Hello, everyone, and welcome to today's one-hour panel discussion webinar, Improving Diabetes Management in Older Adults. We are very glad to know that you've joined us. My name is Dr. Ali Rizvi. I work at the Orlando VA Medical Center and am also a professor of medicine at the University of Central Florida College of Medicine. I have the privilege and honor to serve as the chair of the Clinical Centers and Programs Interest Group of the American Diabetes Association. Just as a reminder, as we go through today's program, if you have any questions, kindly feel free to submit them through the Q&A function in your Zoom control panel. You don't have to wait until the end of the presentation to submit your questions. They will be addressed at the end of the presentation. And the chat function, though, is reserved for other communications. So only use the Q&A function for your questions. I wanted to remind you that as an ADA member, you have access to the Diabetes Pro member forum. Here, you can connect with other members of your interest group and also enter in discussion groups with other like-minded individuals. Now, the American Diabetes Association has been offering cutting-edge learning and educational tools for clinicians since 1940, actually. And the most recent addition to this is the ADA's new Institute of Learning, which has offerings available to members. These include live events as well as recordings, so please be sure to utilize these. And now it is my pleasure to welcome our panel of three eminent speakers. I will go ahead and introduce our first speaker, who is Dr. Medha Munshi. She is actually a professor of medicine now. Congratulations, Dr. Munshi. She informed us that just yesterday she was promoted. She came to know that she was promoted to full professor of medicine at Harvard, where she works. She is a geriatric diabetes expert at Beth Israel Deaconess Medical Center, where she has developed and directs the Geriatric Diabetes Program at the Joslin Diabetes Center. This is a unique program that considers the clinical, functional, and psychosocial barriers that are faced by older adults with diabetes, and then to individualize the treatment strategy for each. Dr. Munshi is extensively published in the area of geriatric diabetes. She has participated in writing panels at the national and international level and has also authored textbooks, so she is a true expert in this field. She also is the founding president of the International Geriatric Diabetes Society. So without further ado, Dr. Munshi. Thank you very much, Dr. Rizvi, for that kind introduction and for this privilege to talk in this webinar. I hope you can see my slides now? Yes. Okay, perfect. So the objective for my presentation today is to discuss the unique characteristics of older adults with diabetes and develop a framework that provides context for the next two very excellent presentations by my esteemed colleagues that you will truly enjoy. But the idea is that why do we need to discuss the older adults differently, and that is what I would like to do in the next 10 to 15 minutes. These are my disclosures. So first question typically comes to mind is that how do we define an older adult, right? It's a cultural phenomenon. It is different in different times that we live in, who is older. And really, typically, one would think that the picture is worth a thousand words, right? So when you think about older adults, some of you might be thinking about somebody that looks like this, while some of us might be thinking about somebody who looks like that. And as you can imagine, these two individuals need to have a very different approach, very different goals, and the personalization of diabetes is considered differently. However, as a geriatrician, I would say that there is a, you know, we still need to consider a little different approach for all elder adults and why healthy older adults still needs a different concept and that different consideration. And that is because of this concept of homeostenosis. Homeostenosis is a progressive constriction of homeostatic reserve. What I mean by that is that if you, you know, as the age increases, there is a decline in physiological reserve. And the physiological reserve is what allows us to maintain homeostasis in presence of environmental, physiological, or emotional stress. So there is sort of a physiological limit beyond which the homeostasis cannot be restored. And that limit actually narrows with increasing age. So you can imagine a magnitude of stressor. And this stressor could be illness, a sort of trauma, or even psychosocial issue. If that happens at a younger age, it may fall below the physiological limit and the person might rebound or recover from that and restore the homeostasis. The same magnitude of stressor. If that happens at an older age, that may fall outside the limit and the body may not be able to recover. And we see that routinely in our practice where, you know, a pneumonia or UTI can be easily treated in a younger person while it can lead to multiple sort of downstream complications and lead to poor outcome in an older person. Another difference between older and younger patients is that older adults have more comorbidities associated with diabetes. And we know that several conditions occur at higher frequency with aging and diabetes. Some of them, like macro and microvascular diseases, we understand that well, and we know how to manage them. However, there is a geriatric syndrome, which includes multiple conditions like cognitive dysfunction, depression, physical disability, polypharmacy, and so forth. They do occur at higher frequency in older adults with diabetes, and they do not only increase the burden of a number of conditions and burden of conditions in the person, but that actually do interfere with diabetes management. And we have shown that in our studies that about one third of our community living clinic coming patients over age of 70 years with diabetes had cognitive dysfunction when we screened them. Looking at, you know, the health and retirement study and how many people over age of 70 had cognitive dysfunction measured there who did not have diabetes, you can see that there is a much higher risk of cognitive dysfunction with diabetes. This is not to say that this is head-to-head comparison because it's completely different assessment. However, there are multiple studies that have shown that there is a high burden of cognitive dysfunction in older adults with diabetes. And we also showed that in people with cognitive dysfunction, there is a poor glycemic, it's associated with poor glycemic control. Similar to that, one third of our clinic population we see as having depressive symptoms, which is much higher than community living older adults without diabetes, not a head-to-head comparison, just giving you an idea that there is a high burden of depression in older adults. And what we found was that people who had depression require more assistance with day-to-day activity. And the reason to understand that is because if these individuals are having difficulty performing day-to-day activity, how much more complexity of diabetes can they manage? So we have to be careful in how we manage them. Another difference is where you see an older person. And we know that there are a large number of older adults who are living and aging successfully in place in the community. There are many people who live in assisted living facilities where they are provided some assistance with day-to-day activities. There are older adults in the hospital, older adults who are going through rehab, and older adults who are in completely supervised environment. And the reason to know that is because the diabetes management challenges are different. So when you think about community living older adults, the patient characteristics include fairly stable population who are high-functioning, and they may or may not rely on the caregivers. For this population, we need to consider that they require frequent education. Complex medication regimens can be difficult for them to follow successfully off and on, especially when they undergo acute illnesses which impact their cognitive or physical functioning. People who reside in assisted living facility, they usually require partial support with their activities of daily living and instrumental activities of daily living. But they are fairly highly functioning and may need additional support from caregivers once in a while. When we are treating their diabetes, we have to consider that these people may or may not have control over the meal, timing, and content. They may need assistance with administration of oral medications, but they are not typically provided assistance with monitoring or insulin injections. And again, when they undergo acute illnesses, there is a high likelihood that they require more assistance, especially if they are on a complex regimen. People who are going to the short-term rehab, these are the people who require partial or full assistance with ADLs, IADLs for a short period of time. So they are still generally high functioning, and the goal is to transition them to the permanent living situation, hopefully their home. And what do we think about them when we think about diabetes? Well, these people may require tighter glycemic control while they are in the facility because we want to facilitate recuperation and wound healing. And they may benefit from self-management education while they are at the facility to improve their glycemic control and medication management when they go back home. That's the goal. And then there are people who are in long-term care facility where they require assistance and or are dependent completely on others for activities of daily living. They are typically low functioning. They have limited life expectancy and high comorbidity burden. And when we are thinking about diabetes care there, we know that they do not have any control themselves over the timing or content of the meals. The ADLs, IADLs are facilitated by the staff, but they do have high risk of medication side effect, and they do have high risk of frequent acute illnesses, anorexia, and cognitive impairment. So again, where they are requires consideration. And finally, you know, thinking about why we need to think about them differently, A1c, which is typically our go-to as a marker for glycemic control, can be a reliable marker to assess glycemia in the older adults. And part of that is because multiple conditions that occur in older adults and frequently seen in our aging population have changed in RBC turnover and RBC lifespan. And we know that hemoglobin A1c normative data does not account if the RBC lifespan is not for three months. And we know that this change in A1c with different conditions are in different directions. Finally, with using more and more continuous glucose monitoring, we have also shown that hemoglobin A1c does not reflect risk of hypoglycemia or glycemic excursion. And you can see the CGM data showing that hemoglobin A1c of 7.2 in one patient may not look same as hemoglobin A1c of 7.2 in another patient. Hemoglobin 7.3% may look very different from this. And then again, 7.5 can be hyperglycemia consistently. So again, A1c perhaps may not be as reliable in older patients particularly. So to summarize this short presentation, older adults are a heterogeneous population with variable health status, which is not reflected with numerical age. They would have multiple comorbidities along with diabetes that interfere with the diabetes management. And they have variable support system that needs to be considered. And then variety of condition may make hemoglobin A1c unreliable measure in some of these older adults to establish glycemic goal. And with that, I give it back to Dr. Rizvi for the next presentation. Thank you very much, Dr. Monge for that nice synopsis of the characteristics of older adults with diabetes, setting the stage for our second speaker, Dr. Albert Huang, who is Professor of Medicine and Public Health Sciences and Director of the Center for Chronic Disease Research and Policy at the University of Chicago. He is a practicing general internist. He studies clinical and health care policy issues at the intersection of diabetes aging and health economics, uses techniques from these and simulation modeling and analysis of real world data to characterize the heterogeneity of the older diabetes population and to follow its natural history. And his research actually has reshaped the international diabetes care guidelines for older people. Dr. Huang got his undergrad degree and his MD as well as a master's in public health from Harvard before joining the University of Chicago in 2001. Dr. Huang, the title of this presentation will be Classification of Older Adults with Diabetes. Thank you. Thank you, Dr. Rizvi. And I want to thank also Dr. Munshi for laying the groundwork for this next talk. And I hope that my talk will set up the talk for Dr. Kukurman-Yafi. So I'm going to dive a little bit more deeply into, you know, what's been happening in the world of classification of older adults with diabetes. Dr. Munshi laid out actually the general issues around heterogeneity down to based in physiology and geriatrics. But I'm going to show you a few tools that have been developed to that can help guide I think goal setting and treatments selection in older adults, those with type 2 diabetes. So I'll first talk about why it's important to improve older patients' classification. I'll feature three different tools. One is the Life Expectancy Estimator for Older Adults with Diabetes, the LEAD model. The second is a different classification system by comorbidity. So Dr. Munshi talked about the kind of the multiple conditions that are associated with diabetes. And this tool will help you see how these comorbid conditions cluster together and actually show different classes of older adults with diabetes. And third, I'll talk about the hyperglycemia risk stratification tool that identifies people at high, middle, and low risk of severe hypoglycemia, talk about some of the implications. And I hope to show you some patient examples and how to use these tools. They're all publicly available, not in one place, but which we're working on, but they're all publicly available. So the reason for trying to improve older patient classification is really that it's this idea that Dr. Munshi referred to, which is that, you know, we are not making decisions about people just on age anymore because age by itself is not really helpful. Trials, the historic United Kingdom Prospective Diabetes Study actually excluded people over the age of 65. But since that time, trials have no longer excluded patients based on age alone, but the sickest patients are still excluded. For example, frequent exclusion criteria is advanced kidney disease. Having data-driven, valid, and replicable approaches to classification has multiple scientific benefits. So it can help us answer questions like, how well do older trial participants represent real-world populations of older adults? And it can also help us answer questions like, do treatments vary in risks and benefits by these subgroups? In clinical practice, having a replicable approach to classification of patients helps us understand who we're taking care of while we're in clinic. So these are all tools on a path to providing individualized care, also we would call it precision medicine. So the motivations behind each of the tools is slightly different and clearly a given individual will be classified differently when using these tools and individually and in combination. So the reason that life expectancy has been a focus for so long is that we know that there's a lag time to benefit for any given therapy. Patients need to live long enough to be beyond the lag time to benefit for any given therapeutic decision. So an example is in the United Kingdom Prospective Diabetes Study, patients needed to live at least nine years to see a separation of cumulative vent curves. That means in that trial intensive glucose control, A1C less than seven, required nine years. That's the lag time to benefit for intensive glucose control. More recently from the CVOTs, a drug like GLP-1 receptor agonist has a cardiovascular benefit effect that we can see in much shorter time frames of three to eighteen months. So we have different times to benefits for different therapies and life expectancy has broader implications for caring for a patient in terms of long-term care planning and for other decisions like cancer screening. Classification by comorbidity is motivated by the idea of trying to figure out what are the major classes and can we anticipate the risks of future events. This can actually help us make drug selection choices. Finally for hypoglycemia risk, I think there's probably no, this is an audience that probably knows this, but identifying people who are at high risk for hypoglycemia can be a way for identifying candidates for de-intensification of treatments and it also can help us make decisions about deployment of devices such as continuous glucose monitors. So the life, I'll just talk to you briefly about how some of these tools were built. The lead model was built using data from the Diabetes and Aging Study. It included a training set of over 90,000 people followed for at least four years and we had three different validation data sets. A 2015 test set, a 2010 cohort, and a 2019 cohort. And we used a technique called predictive modeling based on survival analysis using Cox-Gompertz models. So this model actually generates an actual median life expectancy for a given score. The overall performance of the model is very good with a high C-statistic and it's consistent across the validation samples and it only requires 11 inputs of data. This big table shows you the point system for the lead model that, and the major drivers of life of mortality are actually advanced age, not surprisingly, but in combination with heart failure predicts a great deal of mortality. Other important variables include dementia, metastatic cancer, and you can see the rest. Some of them include durable medical equipment which are indicators of functional impairments such as wheelchair use. And this shows you the scoring. The risk score goes from 0 to over 11 and you can see the median life expectancy associated with each score and the interquartile range. So there's a confidence interval around each of the scores. So I'm going to shift to the second prediction model which is around using comorbid conditions. And in this case, you know, if you approach a patient in clinic, if you look at their problem list, there are a lot of problems, a lot of conditions that a person can have. And so the question we were trying to answer with this model was, can we make sense, can we find the underlying associations between the different conditions? And in this case, we took a list of conditions and used a technique called latent class analysis to say what are the naturally underlying co-occurring conditions and how are they connected. And we found that there were, we found repeatedly that there are three naturally occurring classes. Class 1, which is generally healthy, about 50% of the general geriatric diabetes population. Class 2, which is a little bit sicker. And class 3, which is the sickest and in purple. This graphic shows you the number of conditions that are the histogram of number of conditions from zero to over nine and the percentage of participants on the y-axis. And you can see that class 1, these patients tend to have less than five conditions. Class 2 and 3 have tend to have over five conditions. Class 3 had a lot of cardiovascular disease. Class 2 had a lot of obesity, depression, and incontinence and falls. And I don't want you to look at the numbers, but look at the general patterns of colors. So we've repeated this exercise three times and found basically the three classes across multiple data sets. Again, there's always a class 1, which is the healthiest. Class 2, which tends to have more geriatric conditions and obesity and depression. And class 3, which always has the highest rates of cardiovascular disease. So we found these repeatedly three multiple times. And more recently, we found that the we can actually look forward and see what the complication rates for these three different groups. Class 3 basically has the highest rate of death, hypoglycemia, and cardiovascular conditions. So by doing this exercise, we've discovered that there's a concentration of risks of both of things that we want to prevent in diabetes, but also things that are caused by our treatments like hypoglycemia. All of it seems to be concentrated in this cardiovascular group. So lastly, I want to talk about this hypoglycemia stratification tool. It predicts the risk of an ED or hospitalization for hypoglycemia in the next year. We used a method called recursive partitioning. And again, we had an internal validation set and external and externally validated actually in two separate health systems. This model only requires six inputs. So the inputs are, not surprisingly, a history of hypoglycemia throughout the lifetime. So we know from the pathophysiology of hypoglycemia that hypoglycemia begets hypoglycemia. And so this variable captures that. The other variables are ED visit in the last year, the use of insulin, sulfonylureas, end-stage renal disease, and age above and below 77 years. And if you use these six variables and combinations, you can identify the people who are at high, low, or intermediate risk. And you'll see that the drivers of high risk are really prior hypoglycemic events resulting in ED or hospitalization. So here are where you can go to find each of the models. The lead model, I apologize, is not yet in a web-based form or app. But I want to show you the two others which are available online. The classification by comorbidity is a recently developed app. And the hypoglycemia risk stratification tool is at MD Calc. So I am going to first show you the classification system by condition. So this app, if we, allows you to just simply click whether or not someone has a condition or not. So let's say the person has a history of congestive heart failure. They have a history of kidney disease, end-stage renal disease. They don't have, and it also allows you to, if you don't know the information, it also accounts for that as well. Let's say this person also has coronary heart disease. In the background, this person has a hundred percent probability of being in the cardiac class. So very easy to use and just clicking some buttons. The hypoglycemia risk tool is also available here at MD Calc. Very easy to use. Again, it begins with the history of ED or hospital admission. And let's say this person has no history of ED or hospital admissions. They have had only one ED visit in the last year. Let's say they don't use insulin, sulfonylureas. They have renal failure. They're over the age of 77. They are still classified as low risk. But if they had had a history of ED or hospitalization of hypoglycemia in the past, at least one, they would have been in a higher risk category. So, let's see here. So back to this brief, I just want to finish up. So I'll show you some examples of patients that I have in my own practice. So the first is a 77, 73 year old man with a history of prostate cancer, hypertension, gout, microalvenuria. He is taking insulin. He also has a very high A1C of 9. He has actually a lead model score of only two points. His life expectancy is quite long. It's 14.6 years. He's actually in the comorbidity class of healthy, despite having his multiple conditions. And he has a very low risk of hypoglycemia. I have another patient who's a 91 year old woman with a history of coronary artery disease, heart failure, and stage renal disease. She has a BMI of 40. She's wheelchair-bound, takes no medications for her diabetes. She has a lead score of 10 and a life expectancy of 2.2 years. She's in the cardiac class and she has a low risk of hypoglycemia. So how could this help us? Well, the life expectancy, as I mentioned at the beginning, if you account for the lag time to benefit of given therapies, that first gentleman, he has a life expectancy well over 10 years. So in that case, in patients, older patients with long life expectancy, the general direction of diabetes would still require us to sort of pursue an A1C of less than 7, despite the kind of accounting for the kind of inaccuracies in the measure. But there's a long time to go to acquire glycation end products in his organs. And so pursuing a lower sugar makes sense in that situation. Patients with very short life expectancy, on the other hand, those are cases where we might want to not rely on A1C. And we have some studies that I've conducted that show that the addition of a GLP-1 receptor agonist or SGLT2 inhibitor for the cardiovascular benefits or renal benefits may not be beneficial in that situation because the time to benefit, the time remaining in that person's life is too short. With regards to comorbidity class, those who are in the cardiovascular class clearly would benefit from many of these newer agents like GLP-1 RAs and SGLT2 inhibitors because of their risk profile and their benefit profile. In the case of the geriatric class, the addition of a GLP-1 RA should probably be accompanied by resistance training because of their other risk for falls and for muscle loss and sarcopenia. And finally, for the hypoglycemia risk stratification, that's helpful in terms of deployment of CGM, but also in terms of those who require a glucagon prescription. So I've not discussed other kinds of ways that we can characterize patients like in terms of frailty and functional status. This is not meant to be comprehensive, just these are just three tools that are now available. These tools still need to be validated externally more times to increase confidence in them and to calibrate them. And then for the individual patient, when using these tools, you should be mindful that the actual result may shift depending on additional individual data. But these tools prevent an important step in classification for both research and clinical practice, and much more work is needed regarding how best to incorporate tools into practice and how to communicate the results with patients. Thank you very much. Well, thank you, Dr. Huang, for that nice, succinct presentation, especially the tools we can use for assessing older adults with diabetes, and including just a change in our thought process as to how we approach older adults who have diabetes and comorbidities. I think we're doing fairly good with time. I would like to move on to introduce our third and final speaker, Dr. or Professor Tali Kukraman-Yafi. She is joining us from all the way from Tel Aviv, Israel, where she works as an endocrinologist and clinical epidemiologist. She's a senior physician in the Endocrinology Institute there. She heads the Center for Successful Aging with Diabetes at the Sheba Medical Center, and also heads the Endocrinology and Diabetes Service for Women and Women and Pregnancy. She's associate professor at Tel Aviv University, member of the Herzberg Institute on Aging. She's also a senior international fellow at McMaster University in Canada. Now, the focus of her research has been the challenge of treating, of course, older adults with diabetes, and in this process she has published many articles, given numerous previous talks on this topic. She participates in various working groups and has contributed to local international guidelines on this topic. She's also a researcher, having been the principal investigator of several large cardiovascular trials in people with diabetes, as well as trials of cognitive assessment and abilities in people with diabetes. And recently, her interest has shifted to finding technological solutions for various clinical approaches in people with diabetes in the old subgroup. So, for example, replacing traditional physical capacity assessments with actual technological tools and also studying a digit substitution test of cognitive function in these individuals. So, we're happy to have Dr. Kukurmaniafi. Welcome. And the title, as you can see, of her talk would be Integrating Technology for Better Care of Older People with Diabetes. Thank you so much for that kind introduction, and thank you for the invitation. And thank you, my two previous speakers, for making my talk easier now. So, I will be talking about Integrating Technology for Better Care of People with Diabetes. This is my disclosure slide. And I'll start with the basics. And I think this was alluded to in the two previous talks. I think we all agree that when an older person enters our clinic, what we want for them is that aging with diabetes will look like the people here on the left and not like the people here on the right. And that is challenging. And Dr. Manushi talked about that because people with diabetes have more dementia and have more disability and more depression, and they're more vulnerable to hypoglycemia. And all of these affect self-care capacity, which is a cornerstone in diabetes treatment. So, the question is, how can we do this? How can we help older people with diabetes age in a healthier way? So, it starts off by assessing what I like to term health agility of these older individuals. The American Diabetes Association terms this health status. And many professional organizations, including the American Diabetes Association, recommend that all older people with diabetes over the age of 65 undergo evaluation and accordingly determine the blood glucose, lipid, and blood pressure targets. So, that would mean that these people here on the left that are, no matter what their chronological age is, we would target glucose, lipid, blood pressure targets as for the younger individuals. And the people here on the left, we would probably aim for less stringent control. Now, assessing the health agility status is also important as it helps us finding the people here in the middle. And we have data from numerous trials, like the ones here in the red, showing us that using multidisciplinary interventions, we may slow deterioration to disability and to dementia. So, that's why assessing this health agility is so important. So, the question is, where may technology have a role in this process? So, first of all, we know that insulin scheduling is a major barrier to treatment, both for basal and for bolus insulin. In older people with cognitive deficits, the question of, did I or did I not take my insulin shot, becomes especially cardinal. Older people with diabetes, as we heard, are more vulnerable to hypoglycemia. And severe hypoglycemia may have detrimental effects like falls, like fractures, ER admission, hospitalization, and so forth. And the third place where technology may have a role is in screening and surveillance of this health agility status and possibly helping us create these multidisciplinary interventions that have been shown to be effective. So, who may this be relevant to? On the top of the pyramid, we have the basal-bolus users. Below all basal users or other hypoglycemic agents, insulin scheduling is important for all of these here on the top, and hypoglycemia prevention and management also. And on the bottom here, we have all the 10 million individuals over the age of 65 with diabetes who may enjoy from screening and surveillance of the health agility status and early implementation of preventive strategies. And what technologies do we have or what technologies may help these populations? So, for insulin users, especially the basal-bolus regimen pump, sensor-augmented pump therapy, as well as Bluetooth-enabled insulin pens may aid in insulin scheduling as well as prevention of hypoglycemia. And the new kid on the block, the hybrid closed loops that have entered the market in the last few years, have additionally simplified treatment with a substantial reduction in hypoglycemia. For the larger segment of the population, CGM has an important role. And finally, as I said, for all older people with diabetes, if technology could provide us with a way to screen and have surveillance of the health agility status and provide a means for performing these multidisciplinary interventions, that would be amazing. So in order to illustrate this, what I wanted to do now is share two cases from our clinic. So in our clinic, older people with diabetes undergo periodic multidisciplinary evaluations in which on top of the usual medical and nutritional assessment, they also undergo by a physiotherapist and a neuropsychologist assessment of their cognitive and their physical health agility. So the first patient I want to tell you about is a biology teacher, a retired biology teacher. She's 84 years old. She's had type 1 diabetes for many years now. She also has osteoporosis, osteoarthritis, and has fallen and has suffered from several fractures. When she goes, the physical assessment in our clinic by the physiotherapist, she's found to have low balance and this is with a high risk for falls. I think you would all agree that she may be designated as low physical health agility. She's treated with an insulin pump and uses Libre for continuous monitoring of glucose values. Looking at her glucose values, you can see that she has many hypoglycemic events and these may have devastating effects given the fact that she has high risk for falls given her low balance indices. So what are our treatment goals for her? First of all, we want to adapt glucose control to her physical health agility status, which is we can define it as low. Second, we want to implement physical and nutritional intervention to prevent the falls and maybe further deterioration in her physical status. And while we're doing so, we want to monitor ongoing physical status. So let's start off with the first goal. She's at very high risk for falls and fractures, thus prevention of hypoglycemia and extreme hypoglycemia is a priority. A1c, as Professor Manushi showed us before, has very little meaning here. You can see in her glucose profile that with this technology, we're able to monitor glucose levels, but we're not really able to prevent it. So could hybrid closed-loop be a beneficial tool for her? So what is the evidence for that? So if I had to give this lecture two, three years back, I would say zero. We didn't have any data for people over the age of 65, but in the last several years, several studies have been, trials have been published specifically in this population. And these are two of these studies. One of these studies, the ORACLE study, which used the 670G, actually included frail individuals and cognitively impaired individuals, only a small portion, but still included them. Both studies showed an improvement. This study is the CHAMP-FX device, and both studies showed an improvement in time and range. The ORACLE study showed an improvement in time below range, and the CHAMP-FX study showed an improvement in quality of life. So let's move on to the second case. This is a 75-year-old man. He's a retired engineer. He has type 2 diabetes since 2007, has nephropathy, and A1C of 8. And when we conduct the multidisciplinary evaluation, we see he has normal cognitive agility and low-norm physical agility. He's currently taking a long-acting insulin, a GLP receptor agonist, an SGLT2 metformin, and short-acting insulin in larger meals. So what are the treatment goals for him? Again, we want to adapt glucose control to his physical health agility. We want to implement physical nutritional interventions to prevent falls and further deterioration, and we want to monitor, while we're doing so, the ongoing physical health agility. So let's focus first on the first aim. So he has low normal physical agility indices. That's prevention of both hypoglycemia and hyperglycemia that may cause electrolyte disturbances or further aggregate complications, in his case, are both important. It's impossible to tailor such a treatment strategy based on the A1C. And again, as noted before many times, A1C does not reflect glucose control in older age. However, looking at his CGM, we may enable us to devise a plan that puts the targets that we specified as a priority, as well as providing him of a means, a biofeedback mechanism that may enable him to better control his glucose. So what is the evidence for that? So there have been several studies that included a relatively large proportion of older individuals with type 2 diabetes and have demonstrated the efficacy of CGM in improving glucose control. But recently, we have seen several studies that have been published that actually show an improvement in patient-important outcomes. And this is one of these studies. So it's data from a real-world study from France. And it pertains to approximately 5,000 individuals who started using flash technology. And what we can see here is the incidence rates of diabetes adverse events before the use of flash technology and after the use of flash technology. And as you can see here, on the left, there is a 67% reduction in diabetes adverse events after the initiation of flash technology that is composed of reduction in DKA, hypoglycemia, and comas. You can see here on the right why this is happening. This is happening because we can now actually personalize treatment. We can use precision medicine. And instead of using basal bolus in 65.9% of the cases, it is reduced to 52.7%. Now, as was demonstrated in both cases, a cardinal part of building the treatment plan was the assessment of health agility, both physical and cognitive, both for devising an appropriate treatment plan for glucose, blood pressure, and lipid control, but also in order to employ interventions that may delay further deterioration in cognitive and physical status. But the question is, how do we do that in busy health care systems? Think about the GP, who is usually the one to treat the 14 million people with diabetes over the age of 65 in the U.S. He or she has to consider the physical, the cognitive capacity, the risk for falls, and decide accordingly what are the appropriate glucose targets, the blood pressure, the lipid targets, and do this in five minutes or a little bit more. And the question is, could technology have a role here? So envision a system that continuously collects data from older people with diabetes, preferably passive data where the patient doesn't actually have to enter anything. All this data goes into a data modeling system and gives us the health agility, the continuous health agility status of the patient. This may be added to the glucose and blood pressure recordings. And all this information would be brought up to the physician dashboard in order to determine the appropriate blood pressure, lipid, glucose targets, provide treatment recommendations, monitor adherence to these treatment recommendations, and provide an alert system, and possibly taking it one step further, provide a means for conducting technologically integrated multidisciplinary interventions. Sorry. So our group is working hard on trying to tackle this dream. And up till now, a technological AI-based tool that uses the data collected passively in the cell phone, the accelerometer and gyro data in the cell, in order to assess the physical health agility in a passive manpower, continuous manner has been developed. A digital DSS, which is a commonly used cognitive assessment tool has been developed and validated. And we are now starting a larger validation study. So if anybody's interested in trying these tools, these new toys, please email me. I'll put my email at the end. So to conclude, I think through all these lectures, I think we learned that healthy aging with diabetes is a challenge as recommended by the ADA and many other professional organizations. Glucose, blood pressure, lipid targets should be determined according to the health agility status of the individual. And it's important to assess this in order to determine the appropriate targets, but also in order to implement evidence-based interventions that have been shown to be effective. Technology can provide healthcare providers with better tools to care for older people with diabetes. So I showed you a little bit about the hybrid closed loop, about CGM, and about the vision of having a tool to assess health agility, and the further vision of possibly integrating all of these together into a system that would help healthcare providers provide better care for older people with diabetes. So thank you very much for your attention. Well, thank you, Dr. Kukurmaniappe for that wonderful update, especially touching upon the use of technology in older adults with diabetes. As we wait for some questions to come through, I was going to ask you, how do you feel about the use of technology, the increasing use of technology in older adults with diabetes? And in terms of the information overload, the technology overload, both on the patient and also on the clinician, as these technologies are being developed and fine-tuned, what has been your experience in your patient population as to the reception and acceptance of these tools? I have to say that I have a relatively large clinic of older people with type 1 diabetes, and I find that the hybrid closed loops are a wonderful tool because you still, if, let's say this, that if you have somebody with severe cognitive deficits and no caregiver, you can't use these tools. I mean, it doesn't solve the problem, which is a real problem, a very, very big problem. I think it's a social problem. So if you'd start thinking about what are we going to do with all these people? But if you do have a caregiver, I think the hybrid closed loops are a wonderful solution. And I've been using this actually being very brave and in individuals with dementia and a caregiver, 24-7 caregiver, actually giving them these devices because I think I somehow feel it's much safer than having them on a basal bolus regimen where they might do things that I'm not, I don't actually know that they're doing, but you do need a caregiver to do that. CGM is a wonderful tool. And I think as Dr. Manushi highlighted, in so many older people, A1C doesn't reflect anything. I mean, it's over and beyond the fact that we don't see the variability and we don't see the hypoglycemia and all the things that are so important to older people. It's also about the fact that many times it doesn't represent glucose control. And what is your practice in terms of, there's a question that has come through about using insulin pumps in older adults who are either on pumps or initiating pumps, and especially those that the subpopulation that has dementia or is unable to care for themselves fully and might need caregivers for most of their time. How safe is insulin pump therapy as well as CGM use? How useful? Yeah. So, I think that in people with dementia, CGM may be extremely useful. I mean, I'm actually, I think even in nursing homes, it can actually make the life of the caregivers so much easier if they had CGM, instead of having to prick the patients constantly during the day. Regarding people with dementia and type 1 diabetes and pumps, again, I think it does require a caregiver to be there. We actually just started about a month ago, we started a study with one of the hybrid closed-loop systems, the 780G, in older people. And we're not taking demented patients, but we are including frail and cognitively, and people with MCI in our cohort in order to exactly see that. And it's very, very challenging. So, it means that we as a team spend about a day explaining the system. And afterwards, we have meetings with them every several days. It just really requires a lot of team effort to get them on board and see how it goes. Look forward to you sharing your experience and results with us. One final question, either Dr. Munshi or Dr. Huang, about the use of the new interim-based therapies and SGLT2 inhibitors in older adults. Of course, you know, we are pushing the age limits of using all sorts of medicines for people with diabetes. How has that been your experience in that arena? Do you want to start, Medha? Yeah, happy to. One comment about the previous question that was asked on technology. I think it's important to remember that the heterogeneity of the older population is what you need to think about when you think about how, you know, how do you use technology in older adults. So, before we say that, can we use certain technology in the older adults, the question to be asked is, why are we doing that? Number one, it should not increase the burden. It should help, right? So, depending on who you are using. So, very healthy older adults might be able to use exactly what Dr. Kukerman-Yafi was saying, that they can use all types of therapies. So, it's important to think they can use all types of complex technology. However, that is because we want them to have a good control with a minimum amount of, you know, minimum amount of doing things through the day. Versus when you look at the other side that Dr. Wang described, the people who are, who have limited life expectancy, we may say that, okay, we want technology to decrease the burden. Of monitoring. So, we need to know why we are using before we decide what technology is good for who. That's one. And as far as the second question about SGLE-2 inhibitors, it is still a question. It is, you know, the indications for SGLE-2 inhibitors are more likely to be present in older adults, the congestive heart failure and the renal dysfunction. We do see more problems with DKA and dehydration and fall and hypotension in older adults. So, one of the things, actually, we all get together with our International Geriatric Diabetes Society and and a good way to think about it is if the person cannot drink fluid by themselves, in the sense that they need assistance in even getting to the fluid, it is probably best to avoid that because the chances of them getting dehydrated are going to be much more. Great. Thank you. And Dr. Wang, you want to add to that? I just, yeah, I would. I would just add that it's still, it's an open area where we don't entirely know if these drugs interact with frailty or interact with other, the presence of other comorbid conditions. We might be surprised by when we begin to do trials in those groups. But in the case of GLP-1, I think a frail person who is actively losing weight, that's a pretty good reason not to initiate GLP-1s in that setting. But other, you know, frailty is actually heterogeneous and there may be other frail patients. So, they need the timeline and they need sort of the right conditions and the risk profile to benefit from the addition of these drugs. So, we are kind of at the frontier and we don't, and we are still trying to figure out how best to individualize. Yeah. Well, thank you. And I wanted to thank our panel again for sharing their expertise with us today. Also want to thank all of the participants for joining us. We hope to see you at another ADA webinar in the future, as well as at the scientific sessions in June, right here in Orlando. Thank you. This concludes the session. Have a great afternoon. Thank you. Thank you.

diabetes management

older adults

challenges

strategies

technologies

individualized treatment

homeostenosis

comorbidities

LEAD model

technology integration