 Thanks very much Jennifer. So that's way more of an introduction than I saw. And thank you everyone for attending as we talk about two studies we've done using the data, both to look at aspects of gender and health. So just to make sure that everybody knows a little bit about the CSLA, sorry, CLSA data set, there are approximately 30,000 Canadians age 45 to 85 in the data set, and we use the whole data set. And both the studies that I'm going to talk about, well the study I'm going to talk about and then subsequently the action we'll talk about the second one had findings that we thought were quite unexpected. And then only the parent using methods beyond regression analysis and so there will be some opportunity to discuss methods to uncover such information. Now, when I look at data, one of the questions I always ask of the data is, are there differences between men and women because that's what I do, I study sex and gender and health. And if so, why are the differences, and what might explain those differences, and particularly, are they artifactual or are they real. So one of the gender paradoxes of population health data is that women tend to report poorer subjective health than do men, but in every country of the world, women outlive men. By using survey data, one can look for interactions with sex in the relationship with self rated health, for example. But what we can't do with CLSA data, or with any data set that I've ever seen is to determine whether women and men considered different aspects of life, when they actually were rating their health. So these different aspects might include things such as relative health, do respondents compare themselves to others of the same age. Or do they compare themselves to themselves at an earlier age, or to expectations they might have for their age. And what aspects of life and health are they considering when they make their rating their subjective rating of health. So do they consider mental health or just physical health. Do they consider chronic illness, or only life threatening diagnoses. Do they consider general well being or life satisfaction versus specifically thinking about, for example, physical function and so on. Do they consider large data sets that I'm familiar with ask these questions. So I would Susan is me. This is my hypothesis. I would hypothesize that some of the paradox of women's poor health poor self rated health sorry, and greater longevity is actually artifactual. And it arises from men's and women's different considerations when they're actually ranking their health. This is suggested by qualitative research and some small studies, but needs further study using large and representative data sets and not samples of convenience as was the case with the qualitative studies that looked at this. And now you might be beginning to wonder to wonder why I'm going on and on about this. And it's because, unlike almost all surveys that I'm familiar with or that are reported in the literature. The CLSA seems to demonstrate that women rank their health higher than do men. And so I wondered, or we wondered what's this about. And is it artifactual or is it real. And this is the study that I'm going to talk about has been published. This is a screenshot of the publication and it's in the public domain, it's open access. So if you want to read more, just download it. Let me talk first about the background to the study and the objectives. We know that self rated health is why a widely validated measure of general health. And the questions always ask the same way on all surveys how would you rate your health. Excellent very good good fair poor. So these ratings are often collapsed into a good and a poor rating, though the categories that could get collapsed into good or poor will vary depending on the data set. So, like I said, we know the self rated health is widely validated and a very good measure of general health. We don't although the ratings, how the ratings are formed seems to be dependent on a whole variety of individual and also social factors. Men are thought to focus more specifically on physical well being, and women seem to take a broader view and consider mental health, well being and levels of physical activity and function. And again, I'm not talking about CLSA here and talking about the literature. With the aging, any effects arising from inequalities and social circumstances seem to decline, while the impact of contextual factors like culture or geography where one lives become more prominent in rating of one's health. So we wanted to consider a nuanced understanding of how complex coexisting interconnections amongst sex and social locations or circumstances intersect to shape subjective health and I've introduced the term intersect intersection here. We are not going to talk theoretically about intersectionality but that is certainly a principle or concept behind the work that both of us will present. So the aim of our study was to understand what individual and social factors shapes objective health, and whether such clusters, or sorry whether clusters of these might better explain self rated health of men and women. In other words, we wanted to model statistically the reality that cat the categories of men or women are heterogeneous, and that not all women are the same. So not everyone in the category women is the same. Using multi level modeling, we were able to quantify and differentiate between category variations, but also these within category heterogeneities. We chose multi level modeling to account for the nesting or clustering of individuals within social straight strata defined by, for example, levels of wealth or of education. So we're starting a cluster likely share certain characteristics that shape values and behaviors. And this commonality is important to think about because it actually violates the assumption that each participant in a studies independent of all others. And that's an assumption that's central to regression analysis. So first, we first identified all characteristics that alone were significantly associated with self rated health and the four listed on the screen are what we found sex, education, wealth, and rural urban status. And then we did bivariate analysis, analysis and found that women were 7% less likely to report poor self rated health. Remember, this is the reverse of what the literature suggests. But when we adjusted for education wealth, rural urban state status so the characteristics that alone were significantly associated with self rated health. When we did that adjustment, women then appeared to be 43% less likely to report poor self rated health so that's quite a big difference between men and women. This relationship between sex and self rated health seemed therefore to be strongly intertwined with social factors. However, Poisson regression only allowed us to control for these social factors, and we really wanted to examine them to better understand sex differences and self rated health. But also note that we introduced interaction terms between sex and any of education, wealth, and urban or rural residents, and none was significant. Just to give you a quick view of some of the other aspects of the descriptive analysis. The statistics that align significantly with reports of poor health for both men and women were having more chronic conditions, lower social participation, lower wealth, poor nutrition, depression, impaired hearing and weaker grip strength and grip strength. For those of you who don't know is quite a good indicator of physical function. So significantly actually in both groups in both amongst men and women drinkers rated their this is alcohol drink rated their health as better than did non drinkers and middle levels of income were associated with better self rated health, then we're high income levels and high income levels are greater than 150,000 per year, and that was household household income. There were a number of sex or gender differences in associations that also emerged, and you can see them on this slide. And I'll just let you read them, but in, but in summary so after adjusting for the key confounders. In other words, after removing their impact, sex appears to explain 43% of the variability in poor self rated health. And as I hinted at before, if one wants to study gender. And that's the social opportunities constraints associated with being a woman or a man in a given society. So if one wants to study gender, then the effects of social circumstances should not be removed. And so to consider. I'm sure I'm on the right side here. So to consider gender, we then did multi level analysis. And this is a summary of what we found. So the proportion of variability overall in self rated health was explained by the following so wealth clusters alone explained 21% of the variability. The proportion clusters alone explained 5% sex had very little explanatory values 0.12% and similarly rural urban residents at 0.2%. And we then combined some of these clusters to make more clusters. And when we combined education and wealth. The effect went to 15%. So remember wealth alone explained 21% education alone explained 5%. The intersection of the two as as cluster in class in terms of cluster effects was lower than the 26% that would have come from adding those two effects together. And we're particularly interested in in sex and wealth and sex in general. So, because this is how we get to gender so sex and wealth together explained 15%. But remember again that wealth alone explained 21%. So this dropped this lowered the cluster effect of wealth alone when sex was included as an aspect of the clustering. And we also did three way combinations and with the three locations combined that had the strongest cluster effect were sex wealth and rural urban status. So, I carefully not gone into great depth on the methodology partly because options the methodologist not me, but I want to say a bit about what the meanings of these findings might be. So clusters defined by sex and wealth explained less variability than did well alone. And as I said the effect drop from 21% of the variability for wealth to 15%. And we interpreted this as evidence of a complex intersection of the two characteristics that sex and wealth. That was not additive. And remember that interactions are in a sense a measure of an additive effect. And none of the intersections that sorry none of the interactions that we examined were significant. So we can say that sex somehow mutes the impact of wealth on health, but we don't know why. So I want to go back to the paradox that I started with that women seem to rate their health more poorly than do men, but live longer than men. And perhaps it is an artifact of failing to consider intersections and within group heterogeneity when studying the relationship between sex and self rated health. To me this means it's important to consider gender, as well as sex, and that not all women are the same, and not all women are different from all men. So one method for studying within group variability is what we use multi level analysis analysis and this is certainly not the only level. It's not the only method that can be used to get at either intersectionality or within group variability. It's what we chose to use. And finally, I want to go back to what I said at the beginning and ask again whether some of the paradox might arise actually from sex differences in the factors men and women consider when they're rating their health. And correcting for the failure to consider interactions intersections or within group heterogeneity is a matter of research methodology. And those with greater research skill than have I could make this correction, but understanding differences in what participants actually consider when rating their health requires collecting more data. So I raised the question for the CLSA of whether they might consider this and in future rounds of data collection, ask participants what they thought about when they were making their ratings of subjective health. So the, I want to acknowledge the CLSA of course for the data Queens University, the CI HR who funded the research of which this is a part and future Jen, which is the network. The European Union and Canadian connection that the CI HR funding came through. It was gender net plus actually. And also I want to acknowledge Janelle you who is a student of mine who worked on on this project. I think I will stop sharing. I think I will stop sharing. Wait a second. And then you can share. Okay. Okay, thank you Susan for the presentation actually you made my job a little harder to explain what else we found. What we found for the this is the second study that this is the paper which is under review currently so almost one month which is not a good bad sign so it means people are reading that paper. In fact, we are by the same kind of thinking intersectionally thinking we look at another outcome which is home care access and obviously the data we use is the same data that has been already introduced and most of you already know about it. And some of you might ask, okay, what's the relation to study one first of all the main focus of a study number two that I'm talking about was not looking at gender. So we wanted to identify patterns that kind of or predictors of home care receiving among Canadian population. We had the same idea and we kind of conceptualize it. Okay, so this is something intersectional if you're going on and that's why we put this together. And you will see why it kind of gets connected to the idea of gender. A little bit of the background. So we all know that we are dealing with aging population nothing Canada all over the world. And obviously we need more here. And we also know the pattern of the care is not just so simple it doesn't mean that you know you get a disease you get some kind of health condition and go to doctors nurses and get care. It's some kind of interplay or interrelationship between a lot of characteristics or factors individual contextual and maybe social network. And the good thing is we have a lot of models. A lot of models care people has been developed that kind of describes this pattern. Those models usually identify three kind of groups of factors, let's call it factors for now. The first thing is why people actually need care. So why people start seeking care so I'm thinking about receiving care in general doesn't matter formal informal hospital or home care. The first thing is reason reason for seeking care. So feeling that you are not well. The second thing is predisposing factors or characteristics somehow risk factors. So age, maybe sex, maybe social commitment status, something that kind of is a risk factor for disease and then you will go for to be in the need of the care. And at the same time there are some kind of enabling you should call it circumstances, something at a structural and contextual level so access to care or health care system. These models are very good to when they explain these patterns, but in fact they are not so much useful for prediction. One reason is these models are general the prediction difference for example in here compared to United States. Okay, so maybe we need somehow a new data analysis using CLSA to identify the predictors of the past predictors of home care. That was there somehow objective of the study number the second study and we tried to do it may actually not try we did it separately for informal informal home care use. The idea is identified who is at high kind of propensity for getting receiving home care. That was the idea we wanted. And this is the question we use. So we looked at this question from CLSA, which actually asked simply during the yet last year. How do you receive any professional assistant things like ADL and also ideal and also managing care household, household, and other that this was the definition of formal care. Exactly the same question as but instead of asking from professional doctors nurses, physical therapists, it was if a family member, a friend or a neighbor head. And that was what we kind of defined as informal that was actually the outcome. And as I said we were looking for predictors. And these are it's not an exhaustive list there are other things but actually so we think about socio demographic things like income education sticks obviously, and also family related living and how many generation generation live in in a household mental status, and some of, again, obviously physical and mental care, mental health factors you know ADL safe perception of mental, and also physical health, chronic condition and some contextual factor that we kind of got from CLSA it's some kind of in those indices of material and social deprivation. So we put everything, all of these things in some kind of a statistical procedure which is called recursive partitioning model. So very simply this model is that this kind of statistical analysis is different from regression. So this, I mean the regular regression that you're usually what it does it kind of separates subgroups with higher risk until identify the highest risk groups, not just one. It's able this method is some kind of part of the machine learning procedures kind of identifies or quantifies relationship between a lot, a lot of variables so we know in relation you cannot feed so many values but with this method you can. And output is very, very easy to see and look it's pretty fine. And it's like a tree exactly. But something that is important you identify high risk group at the same time it kicks out on important factors, which is so important because sometimes you think okay, this factor is important for prediction, and it says no it's not. Obviously it's some kind of exploratory. So, there is no causation is some only and only because just gives you the subgroups, it's totally exploratory. And we did it for the outcome of formal care received at home. Before getting there, let me just show you what descriptive results. So obviously we all are familiar with CLSA data we have a high functioning population, mostly because it's in Canada and also because they are not that old and high level of education. What we find is, for both formal and informal care, women kind of receive more, which is again it's no surprise. So as we know women access care or use care more frequently. So, so far, nothing surprising up to this. But when we generated the regression three mother, there are a lot of information them, obviously we don't want to go there. It's one thing that we totally totally expected that sex would be somewhere there. So maybe not the first one the first one is ideal obviously the kind of physical functioning problems, but we would expect sex should be somewhere there, but it was not. So I did a lot of you know kind of a modification and tuning and everything with the mother but you know sex was not really there. We expected sex be somewhere there as I said we did not focus on sex we just wanted to identify different kind of factors and see which group is high risk, more or at more risk or higher risk, and we expected sex would be 100% there but it didn't happen. So it was very kind of unexpected results that we found in this in the second analysis actually in the second study. And let's just focus on one part of this model or regression three C, because it's much easier to look at it as I said sex is not there but what was there, a deal was there. Marital status was there it was one of the important thing that kind of people access for more care, and also age, obviously, age is a factor, but the main thing is again we don't want to talk about that part of it in the paper we did talk about everything in this presentation we wanted to see what happened to sex. It was not there. When we did the same procedure, we actually repeat the same analysis for informal care. In this time, sex was a predictor, but that's so much important. So it's you know the model is a little bit is not more complex is a little bit larger because informal care is a little bit more kind of more complex behavior, asking it needs some kind of social network is a little more complex compared to format, but anyhow we got a little larger model but it's interpretation is not that hard you have just to focus part of the tree and say okay this part of the tree I just want to explore. Anyhow, sex was a predictor, but not so much. Not so important you can see sometimes you see sex there, we would expect to see sex right here. Even there. So, but we did not. But at least sex was somewhere there. And also here it was one of the predictors, but not anymore important. I would still say, this is some unexpected finding. And we really try to make some sense of this data. Because we all know the data is good CLSA is the very robust database and it's representative it's value it's really good kind of database. And this method that we use its exploratory method, but for identification of most important factor it's very powerful. So, maybe it doesn't give you any some kind of ideological information but at least for finding subgroups this is the way to go, but it was not there. So, let's try to make sense of the finding. So what's happening there when sex was not a predictor of former care, what that means, it means men versus women they have equitable access to care for market. In fact, it's a good thing. In fact, it's a reason it's not a risk factor is not something that separates men and women. So it's equity across sex groups. And we also, as I said, we also find the same thing for other social status like immigration status and lower social status, less education and less income. But anyhow, so what we found is some equity. I think most of the thing that you identify in former care model was about really diseases but not for social factors. Okay, so that's something we found, but why. We actually really don't know that is it because of universal Medicare system that we have especially for older adults that some kind of equalize gender inequality, because we do have a Medicare system, it doesn't matter you are a woman or you are a man, so you will get the care that you need. So, we know the system works for a lot, a lot of medical things so we know for access to hospital the system is good so it's some kind of provide equity, but is it the same thing for former care at home and informal care at home. We don't know, but maybe, at least what we observe is there is no inequity there. But maybe as already Susan mentioned, maybe there is something that we really miss. Maybe there is some factors that we have to measure but we don't. What's exactly are those measures I really don't know but there is something related to being a woman versus being a man. So in the previous study we thought okay so how women and men really perceive their heads. So when you ask the same question a man versus a woman, how they think maybe differently. So here, how women versus men in accessing to care behave differently. So maybe that's something that we should measure somehow, but we don't. When you don't measure okay you cannot really see it. The other thing is maybe there are some kind of hidden intersections for the formal care, we were able to kind of tease out or kind of find hidden intersections. And this reason when we included all of those intersections that worked for formal care behavior, sex just disappeared. Why, because you know it was not because of the sex it was because something that we kind of measure by sex, when we include all of them. The sex went away. And we were able to really really tease out intersection, but for formal care, not yet. Maybe if we have more measures, more kind of in factors that created those intersection, we could see the same thing for informal care. And what we observed in for informal care, it was actually because it was a function of that those hidden intersections that we kind of measured these guys in sex, being women versus men. So, in fact, this was some kind of unexpected finding that we found, and we tried to make sense of it. I'm not sure we 100% were able to do it to think that we kind of taught it can explain because the outcome was care is Medicare system actually universal care system in Canada, or these intersections that are there. Yeah, I'm totally totally open to your comment to your question and let's see if we can make a better sense of these findings. At the end, I have to acknowledge almost the same kind of institution that Susan already said but you know I just want to put the name of the co author team, obviously, that's just some feelings was the main author in this paper. And these are the co authors that helped me a lot in kind of conceptualization of this paper, especially because you know the care by itself is not my specialty as a health outcome. And also the health, they helped in kind of writing the paper. And if you want to contact me this is my email address, and now I will stop and would be more than happy to answer any question. You should have. Thank you. Thank you very much to both of you. I do have a barking dog in my background. There's some people in my backyard right now so just for warning. If you can find her if you can post your questions in the Q amp a box. We're trying to get questions focused in that section instead of the chat box, if possible, but I am monitoring both both places. So the first question, there's a question for Susan, and that is how could participate how could the participation rate in the CLSA. Which there's a doctor rena's 2019 ij e papers quoted here as having a 45% overall response was but was that participation rate in the CLSA was about 45% with an overall response rate of 10% so that was quoted. And then how can the participation rate in the CLA CLSA produce a non random sample that could influence your findings such as women reporting higher SRH did the participant participation rate differ by age. And then the second question is, were the sample weights incorporated into the analysis. So, that's a big question from Andrew Patterson and if you wanted to also look at the question Susan that's within the Q amp a box. Okay, thanks. So I also apologize I have a potential barking dog I had to move from one another here. And there's a bit of background noise so I'm sorry. I hope you can hear me okay. First of all, the question about the weights I don't think we adjusted for the weight, the waiting. The question about representational. The representative is the CLSA. So I suppose one could look at the Canadian population age 45 to 85 and compare indicators within the CLSA with with the overall population to get an idea where the representative nature. would not do that. I think that this is a problem with absolutely any research that's done that it one will never have a truly representative sample. And one sort of hopes that the sample is not terribly unrepresentative and carries on from there. So all we can say with both the studies is in the CLSA sample, this is what was found. I hope that adequately answers the question. Okay, great. Thank you and hopefully it didn't if not if there if you want to post a follow up question that's fine too. I have a question in the chat that came from Monica telly, were there differences in formal health care usage in provinces that provide health care without cost versus provinces that use a sliding scale to determine fees for use of supports. Home care, sorry. They're not healthcare. Sorry. No, I get the question. We didn't consider province as a factor. Actually, we did not feed the province in the model. Yes, maybe but we don't know. So the reason is, because you all know, CLSA doesn't include all of the provinces. It's a good, good number of people obviously as a representative sample because of the weighting and sampling strategy, but the question is a very nice. It's a very good question to work on it but maybe if you want to answer the question we need some kind of more census type of data that includes provinces all provinces. We did not in province was not part of the analysis the strategy. So I have no answer for your question which kind of maybe yes. Okay, great. Sometimes we don't always have have the answers to all the questions and that's okay. Another question is for Jennifer can I just add to that. Of course. Also, if one looks at the proportion of the CLSA data set who actually accessed formal care. It's very small. I think it was one or 2% so that to try to break that sample down further by province would have created cells that were just too small to make any kind of valid conclusions about. So the question is for option for informal care, I would say that the sex variable may be influenced by the social network quantity and quality of perceived support from each social type, social tie. And they could be important factors in predicting informal care. Was it possible to assess that using the CLSA data. Thank you for your question. Actually, we did not have all of the social network type of the question how many friends you have and how many times you are seeing your children and the other things. So yes, it can be but again, as I said, it's we were at the mercy of the CLSA data. So when we talked about something that's unmeasured that kind of generates intersection maybe that's one of them. So, because informal care you're totally totally right informal care is relates that accessible having access to informal care having friends having neighbors having family. So that was not the thing that we totally kind of included but actually we include the number of house living arrangements so living in the multi generation kind of multi generation household which was one of the factors for informal care if you go back to the social network and look at it for informal care it was some but not one of the important one, but it was there, but looking at social network the way that we usually do in social network analysis, we did not have access to those data. It actually the data was not there. Again, it can be part of the whole picture of intersectionality that we miss, and it's not there. So, anything Susan. That's fine. I was, I was sitting thinking about what are the policy implications for this and so social connectedness probably is an important local social location to consider, because there are policies that could improve social connectedness, whereas considering something like immigration status. And I would, I would assume that first generation immigrants probably are more connected within their families and more likely to provide care to within the family so providing formal care. But this is not something that can be changed by any kind of policy, short of increasing immigration to change the population demographics. So, it's not a pretty it's an interesting thing to look at but not a particularly useful thing to look at. Great. Um, there was a question that came in in the chat. And that is, if marital status was a significant factor that I'm wondering if the sex of the participants in couples mattered. Maybe you want to take that one on. Yes, obviously so but 100% but the thing is marital status is a kind of measure of having a partner versus not having a partner and sex is just the measure of sex yes it's it is but in fact what we looked at it was intersection between these two. The sex and marital status somehow intersects. First of all, I don't want to say the way that we analyze the data is the way to look at intersectionality by no means I want to say that so there's no standard or gold standard quantitative way for quantification of intersection but what we do, but what we did. If the sex and marital status somehow intertwine and somehow interpolate to generate different rates or different risks for outcome. We have seen marital status and sex, but we did not, we only saw marital status. So because marital status somehow is a measure indirectly of social networks, so when you're married to you have a stronger social network probably. So, but sex was not there. So this question I would say, we measure, I would consider marital status something differently from sex, obviously may and he may consider marital status differently, again, some kind of gender effect that was not really observed. I don't want to say it's not there but it was not as far as I can interpret this kind of unexpected findings. So it's about the time in the presentation that are the webinar that people usually start to drop off so we do have some more time and we will address the outstanding questions but I did want to remind everyone if you can please complete your survey on the before you leave or just after you leave that would be that would be appreciated. So we do have one more question that needs to be addressed for Dr Phillips, and that's the tracking cohort, which is an additional 20,000 participants uses telephone interviews across all 10 provinces, and includes more rural participants. Given that you use the comprehensive cohort, which is the in person cohort. So I wonder if in the tracking cohort of the CLSA, if the results would be different, given that you use the rural urban variable in your clusters as well. I think that's a very good comment and interesting idea, and might move, move this study closer to generalizability, then what we were able to get at with the comprehensive set data set option do you want to add. The tracking cohort doesn't, or didn't have all of the information all of the data that we wanted for this data analysis. Yes, it did have the province and other things that's already here, but for remember what you we just tested for a lot of predictors or I know you're taking with the first study, but to be tested for a lot of factors a lot of variables but you know those are the things that ended up important for us. So that was the reason we use comprehensive because we wanted to use more kind of variety of the data, but 100% when the variety was something important tracking, tracking cohorts would be more important, more kind of use. So I totally agree, but we did not use for that reason. Thank you. And another quick question came in from Kathy PCOR fuller I guess there is no way or no easy way to look at SRH versus more objective measures of health. We actually have just access CLSA data to do some similar analyses, but looking at objective health. I'm very hesitant to call it objective health, because option and I were involved in another large research study using it's the international mobility and aging study. And what we saw from that was that objective health isn't always as objective as you might think it is that women often underestimate their abilities and stop short of reaching their full ability. If one is testing for example grip strength, or standing up from a chair or the various measure objective measures that are available, but we are going to try to look at objective health, and some of the parameters that we've already talked about here. Another question is, does the CLSA data indicate if participants have a regular physician or nurse practitioner. I'm actually trying to think of, think of that as well. I think people who work with CLSA can help. I don't think so. Maybe I'm wrong. Actually, I, this is not the variable that I don't like so much because it doesn't say anything so you might have a doctor but you never go there. We never looked at it, but I'm not sure if there's a data I'm sure that we didn't use that. That's the thing 100%. Even if there was that data we didn't include it in any of these two analysis, but if it's there actually I don't know if somebody can tell us would be helpful. Honestly, I should know this. I know we do ask about healthcare use but whether we ask directly about a regular physician or nurse practitioner, I know we don't use that language so I think it would depend on how the exact question is being answered. Okay, I think probably this may be the last question, but if you do we do have some more time so please feel free to send in any additional questions. More of a statement I think we do have a CLSA base paper regarding self reported sensory problems versus behavioral measures of sensory impairment so that was just a note from Kathy PCORA for. Okay. I don't see any more questions. So I will I think wrap it up. And just, you know, first, firstly by saying thank you again to our presenters we really appreciate your participation in the CLSA webinar series. It really helps to to demonstrate the bring the data to life that we are we're we've been collecting for the past 10 years and we'll be doing so into the future as well. I'd also like to remind everyone that the deadline for the data access applications is the next deadline is January 12 of 2022. So please visit the CLSA website under data access to review what data is available, including note that the COVID-19 questionnaire study data is available as well as details about the application process. I'd also like to remind everyone to complete their anonymous survey upon exiting the session today. So for upcoming webinar, which will be entitled exploring the patterns and impacts of diet and nutrition among older adults in the CLSA. It will take place on November 25 at noon. And it will be presented by Dr. Jacqueline Hurley, as well as Dr. Rachel Murphy details and registration information will be posted on our CLSA website under webinars. If it's not posted already. And remember that the CLSA promotes this webinar series using the hashtag CLSA webinar. We invite you to follow us on Twitter at at CLSA underscore ELCV. So thank you again to our presenters and for everyone who attended today's session.