 So now today's webinar again entitled examining mechanisms underlying the association between adverse childhood experiences and health outcomes and older adults in the CLA say it'll it is the presentation is bought by Dr. Divya Joshi. Dr. Joshi is a research associate in the Department of Health Research methods evidence and impact at McMaster. She holds a PhD in clinical epidemiology and completed postdoctoral training and epidemiology at McMaster. Her research interests focus on examining the effects of childhood adversity and multi morbidity related to issues of healthy aging mental health disability functionality and social participation. And so I will now pass it along to our presenter. Thank you. Thank you Jennifer. You're able to see my screen. Okay. Perfect. Thank you. So in this webinar. I'll be discussing findings from some of our projects that described the distribution of adverse childhood experiences in the Canadian population and discuss the mechanisms underlying the association between adverse childhood experiences and multi morbidity in older in middle adults or adults in the CLA. So I'll start by describing adverse childhood experiences or ACEs. They include some of the most intense sources of stress that affect children directly, such as maltreatment, which includes physical sexual and emotional abuse, neglect, and childhood exposure to intimate partner violence. These are some of the most intense experiences that affect children indirectly through their living environments, such as exposure to parental death, divorce or separation, and living with a family member who may have had a mental illness. So an increasing number of studies have identified the long term effects of ACEs on health throughout the life course. Along with immediate effects on health and educational outcomes, ACEs have been associated with higher risks of health harming behaviors, including smoking, alcohol abuse, and drug use. And so ACEs individually and in combination are associated with a greater risk of additional stressors and are linked to the delay of early developmental milestones. ACEs have also been linked to early and late life psychiatric outcomes, poor social outcomes, chronic physical disease, and premature mortality. Moreover, there's also evidence that the more numerous the ACEs, the greater the number of adverse outcomes experienced by an individual. And research has been suggesting that ACEs may actually pose a threat to healthy aging. So here are some findings from a systematic review and a recent meta-analysis that included 23 studies from Europe and North America to estimate the annual health and financial costs accumulated because of ACEs. And as we can see, and these results are for North America only, but we can see that the cost has substantive totaling almost 38 million disability-adjusted life years and 1.3 trillion per year across the two continents for the four risk factors and the six health conditions or six causes of ill health. And what we see here presented on the bars is the proportion of cases associated with ACEs for the risk factors and the health outcomes. And if we look at the population attributable fractions associated with ACEs for depression and anxiety, and for illicit drug use, we can see that that ranges between 31 and 41%. So chronic physical health conditions had lower population attributable fractions to ACEs than the mental health conditions did. We can see that these still reflect the substantive health and financial burden. And so ACE-related cancer, cardiovascular disease, respiratory illness, and diabetes together accounted for an estimated annual 7 million dailies in North America. And so all of this substantial evidence that early life adversity has negative long-term health consequences, the mechanism through which they can for risk for health problems in later life largely remain unclear. Literature has suggested a couple of pathways. The one mechanism literature suggests involves chains of psychosocial events where one negative factor activates another negative factor. And so based on this mechanism, ACEs have a negative effect on mental health, and it leads to adoption of poor or self-harming behaviors, which in turn then would lead to chronic physical health outcomes. However, evidence also suggests that ACEs may directly affect neurological endocrine and immunological development. And so ACEs are associated with increasing biomarkers for inflammation and shortened telomere length, which in turn may then lead to poor health outcomes. However, this model has not been empirically validated or was not previously validated. The theory is that ACEs become biologically embedded. And when we discuss biological embedding, the concepts of allostasis and allostatic load are useful in understanding how adverse experiences in early life may get under the skin, become biologically embedded and lead to poor health outcomes. What I'm showing here is that typically what happens is that when there is exposure to acute stress, our body will activate the HPE axis, and we will have a fight or flight response. So the neuroendocrine and immune systems will adjust to the environmental stressor, and we can maintain a physiological homeostasis. So this adjusting of our systems in response to a stressor is basically what is known as allostasis. But as you can see in this figure, coordination of allostasis or adjustment to a stressor depends on the brain's evaluation of threat and a physiological response to it. So coordination of threat and how our system responds to that threat is also determined by individual differences such as genetics or developmental factors and experience, as well as behavioral factors such as coping and lifestyle factors or historical factors such as trauma and abuse, other major life events, stressful environments that ultimately determine how resilient a person is to stress. And so allostatic load is basically the variantare our body experiences when there's repeated allostatic responses that are activated, and there's repeated attempts made by our physiological systems to adapt to that stressor. This figure here shows a normal physiologic response. So in times of stress, you'll see a physiologic response, and then there's recovery and in our physiologic response, biomarkers, heart rate, blood pressure, things return back to normal. And in allostatic loads, a few different things can happen. So you could have repeated repeated hits from multiple stressors in our body response and the recover, or you can experience lack of adaptation over time where the systems don't respond to the way they should. Or you could have experienced prolonged response where there is no recovery due to an impaired negative feedback loop. So your physiologic response remains constantly elevated. Or you could actually experience an inadequate response or a hypoactive state where your body just fails to show any response to the stressor. So our systems can adapt acutely, but when there is chronic and repeated stress, over time the physiological systems will fail to adapt adequately. And the failure to adapt will lead to potentially irreversible changes in the biological systems. And this dysfunction is assessed through the neuroendocrine metabolic inflammatory and cardiovascular biomarkers. And in addition to the biopsychosocial model also suggests that social factors such as social support availability and social participation may also mediate the association between exposure to stress and health outcomes. So there is this theoretical evidence for us to believe that social engagement may mediate the association between aces and health outcomes, such as multimorbidity. And as a mediator, social engagement is also programmed early in life, and it's fairly stable over time. You know, as such social engagement is important across the life course. For example, if you look at literature on attachment security, attachment security is a form of social engagement in early life. And the importance of social support and participation for for health in adults is also well established in the literature. And so we hypothesized a priority that in addition to allostatic load, aces would be linked to multimorbidity via social engagement. Both allostatic load and social engagement are programmed early in life and are known to have enduring effects across the life course. And so, although studies have evaluated the impact of aces, we observed from our search of the literature that population level prevalence estimates for a broad range of aces, and especially for for aces, such as emotional abuse and neglect in Canada were lacking. And so before examining association between aces and health outcomes, it was essential for us to gain an understanding of the burden and the distribution of aces and the population. So our first objective was to estimate the prevalence of aces by socio demographic characteristics in Canada. And because we have few radical reasons to believe that aces are linked to multimorbidity by social engagement and allostatic load. And because this model had not been empirically validated prior to our analysis. The second objective was to determine whether these two factors allostatic load and social engagement mediate the association between aces and multimorbidity after adjusting for important covariates. And then we also wanted to examine whether this mediation model varied by sex and by age groups. Most of you may be familiar with the CLC design. CLC recruited over 51,000 men and women aged 45 to 85 years at baseline from the 10 Canadian provinces and so participants are grouped into two cohorts. One cohort recruited over 21,000 participants from all 10 provinces. For this cohort, the questionnaire data collected through computer assisted telephone interviews. The comprehensive cohort recruited over 30,000 participants that lived within a 25 to 50 kilometer distance of a CLSA data collection site. Participants also provided data through questionnaires, but those questionnaires were completed through in-home interviews. In addition, more in-depth physical assessments and biological samples, so blood samples and urine samples were also collected from participants in the comprehensive cohort. I just wanted to point out the timeline in terms of when the different study variables were collected. So, ACES data were collected at the first follow up, which started in 2015. And this and ACES data were collected on all CLSA participants in both the tracking and the comprehensive cohort. So the prevalence of ACES were estimated using data from the entire CLSA sample. However, biomarkers were only collected on the roughly 30,000 participants from the comprehensive cohort. Therefore, the mediation analysis was only conducted using data from the comprehensive cohort. And the epigenetic data were collected on a random sample of 1500 participants from the comprehensive cohort. And for the analysis involving epigenetics, which I'll be presenting later in today's presentation, was only based on this sub-sample of participants. So just to briefly describe how the study variables were assessed, ACES were measured using the short form of the childhood experiences of violence questionnaire. And like I said, this questionnaire was administered at the first follow up. The questionnaire contained 14 items pertaining to experience of abuse, neglect, and exposure to intimate partner violence, as well as exposure to other forms of early life trauma before the age of 16 years. The frequency and severity of exposure to mal-treatment items, which are abuse, neglect, and intimate partner violence, for those items, those items were assessed on an ordinal scale. And so the ordinal scale ranged from never one to two times, three to five times, six to ten times, or ten or more times. Those standardized cut points that have been established in the literature to classify participants as having experienced a specific type of mal-treatment. So for example, there were three questions that assess participants' exposure to physical abuse. If participants reported being slapped or hit three or more times, or being pushed or grabbed three or more times, or being kicked or bit, then that participant would be identified to have experienced physical abuse. Then for sexual abuse, if participants reported any occurrence of forced sexual touch or activity, then that participant would have been identified as having sexual abuse and so on. Other forms of early life trauma, such as parental separation, or divorce or parental death, or living with a family member that had poor mental health, those were already dichotomized as yes or no in the questionnaire. And so we created a cumulative ACE score by summing the number of abuse domains that participants experienced, and the three other forms of early life trauma where participants would have answered yes. And so the overall ACE score may range from zero to eight. So if a participant had a score of eight, that would mean that they've experienced all eight forms of ACEs. Allostatic load, or the biomarkers, were assessed at baseline at CLSA comprehensive cohort baseline. So biomarkers included white blood cells, HPAVNC, albumin, alanine amino transfer is created in hemoglobin, ferroirkin, CRP. We included lipid profile, blood pressure and heart rate, and then body composition markers. And so each biomarker was assessed dichotomously as falling within a high risk category or not. And so we created age groups in intervals of five years for males and females. And then we found the biomarker value of the upper quartile for a biomarker where higher values indicate risk or lower quartile for a biomarker like HDL cholesterol where lower values indicate risk. And so we found the cutoff values for the samples distribution for each specific biomarker. So each biomarker was assessed dichotomously as falling within the high risk category or not. And then the high risk category was defined based on the upper lower quartile. And we created an allostatic load index, which basically was a total number of biomarkers that was falling within the high risk category for each participant. This is a typical way of how allostatic load is assessed in the literature as well. And so that's the method we decided to adopt. So social engagement, we created a latent variable for social engagement based on the social support and social participation scales. Social engagement was also based on data that were collected at baseline. So social support was assessed using the 19 item medical outcomes study and social support survey. And it measured perception of emotional support, instrumental assistance, information guidance and feedback and so on. So for participation was assessed as frequency of participation and the frequency was assessed on an ordinal scale. That was at least daily, weekly, monthly or yearly. Participants were presented with a range of various community related activities. And they were asked to indicate the frequency of participation in those activities. Multi-morbidity, which was the outcome, multi-morbidity was assessed at the three year first follow up. Participants reported chronic conditions that were diagnosed by a health professional with past or an expected minimum duration of six months. And so conditions included musculoskeletal disorders, cardiovascular respiratory conditions, endocrinological disorders, neurological conditions, psychiatric conditions, gastrointestinal problems, kidney disease, eye issues and cancer. And so each condition was assessed for presence or absence and then summed up to create a total score. And so the number of conditions was summed up for each participant to create a total multi-morbidity score. With regards to analysis of the data, so we used logistic aggression to obtain the adjusted prevalent estimates of ACEs within the groups that are formed by the socio-demographic characteristics. So we examined prevalence of ACEs by age groups, by sex, country of birth, education, total income, sexual orientation, as well as province of residence. And in this analysis, each variable was adjusted for all other socio-demographic variables. So when we looked at the prevalence by age groups, we would have adjusted for all the other socio-demographic variables and so on. We also weighted, so the analysis was also performed using the rates that were provided by the CLSA to ensure that the results were generalizable to the Canadian population. We used structural equation modeling with full information maximum likelihood estimation to test for mediation and to manage missing data. The path models tested the direct effect of ACEs on allostatic load, social engagement and multi-morbidity. The direct effect of allostatic load and social engagement and multi-morbidity. And then the indirect effect of ACEs on multi-morbidity via allostatic load index and social engagement. We added a covariance term between the two mediators, and we adjusted for age, sex, income, physical activity, nutritional intake, smoking, and alcohol consumption. And so here are some of the descriptive results that give us an overall picture of the distribution of ACEs in the population. So childhood exposure to physical abuse, intimate partner violence, and emotional abuse were the most prevalent types of ACEs that were reported across all participants. Overall, we can see that 62% of participants reported exposure to at least one ACE, 36% reported exposure to two or more ACEs, and then one in five reported exposure to three or more ACEs. Our results also showed substantial heterogeneity in the distribution of ACEs in the population. So we found that men reported more physical abuse, but women reported greater exposure to sexual abuse and emotional abuse and neglect intimate partner violence, as well as living with a family member who had mental health problems. This figure here shows the prevalence of ACEs by age and sex, and so on the X axis is the total ACE score ranging from 0 to 8. On the Y axis is the prevalence estimate for ACEs. And estimates for women are shown as solid lines, estimates for men are shown as bashed lines for the four age groups. And so the blue lines are for the youngest age group, the red lines are for the oldest age group. And what we can actually see is that the prevalence of ACEs was negatively associated with increasing age group. So individuals in the oldest age group. So those would be individuals in the 1930 to 1939 birth cohort. The red lines, they reported the least exposure to all ACEs, with the exception of parental death of a parent. So they reported the least exposure. In each age group. We noticed that women report slightly higher exposure to greater number of ACEs than men do. And although I don't report here, we also found that there were differences in the distribution of cases by sexual orientation, where participants who identified themselves to be of non heterosexual orientation reported greater exposure to many of the ACEs than those who identified themselves as a heterosexual orientation. In addition to Asian sex, socioeconomic factors, as we would have expected their important factors associated with exposure to ACEs. Here is a distribution of the individual types of ACEs by education groups. And so the blue lines is no post secondary education followed by below bachelors bachelors, and then the green lines are for above bachelors. So we actually see that individuals with no post secondary education or education below a bachelor's degree. So the blue and the orange bars had higher prevalence of all ACEs, compared with individuals who had at least a bachelor's degree. And then this graph here shows the distribution of the individual forms of ACEs by the five levels of income. And consistent with what we've observed with education we see that those who had income less than 20,000 for the darker view bars had higher prevalence of all ACEs, except apparently death when compared with those who had an annual income of at least 50,000. We also found that the exposure to ACEs varied across Canadian provinces. And generally speaking, higher proportions of ACEs were reported for BC, for Alberta, Manitoba, Ontario and Quebec. And these findings are consistent with the results that have been previously reported from the Canadian Community Health Survey, which also reported that child abuse rates will lower for Newfoundland and Labrador in the higher in the Prairie region and in BC. We need further research to actually fully evaluate and understand the observed heterogeneity among the provinces. But overall one thing is clear and that is that the results showed that ACEs are highly prevalent across all demographic groups, although some groups in the population do experience an unequal greater burden of ACEs. It's also clear from the literature that higher prevalence rates of ACEs are concerning, given their potential negative consequences on health and well being. So next, using data from the comprehensive cohort participants we tested the association between ACEs and multimorbidity, and the exam and the role of allostatic load and social engagement as mediators of this association. I included ACEs and multimorbidity as continuous variables in the analysis, but I present them as categories here, just for us to get a sense of the distribution. And so we can see that in the comprehensive cohort, 27% of participants had one ACE, about 16 had two, and 22% at three ACEs. Then with regards to multimorbidity we can see that 22% had one chronic condition, 19 had two, and almost 39% had three or more chronic conditions. The average allostatic load index was about 4.2, which basically means that on average participants had about four biomarkers that were falling in the high risk category. The framework here shows the structural model of the factors associated with multimorbidity in the overall comprehensive sample. The model fit statistics indicate good fit of the models, good fit of our data with the hypothesized structural model. In the overall sample, the R square for multimorbidity indicates that ACEs allostatic load and social engagement together accounted for almost 27% of the variation in multimorbidity. This shows that ACEs were associated with multimorbidity directly and indirectly. So the direct effect was an estimate of 0.12 indicating that greater number of ACEs was positively associated with greater number of chronic conditions. Regarding indirect associations, ACEs were negatively related to social engagement. So increase in the number of ACEs was associated with lower social engagement, and lower social engagement was related to multimorbidity. These were positively related to allostatic load. So increase in the number of ACEs was associated with a higher allostatic load and higher allostatic load in turn was related to multimorbidity. Overall, these results suggest that allostatic load and social engagement mediated the association between ACEs and multimorbidity, and increase in number of ACEs was associated with lower social engagement, higher allostatic load, which in turn, then significantly predicted multimorbidity. Here I present the relation between ACEs and multimorbidity, but stratified by sex. So it's this model, but presented separately for males and females. And what we observe is that the results are largely consistent with those that I showed you on the previous slide with the overall model. However, the direct effect of ACEs and the indirect effect of ACEs through the two mediators on multimorbidity was stronger in females than in males. And this may be because females experience more ACEs than males did. And I, you know, and I mentioned this when we were discussing the distribution of ACEs earlier. Females also experience different forms of ACEs. You know, males experience more physical abuse, female tend to experience more of the other types of abuse and other forms of ACEs. Females also tend to experience greater complexity of co-occurrence of ACEs. So the patterns of how ACEs cluster together in females is different from how they may cluster in males. And therefore it is likely that females are more vulnerable to adverse outcomes than males are. Here I present the association, but this time it's stratified by the four age groups. And so regarding age, the direct and the indirect effect of ACEs on multimorbidity were stronger in the younger age population, but they weakened with increasing age. And so among the oldest age group among those 75 to 85 year olds, ACEs had a direct effect on multimorbidity. But we see that the indirect effect through allostatic load and social engagement was not significant. And my understanding or my hypothesis here would be that this may be because by the time participants are 75 to 85 year olds, there is a time lapse between exposure to ACEs and the mediators. It may also be a result of other intervening events that may have taken place, or that multimorbidity is inevitable with increasing age, regardless of prior influences. And for that reason, we see that ACEs have a direct effect, but there is a no indirect effect through the mediators in that oldest age group. So in addition to the allostatic load theory, the potential of epigenetics as a biological mechanism linking early life exposure to long term health outcomes is also gaining prominence. And the idea is that since aging related changes occur in all tissues and organs, and affect the functioning of all physiologic systems, assessing biological aging using the omics approach, such as epigenetics may actually offer new insights into biological processes. So what I present here is our two people, our two trajectories, one in blue, the other in red, and we can see that both are born at the same time. They will always share the same chronological age as shown in the great arrow and measured in yours. And as you can see in the figure, in early life, the red and the blue are assumed to have the same biological age. However, because of epigenetic and other environmental factors and lifestyle choices, they may progress through functional decline that captures biological aging at different rates. And so we can see that the person in red ages biologically more quickly than the person in blue and is therefore likely associated with an earlier onset of morbidity and perhaps premature mortality compared to the person in blue. And so there's several ways of assessing biological age. One of the methods involved examining and estimating the proportion of methylated DNA molecules at CPG sites. There's accumulating evidence that suggesting that epigenetic mechanisms such as DNA modification through methylation may help explain the lasting effects of early life adversity on health. Studies have consistently reported association between accelerated biological aging and morbidity and mortality. And what I present here are DNA methylation based estimators. They're referred commonly, they're referred to as epigenetic clocks, and they're basically composite measures that have been developed to measure aspects of biological aging. The first generation clocks were the ones that were developed around 2013. They include the original Harvard DNA methylation clock, which was based on assessing methylation at 353 CPG sites. This clock was developed across multiple cells and tissue types, including participants or samples obtained from children and adults. And in this clock is strongly correlated with chronological age. But it's based on 71 CPG sites. And this clock is solely trained in whole blood samples that were obtained from adult populations only. Recently, we have the second generation of epigenetic clocks, such as the phenome age and the grim age. And these clocks were developed around 2018-2019, so fairly recent. These second generation clocks selected CPG sites that were associated with risk factors for disease. And so they ended up incorporating clinical biomarkers of physiological dysregulation. And so these newer clocks have been shown to have greater accuracy in predicting physical functioning, in predicting time to various morbidity, such as time to cardiovascular disease, time to cancer, and also time to mortality compared to the first generation clocks. So there is a lot of evidence that's emerging from studies that have been done in children as well as studies that have been done in younger adult populations that have shown association between childhood exposure to adversity and epigenetic aging. And studies have shown that individuals who had experienced sexual abuse in childhood or had experienced childhood exposure to intimate partner violence had lived in poverty, had lived with parents with poor mental health. They tend to display epigenetic age acceleration, where the epigenetic age or the DNA methylation age is greater than the chronological age. And so these new mental analysis actually showed that childhood exposure to traumatic stress was associated with epigenetic age acceleration, with each additional exposure to a new type of trauma being associated with a six month acceleration. But that was only observed when assessed using the Hanham clock, but not using the Harvard clock. So overall, the literature in this area, most of the work has been mainly examined using one or two types of adversity. It's mainly limited to sexual abuse or poverty, as opposed to a wide range of ACEs. Many of the studies have also focused on the first generation, Harvard and Hanham clocks, and have found inconsistent inconclusive findings. So examining the impact of a wide range of ACEs on later generation epigenetic clocks in the older population are quite limited. And so, one of our other objectives was to examine the association of ACEs with epigenetic age acceleration assessed using the second generation clocks, which were the grim age clock and the phenol age clock. And then we also explored the association using the original Harvard and Hanham clocks, and we also examined the association between each individual adversity domain and the epigenetic age acceleration estimators. A brief description of how DNA methylation samples were obtained in the CLSA. So DNA was extracted from frozen peripheral mononuclear cells, which are basically lymphocytes, monocytes and dendritic cells. And then the proportion of methylation on CPG nucleotide base pairs on the DNA extracted from these cells was measured using Epic arrays. So the Epic array quantitatively measures DNA methylation at a large number of CPG sites and a large number of CH sites across the genome. And then the next steps involves performing bisulfite conversion and processing the DNA on the Epic array. And after the pre-processing processing and quality checks were completed, what we would basically obtain are raw DNA methylation values. And these raw values are then transformed into beta values. So for each participant, beta values can range between zero and one. And beta values basically indicate the proportion of methylation at each CPG loci present in the sample. And it would actually average out for each person what's the average methylation across the different CPG sites. Values that are closer to one indicate greater methylation. And so for each clock, we each clock was derived using the weight and these method beta values that have been normalized using the new normalization procedure. And this is just a standard way of how methylation values are obtained in the literature. So here I describe how epigenetic age acceleration was assessed for grim age the grim age is trained to predict time to mortality. And so it basically estimates mortality risk in unit of yours. Grim age was based on seven age related plasma biomarkers, as well as a smoking related DNA methylation based estimator. And these biomarkers were combined into a single composite biomarker, which is the grim age. Pheno H was trained to predict all cause mortality pheno ages based on chronological age and nine other clinically relevant biomarkers. And so for pheno age Harvard and the Hanum clock, the units are basically biological yours biological age assessed in yours. Epigenetic age acceleration was assessed by estimating the residuals for each participant, where we would regress their biological estimate on the chronological age. And the residuals indicate an acceleration so the residuals are positive, then there's an acceleration where their biological age is greater or higher than the chronological age. If the residuals are negative, then there's a deceleration of your biological age compared to the chronological age. So we could see that chronologically participants bear about 60 years old on average. The average grim age is about 56.3 years and participants were almost two years biologically older, when compared to the Harvard estimator. The Hanum age and the pheno age are fairly closer in estimate and because they're derived a lot using a similar algorithm, they tend to be closer, but the pheno age was lower than all other age estimated from other other clocks. And nevertheless, what we see here are correlation plots for each epigenetic clock that is correlated with chronological age so chronological age is on the x axis. The biological age is along the y axis. And we can see that all clocks are significantly correlated with chronological age with the strongest association found for the grim age with a coefficient of point nine. In the whole linear regression, we examined the association between aces and epigenetic age acceleration measures after adjusting for sex income and poor health behaviors. So smoking physical inactivity, nutritional risk and alcohol consumption. And what we basically found is that the association between aces and the four epigenetic clocks in the unadjusted and fully adjusted model. Is presented with a circle, the fully adjusted is with a square. For the Harvard Hanum and pheno age, there was no association. We found that grim age. Asa score was significantly associated with the faster epigenetic age acceleration. So for every additional new type of ace experienced by participant, there was about a point to your so about three months increase in the are a grim age acceleration after readjusted for covariates. The association between individual domains were elevated, but they did not read, reach statistical significance for for all the clocks, likely because of the small sample size with the epigenetic analysis. However, we did find that co occurrence of poor health behaviors demonstrated a gradient association with the grim age acceleration. So on average participants when engaged in any to any three or four negative health behaviors had an acceleration by about 1.3 years 2.8 and almost four years respectively. Just to highlight some limitations exposure to aces was reported retrospectively, and it's possible that there is recall or reporting biases. However, studies have reported good test retest reliability for individual questions as well as for the overall aces score. It is also possible that the prevalence estimates where for aces were underestimated, given that the CLC sample only included community based participants from the 10 provinces. And that the CLC participants on average also have higher education and income than the pain in population. The cross sectional design precludes any claim claims about temporal associations. For example, temporality between exposure to aces and epigenetic age acceleration cannot be established. And so overall, the results support the theory that multimorbidity is a developmental process beginning with early experience and extending throughout the life course. The findings showed that aces increases stress, low birth social engagement increases static load, which in turn results in increased multimorbidity. We also have evidence to demonstrate that social engagement facilitated at any age for those at higher risk of multimorbidity may reduce or delay the negative health consequences. So aces also provide indication of epigenetic programming from exposure to early life stress. So exposure to aces may induce DNA methylation changes that may be persistent across the life course, and especially in the absence of health behavior and lifestyle interventions. In the next steps, it would be interesting to examine whether aces become biologically embedded and lead to premature biological aging, which in turn that increase the risk of age related morbidity and mortality. So aces that increase awareness of aces and their long lasting consequences that support positive parenting promote healthy child development and overall quality of health environments are needed to prevent exposure to childhood adversity. And trauma informed approaches need to be developed and promoted to assist individuals that are affected by aces. And finally clinicians can play an important role by being cognizant about aces and by implementing trauma informed care to alleviate the harms caused by aces. I just wanted to acknowledge and thank all the co authors on these projects, as well as the participants and acknowledge the funding received from CIHR and CFI. And that's all I have. Thank you. Great. I think you probably need a quick drink of water or take a breath that was an excellent presentation and we've had some positive comments already. We have about 10 minutes left for presentation presentations questions. And I'll start going through these, which are posted in the Q&A section. Just reminder if anybody has any questions to please post them in the Q&A box, and we'll be able to keep track of them. Also anything that's not answered during the webinar we can follow up directly with you. Here we go. So the first question, which came in first come first served is the aces cost slide what was used to create it aces scores have a range and there is a relationship in that higher scores are associated with greater long term adverse events and conditions. Sorry, I can catch that first part but with that sorry. So the aces if you also want to follow along Divya you can see the questions to by going in the Q&A. Yeah. So the aces cost slide what was used to create it aces scores have a range and there is a relationship in the higher scores are associated with greater long term adverse events. Yeah, so that was actually a systematic review that was published in Lancet Public Health. And it included a wide range of aces across different studies, but the more common ones where abuse neglect and more of the household dysfunctional and household adversities. And is it possible to reverse biological embedding or the biologically damaging impacts of aces. Yeah, that's a good question. Theoretically it appears to be possible. However, most studies to date have not actually examined whether preventative interventions can potentially reverse biological embedding. But I should mention that there was a systematic review that looked at interventions related to cognitive based home visiting and parent child skill building and so on. And they have been successfully targeted at cholesterol level outcomes. And so researchers found that cortisol can be modified through interventions. I think it's important to remember that earlier timing of intervention, as well as the duration of the intervention would also play a role in intervention success. But there is some evidence that, or at least as early emerging evidence that biological embedding is potentially reversible. Keeping in mind though that social engagement and an allostatic load are potently influenced by childhood experiences. And it becomes increasingly difficult to reverse with advancing age. And so, you know, again, the timing of intervention is critical. Great. Next question, does the presenter think that the lower aces as one ages is actually lower aces or lower recall reporting, or the impact has lessened in some conscious way. We have actually considered why there is lower reporting of aces in the oldest group. There's literature from the Canadian Community Health Survey that was published in 2012, as well as data from the Genesis project. And they've actually also reported similar results with lower reporting in that cohort that was born around 1930. I think it's important for us to understand the disruptions and the trauma that may have been caused by World War II. But I think it's also important for us to keep in mind that aces may be affected by secular trends. So younger individuals may be more likely to acknowledge and report maltreatment, maybe as a result of increased media coverage. Whereas lower reporting among older adults, maybe because of their reluctance to disclose experiences, which they may have viewed as stigmatizing during the period that they were born. One other thing that we should consider and keep in mind when interpreting these trends across age groups is that in the oldest age group. That estimate may actually be influenced by premature mortality, where participants who experienced aces may have experienced premature mortality and were not around for us to be examined. So I think a number of factors may have played a role for why there's lower reporting of aces in that oldest age group. All right, you've got lots of questions coming in. So we may need to follow up directly with some participants. Next question is, if I understand correctly, allostatic load and multimorbidity are assessed three years apart. And I'm wondering if the temporal spacing is insufficient to assess the link between them. Yeah, it's not really clear what is considered as ideal timing or ideal spacing, but in any case I think that optimal timing may actually be age dependent. So in earlier adulthood, so among the 45 to 54 year old, when allostatic load and multimorbidity are less confounded, the three year period may actually be sufficient to capture the changes that are occurring from allostatic load to multimorbidity. But when you actually look at the findings in the older age groups. It appears that the three-year temporal spacing between allostatic load and multimorbidity may actually not be sufficient as in perhaps allostatic load and multimorbidity almost start to coexist in older age group. So you may want to actually narrow the assessment timeframe to be able to capture if there is an allostatic load, does that lead to increase in multimorbidity? Shortening the assessment period in the older age group may have actually helped us identify if there is a relationship. And I'm wondering if that is actually the reason why we observe an indirect effect in all age groups. But we don't observe an indirect effect in that very oldest age group. So I think that optimal timing may actually be age dependent. Okay, we probably only have time for one more. In the mediation analysis, you were suggesting that level of social engagement, for example, is on the causal pathway to multimorbidity. Is it possible the causal relationship is in the reverse direction? It is possible. I would not deny that possibility of multimorbidity lowering social engagement, especially because multimorbidity may have been present at baseline already. However, just to assess the robustness of this model, we did conduct other analysis where we adjusted for baseline multimorbidity in the analysis. And we actually observed that the estimates were slightly lower, but the overall paths or significance of the paths did not change. And I think this speaks to the stability of multimorbidity that the model is actually fairly, this is not, the model is not driven by history of multimorbidity that despite adjusting for multimorbidity at baseline, we can actually still observe the same association and paths. So it's possible and we do need a true longitudinal design to be able to capture changes. But adjusting for baseline multimorbidity did actually provide us with some evidence that history of multimorbidity is not impacting the results. The model is stable in that regard. Well, we will, I'm not sure how exactly we follow up. We actually haven't had this many questions in quite a while. So congratulations for generating lots of questions and discussion, discussion Divya. But we do have to close our webinar. So thank you again to our presenter we really appreciate your participation in the CLSA webinar series. I'd like to remind everyone that the next deadline for data access applications is on March 30 of next month. So please visit the CLSA website under data access to review available data, as well as additional details about the application process. I'd also like to remind everyone to please complete their anonymous survey upon exiting the zoom session. In terms of our upcoming webinar. It's a history webinar. But we will registration details for our marked CLSA website will be posted we're just in the finishing just putting the finishing touches on the presenter and the titling. And remember the CLSA promotes the webinar series using the hashtag CLSA webinar. We invite you to follow us on Twitter at at CLSA underscore EL CV. And the last thing would be is if if Divya does want to stay on for a few minutes and if there's anyone who specifically wanted to have their question addressed. I don't know if you have the time to stay on but we could also do it that way. I can stay back or I can respond back to the participants if I have a way to contact them that either was fine with me. So Laura is there I noticed a few of our participants submitted questions anonymously in the Q&A can we reply back to those participants. You know, they would have to provide us with their contact information. If they're comfortable doing that. Otherwise, Divya if Divya has time she can, she can take that time to answer a couple more questions, or if she wants to share her contact information that's also a possible way to to be in touch. I'm just putting my contact information in the chat box if you want to stay with me. Yeah, I see most people have left the left the webinar so we can just note to follow up via email for those specific questions. Awesome. Okay. Great. Thank you all. Thank you.