative emotions that are associated with social connectedness that may help us reduce suicide risk. Thank you so much. These are my co-investigators and consultants, and this is a great group. I wouldn't be here without them. Thank you. I appreciate it.

Dr. Martha: Thank you so much, Dimitris. This is really, really interesting and important work. Greatly appreciate sharing it with us. Okay, Yeates.

Yeates: Yes. Great. Also after that, Dimitris, thanks very much. Also thanks to those who are contributing through the chat box. There's some interesting comments and interactions going on there. But let's move quickly on at this point to our next two speakers. First is Kim Van Orden. Kim, would you like to introduce yourself, please and kick it off? Thank you.

Kim Van Orden: All right, help if I would unmute myself and not be socially disconnected while talking to you. I'm delighted to share my work with you all today. I'm going to talk a little bit about my take on the literature on this topic, and then share some of the work that I've been doing with my colleagues. I'm really delighted to be here today, because I'm looking at connectedness as a way to reduce suicide risk in older adults is what's on my research t-shirt.

It's nice to have an entire meeting focus on what I spend every day thinking about. I'm delighted to get to be with you all today. To talk just briefly about social disconnection as a target mechanism to prevent suicide and some of the conceptual work I've done as well. Brief discussion about the evidence base for what we know about targeting social disconnection and then I thought I'd throw out a conceptual model for us to think about together as something we might use to further the literature in this area.

Not new information to anyone here, but important, I think to for us to keep in the back of our mind about how social disconnection fits in the overall constellation of risk for suicide in later life. This is a conceptual model that Yeates developed, thinking about why or how these different constellations of risk come together, and when we might have an older person that greatest risk. You can see how we're thinking about the overlap of these different dimensions of things like depression and comorbid diseases, access to firearms, social disconnection, and disability. When they all come together, we think about an older person as being at greatest risk.

I think one thing that would be interesting for us to think about as well is how all of these different factors are also related to disconnectedness. They're all very much interacting over time. In particular, if we think about, then, the different factors that are associated with social disconnection, you'll see a lot of overlap with those key risk factors for suicide in later life. We know from longitudinal studies with older people that social disconnection is linked with poor mental health outcomes, also linked with difficulties with cognition, physical health, and functional status. A lot of those are bi-directional links. A hopeful thing to take away from this is before able to improve social connection, we may have a lot of positive effects, in addition to improving mental health and preventing suicide.

Also important to keep in mind, I think, is that we know that suicidal behavior is more often lethal and older people. Thinking about more upstream type prevention is essential. Targeting social disconnection is a way for us to hopefully catch people before they get to a more serious suicidal state. Social disconnection and suicide in later life, I try to take some of the stuff from our previous presenters in terms of how they were conceptualizing the different aspects of social disconnection, and use that framework of thinking about structural, functional, and then emotional and quality aspects of social disconnection.

These are the ways that those different factors have showed up in the literature in terms of being empirically linked to suicide deaths, particularly among older adults. In terms of structural dimensions, we know that things like loss of a spouse, or living alone, having fewer confidants, less community engagement are all linked to suicide deaths in later life, as well as functional aspects such as low social support, and then the emotional quality aspects, things like loneliness, and interpersonal Discord. Then I'll meet mentioned briefly, a lot of my work is focused on the interpersonal theory of suicide.

I should mention, I'm a clinical psychologist, and so very much take a psychological lens to this. My beginning work was focused on this theory of suicidal behavior during graduate school. My mentor, in graduate school, Thomas Joyner was developing this theory right as I joined the lab. I got really excited about it. Now, I'm using that work as a way to think about mechanisms for interventions, but the model is on the right.

It's not a perfect model but I think it is one that can be generative. What we see there are two dimensions of connectedness in the red and blue circles. Those are dimensions of connectedness that might fall under that emotional or quality aspect of connectedness. Low belonging is similar, but not exactly the same as loneliness, the idea being that we have this need to belong need to feel connected to other people and to meaningful groups. When that need isn't met, there are a lot of negative health outcomes, including thoughts about suicide, but then as well, feeling like a burden on others.

The idea that when we start to think that others would be better off if we were gone, that can be especially problematic in terms of increasing risk for suicide. I know someone mentioned, I think, maybe Amy in the chat, the idea of, do these types of forms of disconnectedness compound. Do we see a dose-response relationship in terms of risk? That's a bit of what the interpersonal theory is trying to get at that. It's when we have this low belonging and feeling like a burden, that risk should increase. These factors can be especially useful, I think when we think about late-life and we think about some of the risk factors for disconnections such as disability and functional impairment, and how they could impact both belonging feeling like a burden.

The evidence base for connectedness interventions, there's a lot of work going on in the field and it's exciting, but the overall literature is not strong enough. Carla mentioned that the masi et al meta-analysis. I think that our takeaway from that meta-analysis is yes, behavioral interventions can reduce loneliness, but it doesn't really leave us with an idea of what are evidence-based interventions.

The idea of being able to know a specific intervention might work for a population, we don't really know. All of the interventions that were tested in that meta-analysis, none of them have been replicated, for example, and they all have different nuanced ways of doing it and it's unclear. Some of them were group interventions, ranging from reminiscence to cognitive behavioral therapy. A powerful conclusion about, yes, we can likely move loneliness, but it doesn't really give us a sense of what an evidence-based intervention is. It also leaves the mechanisms unclear. When you have a group intervention, is it the time together that's helping, or is it what you're doing in that group.

Then I think one of the most challenging pieces for the field is that a lot of the studies that look at reducing things like loneliness don't actually enroll lonely participants. Not requiring that people have a really elevated level of loneliness to be in the study, which is understandable. It's very hard to enroll people who are disconnected into your studies, I deal with that every day. It's very, very hard to take people who are very disconnected and connect with them.

The question for all of us, I hope we can talk more about today is do effective suicide prevention interventions promote connectedness? Again, the answer is likely yes, but we haven't really tested it. If we look at what we know has been effective preventing suicide deaths in older people. Things like caring contacts, the telehealth telecheck intervention, and things like depression treatments, all of those likely have been impacting connectedness, but it hasn't been directly tested so, we don't know for sure.

This slide shows an experimental therapeutics approach to thinking about suicide prevention interventions that I use in my work and thinking about how we can move these different aspects of connectedness that we know are important and thinking about it in particular for older adults. I know that masi et al, meta-analysis looked at this idea of changing maladaptive social cognition and that seems to be a really important thing to do. We know that in older adults, interventions tend to be more effective things like psychotherapy, when we use the behavioral strategies, as opposed to directly working with cognitions.

In my work, I've been trying to look at how can we help older adults change their behaviors in terms of connectedness, and then have this positive cascade. If we think about connectedness in terms of what the behavior is that we might be able to change, like to think about it as social engagement. We can help people be more engaged in ways that are meaningful to them and in particular thinking about it through the lens of the interpersonal theory of suicide. We want them to think about improving positive connections and contributions to other people. So not just feeling connected but how can we also feel like it's a reciprocal thing we're contributing.

In that work, there's lots of different types of ways we could do that, and I will show you some of them. Things like being a volunteer or doing psychotherapy ways that we can have them change those behaviors. Then hopefully those interventions should change our sense of belonging, feeling like a burden, and thereby have some improvements on clinical outcomes like suicide ideation, as well as depressive symptoms.

Very briefly, I'll go through a couple of studies Dimitris mentioned this one and this is a study that Dr. Conwell led, that I was able to work with him on right when I joined as a postdoc, and it was an early study. Really, really exciting because we thought about how can we take the interpersonal theory of suicide and use it to inform real-life interventions with older adults in the community, and so we looked at pure companionship offered through the Senior Core. The Senior Core is a nationwide program. It's the older adult analogous version of AmeriCorps. It's available nationwide, which makes it very accessible often offer through Area Agencies on Aging.

We took older adults who were at risk for suicide based on the interpersonal theory of suicide, so feeling lonely or like a burden. They were randomized to having a peer companion via our local aging services provider, or care as usual for up to two years. We asked them to do weekly visits, it was quite variable. We asked them to do in person and they get take a supplement by phone. These results just came out in AAGP. The primary outcomes, we found reduced depression and anxiety and reductions and feeling like a burden on others, but very interestingly, comparable levels of belonging to care as usual.

I'd love for this to be something if we have time for us to talk about as a group, when we're trying to do research on connectedness. We need to keep people enrolled in our studies for many, many years, we have to form relationships with them to promote retention and so we found that everyone in the study felt more connected. It's a challenging thing to study. Then what we're doing now in my lab, is we got a lot of feedback from that prior study saying, "This is great, but I'd actually rather be the volunteer, instead of having a volunteer come visit me."

What we're looking at now is exactly that idea. We're looking to see is serving as the volunteer out in the community, is that effective in reducing loneliness. We're comparing it to an active control looking at life review and having people think through their lives on their own, keep them engaged for a year, and see how that does in terms of loneliness. I do want to briefly just mention some of the other work that we're doing and Pat, maybe want to pick this piece up a little bit as well. Looking at engaged psychotherapy, which Pat and George developed to make the problem-solving strategies even more accessible to more patients easier to do.

We applied that to the problem of loneliness and feeling like a burden in my care world trial. We found that older adults were very excited with this approach, they were really excited to focus on social disconnectedness. We found a strong effect on reducing depressive symptoms, and some effect on social engagement, but no movement on loneliness, interestingly. Even though they were changing their social behaviors, after a couple of months we didn't see any movement on loneliness.

I'm going to go through this very quickly, but I just want to throw out the idea of looking at technology-based interventions. I am doing some work with computer scientists, at my university, looking at a virtual agent as a skills coach, as we're looking at now how older adults are able to have those conversations in a very well conversational manner, and how that might help them be able to improve things that can get in the way of connectedness, like making eye contact and smiling, so that's another future direction. I know I'm getting out of time here, so I'm going to wrap up with this idea of thinking about a menu of social engagement interventions.

Carla, was very elegant in her presentation today, helping us think through the idea that if we don't think about the barriers to disconnectedness, we're probably not going to be able to help older adults. That's what this slide is getting at as well, which is we can't just throw any intervention at an older person, it has to be personalized. A one size fits all, is not going to work for social disconnectedness, especially if you're looking at a population that's heterogeneous. A hypothesis I throw out is on the left, so something I spent a lot of time thinking about is how those of us who do psychotherapy and behavioral activation with older adults, no you don't necessarily have to get super precise in terms of what activity they pick they can pick something that they want to do, and that likely will improve their mood.

That doesn't seem to work quite as well in terms of social engagement. So if you have someone-- it seems like you need to be a bit more precise in terms of picking something that really targets their problem with disconnectedness. Say if you have an older adult who feels lonely, but is not home-bound and has some social interaction, you have to think about what's causing their loneliness and be a bit more targeted, but again, very much a hypothesis for us to think about and I was really excited to read Matty's paper with Mem K and Renée looking at behavioral activation and how that improved loneliness. I think it would be an interesting thing for us to chat more about in terms of how that may be a little bit different than what I've done with Engage. What I have up here are just some thoughts about what I would love for us all to come up with together, is thinking about how we might map interventions to the causes of someone feeling disconnected, things like health technology that Pat's going to talk about now and who do we want to target that for.

I have that. Then just my last slide just so that you all have the references, where I would say we're at with the field, just that positive connections that involve carrying contribution may save lives. I'm really looking forward to us being able to directly test that mechanism and then be able to develop tailored strategies for the different causes of disconnection. I will stop there.

Yeates: Thank you, Kim. Very good. We will just move ahead with anticipating probably taking 10 minutes or something from our lunch break, so we're keeping the same schedule and welcome now Dr. Pat Arian. Pat, are you with us, and can you share your slides and introduce yourself? Good. Thanks.

Pat: Yes, and I just needed to turn my microphone on, and I will get this going. Hold on. I've got Zoom menu bars in places that are not helpful to me. Here we go. All right, so I'm Pat Arian. I'm a professor at the University of Washington, and I direct an alacrity center which is funded by NIMH. Our focus is on the intersection of computer science, human center design, and implementation science to modify psycho-social treatment so that they're more accessible. I also direct the creative lab where we're using a lot of this information to inform technological advances and increasing access to mental health services in underserved populations, including old adults. I'm also a geropsychologist and I've been working in aging for like 30 years.

Let's see. Today what I'm going to do is I'm going to and mix things up a little bit for people, my slides are a bit different than what I had submitted. I thought what I'd start first is talk about some of the work that we've recently done to use NLP or Natural Language Processing and machine learning to use Google search data to detect proximal suicide risk. I wanted to start there because it really introduces a novel way of thinking about how we could rapidly deploy interventions that could prevent suicide attempts, and then potentially create targeted interventions that are easy to access and acceptable to older and other populations.

I will be talking a little bit about the benefits and challenges of digital interventions in older adults, as well as future directions. Let me start with suicide risk detection. Why is it so hard to assess risk? Although we can identify who may be at risk, it's very, very hard to detect when people will be at risk. All the work that has been done through, for instance, Greg Simons' work, Michael Schobomb’s work, looking at, for instance, electronic health record data to be able to predict timing and event has actually not borne out, it's very hard to do. At best, you present, or you can predict 1% of all attempts using data that way.

Part of the challenge is that risk fluctuates overtime on the order of minutes, hours, and days and risk profiles may be very unique to the individual. Some of the work that Courtney Bagge has done at the University of Michigan. She was actually originally at University of Mississippi or Mississippi Ole Miss. You can identify there are 10 warning signs of a suicide attempt, but those warning signs do fluctuate over time, their intensity fluctuates over time and which warning sign is the most predictive also fluctuates over time.

Further, a few people seek care when they need it, and so when they're in early processes of risk or even when they're high risk, they're not seeing mental health providers or providers who are in a position to evaluate whether or not they need help. Our group was actually approached by a very large information science company in Silicon Valley to investigate whether or not we could use digital data or search history to predict proximal risk.

When you think about the potential of this, 88% of people between the ages of 55 and 64 use the internet and search primarily for health reasons but for other information gathering reasons. 73% of the 65-plus age people use the internet. This cuts across ethnic and social-economic classes, although, obviously in lower-income populations where access to technology is a little bit more challenging, you'll see lower percentiles, but however, it is a large proportion of older people who do use these technologies. Search data is different from say, for instance, social media posts.

There's been a lot of research looking at Facebook, Reddit, and other kinds of social media data sources for predicting suicide. However, when you're posting something on Facebook, for instance, of which a lot of older people do use, the purpose of Facebook is to actually post successes or a particular persona. However, when you're searching with the internet, you're as close to as true self as possible because you're actually seeking information based on a mood-state or an information state that you're in.

Interesting and one of the things that we learned is that search histories are stored in perpetuity from the very first day you've ever surfed the net, so to speak. The first day that you started using Google, or Bing, or yahoo to search information on the internet, that data is stored. Somebody like me, there's like 15 years at least worth of data. Interestingly, too, search data is accessible to you as an individual, I'll talk about that in a second, but it can also be easily shared with other people.

For instance, Google search and Microsoft also has the same procedure, it has a tool called Google Takeout, I can show you how to use it. You yourself can very easily go on to Google Takeout, and for any of the Google tools that you've ever used, you can download that information in a CSV file and that can be shared with researchers, physicians, whoever. In fact, we have built apps that will allow an individual to share that data if they would like to with researchers.

Search data constitutes all sorts of really rich information, everything from how you search, like how often do you search during the day, when you search during the day, when you search how many different sources of information do you click through to. It also includes sentiments. There's what we found in some of our preliminary work is, there's a lot of information about what you're thinking, what you're feeling, what you're interested in when you look at search data.

Very quickly, so I can move on to interventions, which is the purpose of my talk. We conducted a descriptive cohort study with Kate [unintelligible 01:18:06] and [unintelligible 01:18:09] lab. Kate was conducting a study of about 150 people who had recently attempted a suicide and was already collecting information about life events that had led up to those attempts, and with the eye towards being able to come up with proximal risk factors for a suicide detection.

67 of the people in her study were available to participate in this sub study where we surveyed these participants about their comfort in using search data for risk detection and prevention. Then, out of those 67 people, 30 volunteered to download their search data and share it with us. This is all very descriptive and all very preliminary because search data on 30 attempts using machine learning and natural language processing is a very small. What we did find which is very promising is that we could map, for instance, search terms onto the 10 warning signs of suicide risk, and that search behavior was also very greatly associated with certain periods prior to an attempt.

There were four periods we looked at, 60 days prior to attempt. I don't know if you can see my cursor, but the column on the right represents the search terms, represents suicide risk period of 60 days before the attempt, 30 days, 15, and then 7 days. On the Y-axis is the actual suicide attempts, and on the X-axis are the risk factors, so search term, information around alcohol anxiety, burden, emptiness, isolation loss, social communication, that's basically talking about dying. The searching for suicide methods and then search patterns and then over for each of the time periods on the right-hand side is the overall number of basically searched risk patterns that were evident before suicide windows. The takeaway from these graphs is basically that first off, everybody's signature is a little different and so there is going to be, need to be a lot of individualization in terms of coming up with a risk pattern or risk prediction for individuals.

However, you can identify risks pretty robustly as early as 60 days before an attempt, up to seven days before the attempted.

The signal's a little bit clearer at seven days than it is at 60, but for a number of suicide attempts, we were able to tell in that period of time or see huge differences in search behavior search terms that lead up to the attempt. What does this mean? I think what's one of the things that's really promising here is that using technology, there is the potential to intervene.

These search information could be used very similarly to what we use to extract information for health apps that are available on people's phones which already contain information. You can allow Health Kit or Google Fit to access electronic health record data, so it's all stored in one place. It could be utilized for personalized interventions through technology. Digital tools are really well equipped to provide timely intervention. If you can get down to this person is at risk at this period of time and start offering a menu of strategies that could be helpful for preventing suicide and also using these search terms to personalize those strategies.

There's a lot of promise there, but the key question is currently for what we have, are these digital tools effective? Are they targeted and are they accessible? When I talk about digital tools and I'm realizing I'm starting to run out of time, there's a variety of different things that could be utilized. There's telemedicine, mobile applications, online communities, and some more fringy work around video games, digital assistance, and avatars.

When you look at the suicide prevention digital tools, there are very many of them that range from accessing help right away, to a virtual hope boxes that provide you with strategies to either distract yourself from suicidal thinking, to inspire you, to engage in activities that would give you some meaning, to coping tools. This is Virtual Hope Box from the VA, but there's also a number of other tools.

This one's actually-- the DMHS tool is actually targeted towards older adults and was designed with older adults in mind. You can see it's a little easier to access some of these things like stress busters, problem-solving, safety planning, and tracking your mood to very slick stylized things like stay alive, where they offer very much the same services. In the context of older adults, a recent Statistica study found that 57% of older adults own smartphones and tablets.

It's still not highly penetrated and likely in lower-income communities, you won't see quite as many smartphones. However the numbers are growing and if you can think of older adults as not being one stagnant population, but quickly evolving population, that number is going to increase over time as people in our age groups or younger start to age.

Interestingly too, I was very interested in the use of video games because there's an opportunity to target cognitive remediation strategies through these technologies.

40% of older adults play video games. What's interestingly and which is interesting is that most of them are women 48% of those who play video games are female with 24% being males, because we were interested in the impact of COVID on social isolation and mood and older adults and how technology might be utilized or mobilized to increase access to intervention, Patrick Rallway, and I conducted a national and local survey of adults 60 years old and older.

I'll talk about, and the reason why we did a local survey is because obviously, when you're doing an international survey, you're going to be sampling people who are already very familiar with technology. We wanted to do a gut check with a local group of senior housing communities, some of which are federally or are state-funded and also are, I think, they're called permanent supported housing.

This is just the demographics that the average age was a little over 65 years old, 98% of people were practicing social distancing and with half of them not socializing or seeing their family, 26% reporting that they were lonely, our rates of depression and mood were a little lower at 12%. In terms of comfort in using technology, 89% felt very comfortable with it at the national level, although at the local level, it was really more 42% and majority of people felt very interested in learning more about technologies for those who were not comfortable with using technology.

90% nationally had a smartphone, 65% locally so still a little above the average of what the recent Statistica survey revealed. Very interesting, though, to us is that 87% nationally and 60% locally were using text messaging as a way of support. This plays in very nicely to some recent work we've been doing with a technology company that offers psychotherapy over text messaging, where we looked at data from 2000 users of this company of which a little over 300 were over the age of 50. That cut-off based on the company's demographic measures, they basically stop counting anybody over the age of 50 in terms of age bands. This constitutes probably a younger spectrum, of "older adults." They-

Yeates: If you could wrap it up in a minute, it'd be great.

Pat: Yes, I can. Just to show you that 81% of older adults use the text messaging platform only for depression, anxiety and other stress. Just as a take-home message and I'll stop here is that when it comes to looking at digital technology for older adults, there is more work needed to develop innovative tool for social isolation and suicide prevention Kim had already mentioned that. However digital technology, digital medicine can be very effective in older adults.

Namki Choy (sp?) has done a lot of work in this area using a video telemedicine in older populations and her populations have been ethnically diverse. This is also a really good study on proving social support for older adults using technology, prison is a randomized clinical trial of that that's accessible. Apps are another potential avenue however surveys are showing that while there's strong interest and motivation to use applications, designs need to be modified for an older population to use.

Online communities are also promising, but they tend to be difficult for most older adults to navigate. I can talk about that a little bit in the discussion I have more, but I'll stop here just to save time.

Yeates: Sorry to have to ask you to do that. There's so much promising new directions and opportunities that we can take, and this is wonderful. Thank you. At this point though, let's press on and I would ask then Joanne to take over and help moderate the discussion with our four speakers and a reminder that the Chat box is there and the mic's open for invited guests.

Joanne: That's great. I'm actually going to just take off our discussion with a couple of thoughts. I'm a clinical psychologist at Weill Cornell Medicine, and we covered a lot of ground, what a great collection of speakers so thank you. I just wanted to raise a couple of points and a couple of questions. One was, if we're bringing interventions with well-defined populations and targets and outcomes, can we bring them to the community?

Our experiences is that if you can do that, you can actually increase the impact that you can provide and also potentially sustainability. We've found actually that research and practice partnerships are critical and I know that a number of the presenters today have these kinds of partnerships and that really being able to build interventions and test them with our partners, gives us an opportunity, I think, to access older adults who wouldn't normally see us or come to us we've done that with elder abuse victims who are depressed. The other thing is a number of the presenters have raised this question. It started yesterday with this question about the different votes we have is that really one intervention does not fit all. Heterogeneity is a real opportunity for us to personalize what and how we deliver, who needs what to test the real-world usefulness of our interventions. Patrick already started to talk about some work she and George Alex are doing bringing an evidence-based practice to a tech company. Also, the opportunities to really diverse an often ill-served populations, hard-to-reach communities to mention that.

We really want socially isolated older adults and yet they are socially isolated. The other thing is as we're doing this to really keep in mind that part of our mission, they include being able to promote detection of need and helping these populations figure out the mental health needs that they have and also helping them get their individuals into treatment. We develop doing engagement intervention to improve initiation of mental health services. Actually, we had a more effective engagement rate even amongst those older adults with suicidal ideation, and also to think about other treatments and how social disconnection fits in.

As I was thinking about this talk and this presentation, I was thinking something you've all begun to do which is how can we identify well-defined targets constructs? What's modifiable? What can we change? How do they relate to each other in suicide risk? Something else that it's come up already is how much intervention and how many sessions is enough to reduce risk? We also know that we need to serve those who are least visible. We actually find that there are a number of older adults who are not using technology. Part of our data matches yours actually.

We find about half of the individuals that we try to serve in Psychotherapy have smartphones that they use but half are still using the old Obama phones. Then something else that came up today is who should be offering these interventions? How can we address these workforce issues? I was excited to hear about peer presentations and peer interventions. Last but not least, can we track who participates in our interventions, who uses them, how they use them, and whether they drop out? Those really have to be the tests of how effective our interventions are. With those thoughts in mind, I thought we could go to the discussion. I'm really excited to hear what our presenters have to say, and I'm going to unshare. I like to open it up. Thoughts, Kim, Carla, Dimitris?

Kim: Well, I guess one comment I have is that I think one of the things is really interesting is that I think sometimes when we were talking about older adults, we have this one mindset. It's actually a super diverse group. It's always changing. I'm surprised, I haven't gotten this question but, in my workaround, technology get a lot of skepticism about would my older person actually use this? Based on what I see in front of me, nobody uses technology. There's probably bias in the survey and so in surveys and such. Even if there aren't using technology, there's a strong interest in using technology and for us not to consider that given that I'm about to enter into the geriatric cohort.

It is at one point going to be a part of how we deliver services. There are challenges when you're using technology in low-income populations, but there's a lot of us who are creating workarounds around that. For instance, one of the challenges we have with low-income communities and their use of smartphones is first our phones tend to be quite old. They're the ones that you can afford to buy. The other thing is that they tend to be shared. The privacy issues get to be challenging but you can design around those things. You can come up with strategies to make a mental health or a socially isolation tool accessible in a way that it doesn't look like that's what you're trying to do.

There's a lot of work on low-tech interventions like text messaging or SMS that is available ubiquitously regardless of what kind of phone you have. Or even if you just have a computer that could be very timely could theoretically be used around to help with social connection and social support. I think we need to start to think outside of the box in terms of reach, in terms of what people can offer. I'll end by saying that in very poor countries, it's not consistently but in very low-income countries. There are models where we use health workers who have access to technology to provide intervention that we're leveraging here.

When I heard Dimitris saying that offering people iPads, I became very excited because that is what we need to do. We need to start looking at what our Global Mental Health people are doing and using their lessons to figure out how we can increase access to technology and services for our are older and low-income populations.

Carla: Thank you. As a follow-up, I think that the point about older adults being heterogeneous is incredibly important, and that's in terms of their comorbidities, their language, their socioeconomics. As I mentioned in this way that I've been doing during COVID. We had a much lower socioeconomic status, which I think is why we have such slower internet and technology use. I also think about as we're researching and studying this, are we looking at the older adult population that is now or--? Patricia as you said you're entering this group. Your generation is going to be very different from the adult we have now in terms of adaptability.

One of my big pet peeves if I'm allowed to say is that the technologically based interventions that we’re using are being adapted for older adults and not created for older adults. Drives me nuts that Apple tablets are not very user friendly for people that have hearing impairment, visual impairment with their tiny little buttons, with anyone that has any remote amount of executive dysfunction even that may be normal for aging. That's very complicated. I’ve said this before in another talks, but I've worked with this company called Grand Pad. They have been piloting some of their tablets because they’re designed for older people and they have big icons and there's four and not 500 on them.

My mom who's not the most tech-savvy person, but the number of calls I get because her WhatsApp icon moved from her home screen and she thinks WhatsApp is off her phone and you think about-- These are some of the barriers that are there. They're going to continue to be there unless we really start creating interventions and technology for older adults and not just expect them to adapt to what we're all using in our younger ages.

Joanne: Still we’re getting a couple of questions through the chat and I want to share those with you. The first one is how do these interventions compare with findings on carrying letters interventions? What mechanisms do the interventionists think are effectively the same or different to caring letters? I don't know if anybody can address that question.

Pat: I might because I work with Kate so she's carrying letters lady. She's been actually using carrying letters as a technological intervention. Again, this is not specific to older adults, although she does have some older people in her trial sometimes. I think she's finding some similar outcomes that-- The point of carrying letters is to have that person have a connection and to think after they've been hospitalized and think about like that there's somebody out there thinking of them, and that please seek services or please follow through with your referrals and such.

She would hypothesize that it is that having that connection or feeling like somebody is caring about you. In SMS, you can do that fairly easily as long as somebody has access to that, but I don't know that it's been studied quite as well. There's a little difference between getting an email versus getting a text that might have a different level of annoyance factor that would need to be explored.

Kim: If you don't mind, I'll jump into you. I think carrying letters is a really important thing to think about because it was originally designed for people who didn't want treatment. The original moto study was people who refused additional outpatient treatment. I think when we think about looking at connectedness in the context of suicide prevention, it's a slightly different group of disconnected people than the folks who might say go to a senior center because they're lonely.

A lot of the people who are in this group of at-risk for suicide disconnected, not all of them actually want to work on social disconnectedness. You're trying to help people with something they don't want help on, which makes it really, really tricky and hard. The carrying letters intervention, I think, is especially compelling because it was a group of people who didn't necessarily want help, and it was very effective.

Joanne: Great. We have more questions coming in. Somebody asked, in terms of intervention, how many of you are connecting with youth organizations that connect virtually with them read to them tape record? This person mentioned Senior Planet, which is based in, I don't know, we have it in New York City. The question is can we help those people who want to use technology become more tech-savvy, and any thoughts you have about that?

Kim: Yes. That's such a beautiful idea. I have that written down to bring up intergenerational programs. I found a lot of times that people in our trials, where we talked about asking people just serve as a pure companion for another older person, a lot of them said, "Well, I don't want to spend time with old people." They said, "I just don't want to do that." I think, for some people in intergenerational program, can be really, really important and really effective.

I know that the foster grandparent's program, which is the Nationwide Senior Companion Program is looking to start making that virtual. Kids aren't able to go to schools, old adults are home during the pandemic and so they're looking, how can we figure out making these intergenerational connections, say, via tablet?

Carla: I think I love the idea, there's a lot of good work being done on intergenerational work. I think we need to return back to this concept of who are we working with and why are they lonely or isolated and do they want that? I've given the example, before that I had a patient who had a friendly visitor, and she didn't particularly like the visitor, but she just complied. She's like, "I think that the students getting more out of it than I am, and I'll just do it." I think this is this type of thing that we don't just want to throw being [unintelligible 01:42:46] but throw an intervention or a person out the older adult, like, "Here, let me solve your issues," without really understanding what they need.

Just something to think about as we think about the interventions. I saw a comment in the chat about payment, too, which is my overarching goal in the world, among many others, is to really encourage payers, whether that be federal or state level or private payers to really think about payment of these things in the same way we think about we pay a lot of money for statens and for cardiovascular interventions, and if we're seeing health effects and financial repercussions, then we should start to have our payers pay for these interventions if we can show that there's evidence, and that's the work ahead of us.

Kim: I think in the payment world. It's interesting, I've been involved in that in terms of digital health. I think there are billing codes now in some sectors, maybe in federally qualified health centers, for instance, can bill for peer services. It's not a lot, but they can do that. Because of COVID, COVID has done a really interesting thing because everything's gone online because you can't see anybody in-person, that there were some temporary suspensions of some of the rules and regulations around billing and [unintelligible 01:44:06]

There's a ton of work going on right now to see if they can continue those extensions or make that a permanent thing. I think one of the challenges I was talking to my chair [unintelligible 01:44:22] who does this all the time. This is sort of his world, was saying that, yes, there's a lot of movement as the CMS level for telemedicine but CMS is also recognizing how expensive it's gotten to be. Because what happens is that when you make things more accessible, people use them, and that cost the government money or even health plans money.

That's the challenge you're going to have is that and this is bears out in the literature around access, like utilization of psychotherapy vied with telemedicine, you see much better adherence. People use more sessions, and it's unfortunate catch 22.

Joanne: I think with the access too what's interesting what I've noticed in some of my homebound population and again, specifically around COVID is that some of my patients and their caregivers have smartphones and have the capability, but they don't have data plans. This is another thing that highlights some of the socio-economic differences that or we were having some challenges also with and which we've been working around, but with non-English speakers and hearing impaired with being able to do that over some of the digital technologies to connect to services.

Those are some things that I think we still need to work through and figure out whose responsibility is that. Mine is at UCSF that should be responsible to having this our patients have access to us or is it the health plan or is it the individual patients and families. It's incredibly complex.

Dimitris: I also want to point out, I want to comment that I agree totally with Carla about modifying the interventions for older adults, and especially I would say, for cognitively impaired older adults, because that's something that we see more often. What we're trying to do at Cornell is we take this well path intervention and in collaboration with Sarah Sazer, Dr. Sazer, we got an administrative staff, and we will try to see whether we can improve this intervention for cognitively impaired older adults. Then we may also use some of this intervention for caregivers, and how to deal with people who are cognitively impaired.

Joanne: We have a few more questions. Amy Byers asks, if presenters would comment a little bit about how to navigate bridging interventions for social disconnection with other colleagues, and other services behind mental health. I think in line with the idea of how do we get access to those individuals who might not come to us? What you would recommend? How do we integrate our services?

Dr. Martha: I'll jump in. What we've done is, as you know, is that we've worked with the organizations that provide services to older adults who have a high likelihood of having socially disconnected for example, people need home health services, or even more so home-delivered meals and you've worked for that population too, and have worked with the organization to help identify the people who report loneliness. Then we've moved in from there. We're still trying to work out where we could even make that even grow that approach. I think piggybacking on the service providers that are already doing this is one way to do it.

Joanne: I will say [unintelligible 01:48:04] that too, because I think with older abuse victims, I think the key is the partnership, because they had seen depression. They've seen the loneliness. Social isolation is both a risk factor and a byproduct of elder abuse. They were very happy to partner. Again, I think with a partnership, it's a reminder that we always have to figure out what they want out of it as well.

Dr. Martha: Well, Renee and I went to our aging services, people said, "Oh, we'd like to do depression of your remote teledepression, blah, blah," and they said, "Well, that's fine but we're really worried about loneliness." It was they that drove that entire, that initiative that we did.

Carla: I think this is why it highlights so much of from a public health or population health perspective, why it's important to really think about how are we evaluating risk and looking at things upstream before someone is already disconnected and reaching because it challenges our assumptions. I think if I have the assumption that many people do that all older adults who live alone are lonely, that's a whole different thing than what the reality actually is.

It's identifying people upstream, but again as I showed in some of my slides and didn't get to finish them, but clinicians want to be able to address this, but that means that if we're evaluating or screening or looking for this, then it's being able to know who are the partners that I have in my community where I can refer so say Miss Jones has recent grief and that is why the reason why she's lonely, what can I refer her to, versus my patient that lives in the Tenderloin who is incredibly isolated because they're afraid to go outside.

If I don't, as a clinician, really evaluate that upfront, then I'm going to miss people, and then if I don't know what to do about and don't have options in the same way that I know for cardiovascular disease, these are the options I have for my patients and so we need a similar thing although, again, it's not a one size fits all, but it's an individualized framework to deal with this.

Kim: I think too if we're thinking about social disconnection and late-life suicide, we can't forget about primary care. Then that's going to be probably where we're going to find the older men who are just connected and not connected to any of those services as well, thinking about the work that Carla was talking about measures we might be able to use within the health system I think will be essential.

Joanne: There was also a question about how people navigate the ethical issues related to using digital technology and protecting confidentiality.

Pat: Yes, there's a lot of work on that it's a huge concern. I didn't get a chance to get to was the results from our survey around using Google search and it's interesting because many of the participants they were very much like, "If this is going to be helpful it sounds like a great idea, it's a natural behavior I do that I don't have to make myself do," but they had two things one they were like, "But it's got to be accurate. I sometimes search for suicide because that's my coping mechanism," which isn't a great coping mechanism obviously, but still it was this like it has to be very accurate and then they're like, "I don't want anybody else to use this. We have to talk about how would this be something that was designed for me would it be part of the healthcare record."

There is a lot of work around making sure data isn't shared, there's John Torous who's a psychiatrist at Harvard, he has been really pushing companies and fields to be much more accountable around their terms and conditions. He's published quite a bit about how there's a disconnected of a way what people say and what they actually do. Some of that is also educating people about which apps are or tools are private versus have poor security standards.

There are currently 50 evaluation programs for digital mental health. The most popular ones are Echo and also Cyber Guide which Stephen Schuller at UC Irvine heads up. It evaluates all of these tools based on the evidence base, security privacy, user center design, and utility based on consumer ratings. Yes, it's a huge concern. There's a lot of work around that going on right now.

Joanne: Great, there was also a follow-up on the tablets that are adapted for older adults, can somebody mention that just give the name of the tablets as well the company that makes those?

Carla: I think there's more than one, the one that I've worked with is one called Grand Pad. It's a nice story because it's literally a father and son who created this for their own mother and have adults on their board, so they were developed for them rather than just using a straddle iPad.

Dr. Martha: There are smartphones that have been adapted for older adults, and just as an aside, we've got my mom one who's 100 and it's called Jitterbug and all the icons are big, but she can't swipe. She finally sent it back because she could [unintelligible 01:53:39]

Carla: Well, it's interesting Martha because that's one of the things that Grand Pad worked on was the swiping thing because it has to do with the touch and the moisture on your finger and that's one of the biggest barriers for people. It's often joked about how older adults are constantly pressing and it makes it seem like it's their fault but it's actually the people developing the technology that are the problem, I said it [laughs].

Pat: Yes, [inaudible 01:54:03] some of the tech companies to improve those things like Facebook has a tablet and their engineers are working on how to make these tools a little bit more accessible. Samsung for a long time was actually partnering with UCSF around potentially accessibility around geriatrics because they really saw a use case for that, but I don't know what happened after I left UC around those issues.

There are some companies are super interested in it, some companies admittedly aren't because they don't see it as their customer base and so it's hard. We have to push organizations that do care about aging to get companies to-- Because even I think it's really interesting this whole thing about swiping, I didn't know, but even I'm actually starting to have a difficulty using the keyboard on my phone. I have a giant phone, and it's probably the moisture issue, but I want to use this. I use these technologies all the time but there are just things that happen when you get older that make them inaccessible.

Joanne: George for example has pointed out [unintelligible 01:55:12] that in the Cornell data there was 27% of health workers who endorsed depression during COVID and your population of individuals around the same age was about 12. One of the questions is how do we think about the different experiences during this kind of pandemic and the impact on the risk for depression and suicide? I don't know if you wanted to just comment on that.

George: My assumption is that the different exposures and the different environment and what they have to learn and deal with is what makes a difference but again, I wanted to point out that fear of exposure, real exposure, and concern about finances were not associated with depression or suicide ideas. In contrast, it was the experience of loneliness that did. To me, this is counter-intuitive because they were lonely in their own stressful environment.

Yates: George that's in some contrast as I recall to the work that [unintelligible 01:56:38] published after the SARS experience in Hong Kong and early 2000s in which, retrospectively, lots of methodological issues, but it was fear of infection which was driving to some extent suicide in the older person.

George: Not in the Cornell data, it was literally no association with the suicide item and the PHQ 9 or the GID or the PTSD measure that we had.

Dr. Martha: Joanne, we could talk all day but maybe you wanted to finish us up another question or so?

Joanne: I'm going to do a quick look and let's see, somebody commented also on the way in which when they work with tech companies the timeline is different than the timeline that we as researchers often work within and I don't know, Pat do you [unintelligible 01:57:52] I see you're unmuted. I bet you want to comment on that.

Pat: Yes, I mean it depends on the tech company. Some are very interested in developing an evidence base, and they already have a product, and they're out there. They're okay with going back and show the evidence. Then some companies aren’t, and I can go into a whole lecture as to why that is. I know that at NIMH, and I'll le NIMH types talk about this but there's been a lot of discussion around how the timeline for research isn't nimble enough to address the changing environment. There are ways to work with tech companies. There's also new, I think, some tolerance in the tech research world for testing principles rather than products, so to speak like knowing that products change over time like if you work with a company and they fold, which is happens a lot, but you find another company that's very similar, it's not about the company, it's about the principle that you're trying to test. There's a lot of strategies around that. There's a lot of strategies around knowing which companies are going to be working with you and which are not going to be working with you, that helped that, but there are companies out there that are interested in generating an evidence base.

Jovier: I was just going to chime in that we do have small business innovation research grants where researchers can partner with companies to work on some of these technologies, but I also want to mention that there are other, I think, older adult kind of. I guess I should say lobbying groups because I was asked a couple of years ago by AARP, they wanted to hold a competition around technology companies that provided ways to help older adults deal with social isolation, and it was a competition question.

We couldn't be involved because, of course like Pat said our definition of evidence was different than I think what AARP would be interested in, but they're huge organization which has millions at their disposal, and they were totally on board with the idea of how to think about providing services to older adults. It gets back to the first point you all made, that maybe things should be developed with them in mind. Think outside the box about other kinds of foundations or people who might be able to help with some of those questions. They're on TV here almost every other day for something. They take up a whole block in downtown DC. There's a powerful wing of people who do those sorts of things. I want to thank you all for such a great and rich discussion, but we do need to try to catch up with time because it's already noon. I think with that, I will try to go ahead and introduce my boss, Holly Lisanby, he's going to provide a brief commentary from NIMH. Holly, are you there?

Holly: I am. Can you hear me?

Jovier: Yes.

Holly: Great.

Jovier: If you share your screen, I'll turn it over.

Holly: Yes. Great. Well, I'm going to start with a few remarks and then I'll show a few slides. I'd really like to thank the workshop chairs, doctors Bruce and Conwell, also the organizers, Doctors Liz Necka, and Jovier Evans and all of the members of the NIMH Planning Committee, all of the presenters that have given wonderful talks, and also the many attendees. We're really delighted to have so many people attending online, joining us for this important discussion about how social disconnectedness and its relationship to late-life suicide may be important.

As you heard from Jovier, I’m Holly Lisanby, I direct the division of translational research at the National Institute of Mental Health but importantly, for this discussion, I'm a geriatric psychiatrist. I care deeply about advancing research to help prevent suicide and improve quality of life for seniors with mental health needs.

Rigorous research into modifiable targets for intervention to prevent suicide in late life is a high priority for NIMH. We are so delighted to have been able to assemble such a distinguished group of experts, represented on this workshop, to help us think through ways that research can make a difference. To provide some context to this discussion, I'm going to share my slides now and review a few points about NIMH progress and priorities in suicide research. Let's see if this works here. Are you able to see my slides? Yes. Okay. That's my title slide. Let me move on here.

Suicide is the 10th leading cause of death in the US, and it's on the rise, which is the wrong trajectory that we want to see. Life expectancy has been decreasing in the past few years in the US, and data suggests that this is due in part to an increase in deaths in mid to late life. Here, you see the suicide rates for women in the light blue and men in the dark blue, broken out by age group. You see that for men, the suicide deaths peak around over age 75. Suicide attempts are more likely to be lethal in later life, about a quarter of attempts in late life are fatal compared to an estimate of 1 in 200 attempts in young adults.

As has been discussed in this workshop, COVID disproportionately affects older adults, both the disease itself and the policies in place to prevent transmission of the disease such as social distancing, limiting visits to assisted care facilities, and many of the opportunities for social interaction have now gone virtual. As has been discussed today, those technologies are not always accessible to older adults, and they may not be designed with them in mind.

Nearly half of people who die by suicide are over the age of 50. They tend to exhibit characteristics that intersect with age, such as veteran status and owning a firearm. Importantly, most people who die by suicide have seen a healthcare provider within a month prior to death. This means we need to make the most of every clinical visit because it may be the only opportunity, we have to save a life.

The CDC report using data from 1999 to 2016, looking at over 20,000 US suicide deaths found that 46% of these individuals had been diagnosed with a mental disorder. This suggests that the rest went undiagnosed or lacked access to care in order to be diagnosed. Men, in particular, were less likely than women to have received a mental health diagnosis and were more likely to use a firearm to kill themselves.

Screening instruments like the Patient Health Questionnaire or PHQ-9 can improve risk detection. You can see on these lines here that those who reported depressed mood, some several days, more than half the days, or nearly every day, had an increased risk of suicide attempt over time. We have other tools to screen. The electronic health record can be used to apply algorithms to identify suicide risk. This is the VA’s REACH VET program that uses an algorithm each month to identify the top 0.1% of the VA’s patient care population with the highest predicted suicide risk which amounts to around 6,000 individuals in a given month out of a patient population of around 6 million. A dashboard for providers indicates patient risk, and providers are encouraged to consider ways to enhance care. We hosted a meeting on the clinical applications of risk algorithms last summer. In addition to NIMH researchers, we included representatives from the VA, the Department of Defense, the American Foundation for Suicide Prevention, and PCORI.

Other research shows that universal screening for suicide can actually double the rates of suicide risk detection. You can see here the increased detection rate when universal screening is implemented. However, the ED safe study pointed out that suicidal ideation and behavior among older adults seeking care in the Emergency Department is often missed.

Now, there's been progress and you've heard throughout this workshop progress in different types of interventions. One example was the NIMH prospect study which was led by members of this workshop. Given the fact that older adults who die by suicide are more likely to seek care in primary care settings than in mental healthcare providers before death, implementing these interventions in the context of primary care, can be especially impactful as was discussed earlier today. We've actually known for 20 years now that collaborative care for depression based in a primary care setting can reduce suicide risk more than treatment as usual. We're hoping to see how the successful prospect collaborative care findings can be used to improve risk identification and suicide safe care now that collaborative care can be reimbursed, and patients can be tracked using these algorithms through the electronic medical record.

As was touched on earlier today, you heard a discussion about the engaged intervention. We have effective psychotherapies for late-life depression, which can also be delivered through telehealth and studies such as these led by members of this workshop, the Engage Intervention for late-life depression was developed to match the skill level of community clinicians. Based on the theory that dysfunction in the research domain criteria construct of the positive valence system or the reward system is a critical mechanism for late-life depression, this intervention uses reward exposure to change and enhance behavioral activation, which is an index of positive valence system function as its principal intervention. You can see on the graph on the left, showing that it does engage this target in terms of activation.

Earlier this week, I presented a new concept during the public session of the NIMH Council on the identification of positive valence system targets to develop novel suicide prevention interventions and this is one such example. The graph on the right shows that this intervention was non-inferior to problem-solving therapy, but since it is designed for trained community social workers, it can be implemented more broadly and increase the number of therapists who can reliably treat late-life depression.

Now, we also have developments in other forms of modalities such as psychosocial interventions, but also medication interventions. Ketamine is associated with rapid within minutes to hours, a decrease in depressive symptoms and in suicidal thoughts. The FDA recently approved esketamine nasal spray in conjunction with oral anti-depressants for treatment-resistant depression in adults. NIMH supported both intramural and extramural research helped to lay the foundation for the development and subsequent approval of this medication.

The Montgomery-Asberg Depression Rating Scale or MADRS score, on the Y-axis here, in this figure decrease from baseline up to four hours after the drug administration during the esketamine group in the green line compared to placebo group in the orange line. This difference continued after 24 hours. There were significant group differences observed at the 24-hour mark at day 3 and day 11, but not at the double-blinded endpoint which was day 25. We need to know more about the safety of interventions like esketamine and its efficacy in older adults. Highlighting here, a funding opportunity we have on safety and feasibility trials for rapidly acting interventions for severe suicide risk which encompasses not just medication approaches, but also psychosocial interventions, as well as devices.

Now a bit about devices because many of you know I have a background in brain stimulation research. We also have new ways to make known effective treatments like electroconvulsive therapy, or ECT safer for the older population. I lead along with other members of this workshop the NIMH supported pride study, which was prolonging remission in depressed elders, in which we found that a particular type of ECT right unilateral ultra-brief ECT, to be rapidly effective in treating seniors over the age of 60 with a major depressive episode.

The orange bars here show the cumulative percent of patients responding in their depression after each individual ECT treatment. You can see the onset of improvement was quite rapid and importantly, it was rapidly effective in resolving suicidal thoughts and behaviors in these patients. On the blue bars, you see a histogram showing the distribution of the scores on the suicide item on the Hamilton Depression Rating Scale. You can see a significant percentage of these patients were actively suicidal or had a serious attempt. After treatment with ECT in the red bars, you see that most patients had complete resolution of their suicidal thoughts and behaviors.

We also found this intervention to be long-lasting, in other words, preventing relapse. Using continuing the ECT in a symptom titrated algorithm-based fashion, which is called the stable algorithm shown here, we found that only 13% of patients relapsed after six months of stopping their acute course compared to 20% of patients on medications alone. Even more importantly, we found this intervention to have an excellent safety profile, which is highly important for its adoption, especially in an older population.

Now, to learn more about the NIMH priorities and suicide prevention, you can take a look at this recent JAMA Viewpoint article that highlights our priorities. They're listed here on this slide. This workshop today and yesterday on social disconnection in late-life suicide seeks to take this a step further and deeper into this particular at-risk group. We really want to hear from you what our key research gap areas have special relevance to late life, and we want to hear how and where can research investments accelerate the progress that we so desperately need. It literally is a life or death matter.

Now, a bit about how social disconnection fits within our strategic plan on our website, you can find our most recently updated strategic plan for our research priorities at NIMH for 2020. It clearly recognizes that social and environmental factors play a critical role in health and health disparities.

A central cross-cutting theme and our strategic plan for research is understanding how environmental influences that vary within and across populations, including social factors such as social disconnection, how these interact with or affect biological systems important to regulating mental processes. I'd like to thank you for your time and efforts here over these past two days. We're just so excited to have you joining us and I'm happy to take your questions.

Dr. Martha: Holly, thank you. This is both informative and inspirational and gives us a lot of us in the research world a direction to go. We greatly appreciate it. Thank you.

What I think we'll do is let people text us some questions that we might be able to address later. I think it's time for us to get moving on to lunch because hunger takes over here. At the same time, we'll have the participants return at one o'clock when we're going to be breaking out into our breakout groups. The public who are not part of the participant group will be returning at 1:45 PM. Let me see, there were some other things. When we break out into our breakout groups, the group moderator will begin with a very brief recap of the discussions and then you'll go on from there. Jovier or Yates, is there anything else we need to say?

Yates: I think just as my understanding is when we come back, following our lunch break at one o'clock, we will be automatically routed into the particular groups to which we were assigned. Also, I just didn't want to think ahead of time the group leaders in the synthesis session, we ask at the beginning that each of you provide just a very brief recap of the discussions that occurred in those breakout sessions. Thank you.

Dr. Martha: Like yesterday, we really recommend that you do not log off that you stay on the Zoom, but when you get back at one o'clock or 1:45 PM, you'll be placed in the appropriate place. Meanwhile, thank you, everyone. This has been a terrific morning, building on all that work yesterday. We're looking forward to the breakouts, and then the gathering together at 1:45 PM for our synthesis session, so have a good lunch. Thank you.

Jovier: Thank you all. Thank you very much.


Jovier: I think we are just waiting for everyone to get back. Maybe we can start with maybe a brief summary of the discussions in each of the three groups. Then we'll open it up to our distinguished panelists to talk about the way forward. Hi, Chip. You're big on my screen, so I can say hello immediately. Who would like to go first?

Dr. Martha: We were going to have the moderators do this brief summary.

Joanne: I will.

Yates: No more than a few minutes, I think.

Dr. Martha: Like very few minutes.

Joanne: The session number three, or the intervention session, we covered a number of different topics. I'm going to give you a sense of the things that's on my mind and then I ask anybody else to jump in. We talked a lot about who should be providing the interventions, and what it meant for the older adults that feel lonely and who they wanted to connect with and the importance of that. Was it a mental health provider? Was it a family member? Was it a peer support, and how to think creatively about that?

We started our discussion talking about dose-response and trying to think about how the amount of time required or the way in which we could think about changing loneliness, but also in terms of dose-response in terms of suicidal ideation or suicide risk, and the question about whether or not more intense intervention was needed earlier on, and it could be tapered off after time. It was an interesting discussion.

We also talked about how to think about weaving suicide risk interventions into other settings, like emergency rooms, where they don't have the luxury to do a 45-minute session for eight weeks in a row, but where there are opportunities, and opportunities that we're more than just referring out because I think there was a lot of consensus that being referred wasn't always what people were looking for, and out often wasn't ideal either. Last but not least, some interesting questions about the cultural context for understanding belonging and loneliness, and how to both talk about it linguistically, because it's a language as it doesn't translate as well, but also to think about how that gets defined and to flesh that out a little bit more.

Dr. Martha: Thank you.

Yates: Thank you. George, you want to tell us about this?

George: Of course, thank you. Our group discussed the heterogeneity of people who commit suicide or who attempt suicide. When it comes to loneliness and its relationship to suicide or mortality overall, there was some consideration to the fact that loneliness may increase inflammatory endocrine and other responses related to mortality, but it may do that in a reciprocal way with mental or mental disorders, physical disorders, or disabilities that may promote loneliness and independently promote some of these processes that increase mortality.

Much of the discussion was on the heterogeneity of the groups who commit suicide with a particular focus on old white men who have by far the highest suicide rate, and it continues to increase. Also, the issue of white veterans, in particular, commit suicide more than other groups in that age category had to do with socioeconomic level, educational level of the veterans the voluntary force now. What bring with them which is a certain lack of personal competencies along with substance abuse and other related aspects in criminality.

The question then is, if you focus on this group, what would be the type of intervention that would be meaningful and potentially effective? The discussion further expanded to what happens with cognitive abilities and suicide? Yates mentioned a Taiwan study of people who had the first attempt in late life and then on follow-up, they had seven times higher likelihood to develop dementia. The question of what happens to men at the end of life, and what is their expectations, and what are the losses that they experience as they're removed from their work environment where they had competence while they may have failed to develop competencies in other areas.

Finally, it was a comment that suicide attempters in late life almost never volunteer to help other people. Speaking to effect that may not be a fruitful way of trying to prevent suicide attempts. I think that pretty much summarizes the various themes in our discussion.

Dr. Martha: Vanetta, do you want to talk, or do you want me to try to summarize our talk?

Vanetta: Yes, I think you could give a better summary.

Dr. Martha: We'll try.

Vanetta: I can jump in if needed.

Dr. Martha: That's great. I think that our group, I suspect, covered territory that the other groups covered. The conversation got very much into thinking about where do we cut into this problem. There's multiple answers of course but one being do we think of this something as more societal and should we be thinking of this the way, can you do something about our society that would make people more integrated, more connected in the same way you make the risk of car fatalities lower by having safer cars and the legislation that has contributed that? Do we try to identify those highest-risk people, the people who become either through transitions or circumstance, they are homebound, they have become bereaved, they are making transition into nursing homes, they're making a transition out of prisons. These are these high-risk moments so we have more targeted populations, or do we better understand the phenotype of the person who will die by suicide, and can we identify that risk? You can see probably these three, whether you go from the public health model to the very strongly clinical model, all have their advantages, and strategies, and understanding the mechanics that goes with them. Is anybody else from the mechanism group who wanted to say anything that was really missed in what I summarized?

Okay. Those are the three groups. They're it seems like there's quite a bit of overlap. What we're going to do now is we have a panel that includes Yates and me, so we get to talk. We also have doctors Caine, Pearson, Reynolds, and Stanley. I want each of you four to introduce yourself briefly. Then we're going to have a conversation that we're not going to come to the answer obviously, but I think we can begin to continue to say, "Where should our research be going? Where could science be the most fruitful perhaps? What are the barriers to that? What are the opportunities?" and so forth. Eric, do you want to introduce yourself, and then we'll go to Shane, Chip then Barbara?

Eric: I'm Eric Caine from the University of Rochester. I'm pleased to be here and working with Yates, and Kim, and Caroline. If I'm leaving out anyone else from Rochester, I know there are several postdocs. It's really a pleasure to be here. I'm going to try to go through five major points quickly. If I talk too fast, I apologize because Marty is--

Dr. Martha: Eric, I wasn't saying two and a half minutes for you to talk. I just want you to introduce yourself.

Eric: Okay.

Dr. Martha: Yates, how should we do this? Yates?

Yates: Why don't we just carry on, so each person introduces briefly and then makes their points. Maybe 10 minutes or less for that, though.

Eric: Okay, I'll talk a little bit more slowly.

Dr. Martha: Anxiety has gone down immensely.

Eric: No, just the amount of breath I was taking in to see if I could get it all out in one sentence. This has been a really interesting conference. I've learned a lot but I'm also going to be critical because I think that we have to really come back to ask the question, "Has this conference given us guidance on reducing suicide in later life?" The answer is I'm not really sure it has. Let me challenge us all to think about that as we go through it.

One of the things that is really cool is, of course, we started with Durkheim. The two late 19th-century figures out there, Durkheim and Freud, were battling, "Is it the person or is it the society?" Of course, it's almost like we're back into, "Are we making interventions at the broad social level, the population level, or the individual level?" One of the things of course that comes out over and over and over again is whatever we do has to be culturally attuned. That's a phrase I use. It has to fit the people we're working with. It can't be something that's just driven from outside.

The second point is we think about going from culture, and society, and groups, and populations to individuals. What are the mediating forces, what are the mechanisms? We've talked about mechanisms, but I remember back, this starts from my age to a conference in 1990 where I was asked to give a talk about caregivers of people with Alzheimer's disease. Everybody was interested in immune processes and immune functions. I was interested in [unintelligible 02:28:37] just done some studies on depression and immunology with a fellow named Jeff Linus, and immune responses didn't show much at the time. I kept saying, "Well, this is really not about old people. It's about aging people.

If we think about aging in a developmental perspective, which has been one of my things over and over again, then we're thinking about people who bring their habits of living, their interpersonal style and other things into their later life. In doing that, we always have to think about personality. Here I was, the psychiatrist, mostly neuropsychiatry at the time, hectoring people about doing the NEO PI and looking at personality. I'm struck again that here we are, talking about people who are coming into late life, and I don't think I've heard the word personality during this conference. All of the uptake to interventions, all the ways that we think about it, are tied to that. We started collecting NEO data. Paul Duberstein did some really great stuff, looking at personality and using informant inventories for the NEO. It's really clear that if we're going to make interventions that are going to be meaningful to people and changing their behaviors, we have to think about their emotional openness, their flexibility, and their willingness to interact. So much of personality reflects their interpersonal coinage and how they deal with it and all those sorts of things.

I think we have to really think in a little bit more clarified way about what we mean by mechanisms. I was really taken in, please, I had come into the notion of connections as being like the word ego, which was used in forever, not so much anymore, as a fill-in word that everybody knows what they think it is, but nobody really knows how to measure it and what it is. One of the things I've come away with is really the developing notion of measuring connections and disconnections, but yet it's still at the frontier. I really think that this is one of the areas where we're going to see a lot of growth and just really impressed though.

I want to switch. I'm trying to go over my points quickly. In 2018, there were 10,477 men who killed themselves and 2,413 women. Of that 10,477 men, 9,940 were white, 283 were black. Of the 2,413 women, 2,256 were white, 77 were black. Now, I don't think that black women don't suffer from distress, depression, resource limitations, and all the other frailties we have in life. I've always wondered, "Why don't they kill themselves more?" We speculate about that. We talk about religiosity in meaning and social connection, but we really haven't seen that studied. Clearly though, if we're going to make a difference in the suicide rate in the United States for people 60 and over, it has to be culturally tuned to white men. There's really no way around that. So much of the group discussion has been around older people, and I think that's great research. I'm not disagreeing with that at all. I think that the notion of cultural differences and dealing with very diverse groups and communities and socio-economic factors are central to understanding aging and premature mortality. I certainly use a socio-ecological model, but if we're really talking about suicide prevention in later life, it's that 9,940 people that we need to be focusing on. The number of American, Indian, Alaskan, Native elders, the number of people from other ethnic groups is de minimis, whatever it is, tiny. We have not really taken up that struggle. Now, these are probably the most cantankerous, unwilling, difficult to recruit, difficult to keep into studies populations, but that's where the action is if we're talking about suicide prevention. I'm being very limited to that.

The epidemiology needs to help drive us. Now, in the late 1990s, when Gates was emerging as famous, leading aging researcher and all this other stuff, I said that he had to do a study which he's never followed my advice on. It's still out there to do which is he ought to recruit 1,500, 2,000 grumpy old, comorbidly-affected, socially isolated white men and try to discover why they don't all kill themselves because even with the suicide rate of 1,000 per 100,000, that means 99% of them don't kill themselves. Of course, 1,000 per 100,000 is a stunningly high suicide rate when we consider even old white men.

We have no idea of what's protected in people who are fraught with all of what we would call the population-level risk factors. Now, I don't want to make an ecological fallacy thing and say, "I can pick out every one of them and apply it to each of them," but at a population level, we are totally ignorant about what keeps those people alive even when they're socially isolated.

This brings up a whole lot of things. If we think about our psychotherapy studies in late life, if we think about prospect or impact or any of these, those are majority women studies. Women are much more willing to use psychotherapy than men. I'm not going to get into how much smarter that is, but it's just suffice to say that men are remarkably recalcitrant to psychological interventions.

I don't know whether any of the results in any of those studies is applicable to preventing suicide. I would challenge people to think about that as we think about interventions. We have to have new ways of thinking, new ways of recruiting and new ways of walking up to what is a-- I don't know how heterogeneous this group of old white men is, but I do know that they're-- I have to be very careful about making pejorative comments about a group I'm in, but nonetheless, I would say that they're extremely difficult to engage and do research on. Places like the Emergency Department, primary care, and diverse community sampling is going to be really important if we're going to do the kinds of intervention research that we think is necessary to actually change suicide. It may come back to broader social interactions and broader societal things because we know that those are huge drivers of suicide rates and socio-economic factors and social factors are. I'll stop at this point and hope I provoked a few thoughts or ideas.

Dr. Martha: Thank you. Dr. Pearson, would you like to introduce yourself. I suspect you have a few comments.

Pearson: And you're going to make me follow that? [laughs]

Dr. Martha: I know. Maybe I should have found someone well…

Eric: Hey, Jane, you've been dealing with me for over 40 years or so or, it's 30. You're good at it.

Pearson: Actually, I had similar thoughts to Eric's in terms of really trying to understand the association between the social disconnection and suicide risk given all our challenges in both of these areas right now. I would argue both areas are still really new. The good news is here, we've got a lot of really smart people working on this, they've got some really interesting angles, then the challenge is trying to find the place you have the most leverage.

Eric just put it out there, how are we going to find these men to study because those are the highest risk for people who are going to die by suicide? You could argue that there's plenty of women with ideation and men and women who attempt. Those are both very costly and have a lot of morbidity. If you were writing your grant, you might be thinking about how to do that. I know Lisa Brenner is here and if you think about the men in that category, what percent of them are veterans? We figured out something through serving veterans that seemed to be working well, that could be quite promising for the larger population numbers.

You could start thinking about this in the way Carla did and some of you have done your own research, you go to a setting and you figure out who can you help there in understanding those opportunities? It might not be the first thing you want to do to address the population, but it might be the easiest thing you can do right now because it's a gateway to understanding what it is.

We had Emmy Betz in our session and as Joanne mentioned, you might have a few minutes in the ED to engage in somebody, how do you deal with that? We do know men and women show up there. I think for older adults closer in gender. If you're in adulthood, it's equal, male and female percentages of people attempting who are seen in the ED. It might start drifting to higher females in later life. I don't know, maybe Emmy could correct us there.

What got me thinking about how do you engage with men, especially in healthcare settings, is we've been trying to help a number of the different associations dealing with unemployed people right now under COVID. You've got job centers that are like one-stop shops, where many have figured out, "We have to check on a lot of things for people." Is their electricity on? Do they have enough food? Quick mental health check. Some of the more effective interventions in those settings talk to men about what they want to get accomplished. It's not therapy even though there's a lot of CBT things covered, it's how do you get the job done? What do you want to engage in? What do you want to complete? It's a version, I would say, of CBT or problem-solving therapy that keeps them engaged.

I heard a couple of folks talking about the benefits of groups and peers helping them. That setting is all set up for that, where you can move some things out. I would just encourage people to think about, "We've got all these really interesting interventions that hold promise, either from technology or from work on adulthood that could be applied, but what's the setting where you're going to put your effort in to try to test some of these things out and see what they might be able to shift in terms of outcomes?"

Some of these outcomes are going to be pretty long. These are long risk trajectories.

Again, this is a challenge for this area, and what are good proxy measures for looking at outcomes. Then there were so many different approaches to the mechanisms that I'm not even going to try to cover them. All I'll say is that there's plenty of stuff to choose from. As you're building up your applications, because I'm going to be wearing my NIMH hat, because we want to see more research in this area, you've got a lot to choose from. Depending on which setting you're in and how pragmatic you have to be, there was a lot discussed, I should say here, and a lot of the interesting technologies that could be brought to this. I'll stop there and look forward to other comments. Thanks.

Dr. Martha: Thank you. Chip, do you want to introduce yourself and give your comments?

Chip: Sure, thank you, Marty. Chip Reynolds, Emeritus Professor in Geriatric Psychiatry, of Pittsburgh, and editor of the American Journal of Geriatric Psychiatry. Friends think for a moment about the word social, that we've been using and think about where it comes from. It actually comes from a Latin word, sokious, which means ally, or confederate. Indeed, that brings to mind a metaphor, I think, for what this workshop and what all of us are really about and that is to say that we are all allies or confederates, if you will, in the campaign, where dare I say, at the war against suicide.

Think about the immense social and intellectual capital that's represented by the participants in this workshop. Think also, though about Marty's comment yesterday, Marty said, "I miss mingling with you guys," which I think underscores the importance of the social dimensions of the science that all of us love and are committed to doing. I think for me the question becomes, how do we invest this immense reservoir of social intellectual capital? How do we attract additional, if you will, financial capital from entrepreneurs, who are interested in doing high risk, high reward research? That for me becomes I think, a very interesting question.

Well, let me suggest a general approach. I think this may align to some extent, Eric, with your comments, and Jane with yours. If you look at the immense public health impact that cancer medicine has had, or that cardiovascular medicine has had, these are parts of medical science that have followed the strategies of what is becoming known as convergence science. Transdisciplinary, I think is another word for convergence.

The word convergence comes from the Latin, and it means turning together. All of us, if you will, are turning together in the service of bridging science and service to help our patients who are at risk for suicide. I'd like to suggest that this workshop has been more than a conversation in interdisciplinary science. Indeed, we are taking important steps as a field towards true transdisciplinary or convergent science, although we have a good way to go.

In that context, though, why is it important that we do so? The answer is simply that understanding suicide and preventing it is such a large, complex problem, that it is only through convergence or transdisciplinary science, getting ourselves out of our silos, that I think we can make progress. To me the conversations that I've been privileged to hear from you guys in the last couple of days certainly brings home that point.

Again, in that context, as the temporary steward of The American Journal of Geriatric Psychiatry, I'd like to offer the journal as a forum to continue supporting this conversation and hopefully catalyzing more collaboration, that is to say, working together. Indeed, Marty, I'm grateful to you for your current service as editor of a special forthcoming edition of AJGP on social isolation and loneliness and indeed, a number of the speakers that you guys have heard in the last couple of days are making contributions.

The other thing that I would note about convergence mental health science is that my colleague, Harris Air and another Aussie, Michael Burke, Helen [unintelligible 02:45:21] and I are serving as co-editors of a volume on convergence, mental health science that will be published by Oxford University Press this coming November. I think you will find that interesting, in a very pragmatic as well as scientific sense for helping our future move forward. I'll add very quickly, if I may, that any royalty income that I happen to realize from the publication of that book I will donate entirely to the American Foundation for Suicide Prevention.

In terms of what we need, I think that in terms of the field going forward, we need a greater emphasis on the social systems of caring for suicidal people. We need to embed effective treatments. Whether it is the use of lithium as a suicide prophylactic agent, or clozapine, or ketamine, or whether it is use of a learning-based psychotherapeutic approach like engage, or problem-solving therapy, we need to engage these treatments within broader systems of care. There are examples of this that I think can provide guidance to re-engineering systems of care that will help us have a true public health impact on rates of suicide.

Hampton wrote about the Henry Ford Hospital experience about a decade ago, in a paper published in JAMA, and Kapoor and colleagues more recently, and Lancet psychiatry, wrote about mental health service changes and organizational factors associated with reductions in the incidence of suicide. A lot of this comes back to what Jane was saying earlier on about impact and prospect. Namely, these are studies that provide guidance to coordinated or integrated care, mental health services, behavioral health services in primary care. Again, Eric, to your point about reaching the demographic that is most at risk, older white men, but also to Pat Ariens earlier point, Pat's wonderful talk. Pat was a principal in the impact study, given her work in problem-solving therapy. Today, Pat's wonderful talk on the use of digital tools to recognize more acute or proximal risk factors for suicide. If we can embed these within systems of care, I think we may have a great impact. Marty, do I have another minute, or should I stop?

Dr. Martha: 20 seconds.

Chip: 20 seconds.

Dr. Martha: 20 seconds.

Chip: Well, I think I'll quit while I'm ahead. Thank you.

Dr. Martha: Thank you. There'll be time for discussion, but I do want to have time for Barbara to talk.

Chip: Thank you.

Dr. Martha: Thank you. Dr. Stanley, would you like to introduce yourself and we want to hear from you?

Dr. Stanley: Sure. Hi, I'm Barbara Stanley. I am a professor in the Department of Psychiatry at Columbia University, and I'm a research scientist at New York State Psychiatric Institute. I am a suicide researcher. When I first got this invitation to this, it's like I was a little puzzled, because it was about social connection, which I don't really write about and it was about late life, which I don't really do research on.

I have to say, I am so pleased to have been a part of these two days and I feel like I learned a lot. I have to say, it will change some of the ways that I do my research on suicide. From that perspective, it's been wonderful. As I've gone through these two days, I've looked at it through the eyes of a suicide researcher and I've kept the four definitions in mind all through the day, the two days, so social connection, social support, social isolation, and loneliness.

I kept coming back to them and I kept thinking about, "Okay." If you hear a little squeaking in the background, that's my social support playing with the dog. I thought, "Okay. What do we know about social connections from suicide research?" Well, actually not that much. Social support and suicide research, so we have done some research on that. Isolation and suicide. There's not a lot out there, and loneliness, it was mind-boggling as I listened to this, it's like, "Well, could I pull up in my mind three articles on loneliness and suicide?" I actually couldn't. I see Eric looking at me there, but maybe he can.

I thought, "Okay, so this can really kind of reshape the way that I think about things and maybe that others can think about things. The model that we frequently operate from is the stress diathesis model of suicidal behavior, right? There is the diathesis, the underlying conditions like depression and bipolar illness and family genetically, and then there are the stressors.

When we think about the stressors, what we typically think about are negative life events. It might be negative interactions, that's kind of the most that I could get out of what we talked about in terms of over the last two days to fit into the stress diathesis model, but why wouldn’t' chronic loneliness be considered a stressor? I don't think that we have looked at it that way, at least not in the work that I have, I've seen.

Just a couple other things. One is that, when I thought about the definitions, one of the things that I came up with, is that-- Actually, it's really interesting, Eric, very similar to what you said about personality, personality is really important here because I was thinking about social isolation versus loneliness. Social isolation is a physical property, but, of course, the idea that-- Loneliness is so much because perception is within the context of the definition, is in the eye of the beholder.

I was reminded of a conversation that my grown son and I had a few months into the pandemic, and we were talking on the phone when everybody was starting to get antsy about being alone, and I said, "Tom, I have to say, I'm okay, I don't mind being alone." I said, "Maybe it's because I have all of these Zoom calls, and I'm happy to close the computer at the end of the day." He said, "Mom, I'm exactly the same." I was isolated, but I wasn't in any way lonely, and so to me, it feels like maybe we shouldn't be talking about isolation, we shouldn't be talking about loneliness instead of isolation.

A couple of other things. I do think that I was very piqued, my interest was very piqued about what are the potential biological underpinnings of, for example, loneliness? What could we do to kind of alter them or their potential treat? I mean, crazy as it might seem, treatment targets. Another kind of crazy idea that I had over the two days is we were talking about social connectedness, but what about-- Corollary of that is physical connectedness, and so that's when I was thinking about the biological processes, I was thinking about what does physical connectedness do in terms of oxytocin and endogenous opioids and so forth?

The last thing I'll say is, kind of turning to sites of intervention, I do think that all a lot of EDs now are doing routine, we are doing the PHQ-9, they are doing some sort of suicide screening, and the same is true for primary care. It's kind of built-in the sauce now, but I was thinking, maybe it's not enough for this age group to just ask, do you have suicidal thoughts, but what about are you terribly lonely. I know that it will be hard to add anything in a PCP practice, but to me, it feels like if we can-- That suicidal thoughts are kind of acute, and loneliness is kind of maybe the process that leads up to suicidal thoughts.

Then the last thing I'll just mention here was I think, also underscores the connection that Eric made about focusing on older men. I think that this is a correct memory of the statistics, that widowed older women don't necessarily go up and suicide, but when a man, an older male is widowed, their suicide rate bumps up, and so there's something then about the loss of the social connection in that age group. To me, that is a potential way to be thinking about it. Thank you.

Dr. Martha: Well, thank you. This is great. Yates?

Yates: Yes, there's so much here. It's such a wealth of knowledge, it needs to be knit together in various ways, and there's obviously limits to how tightly we can, it would be artificial to knit them together too tightly. Sometimes in the face of that kind of complexity, I'm comforted by just thinking of simple thoughts. One of them is just thinking about our work with psych autopsies and interviewing primary care doctors over the years who had lost one of their patients to suicide. It was a devastating thing for some of them, it was really, really hard. That says to me, something that I experienced in my own relationships with primary care, in my own care and others that these are men and women who really want to do the best for their patients and for older people, and that the older people really warm to that.

Kim noted that in our senior connection study that one of the real challenges we face, and it's part of the methodologies of community-based studies of connectedness is that when you follow people or you even engage an older person in a research study, you're giving them something very special by way of a meaningful experience, and that those who are randomized to carry as usual, it's really extraordinarily enhanced simply by being able to contribute something that way.

Those two things, I'm not sure exactly how they fit together, except to say that I think that in primary care what's always struck me, and this reflects the conversations Eric and I have had over time, that there's an extraordinary amount of suicide prevention that goes on in primary care just on a day-to-day basis, because of the caring that people experience when they come and talk to the receptionist or talk to the talk to the primary care doctor. I think that happens all over the place, and what we need to kind of capture is something about why people don't kill themselves, and that my hypothesis would be they don't kill themselves because of simple, caring interactions that they have.

Then the final, what do we do with that by research, I'm not sure, but one of the systems aspects of that, I think, is our need, which is happening around the country, to really take much more seriously the meaning of patient-reported outcomes in our delivery of care, when we are listening to the older person. It's part of the age-friendly health system movement, the four M's, their mobility and mentation and medications and their meaning, what is meaningful in the life of that older person?

If we can organize our care around things that are meaningful to that older person, even as simply as being valued and being heard, there's an intervention in there, I think, that would be well upstream, doesn't apply or need to apply. It shouldn't just the older white men, it should be to everybody. It's a movement that I'd like to learn and think much more about, I think it has relevance here. Thank you for letting me get that off my chest, Marty. Are there other thoughts you have?

Dr. Martha: I do have a thought. I want to go back to the definitions because we've been doing some work with May Pepin and Matthew Chei and I on loneliness and as an outcome in older adults, they're identified, as I said, the aging home-delivered meals programs, Meals on Wheels programs. The first thing is that up in New Hampshire, and those of you who live in rural areas, know that there are these people who live out in the mountains and they are by themselves and they are isolated, and they are not lonely, and they want nothing to do with any of this. Then there are people who, like Eric, is able to mend some of them, and then there are people who are at home with a spouse, and immensely lonely. We really, I think, Yeates, you started by saying we need to differentiate these constructs and I think that's really important.

We also found that an intervention that compared friendly visiting, this is over televideo friendly visiting, Zooming like we are now, compared to something that gave people more skills on how to take their lives, which were quite restricted functionally and make them more-- Well, to address the issue of loneliness, that we actually found changes, when it was not enough to have companionship, it was something else that people needed to help them take over their lives a little bit more and become less lonely.

That did have also an impact on these people who were not depressed, but it did affect their mood. I think we need to-- I think there are some findings there that are important that it's not just companionship, it's not just another person, it's not just another person to talk to but there are something too that will help people engage in their life to make it, I would say, more meaningful or at least feel like they have more control over their lives or something. We have two, at least a couple of things that people want to talk about. I think Robert Joseph Taylor wants to make a comment. Then Jovier has a question for the group. Robert, do you want to make your comment?

Robert: Yes, sure. I guess I want to make a comment about Dr. Kane's statements. In particular, he had to be really careful about giving round numbers and saying that, because of these round numbers, these are groups we don't have to worry about. American-Indians and Alaskan natives have historically high levels of suicide, much higher than the general population. I don't know if they've been compared to white men but considering if we say that the highest rate is among white men, you compare with the general population, then you would certainly argue, again, have a look at those numbers.

It only makes sense. I mean, American-Indians have a history of, both genocide, and if people know anything about the boarding schools of our cultural genocide. Again, I just think that this really, you have to be really careful about saying, the numbers of suicide among these population is so low, we don't have to worry about it.

Eric: No, I wasn't saying we don't have to worry about it but actually, with American-Indians and Alaskan natives, elder suicide is relatively low. It's the youth and the`- It's like 18 to 40. There they're stupendously high. If you were going to do suicide prevention in those communities, you would again target it to the populations that are dying. I'm not disagreeing with you about the need to pay attention to LGBT youth, or African American youth who die of other causes, but suicides are rising or American Indian, Alaskan Native, absolutely, but if I'm talking about elder suicide, they're very low, even within their own communities compared to their youngers.

It's tailoring the suicide prevention efforts and the suicide prevention research. Then we're always between-- There's another issue, which I think is really complicated. The issue of rates and communities versus overall rates and burden in the country. When you say, "I want to reduce suicide in the United States by 20% by 2025," then inevitably, you're going to have to go to the groups that contribute most numerically, collectively, but at the same time, you have to say, "Which groups are most vulnerable?" Those may be small numbers or smaller numbers.

You take American Indian and Alaskan native deaths, and they're about 700 a year. Nothing that we would do there, even if we reduced it by three quarters would change the big number. There's really an important social discussion of how do you pay attention to disadvantaged groups, people who've been exploited, people whose lives have been historically ruined and attacked, and the bigger number of the national priority? Dr. Taylor, you and I don't disagree on this. It's just I would target it explicitly.

Well, I've got the floor for a minute. The other thing I would bring up for people to think about is there's a tremendous amount of difference between preventing people from becoming suicidal and taking care of the people who are suicidal. If you look at heart disease prevention or cancer prevention or other major prevention programs, HIV prevention, most of the impact has been preventing the person from coming to the, we'll call it the disease state or the, whatever you're trying, and ultimately less impact numerically on once someone's suicidal. Most people who die by suicide, especially men die on their first attempt. They may have seen their primary care doctor or others, but often are not identifiable because at the time they see the primary care doctor they're not suicidal.

They may be sad, alone, have pain, have all the other things, but they're-- If you said, "Are you suicidal?" They may say rightly, "No, I'm not suicidal." Again, we have to think in a very creative way, and this is why what Yeates was saying about primary care and what's protective there, is so important. Coming back to the original point, I don't disagree about underrepresented groups in terms of the big population, it's just, where do you put your effort for which groups at what time in their life course development?

Chip: Marty, can I address that?

Dr. Martha: Yes, I was going to give you your minute back.

Chip: Okay. Eric, and Robert, another reason, though, that I think we need to focus on older black people who apparently have lower rates of suicide is that we're really talking about a deeper understanding of the risk and protective architecture for suicide. It may be that we can learn a great deal about protective factors for more in-depth study of older black men and women.

Then, Eric, coming back to the point about cardiovascular disease and cancer. I agree with your point, we learned about the importance of stopping smoking, reducing smoking, we learned about the importance of controlling blood pressure, and levels of serum lipids in terms of preventing the transition to Frank disease. I would say, in that context, that may be one important avenue will be to learn more about preventing depression.

We know something about preventing depression based on studies of indicated and selective prevention of depression. Since that is a major risk factor for suicide in older adults, together with other things like medical comorbidity and isolation, that may be one way that we can have a dent. By preventing depression, we may prevent some of the terrible downstream sequelling which include not only suicide but also depression, as you and others know is a risk factor for dementia.

Eric: Right, I agree completely. I mean that there are only, I'm putting "only", in 2018, only 77 women, African American women, Black women as it's put in the CDC data, 60 and over who die by suicide is stunning in the face of adversity. What keeps people alive? We tend to always look at the pathology as opposed to the protection. Here that, Jane?

Chip: Eric, I think the other policy issue here, and we've learned this from our friends in Australia, particularly, is that by limiting access to firearms, we can greatly decrease rates of suicide. I know that that's a controversial issue in our country but again, limiting access to means whether it be in the clinical setting, such as you and I work in, or more broadly in the public health setting, that also could have a major impact, as we have learned from our colleagues and friends, Down Under.

Eric: Chip, I think actually, if you look at the data in the 2010 to 2020, actually, the suicide rate in Australia has gone up. There's been a shift, originally it went down after limiting, but there's been a shift to hanging. Where it has been most interesting but also challenging is looking at pesticides. If you try to do a home-based pesticide limitation, putting double lockboxes and other things, it doesn't reduce suicide rates in places like Sri Lanka and other things. However, if you say at the national level, we're limiting production, distribution, purchasing it has a stunning rate on reducing suicide rates. This poses a major challenge for us because we don't have broad institutional ways at this point, policy-driven ways of reducing firearms. There's a VA bill that's working through Congress right now and I think yesterday, the House Senate conferences or maybe it was the house bills took out the firearm part of the legislation. Again, I absolutely agree that we've got to figure out how to reduce firearms, but we have a tremendous cultural challenge in dealing with that. [crosstalk] I'm not disagreeing with you. It's just, how do you do it in a complex culture like ours?

Chip: Thank you.

Yeates: Barbara you had-

Barbara: Yes, just one quick comment. I get the idea that prevention is important and can have a higher net yield, but in cancer we don't turn away people who are sick with cancer and so it's really important for us. I get a little frustrated with these conversations. Is it prevention or intervention? It's both. It's just like end of story, it's both. The other thing that I feel is, I love the idea of studying protective factors in people who seem to be more resilient, but who in the world is going to fund that? How do we think that that is going to work? In the olden days, we used to be able to get funded for research like that, but I don't see it anymore unless we really do somehow focus, embed it within prevention.

Chip: Barbara, your comments really raised the importance of staging approaches. We've learned a great deal from our friends in cancer about the importance of staging approaches to disease. The same is no less true, I think, with respect to mood disorders like major depression. What we've learned there is that early interventions as indicated prevention may prevent the evolution of a more malignant and difficult to treat form of depression, which entails a greater risk for suicide. Again, I think we have to import from our colleagues in oncology, the notion of staging of illness and to think about interventions in terms of the stage of illness, where our person or our patient is.

Eric: Barbara also, I think this goes to institute priorities. Rumor has it, there's some very large studies of children that are following them longitudinally from their prenatal state into adolescence and maybe beyond. Institutes have their priorities. If you look at the strategy right now for NIMH, at least as I've seen it, it's really focused on the interventions for suicidal people. I'm not disagreeing with you, it's all prevention, but there are different strategies for preventing people from becoming suicidal versus dealing with people who are suicidal.

The other side of it is that that our institutes are bifurcated. If you had to pick one group right now, generally speaking, who you would go after, it might be people with substance abuse, alcohol and drug abuse, because they have stunningly high suicide rates. Going into later life those are different institutes. The reality is that the people with the problems don't necessarily fit into the institute related boxes.

Chip: Eric and Barbara, Barbara, I think that one approach to your question about sources of funding for studies of resilience and things of that nature, positive aging really has to do with what I think of as venture intellectual capitalists, entrepreneurs who are interested in investing in high-reward, high-risk intervention research. I think the kind of work that we see unfolding on the West Coast, for example and [unintelligible 03:10:33] Justice Lab, positive aging is a good example of that, where philanthropists and entrepreneurs have been willing to invest, if you will, in positive aging and resilience which is a strong foil to the development of depression and suicidal risk.

Jovier: As one of the funding people, I just wanted to quickly chime in and say that we do recognize that people are complicated and we have the ability, I think, to focus questions around research priorities for people in need. I don't think you have to bifurcate a person to say, it's an alcohol problem or a drug problem or a mental health problem. I do think there is interest across institutes to deal with some of these questions, particularly around that whole notion of deaths of despair and rising rates of suicide alley as they come through.

I don't know the best way to say this, we're dividing church and state here. You guys are the state, I'm the church. There's some things that I can't talk about, you know what I mean? Please continue, I'm enjoying the conversation immensely.

Chip: Jovier I do think that your comment, it really underscores the need for transdisciplinary, for convergence mental health science, that can deal with the complexity of the risk and protective architecture for suicide. As you were pointing out quite correctly, my friend, it transcends the boundaries across NIH institutes and centers so it's reassuring. I have benefited from dual programmatic assignment. Thank you more than once.

Jovier: I think my take home message here is there's a lot to do.

Dr. Martha: I want to bring us back to social connectedness for a second and to remind us that in all of this, there are social relationships and we talk about, first of all, should primary care be screening for loneliness? The challenge of primary care, and I don't know if they could even do that, what are they going to do about it? Or is it really one more way of looking for risk factors?

Are we looking at social connectedness indicators to set that we can as an indicator of other problems? Or is it something that we're going to try to be an indicator that is part of the problem that we're going to try to fix? Or is it something that we really just want to reduce the impact of loneliness because people are lonely or socialization on people? Or are we really going to step back and see if we can change, whatever you step back to, change the circumstances that may put people in situations where they're isolated and lonely? All of that is related to suicide risk, which I think is the conversation we've been having. Thinking through how we can study that as pretty important.

Eric: Marty, I think there's an empirical question, which is, one phase of my life I was doing a lot of geriatric work. One of the things that I found working with older people in nursing home settings, working with primary care doctors, working on psych floors was, it was very hard especially with men to say, "Are you suicidal?," Because I often got, "No," or, "How are you?" "Fine."

Older men actually sound like adolescents. "How are you today?" "I'm fine." "Things going well?" "Yes." My son used to say, "Normal," when he was 15, I'd almost hear the same thing from the older adults I was working with at the same time. The conversation changed to, what are you doing in your life? How are you spending your time? Who are you interacting with?

Then in the context of that, hearing that someone's life had shrunk and it's almost like all had been pruned back, then be able to start talking about how demoralizing this is or how distressing it is. I think one of the empirical questions is, does the interpersonal engagement that goes on with it, is that more effective than a screening questionnaire where you just put in two or three items, which of course is what we've been talking about? This putting in screening forms will, especially this group of people answer, honestly or are they self-regulatory or even self-aware in a way that it brings up. I don't know the answer to that, but it's a question I think that's testable.

Dr. Martha: One of our homebound meals on wheels clients had a fairly big old rambling house in town and because of life and freezing weather and economics had reduced his life, and you've seen this, to one very small, cluttered room and no connections with anybody. The challenge was, metaphorically to move out into the other rooms but in reality, to move out and get back engaged with life, and that's a challenge.

Eric: I always analogize it to a big tree and the branches start to fall off and it gets narrower and narrower and narrower and smaller.

Chip: Eric, your point is so important because it gets at the importance of not using stigmatized and stigmatizing words like suicide or depression. It's much more important to talk to people in terms of the lived experiences that they're having, and to use words that emerge from the particular culture or setting. I learned in Goa, India, for example, not to talk about depression per se, because it's a stigmatized word, but rather to use the Konkani equivalence of words like tension. You have to use the right words, and you have to be willing to listen to capture what's really going on. I think that's the point I hear you saying. Marty, to your point, we can screen and that's important, but screening in the absence of a bridge to services.

Dr. Martha: We're coming to the end of our time, and I will start by saying Liz Necka and Jovier Evans have really been wonderful putting together this group of people. Thank you, and thank you, Yates, it's been great working with you. It's been great seeing everybody and working together. You’ve got the microphone, you can talk.

Yates. Well, I think you've expressed the sentiments. I've learned a lot, I've had a lot of fun with this. Some of the time went by very, very quickly. I wish we had more, and I wish we had different ways of interacting in person and that we will look forward to another opportunity to do that. That said, I think this has worked really well. The technology, thank you all behind the scenes for making that happen and thank you all for being here on the screen and participating so actively. I hope that the discussion doesn't stop here, that there is a way for us to continue to share these ideas, develop and enrich the contacts that are possible towards the concept of transdisciplinary science. What was the term you used, Chip? It was a-

Chip: Convergence.

Yates: Convergence.

Dr. Martha: Convergence.

Yates: I love that. Let's converge again and keep this up.

Jovier: On behalf of Liz and myself and everybody here at NIMH, I really do want to thank you all for bearing with us for these two days on your computer, staring at us on a screen. We really do appreciate it. I've enjoyed all of the conversation and the comments in the chat room, and I apologize to all of you that you couldn't put in your chats. We didn't know that until this morning. It's a work in progress, Zoom, but again, we have a lot to do. I think that was evident from the conversations these last two days. We need you all to think about ways to do it. We are open, we are available, and I guess in the words of Ghostbusters, I'm with the government and I'm here to help.


Thank you all. Please enjoy the rest of your weekend. I need to talk to Marty and Yates for a couple of seconds but thank you. Again, I would like to publicly acknowledge Liz Neca's tireless contributions to putting this all together because she did a wonderful job organizing this and working through all these things. Any good things of this are hers, any problems are totally my fault, but other than that, thank you very, very much.

Yates: Thank you, Liz.

Liz: Thank you.

Evans: Thank you.

Dr. Martha: Bye everybody.

Yates: Bye-bye.