 Great, well, why don't we go ahead and get started. I want to welcome everyone to the CLSA webinar series. I'm Lauren Griffith. I'm the CLSA, one of the associate scientific directors at McMaster University. And I have the great pleasure of introducing some of our presenters for our seminar today. So I would like to introduce our presenters today. The presentation is going to be definitions of social isolation. And this is a pilot study using some of the CLSA data. We're very excited about our presentations now, actually, including some of the results using the CLSA data. And this is one of the first studies that has really been moved along and has used some of this important data. So the two presenters are Verena Menek and Nancy Newell. And Verena is a professor in the Department of Community Health Sciences in the College of Medicine at the University of Manitoba. And her main research interests lie in the area of healthy aging, determinants of healthy aging, age-friendly communities, and health care utilization among older adults, particularly at the end of life. And she is currently the Manitoba lead of the Canadian Longitudinal Study on Aging. And the Manitoba sites claim to fame as they have one of the lowest cancellation rates, even in the dreary winter snowstorms. Nancy Newell is an assistant professor of psychology at Brandon University and a research affiliate with the Center on Aging at the University of Manitoba. And her work has examined some of the causes and consequences of loneliness for older Manitobans in terms of health and longevity. And most recently, she has turned her attention to exploring what types of service or interventions might help people become less lonely or isolated and more socially connected. So without any further ado, I am going to pass these slides along to Nancy to start the webinar. And welcome, Nancy. Hi, everybody. I hope you can hear me. If not, I'm assuming someone will type that in and let me know. Thanks for joining Verena and I online today and greetings from Manitoba. This is our study title, but I just want to start with this one right here. We're going to put it right out there front and center. Verena came up with this great mash-up here of terms. And basically, what we're starting out is saying that defining isolation right now at this juncture is confusing. And it's quite honestly frustrating. We're grappling with definitional issues and other researchers are doing the same. And of course, people are trying to apply program planners, trying to assess isolation in their clients are probably just as frustrated as we are. So we wanted to start and put that right out from the get-go. And so we're going to talk about that a little bit and we're able to continue to grapple with this and try to disentangle some of these issues with the CLSA, so we're excited about doing that. So as I mentioned, it's confusing right now. There's no consistency in terms of definitions of social isolation. And from my point of view, the second bullet there was kind of where I entered this. I studied loneliness for a number of years and I began working with a community organization in Winnipeg. And we really wanted to find out if one of the programs that they were offering, if they were reaching their target group and one of their main target groups for their program were the socially isolated. So I went into the literature intending to find a scale that I could find a cutoff to identify people who are isolated or not. And that's when I discovered, to a great extent, the confusion around definitions. And there's, in terms of scales that have established cutoff so that we can identify people who are isolated or not, they are few and far between. So that's something that was a bit of a frustration for me trying to simply ask and answer that question of are we meeting our target group there? There's no gold standard instrument and as the mishmash here of the isolation term shows, terms are used almost interchangeably and certainly inconsistently at this point. One of the things that we, as researchers anyway, seem to agree on, although there's still some confusion around this, is distinguishing isolation from loneliness. So I've put the two definitions here around social isolation and loneliness on your screen. So there's general agreement that social isolation on the left there is more objective. So it's more of an objective situation. Here's a definition that it's an objective situation of a person and refers to the absence of relationships and contact, so absence of contact and contact and places social isolation actually in one spectrum and social participation at the other. And we oftentimes contrast us with more subjective experience, which is loneliness. So that's one kind of way that we agree in terms of definitions. There's a lot more agreement in the loneliness literature on how to define loneliness by the way. Probably the most common definition is a cognitive approach, which suggests that loneliness stems from a mismatch basically between the relationships that you want compared to the relationships that you have. So loneliness is basically a perception that either the quantity or quality of the relationship are insufficient in some way. So looking at the next slide here, what's important here is that a picture like this, we've got Elizabeth on the left and Dawn on the right, and in our relationships there's social network structure. People in their inner circle versus people that are close but in their outer circles. This can tell us about relationships, maybe numbers, maybe even frequency of contact. We could ask people about their contact with these relationships, but we really can't say anything about loneliness based on these schemas because, and this is a nice contrast here between the objectives. Elizabeth has a lot of relationships objectively on paper, but she may be lonely or she may not be lonely. Similarly, John, who has maybe fewer relationships, we really don't know until we ask John whether he is lonely or not. This doesn't also, this kind of social network type idea doesn't tell us about the role of these relationships. Are people providing them with tangible support, emotional support, et cetera? So it doesn't tell us about social support either. So there's a few dimensions here that we're trying to disentangle and this helps us with definition and measurement. But basically when we contrast isolation and loneliness, we can see that this explains groups of people who might have a lot of relationships like Elizabeth and might be lonely. And it helps explain people who might have few relationships who might be more socially isolated but also might not be lonely. So I'm gonna return to kind of definitional issues and Brian is gonna talk about what we're doing with CLSA data. Before doing that, I just wanted to talk a little bit more about what we know around isolation and loneliness and kind of where the literature I think is going in terms of looking at interventions to try to help people overcome their isolation and loneliness. So what we know, interesting for me anyway, I've been studying loneliness for a number of years and we used to have to spend a lot of time on a slide like this. We used to have to justify why study loneliness, why focus on isolation. And in the last five, six years, I've seen that we don't need to be spending enough as much time. This seems to be more and more accepted that isolation and loneliness are important health risks in particular, there's been a real sense here that's changed them. And I think there's been a couple of game changers in particular, I haven't put references here. Sorry, if you want any references, please contact me. Citation, please contact me after. But one of the game changers I think was Cassiopo and colleagues out of Chicago. And his lab and his research really helped to show how loneliness in particular could directly cause health difficulty that had direct impact on our body. So he looked at things like immune system response, blood pressure, sleep. And so I think his research and his lab and others who studied the direct mechanisms was a real game changer in terms of recognizing loneliness and isolation as health risks. And there's other studies here listed that have also shown the relationship between isolation and health. I think another game changer was this study here from Holt Lundstad in 2010. And if you haven't read this paper, I would recommend it because I think it's one of the most strongly written documents that I have read. Normally researchers, we like to say, more research needs to be done. We don't quite know or we have tentative conclusions. This was a really strongly worded document. It was based on a meta-analysis of over 100 studies that looked at social relationships and mortality. And she basically said, listen folks, we've got enough evidence here to show that social relationship factors like isolation and loneliness are important risk factors and they're comparable or even more important than other established risk factors including smoking and obesity. So I think this was really a game changer in terms of saying, okay, this is an important health factor that we need to be paying attention to. We also know that isolation and loneliness is relatively common in older and later life. Keeping definition of limitations in mind, a couple of studies out of BC and several studies out of Europe have shown that up to 20% of older adults are socially isolated and the number of studies on loneliness suggests that between 20 to 40% reported at a given time being moderately to severely lonely with around seven to 9% of older adults reporting being severely lonely at a given point in time. Of course, these findings differ depending on definitions and samples and nature views. We also know a lot about risk factors. So a lot of literature is focused on risk factors for social isolation and loneliness and let them overlap. We've taken an attempt here at grouping them. So social groups and demographic factors including things like gender, marital status, social groups, including groups like new immigrants to Canada, for example. We know that they're at risk for being isolated and lonely. Life transitions and events, life events and transitions we mean here are things like bereavement, losing a spouse, retirement, losses in certain roles. And as a psychologist I spend a lot of time looking at the kind of personality or psychological factors, perceptions of control over relationships, expectations around our relationships, for example, health related, mobility, environmental factors such as transportation, those of you who are involved in age firmly, these are the kinds of factors that we know. We place people at risk of being isolated or lonely. And all of these factors are really entry points for intervention. So for example, if someone is not able to be as socially active as they want to, due to transportation, that's an entry point, okay? That's where we would need to intervene. If it's more psychological, maybe they've got different expectations around their relationships, that's another entry point to intervention, which is really where the literature is going. But unfortunately at this point, we know more about the risk factors than we know about how to intervene and how to help people overcome loneliness and isolation, which is basically what I state here on this next slide. So that's, I see now that we've kind of set the tone in terms of the importance of isolation and loneliness to our health, the literature is really going towards well, what can we do about it? How can we help people? And unfortunately the literature just is not there at this point and it comes back to definitional issues in a lot of ways. If we're trying to target people, trying to identify people who are isolated, we need to have good definitions and good measures. So some of the problems as we see it is right now, we talk to organizations here in Winnipeg, for example, they say one thing that they really have a struggle with is how to find people. How do I identify people who are socially isolated? The UK came out with a report they called the Hidden Citizens Document, talking about that issue is how to identify isolated individuals in the community. Another problem that we have is we need to know more about how we can target interventions that people at risk of or who are experiencing isolation and loneliness and really importantly, what works best and for which groups of people? And again, this comes back to measuring in order to target groups of individuals who are isolated or not and to be able to track change over time, we really need better measures and better agreements around definitions around social isolation. At this point, when we take a look at the measures and Brayn and I had a lot of discussion around this, there's a lot of confusion and we and others have looked at it in terms of kind of three features that scales or measures are either including one of these features, two of these features, three of these features, there's a real mix in terms of measurement. So structural idea, measuring the number of people in a person's life, the frequency of contact, measures sometimes include functional concepts around social support to what people in people's lives are doing for them, tangible support or emotional support, things like that. And loneliness is sometimes included in definitions and measurements of social isolation. And in some studies, they're including simple indicators like living alone as a measure of loneliness or measure of isolation. So there's a real mixture here. Interestingly enough, Valtorta et al. just came out, if you haven't seen this paper, 2016, BMJ. And she, again, just like we're trying to do, we're trying to make sense of the measures that are being used. So she included, I don't know, about 20 measures or more maybe of isolation and loneliness and she's trying to understand how they line up with different dimensions and she's chosen two dimensions, structural measurements that are more structural in nature versus functional and those that are more subjective in nature versus more objective. And she's got a great graph that you can check out that has placed different measures that we're using right now in the literature and where they stand. And so Brian is gonna be talking about the MOS social support survey, for example, that's in CLSA and Valtorta has classified this one more as being a functional measure, tapping into more of the functional side of relationships, social support and being more on the objective end of this spectrum in terms of measurements. So we come back to this graph in a lot of ways, different dimensions that we're able to try to understand from people, social network versus function of relationships versus loneliness. And so this is kind of what we're trying to do and we're able to do with CLSA data which is what Raina will talk about next. So my task today is to talk a bit about some results using CLSA and again, I hope you can hear me and if not, I'm sure somebody will let me know. We're using the CLSA tracking cohort so ages 45 to 85, 21,000 of people and some of the analyses that I'll show you are based on the 21,000 people and some are only including older adults and some of the analyses are weighted so then they would be weighted up to the population. Just in terms of measures and this really takes what Nancy was talking about a step further into what is actually in CLSA, we have social network size variables so people were asked about the number of children, siblings, relatives, close friends and neighbors so we can use those as a measure of the network size. We also have measures of the frequency of contact with each of the network members and that ranges on a scale from a lot to not much really and you have the details there on the slide. We also have social participation measures so it's eight activities that people have participated in how often do they participate in those and we can think of those as contacts with people not with close people, not people close to the person but still contact with people so they kind of are if you will at the outer edges of a network an indication of the network size. As Nancy mentioned, we have social support variables there's the 1999 medical outcomes study survey and there are four types of social support in that scale. One is affectionate support and I've given you examples here so it's someone who hugs you so you can imagine it's really somebody who gives you that direct support so emotional support. Emotional support then more specifically is a little broader so for example someone you can count on to listen to when you need to talk, still emotional support but not as close as that affectionate one. The first type, we have positive social interaction that's kind of having somebody to go out, hang out with have fun with, relax with and then lastly tangible support somebody to give specific help like you with in confined to bed so this would also be called sometimes in other literature, instrumental support in this scale it's tangible support. Now, when we start to use the data we actually typically can't just use the variable so there's quite a bit of a recording going on and what it's meant by the cartoon here is simply and if you can see it maybe not that well in the middle then a miracle happens so we do a lot of recording of variables until we get derived variables. I don't have time to talk about that but if you have questions I'll certainly be happy to answer those. What I wanted to next is give you a bit of a comparison of the prevalence of social isolation using different definitions and really the purpose here is to show you that depending on how we define social isolation and particularly how we cut off those cut-offs of when do we call somebody socially isolated versus not we get really quite different results. So let me take you through this table. So overall, so this would be based on all tracking CLSA participants so 45 to 85. And in terms of overall one definition that's been used off social isolation is living alone so we have 23% of people who live alone that could be one definition of social isolation. When we go more specifically into frequency of contact though we get quite a different pattern. The first, so the second column there, no contact with social network members in the last six months of year that would probably we would call those as extremely socially isolated. We only have 1.4%. Then when we move into a more a different cut-off different cut-offs a little more lenient cut-offs if you will, we have prevalence of 8.5 versus 26.8. Again, the point being here, depending on your cut-off different depending on your definition you get quite different results. The second point I wanted to make here is that our characteristics that are related to social isolation very depending on our definitions chosen. So take a look at the age effect and what we get under living alone we have a not an unusual pattern here not a surprising pattern that quite a lot more older people age 65 plus are living alone but there's really no difference when we look at our other definitions. So the prevalence rates are pretty similar across the age groups. The same thing happens when we look at gender and we could look at these other factors too but here I'm just showing gender living alone mostly well quite a lot more women live alone. Again, that makes sense. We have that older women without women effect happening here but that's not the core case for the other definitions and I would like you to keep in mind the gender effect that flips to the men that men are slightly more likely to be socially isolated given the other definitions than women. So keep that in mind as I show you some other results. The next piece then I wanted to show you is looking at the relationship then between social isolation or what we now call social network groups and social support. So how does that network size, the frequency of contact how does that relate to what people what the kind of supports people actually get? So first of all, we conducted cluster analysis to group people and I'll go into a little more detail in a moment. We then compared them these network clusters or groups based on socio demographic and health variables and then we look at the relationship between those network groups and social support. I can also mention that in additional analysis we're looking at health outcomes but I really don't have time to talk about those right now but we were moving that way as well. So just a word on cluster analysis. Cluster analysis tries to group people. So it's not like factor analysis which groups variables cluster analysis groups people and we're trying to find groups that are similar groups of people that are similar to each other but different from the other groups. And in this case, our clustering variables are social network size, frequency of contact with no sort of network members and the social participation variables. The image you have here is simply a hypothetical example which shows you that we're trying to identify these groups of people that are similar to each other. Now, if you've ever done cluster analysis or factor analysis, you realize you know that this is not a perfect science. You have to interpret clusters. You have to make judgment calls as to how many clusters or factors and factor analysis you want to include and the real challenge is how to label them. So actually our team has gone back and forth many, many times as to the labels we attach to clusters. So let me give you the results. We're finding six clusters and just as a general comment, they generally move from larger networks to more restricted networks. So let me take you through this. They're ordered by the way in terms of their prevalence. So the diverse cluster, this is people who have large and diverse social networks. They have a lot of family, they have a lot of friends, they have a lot of social participation. We have 25% of those. We have a second cluster with, which really is very similar to the first one but just doesn't have many siblings. So 24% roughly. So still very diverse, very socially connected. Then we have a family friend-focused cluster. They're less likely to see neighbors, less participation in social activities, also generally smaller family circle. But because they have kind of less of that broader network, we've called them family-friend-focused. Then we have another cluster, about 14% that have few children but relatively higher on contacts with neighbors. We have a few friend cluster. We have with few close friends, hence few friends. And they don't really participate in social activities. And then we have a restricted cluster with few friends and really they're deficient on pretty much everything. They're quite restricted. When you add up these percentages, you can see that about half of the population then is really has very diverse social networks. Then there's some in-between folks and then at the bottom we have roughly 21% that have rather restricted social networks. So when you look at, and now I've added this arrow on the side, really we have a continuum from more socially isolated, the two clusters at the bottom to more socially integrated, the two clusters at the top. So what we can do next is see, how do these groups differ in terms of social, demographic and health factors? And we looked at a number of those. I'm just gonna mention a few things. Our diverse clusters relatively young, by the way I should have mentioned and I failed to do so that this analysis only focuses on the older people, so 65 to 85. So when I say young, this is 65 to 74. So they're relatively young and healthy. Our diverse low sibling cluster is older, so 75 plus. And that kind of makes sense probably we can imagine that these people because they're older they have lost siblings. Some of their siblings may have died. Our family friend focused cluster is kind of in the middle with just nothing really unusual about them on any of the characteristics. Our few children cluster tends to be more single. Again, that makes sense. A lot of these people probably were never in a partnership. They will have not had children. The few friends one is an interesting one because there tends to be more men and also quite a high proportion of married individuals in there. And then we have the restricted cluster which is that just a female more single cluster. Now, when we start to compare them with the social support types that these groups of people get, let me just explain here that we're using our diverse group as the comparison group and we're doing multivariate analysis. We're controlling for socio demographic factors as well as health factors in the analysis. And to see, so how does each one of the clusters differ from the diverse which is the most socially integrated cluster group? When we look at the diverse low sibling cluster there's no difference. So they're really very similar and that kind of again makes sense. Both groups are very, very socially engaged. They have a lot of people around them, a lot of friends and social activities. Where it gets more interesting is in terms of our middle two clusters, the family friend focused one. So again, they tend to be a little more, just have a closer smaller network more again as the title seems to suggest family friend focus rather than that one network. It tends to have less emotional support and positive social interactions. No difference in terms of affection and intangible support relative to diverse cluster. The opposite happens when we have the few children cluster they differ on terms of affection and intangible support. So what we think is going on here is that once you start to lack broader social networks you get some deficiencies and supports but more in terms of those broader, that positive interaction or having fun with other people, getting out, hanging out, having somebody to talk to rather than those, the more closer the activities like the tangible support being activities of daily living support. The few friends cluster and the restricted are not very different. They lack relative to diverse cluster in all types of social support. So we have those 21% odd people who really are quite lacking relative to diverse cluster in all four types of social support. So what does it mean then? What do we take away from this? First of all, we take away from it that we do have a continuum from social integration to social isolation. So yes, we have this effect that the most socially isolated people are, so those with more restricted social networks, that those individuals have fewer social supports, all kinds of social supports. And that fits very much with that literature that Nancy was describing in terms of those other people that are socially isolated, those are our concerning people there, they're lacking in the social supports and therefore also have health issues. What is perhaps more interesting and I think personally is that we have these moderately restricted social networks that also have some problems associated with and those individuals as well may have certain social support needs that are not being met. And that really leads us then back to the issue of what about interventions? How do we target interventions given what we're finding? And what we're thinking is really that there need to be certain intervention for certain groups of people that we need to target interventions at certain deficits. So take somebody who was in that middle group who lacks a wider social network, that individual may really benefit from attending activities, maybe a senior center to get those positive social interactions that at least according to our analysis, they may be lacking. On the other hand, take the person who has a small close family network who may still feel some lacking affection and support. Let me just highlight here again that interesting cluster of men. Men who lack a wider network, yet a lot of them are married. What is going on there? And it leads to some interesting hypotheses around, first of all, perhaps marital relationships, but also what about masculinity? Men not wanting to reach out to the broader network, not getting engaged. What can be done for them that targets their needs and ultimately their social support needs? So ultimately we're saying that certain interventions may really need to be targeted as people with specific needs. So let me throw up this slide again. This jumbled really mess of definitions and where to go from here, but what do we think where we need to go is, first of all, we feel that it is important to separate the social network structure. So what people there are in a person's social network and how often people see them from social support? Because we see different gaps depending in social support. So there are gaps in social support depending on the network structure and the specific constellation of people in a person's network. We do still have this really big problem of cutoffs. What cutoffs are meaningful in terms of identifying extremely socially isolated people? Those who are socially isolated, but also those in between groups. So they have some restrictions in their networks. Where are those cutoffs? So that's one place we really need to go. And again, as I mentioned, we have now, we also have analysis looking at health outcomes and maybe that will help us identify some cutoffs and glaring that will help us ultimately in targeting interventions of people with unique needs. So that concludes our presentation and we'd be happy to take questions. Great, thank you so much, Nancy and Marina. It's a very, very interesting presentation. I'd invite all of the participants to submit questions in the chat panel. There's a few things that have come up quickly. One is just a bit of information from Joanna Trimble saying that there's an excellent documentary on this issue called, I believe it's the Remaining Light by the BCCCPA office. So I'm not sure if either of you are familiar with this. No, I'm not, I'm just asking. Yeah, we can certainly post this. I think it'll go along with the slides. We can post the link to this because I think this could be quite interesting. We have a question from Beryl Cable Williams. Did people who were identified as having low social, or having low networks report having unmet needs or being discontent with their situation? I can take that one, sorry, I'm getting an echo here. In terms of our restricted networks, yes, those are the groups that really have the least social support in terms of all kinds of all types of social support. How happy are the discontent with the situation? That is an interesting question and that's really in a way that's something we don't have in CLSA. We don't have specific variables around the quality of the relationship. So you might have a lot of people in your social network, but they're really not positive relationships. So you might have one person in your social network and the other person in the social support. But in terms of social support, the more restricted, the less social support. In terms of, I just want to add in, we started off and Nancy talked about loneliness and then we didn't talk about loneliness anymore. Loneliness needs to come back into the discussion because it also relates to the quality of having, well, of the relationship in a sense of what Nancy was saying. That's who you have in your relationship actually is what you want in your relationship. So that's for another presentation. We're not, that was just beyond the scope of this one. Yeah, and just to follow up with that, in this particular study, we didn't include loneliness in it. So that would be, I think, a great measure in terms of the actual, if people were satisfied with the relationships. There is one item in CLSA that's part of the CEDS scale, the depression scale that asks about loneliness. And we've kind of discussed whether, particularly not as an outcome in future studies. So it's a good question. Great, there's another question about communication challenges due to sensory losses, vision and hearing. Was this one of the causes of low social supports? We did not look at those variables and partly, well, mainly because there's another project underway that looks at that very issue. When I mentioned health measures, we were only looking at function and chronic conditions as correlates of the social network clusters. When we're looking at then, the clusters and other outcomes, if you will, realizing it's cross-sectional data, we're looking at mental health. So I can tell you that there is a very, very strong effect between social network size. So our restricted clusters have a much, much greater likelihood of having mental health issues, including depression. So that's an aside, but we did not look at vision or hearing. Great, and Natasha Curran, thanks you for your presentation as Dewey. Did the emotional support measurement account for pets or therapy animals or only human interactions? It does not include pets. There is a variable in CLSA around pets and in fact, there's a webinar coming up on pets, right? So the emotional support is only on human interaction. So there's a question about the traditional prevalence of social isolation of 20% is high based on CLSA data. It's more like 8.5% plus 1.5%, it says. So is this in terms of what you found, is this what you expected? I can take that one if you want, Verena. Yeah, the prevalence that I showed there of 20% was actually using what's called the Lubin scale, developed by Lubin, and it's been used as they say in BC and by researchers out of Europe and that's right about the 20% there. This is a scale that includes friends and relatives, including your spouse, how often you see them. And it also includes though, so that's kind of a structural idea and then includes more of functional. Do you feel close to them? Can you talk to your friend about private matters? So I think that what we're finding there, I guess they don't wanna leave you in the wrong direction, is again relating to definitional issues. So using that scale, the Lubin scale, that includes structure and function, that's where we're seeing a higher prevalence, I guess of isolation. And the percents that Verena put there on the slide, we're still looking into that by the way, but she just put some measures that we came up with around structure, how often people are seen individuals in their network. And so that's where I think we're getting the lower prevalence. If you wanna add on to that Verena? Yeah, just to add on, first of all, you can't sum 8.5 plus 1.5 because the 1.5 are included in the 8.5, it's just a different cutoff. But if you went back to the slide of the different definitions, the most relaxed one that we put up, it would be around that actually was in the 26%. Not that far off, but it depends hugely on your definition and your cutoff. It depends on the measures. So even if we go back to our cluster analysis and just our two most restricted clusters, if we add them up, we're at 21% of the population of older adults. The other ones were including the younger. So it depends on definitions, variables, and cutoffs. Great, so we have another question from Ann Tully, who's actually gonna be our presenter, our next CLSA webinar presenter. And she says, she's curious if you saw any interesting patterns in relation to SES. And she says she realizes that loneliness and social isolation may not play out in the expected direction. There was surprising, little interesting with SES and I'm defining your SES in terms of education and income. In fact, the paper that is on the review, we ended up not having income in there because it just, we didn't add anything surprisingly, actually. But, and even education, which is still, we still have that in the analysis. It's nothing exciting there. Okay, from Joanna Trimble. As we see the results of social isolation of our elders at the same time, we are seeing cuts in social programming for this group. I think it's because the programs are seen as frills. How do we get the word out about the importance of such programs to the people who fund them? I think this is a great question for sure. I'm not sure if people are following the UK campaign to end loneliness. It's a really bold campaign, campaign to end loneliness. And I think they're kind of leading the way in a lot of ways internationally. They recently came out with a document that was aimed at program planners, service providers, urging them to measure and evaluate their programs. And they actually included measures of loneliness, five measures of loneliness that the program planners could choose from. And the argument here is that, yes, funders are wanting to show that what they're funding I guess especially in the UK or at this juncture makes a difference. And so they're needing that evaluation component. So I guess I would answer that in terms of if we're doing great things, we need to show that they are actually working and for who they're working for. So learning more about intervention and what's working I think will help us with that. Okay, and one more question about would you have any ideas on which variables would be most useful to target isolated caregivers in the community? I'm not sure if I, is that, it's the question, maybe we could have a clarification on what that question is in terms of CLSA, which variables. So if you identify the caregivers in CLSA, which ones of the variables you would use, is that the question? So maybe we can... Yeah, so maybe we could see if there's a clarification to that. In the meantime, a question about why you only included the tracking telephone interview cohort. If the cohort with the clinical exam was included, where you get the same six clusters and is that something that you plan to do? Well, that's why we labeled it a pilot study. We've used tracking cohort people because that's where the people, that was the cohort that was available when we requested the data. Now, as the commenters indicates, we have the comprehensive cohort as well. Yes, I mean, the idea is to look at the other full sample which we get different clusters possibly. I mean, cluster analysis, just like people who have done factor analysis, cluster analysis is very much dependent on, well, people who you have. You're trying to group people in your sample, but also very much dependent on the variables included in the cluster analysis. So yes, the idea is to move towards the comprehensive cohort again. The disadvantage being is now it's no longer nationally representative. But yes, gosh, that was a long-winded answer to saying, yes, we will use the other sample as we will. At the same time, I don't know if we would expect that they would differ greatly. There's other literature that looks at some samples and found not the exact same clusters, but similar clusters. So it's centered within a larger literature that we probably wouldn't find too much variation, but there's only one way to find out, I guess. So we did get some clarification here. She says that she's simply interested in any useful findings applicable to caregivers in social isolation. Well, I guess, you know, we, as a team, we've talked extensively about definitions of social isolation. So I will give my opinion. I do think it's important to separate the structure from the function. So, because I really, so in other words, the size and the frequency of contact is one issue and what it is that these people do is another issue, rather than merging the two. And when you look at the social isolation literature, it is very, very messy and some definitions will merge the two and Nancy was mentioning the lupin scale. And while I like the lupin scale in one sense, it still merges the two issues. I would suggest to keep the network size separate from the function, just to get a sense of, so you could have one person and that person could be a super helper versus one person who doesn't help at all. You could have a spouse who is not sufficiently supportive. And I think we've seen some of that perhaps in our one of our clusters. These are married people, but yet they're lacking in social support. So I guess my suggestion would simply be, keep the support variables separate from the network size variables. Okay, there's another comment. I think many elders self isolate because they're ashamed of their infirmities and don't want to feel maybe a burden on others. Have you looked at self image of those who are socially isolated? We can't do that with CLSA, but it's a very big issue that the self isolation and why people are socially isolated. So are people, do they want to be alone? Do they, are they just happy to be alone? We certainly said people like that, are they alone, but they feel the stigma, for example, of being lonely, admitting that they're lonely. So all these issues actually we are thinking about and Nancy can elaborate on if you want, but CLSA, just in terms of the specific question, CLSA does not allow us to look at that at this point. I think it's a great question and just kind of gets at that idea of, especially I'm looking a lot right now at the extremely isolated group of individuals and people often say, well, some people just want to be alone, but others, as you say, they might be self isolating for different reasons. And I think those are important questions that you raised. So I'm going to exercise my moderator's prerogative and take the last question as there's no additional ones here in the screen. I'm wondering if it's interesting that you look at in the 65 to 85, that age is one of the things that defines these groups. Do you find that social isolation is changing over time? Is it specific events? Is it something like, has anyone kind of taken a life course perspective to this? And thinking about the next wave of data, what sorts of things do you think that you'd like to look at in terms of longitudinal analysis? Well, people have looked at life events and Nancy alluded to that for sure in terms of the important transitions over the life course and the big transition would be widowhood where your network changes very abruptly, sometimes more predictably, retirement is another one, divorce. So those people have certainly looked at that where I think we can go with CLSA longitudinally, we'll see how these network structures change over time. So for example, so we had six clusters, well, really five. I mean, if you think about the two diverse, they're really kind of similar five groups of people. Do they change over time? How do they change? What makes them change? So do people become more isolated or do they become less isolated and why? So I think the longitudinal piece will be hugely important and will be eagerly awaiting future phases of the data. And I think the person who brought up caregiving too, that's a really important piece because we're finding and we're often asked, can you be lonely or isolated when you're married? And of course, we're finding that, yes. And so what we're potentially not disentangling there is whether some of these individuals who are isolated are also caregiving. So they have a spouse, but they're caregivers. And so that's explaining what's going on for those individuals. Great, well, on behalf of all the attendees of the webinar today, I just want to thank you both Nancy and Verena for a very interesting presentation, which stimulated a lot of questions and interesting thoughts on future ways that these data can be used. So that was excellent. So thank you very much. I just want to, as well, announce the next webinar, which is going to be December 6th, again, from one to two. And that is going to be by Anne Toohey. And it's kind of an interesting progression in terms of going from this particular webinar, looking at aging in place with pets, is pet ownership relevant to social participation and life satisfaction for older adults in Canada? And again, in seeing all the interest in the data and the questions that have come up in the chat window, I want to remind people that the CLSA data are now available, both the tracking and the comprehensive. And I would invite people to come to the CLSA website and look and see what is available in terms of the broad range of data that we can interrelate in terms of social isolation and loneliness, but all sorts of other areas. So again, thank you for your time today. And we hope to see many of you back here on December 6th.