And so I'd like to welcome our first speaker, Dr. Camille Poe, who's a physician investigator in the diabetes unit at Massachusetts General Hospital. And she's an assistant professor of both medicine and obstetrics, gynecology, and reproductive biology at Harvard Medical School. And she'll be talking to us today about the current paradigm for gestational diabetes mellitus diagnosis. Thanks so much. Thanks to the organizers. And thanks to the audience for coming today to our symposium. So I'm going to be talking about the current and I've added future paradigm for gestational diabetes diagnosis in an effort to set up the next two talks, which will focus on risks of gestational diabetes to the offspring. I have no disclosures. So here's the outline for my talk today. First, I'm going to give you a little bit of background about gestational diabetes, which may be familiar to you. Then I'm going to talk to you about our current diagnosis paradigm, which will include some data from the most recent randomized control trials in this space. And also, then I'm going to talk about our future, what I think our future diagnostic paradigm should be. So background. So gestational diabetes is common. It affects now more than one in 18 pregnancies in the United States. And it's associated with a variety of adverse pregnancy outcomes. And these include adverse and these include things like fetal overgrowth related to that, birth injury to the neonate, and in addition, neonatal hypoglycemia, which you're going to hear a little bit more about later in this symposium. In addition, the mothers with gestational diabetes are at an increased risk of preeclampsia and of cesarean delivery. And then long term, gestational diabetes confers increased risk to the mother in terms of both diabetes and future cardiovascular disease. And in addition to the child exposed in utero have an increased risk of obesity and diabetes later in life. Many of the complications of gestational diabetes can be ascribed to fetal hyperinsulinemia. And this has been described by something called the Pedersen hypothesis that I'm just going to take you through here. So glucose in the mother when it's elevated is readily transmitted across the placenta to the fetus. And then once the fetus can make its own insulin, the fetus will make more insulin in response to those higher levels of glucose. And insulin is a potent growth factor for fetuses in fetal life. And that can lead to the excess growth that we see in these babies and abnormal fat distribution. And then if the hyperinsulinemia persists postpartum, that can lead to neonatal hypoglycemia. So we know from seminal randomized trials that treatment of gestational diabetes, in other words, lowering the blood glucose, lowers the risk of adverse pregnancy outcomes. So I just want to mention two of those seminal studies. So back in 2005 in the New England Journal, this study was published. And in this study, they included women with a fasting glucose less than 140 who had a two hour post 75 gram glucose load glucose between 140 and 198. And they randomized the individuals to either no treatment or treatment to target these glucose levels here. And all the participants in the intervention group were checking their blood sugars, both fasting and postprandially. And they found a decreased risk of serious neonatal morbidity. And this composite outcome really consisted of things that were quite serious, death, shoulder dystocia, fracture, nerve palsy. And that incurred in 1% of the intervention group versus 4% of the standard of care group. There was another trial. So that trial was conducted outside of the United States. There was another trial conducted by the Maternal Fetal Medicine Network in the United States, which was published in the New England Journal in 2009 by Dr. Landon. And in this study, they were doing similar testing to what we do in the United States. And they only included individuals with a fasting glucose that was under 95, so normal fasting glucose. But these individuals had elevated post-load glucoses. And that's why they called it mild gestational diabetes, because they felt like they had clinical equipoise to test this. And so again, the individuals in the study were randomized to either no treatment or treatment to target these goals shown on the screen. And you may recognize these, because these are the same targets that we use today in clinical practice. And they actually did not find a significant reduction in their primary composite outcome, which I've listed for you here, although there was a trend. But they did see a decrease in several very important secondary outcomes, including macrosomia, large for gestational age birth weight, cesarean delivery, shoulder dystocia, and preeclampsia. So even though this was technically a negative trial, it did form the basis for supporting treatment of these individuals, even if their hyperglycemia was so-called mild. So in summary, gestational diabetes is common. It confers a risk for both short and long-term outcomes. And treatment, we know from randomized trials, improves the short-term pregnancy outcomes. We actually don't know or there isn't evidence that treating gestational diabetes improves long-term outcomes for both either mother or neonate. So what about the current diagnosis paradigm? So many in this audience will be familiar that in the United States, we practice what's called two-step screening. And we do it universally in all pregnant women. This is very different from what's done in most of the rest of the world. So if you have no, and I'm not going to get into early screening now, but I'm happy to answer questions about this. But I'm just talking about routine screening in the latter half of pregnancy. So if there's no known diabetes at this 24 to 28 weeks gestation, all pregnant patients in the United States undergo this non-fasting GDM screening test. And that includes a one-hour 50-gram glucose load. Again, it's not timed. You can do it any time of the day. You don't have to be fasting. And then different hospitals or centers or practices have different cutoffs for what normal GLT is considered. So it can range anywhere, even in the same city, between 130 and 140. And people argue about what the cutoff will be. But if you sort of have a normal test, then you're done. If you have an abnormal screen and your glucose is elevated, again, either greater than 130 or 140, then you would go on to have the real diagnostic test, which is a fasting three-hour oral glucose tolerance test where we give 100 grams of glucose orally. And in order to meet criteria for gestational diabetes, you would need to have two abnormal OGTT values. Now, there is this intermediate category that we and others have described that has been of interest to some people. And we're calling that gestational glucose intolerance. But I just want to emphasize that this is not a recognized disorder in the United States. Most women who are tested are either told you have gestational diabetes or you don't have gestational diabetes. There's not a gray area that I've indicated on this slide. And then the most commonly used cutoffs here for the OGTT are shown here. And these are the Carpenter-Couston criteria. And I just want to mention what these are based on. So these are based on studies conducted in the 1960s in which they looked for glucose levels that were greater than two standard deviations over the mean. And then subsequently, those weren't the Carpenter-Couston criteria. Those are known as the O'Sullivan criteria. But subsequently, the way we measure glucose in blood has changed, so gone from whole blood to plasma. And then modern assays using hexokinase have been implemented. And so these numbers needed to be adjusted to account for the differences in glucose with those measures. And that's how we ended up at Carpenter-Couston. Let's contrast this with what's done in most of the rest of the world, which is now one-step screening and diagnosis for gestational diabetes. So outside of the US, sometimes everyone and sometimes people with selected risk factors undergo a single test for screening and diagnosis. And that is a 75-gram, two-hour oral glucose tolerance test. And on this test, to be diagnosed with gestational diabetes, you just need one abnormal value. So if all your values are normal, then you're normal. If you have one abnormal value, then you would be diagnosed with gestational diabetes. And here are the thresholds that the International Association of Diabetes and Pregnancy Study Groups has set and have been endorsed by the WHO. And they're shown here. So what are these based on? So these are actually based on the HAPO study, which I'm sure you'll hear more about later in this symposium, which was published in the New England Journal now more than a decade ago. And what this group that came together did is decided that we should make the cutoffs for gestational diabetes, the thresholds at which there would be a 1.75 odds ratios in adjusted logistic regression models for the risk of primary adverse outcomes. And I'm just showing you here that this is a bit of a judgment call. Because the main results from the HAPO study showed that the risk of any of these adverse outcomes increased continuously across the whole spectrum of glucose. So here I'm showing you large for gestational age birth weight, or birth weight greater than the 90th percentile. And you can see in these different color lines are the fasting, or the one hour, or the two hour glucose. But it doesn't matter, because you can just see that the risk increases continuously across the whole spectrum. So they decided that they would choose the 1.75 odds ratio, but that was a judgment call. So when we compare these two different criteria, because it's very controversial, what is really the difference? And I just want to point out that they're pretty similar. I think the fasting glucose being 92 in the IADPSG criteria is going to result in more diagnoses on its own. But the main difference is the load of glucose given the length of the oral glucose tolerance test, and then how many values you need to be considered abnormal. So it's really that fact that you only need one value to be abnormal that accounts for the increased rates of gestational diabetes that will be diagnosed if you use this one step method with the IADPSG criteria applied. So there have been two recent randomized control trials conducted in the United States comparing one step testing versus two step testing. And I think it's worth thinking a little bit about those. So the first one to be published was published in the New England Journal of Medicine in March of 2021. And this was a huge trial, a huge undertaking. They included over 23,000 pregnant individuals. This was a pragmatic trial. It was completely unblinded. And in fact, there was a waiver of informed consent because they felt clinical equipoise. So they did this across a whole health system without consenting individuals. So that's one reason why they were able to do such a large trial. So they compared the one step testing to the two step testing. And as expected, they found a much higher gestational diabetes diagnosis rate with one step testing. And I've given the percentages here. It was about double. Their primary outcome, large for gestational age birth weight, was not significantly different. So this is contrary to their hypothesis. They thought one step testing would lower the risk of large for gestational age birth weight across the population because more people would be diagnosed and treated. And that did not pan out. There was actually an increase in neonatal hypoglycemia with the one step test. And I'm sure you'll hear from the next speaker that this was likely due to ascertainment bias. Some limitations. So one, there was a lot of crossover. So you can imagine this was conducted in the United States. People are used to doing two step testing. For one step testing, you have to show up fasting. A lot of people probably didn't show up fasting. So they went ahead and did two step. Or the clinicians were just used to doing two step. So there was a lot of crossover. And then there was a safety parameter where some people, even randomized to two step, would have treatment if they had fasting hyperglycemia. And so that could have caused the groups to be more similar. And some critiques of this study have suggested that it may have actually been underpowered, even though it was a large study to show the difference that they wanted to show. There was another study published later in 2021 that was really quite elegant, I think, that was conducted testing almost the same hypothesis, but had a very different study design. So this was a smaller study, but the study was actually blinded. So everyone in the study went in and got that non-fasting 50 gram glucose load. But they didn't use the information, they kept that secret. And then everybody, they took people, depending on which group they were randomized to, and gave them another, the second test, either the 75 gram or 100 gram. But they didn't tell people what kind of test they did. And they just told the clinicians, like either this person has gestational diabetes or not. So it was really kind of an ingenious way of blinding the study. And then they only sort of in their analysis, they were able to sort of ignore the 50 gram test in the people randomized to one step, if that makes sense. So again, this showed pretty consistent results, increased gestational diabetes diagnosis rate in the one step group. And then because of that, there was a lot more meds and a lot more fetal testing in that group. And again, no significant difference in large for gestational age birth weight. And again, increased neonatal hypoglycemia. One interesting thing about this study that I think doesn't get mentioned that often is that there was a lot less oral glucose tolerance test associated nausea and vomiting in the 75 gram test group, which is interesting. I think it, you know, it's, we need to think about the patient experience of testing as well when we're thinking about what diagnostic criteria to use or what type of testing to do. So I think that was a notable finding. So we wrote an editorial when the second paper came out, me and Ebony Carter from Wash U. And what Dr. Carter and I said is that these randomized control trials when taken together suggest a significant increase in both gestational diabetes incidence and healthcare utilization using the one-step testing method. And there's not a commensurate improvement in adverse pregnancy outcomes as had been hypothesized. So right now, I think the evidence supports continuing to do two-step testing in the United States, but that's very controversial. We did note two things. So one, we think there needs to be an effort to develop novel gestational diabetes screening strategies that decrease the patient burden associated with testing, you know, sort of related to this unpleasant experience and the nausea and vomiting associated with doing these large glucose loads. And secondly, these studies say nothing about long-term outcomes, which are really important. And so we don't know if treating according to the one-step would have an effect on long-term outcomes in either babies or mothers. And that's something that needs to be studied further. So in summary, in the United States, we practice universal two-step testing and most places internationally do one-step screening. The recent US-based randomized control trials do support this practice, but we need to address the patient burden associated with OGTT and long-term outcome data with these two strategies. So where do I think we should move from here? So after we diagnose someone, we gave people with gestational diabetes the same initial treatment plan, and I wanna take you through that so that you understand why I think this may need to change and why we might need to change our diagnostic paradigm. So this is an example of a log from one of my patients. So after we diagnose someone with gestational diabetes, we tell them, okay, now you have gestational diabetes, you have to write down everything you eat. So this patient has diligently logged all of their meals and snacks over a week, and we ask people to do this, it's a lot of work. Then we ask them to count and limit their carbohydrates, but we say, limit your carbohydrates because of the glycemia, but not too much because the baby needs them. So this is very contradictory information that we give to patients and I don't think they like it very much. We ask them, and this is the biggest thing, to limit their blood sugars four or more times a day, depending on what medications they're on. And this is a lot of checking for people with relatively mild hyperglycemia that we, of course, know impacts the fetus and that's why we do it, but it is a large patient burden. And then we ask them to modify their diet to achieve strict postprandial glycemic targets using trial and error, essentially, with some expert nutrition advice, but again, individual responses to foods are variable, so we ask them to do trial and error, and otherwise, our first line of medication is insulin, as was shown in those two randomized trials that I showed you that improved outcomes. And then weeks after delivery, we ask them to come back and do another oral glucose tolerance test to make sure that they don't have continuing diabetes, and you can imagine that after the experience of doing the first oral glucose tolerance test, few people show up for this test. So I hope that I've conveyed to you that I think gestational diabetes treatment as currently practiced is burdensome, crude, and applied as if all gestational diabetes is the same. And I think our diagnostic paradigm needs to change because we know that diabetes is heterogeneous. And so when we first are treating a non-pregnant individual with diabetes, we always ask what kind of diabetes they have, and here, if you don't have features of type 1, if you don't have features of a more rare cause like pancreatic diabetes, monogenic diabetes, we sort of put you in this type 2 bucket, and I would pose the question, is type 2 diabetes one disease, or is it more than one disease? We do the same thing with gestational diabetes, but I think even a little bit more narrow where we say, do you have known type 1? Do you have known type 2? And if not, we put you in this gestational diabetes bucket. So is gestational diabetes one disease? And so I have this vision where at diagnosis, we could be a little bit more precise, and we've been working on physiologic subtyping at diagnosis, and we think that by using this approach, we may be able to provide more tailored perinatal risk assessment and therapy selection, and potentially more tailored type 2 diabetes risk assessment and prevention. And so we've described these physiologic subtypes in diabetes care in 2016. We're at diagnosis based on glucose and insulin levels. You can categorize people as either having more of an insulin-resistant gestational diabetes or more of an insulin-deficient gestational diabetes, and I'm just gonna show you this paradigm here where about half of individuals with gestational diabetes have excess insulin resistance, and the other half either have more less preserved insulin sensitivity, so less insulin resistance, and have lower absolute levels of insulin, and then the other 20% sort of have a mixed picture. And there's a few people that you can't classify under this paradigm, but what we showed here is that those with insulin resistance have larger babies and a greater risk of adverse outcomes. And so if we know this at the time of diagnosis, maybe we can move beyond these simple classifications to better predicting risk and targeting our therapy to the right individuals. Now we define those characteristics based on the one-step testing method because that was an international cohort. But I mentioned this other category of gestational glucose intolerance here that we get with this two-step testing. And one thing we're working on is figuring out if because we're gonna stick with two-step testing most likely in the United States, if there's anything we can do for this intermediate risk group, because we know they do have an increased risk even if in the large randomized trials there wasn't a population benefit of treating individuals with milder glucose intolerance and treating these individuals wasn't directly tested anyway. And so with two of my mentees, Dr. Maya has looked at large for gestational age risk in this intermediate group, finding that if you have one or zero abnormal OGTT values, you have an increased risk of LGA, which has been described before in smaller cohorts. And another endocrinologist, Dr. Daryl Seelan, has looked at maternal diabetes risk in our cohort, finding that even when you're in this intermediate risk category, you have an increased risk of maternal diabetes. And we've taken our physiologic subtyping paradigm to this intermediate group, showing that those with increased insulin resistance have a greater perinatal risk, a greater risk of large for gestational DH birth we, and hypertensive disorders of pregnancy. But it turns out that everyone in this intermediate risk group has an increase in future prediabetes risk, regardless of the subtype. And so I think this is important to know as we're building out this paradigm of more precise and refined gestational diabetes diagnostics. And so I'll leave you with this vision for gestational diabetes, more precise diagnosis, precision medicine, if you will, based on physiologic subtyping, and we have a lot more work to do in this area. So to summarize, in the future, we need to look to new approaches to decrease patient burden of testing, and I didn't have time to get into our work in that area, and that physiologic subtyping may be a way to guide our management in the future. So thank you very much for listening. I think we're gonna do questions at the end. Thank you. Thank you, Dr. Poe. And next, Dr. Sabatama Sen will be speaking to us, and she is a practicing neonatologist, assistant professor of pediatrics at Harvard Medical School, and maternal infant health researcher at the Brigham and Women's Hospital, and she'll be speaking about continuing the conundrum, detecting and treating hypoglycemia in neonates. Thank you. Thank you so much for the introduction, and thank you for inviting me to speak today. I have no disclosures. And the objectives of today's talk are to review risk factors and underlying mechanisms of neonatal hypoglycemia, to discuss the current state of knowledge regarding outcomes, to examine data underlying our operational thresholds, to review common approaches to neonatal hypoglycemia evaluation and management, and throughout this talk, I hope to highlight some gaps in our understanding of neonatal hypoglycemia. So just to provide some historical perspective, until about the mid-1960s, infants had their glucoses checked only if they were symptomatic. And the concerns with this approach were, of course, that we were missing many babies who were asymptomatic, and we now know that most babies who have hypoglycemia, in fact, are asymptomatic, and that the symptoms of hypoglycemia are quite nonspecific, including jitteriness, lethargy, and that there's many diseases that also present with overlapping symptoms. The pendulum then swung the other way in the 60s to the 80s, in which, during which time, universal glucose screening was instituted. And in many institutions, glucose screening requires separation of the mother and infant, so there was concerns about decreased rates of breastfeeding as well as bonding with this paradigm. In addition, there was misclassification of normal transition as hypoglycemia, and there were concerns about this misclassification because of the lack of normative data. And then lastly, of course, the healthcare system burden was a large concern. In our current era, we use risk-factor-based screening. And the concerns regarding this approach are that, of course, in order to identify the risk factors, we require contemporary data on maternal metabolism and hypoglycemia. We may miss infants who have no risk factors, and misclassification is still a concern without outcome-based hypoglycemia cutoffs, and there's still this ongoing concern that we're treating a number without these outcome-based data. So what are the established risk factors for screening infants for hypoglycemia? Here in the US, we use the AAP, the American Academy of Pediatrics guidelines, at least in the first 48 hours of life. And the American Academy of Pediatrics suggests that infants who are late preterm or who are small for gestational age, infants of diabetic mothers, or large for gestational age are screened for hypoglycemia in the first 12 to 24 hours of life. I've also put up the risk categories from other institutions, both nationally and internationally, to show that there is certainly a variety of additional indications that other organizations recommend as risk factors. Overall, in the US, about 30% of neonates, or about a million infants per year, are then screened for neonatal hypoglycemia. What proportion of these infants actually develop hypoglycemia? In this prospective cohort study, about 500 infants were examined to see whether they developed hypoglycemia based on their risk factors. What they found was that about 50% of at-risk infants developed hypoglycemia, and you can see here that they illustrated it by risk factor. Of these, about one in five infants developed recurrent or severe hypoglycemia, and about one in 10 developed prolonged hypoglycemia. The burden in the US is about 500,000 infants per year who develop hypoglycemia, and about 100,000 infants per year who require NICU-level care for the treatment of hypoglycemia. The burden on the system from hypoglycemia requires that we really understand the underlying mechanisms of hypoglycemia and understand the outcomes and repercussions for hypoglycemia. So the underlying physiology, why are these groups at risk? There's three main buckets or categories that predispose infants to developing hypoglycemia. The first is decreased stores of glycogen. Infants who are small for gestational age have decreased fuel supply in utero, and this leads to decreased glycogen stores. When infants are born preterm, their ability to accrue glycogen is truncated because the majority of glycogen is actually accumulated in the third trimester. In addition, infants who are born preterm are at additional risk because the enzyme that's needed to mobilize glycogen is found in lower concentrations in preterm infants than in term infants. The second bucket is increased insulin secretion, and this is by far the largest bucket for us in the US. This includes babies who are small for gestational age and growth-restricted, and this is related to the catecholamine release in utero where infants who are small for gestational age have high catecholamines and low insulin in utero, but there's a compensation that occurs after delivery in which the switches, the catecholamines decrease, and their insulin levels increase after delivery. Maternal beta blockers, which are commonly used for treatment of maternal hypertension, readily cross the placenta and lead to fetal sympathetic blockade, which then increases fetal insulin. There are syndromes that we always have to consider, particularly when we see severe hypoglycemia, syndromes such as hyperinsulinism and Beckwith-Wiedemann that are rare, but certainly lead to increased risk of severe hypoglycemia. And then I'll lastly discuss on the next slide the physiology of infants of diabetic mothers as well as large for gestational age infants and how they are predisposed to hypoglycemia. And then the last bucket is increased insulin sensitivity, and infants who are small for gestational age or born preterm are also at risk for hypoglycemia because of this mechanism. So Dr. Po beautifully talked us through the Peterson hypothesis, but I will additionally explain this in the context of neonatal hypoglycemia. So maternal hyperglycemia or maternal glucose crosses the placenta by facilitated diffusion, which leads to the stimulation of the fetal pancreas to produce insulin. And this fetal hyperinsulinemia in utero is matched by the maternal glucose supply. However, when the cord is clamped and maternal glucose supply to the infant is then truncated, this can result in hypoglycemia in the neonate. In addition, exposure to excess insulin in utero stimulates fetal growth, leading to large for gestational age and macrosomia. Of importance and of note, as Dr. Po alluded to, infants who are born to mothers with a diagnosis of diabetes are screened for hypoglycemia. However, the HAPO study clearly showed that there was a continuous association between maternal glycemia and neonatal outcomes, including hypoglycemia and large for gestational age infants. So many of our studies that have been conducted, the denominator is challenging to interpret because only infants who are born to mothers with a diagnosis of diabetes are screened. And so this ascertainment bias is something that has been challenging for us to disentangle in many of our population-based studies. What about other brain fuels? There's been an ongoing suggestion that healthy infants are able to mobilize alternate brain fuels to support their brain energetic needs during their initial period when they experience hypoglycemia. The GLOW study was a study of 67 term neonates where investigators longitudinally measured lactate and ketones beta-hydroxybutyrate over the first five days of life. What they found is that lactate, which is shown in the top graph, contributed about 25% of energy on the first day of life and then tailed off over the subsequent days. Beta-hydroxybutyrate or ketones did not actually kick in, did not actually start contributing to the infant's total ATP until two to three days of age and then started to tail off by about four days of age. What's interesting is that this study really suggests that the majority of an infant's fuel is obtained through glucose, but also these wide intervals, as you can see here, between, for example, the 25th and 75th percentile also suggest that metabolic heterogeneity in neonates may contribute to the variability we see in neurodevelopmental outcomes in infants exposed to hypoglycemia. So since glucose is really the main fuel source for the neonatal brain, how might hypoglycemia impact the neonatal brain? Initially, there are compensatory mechanisms that are elicited. There's an increase in cerebral blood flow, there's increase in glycogenolysis, and there's increased lactate production. What we're trying to avoid, of course, are the effects of severe and prolonged hypoglycemia, including failure of the sodium-potassium neural pump, accumulation of intracellular sodium and intracellular calcium, which leads to calcium-mediated accumulation of excitatory neurotransmitters and free radical and eventually cell death. So what do we know about the outcomes related to hypoglycemia in children? Historically, severe prolonged hypoglycemia was found to be associated with neurodevelopmental impairment, and there were mixed results for mild or moderate hypoglycemia. However, this data had its limitations. First of all, the population often lacked controls, and many of the studies were done in very specific populations, which were then generalized to the overall population. There was limited maternal data, and I will also underscore that there still is limited maternal data in our investigations of the associations of hypoglycemia with long-term outcomes, and there was also a lack of functional outcome ascertainment. Recent studies have been designed to provide more rigorous and generalizable evidence, and I'm gonna walk through two of the more recent studies that fall into this category. The first is the CHILD Study, which stands for Children with Hypoglycemia and Their Later Development, and this was a study that was conducted in New Zealand in 2006 to 2010. There were 616 infants, so these were all preterm, greater than 32 weeks or term infants, but all of these infants were considered at risk for hypoglycemia by standard criteria. Infants had blood glucoses measured intermittently by heel-stick sampling, and also had continuous glucose monitoring instituted, and the definition of hypoglycemia that was used in this study was less than 47 milligrams per deciliter. Investigators measured neurodevelopmental outcomes at two years, four and a half years, and nine to 10 years. What were their findings? At two years, hypoglycemia was not associated with overall impairment. However, they found that spending more time outside of a central band, which was a statistically calculated central band of 54 to 72 milligrams per deciliter was associated with impairment. Among children who had hypoglycemia, children who developed impairment were more likely to have a steep rise in blood glucose in their first 12 hours of life, and this association was even stronger in children who were treated with dextrose, suggesting that blood glucose lability or variability may contribute to adverse neurodevelopmental outcomes. At four and a half years, once again, hypoglycemia was not associated with overall impairment, but they did see that hypoglycemia was associated with deficits in specific domains. In particular, they found it was associated with low executive function and challenges with visual spatial integration. And prolonged and severe hypoglycemia further increased this risk. Interestingly and surprisingly, at nine years, hypoglycemia was not associated with educational attainment, executive function, or visual spatial impairment, and we'll talk a little bit in the next slide about why this might be. So about 4% of the infants in this cohort were referred for services at four years. So if they were severely affected, they were referred for services. And in this cohort, blood glucoses were maintained to achieve greater than 47 milligrams per deciliter. So both of these aspects may have biased the results towards the null. 18% of children were not assessed, some because of loss to follow-up, but some also because of COVID-related challenges. And perhaps most importantly, rates of low educational achievement, which was defined as two grade levels below the norm, so pretty severe educational achievement criteria, were about 50% whether children were exposed to hypoglycemia or not. So this suggests that perhaps there's contributions to this adverse neurodevelopmental outcome from some of the factors that actually predispose to hypoglycemia. So our traditional paradigm has been examining the associations of hypoglycemia with neurodevelopmental sequelae. And I propose that this is actually looking at this association with one eye open only, like this baby is doing. And in fact, maternal dysmetabolism is thought to impact both hypoglycemia and reprogram the fetal brain. And so we really need to be looking at this association starting in pregnancy, characterizing the maternal metabolic environment, and examine neonatal hypoglycemia as a mediator of the association between maternal metabolism and neurodevelopmental outcomes. The second study that I wanna highlight is a study done by Dr. Kaiser. This was a study that was done in the setting of universal glucose screening. It was done in Arkansas, and the babies were born in 1998 when universal glucose screening was in effect. They categorized their exposure, their glucose levels as less than 35, less than 40, or less than 45. And their outcome was proficiency on fourth grade state math and reading achievement tests that they were able to match through the state hospital records and state educational records. What they found was that at the different cutoffs, that hypoglycemia was associated with lower odds of proficiency in both reading and math. They were able to adjust for many maternal sociodemographic factors. However, they were not able to obtain any information regarding maternal metabolism. So this study also was not able to include some of these preceding factors. So when we pull the data, what we find is that hypoglycemia is associated with early childhood neurodevelopmental impairment, specifically in the domains of visual motor impairment. There's an increased odds of adverse outcome and increased odds of executive function impairment. In mid-childhood, we can see that there is an increased odds of neurodevelopmental impairment and low language, literacy, and numeracy. So just to summarize the seminal study findings, hypoglycemia is associated with deficits in executive and visuospatial function in early childhood, which may contribute to lower school performance. The role of rapid rise in blood glucose after hypoglycemia in outcomes requires more investigation, but judicious treatment of hypoglycemia is warranted based on the current evidence. And the contribution of maternal metabolism and fetal infant metabolic demand have not been adequately considered and require rigorous future investigation. So moving on to talk a little bit about how we approach hypoglycemia evaluation and management. As I mentioned before, we evaluate infants who are considered at risk for hypoglycemia. We initiate preventive measures of birth, including skin-to-skin and early feeding, and then we screen blood glucose starting at 60 to 90 minutes of life for between 12 and 24 hours, depending on the risk factor. Most commonly, point-of-care glucometers are used, and these are the same point-of-care glucometers that are used in adults. So one of the challenges that we have in neonatal hypoglycemia is that the FDA allows a 12% error in these devices because they're mainly created for adult diabetics with hyperglycemia, but of course, at the lower end, when we're treating babies with blood glucoses of 30 to 50, this is a major challenge in our approach. And then we treat infants based on hour-of-life norms. So this is a pictorial of how the hour-of-life norms differ by different committees. As I mentioned before, we often use the AAP guidelines in the first 48 hours, and then we use the PES guidelines at many centers, which is the bottom line in green after 48 hours. A recent study, the GLO study, also quantified how the variability in these guidelines impacts classification of infants for hypoglycemia, and they found that, for example, in the four- to 24-hour period, an infant who was classified using the AAP, there would be a 3% risk of being classified as hypoglycemic using the AAP guidelines, but a 40% risk classified using the PES guidelines. So why this variability in norms? Well, I think many of the studies are based on very small physiologic assessments. In addition, as I mentioned previously, some of the populations from which this has been extrapolated are very specific, such as preterm infants. And so there's still work to be done in terms of determining what these norms are, specifically in the context of outcome-based studies. And in particular, the way to answer this question is to really try and understand at what threshold does treatment of hypoglycemia improve outcomes. So really using outcome-based guides to determine the thresholds for hypoglycemia evaluation and treatment. And the hypo exit trial tried to answer this question. This was a study of 689 infants at 17 European centers. Infants were greater than 35 weeks with an established risk factor, but importantly, these were all babies with moderate to mild hypoglycemia. They had to have an initial blood glucose of greater than 35 milligrams per deciliter. So the findings really don't apply to infants with severe hypoglycemia. They randomized infants to treatment at either 47 or 36 milligrams per deciliter. And then each infant was treated based on standard of care at their institution. So there was some heterogeneity in terms of the treatment approach. And then their outcome was Bayley or overall neurodevelopmental outcome at 18 months. What they found was that treatment at 36 was non-inferior to treatment at 47 milligrams per deciliter. So as you can see here, the 95% confidence intervals crossed zero for the treatment groups in terms of stratified by all of the different risk factors, preterm, small for gestational age, large for gestational age and infants of diabetic mothers. So really across risk factors, it seems that treatment at 36 was non-inferior to treatment at 47. However, the study had some limitations. The first is that the outcomes were measured at 18 months. And as we've discussed based on the previous data, many of the outcomes that are associated with hypoglycemia really can't be assessed until school age. So four to five years old at least. And so it's been difficult to move this study into practical application because of that primary limitation that they weren't able to really evaluate the domains that are thought to be affected by hypoglycemia. In addition, the infants of diabetic mother sample size was not achieved, which is our largest population. And there was a 16% loss to follow up. So how do we treat infants when they're diagnosed with hypoglycemia? So our first line of treatment is enteral measures, which includes dextrose gel and feeding. So dextrose gel was implemented based on a randomized control trial of dextrose gel versus placebo gel in neonates who developed hypoglycemia. And in this trial, what they found was that dextrose gel decreased treatment failure. It decreased the need for NICU admission for IV dextrose. So we at the Brigham implemented dextrose gel about a year after these findings and found that the proportion of infants who received IV dextrose, which is represented by the red line for hypoglycemia, decreased after implementing gel in our NICU. And the implementation time point is demarcated with the dotted line. In addition, there was minimal data on how to feed these babies. So what kind of supplementation to provide or if we should be providing supplementation at all. And so we conducted a study where we examined pre and post blood glucoses by feeding type in about 100 infants. Here you can see that in the dark gray bars are pre-gel blood glucose and in the light gray bars are post-gel blood glucose. And it's stratified by different feeding types. And what we found was that supplementing breastfeeding with formula or donor milk increased blood glucose more than breastfeeding alone and decreased the need for repeat gel and IV dextrose. So as you can see in the graph, infants who were supplemented, their breastfeeding was supplemented with formula or donor milk had a slightly higher increase in their blood glucose post-intervention than those who were just breastfed alone. And this decreased the need for recurrent treatment for their hypoglycemia. In addition, we found that donor milk was similar to formula in preterm infants in terms of effectiveness for decreasing the need for later treatment, but this was not true in term infants. In term infants, formula seemed to be, seemed to lead to a decreased need for repeat treatment, but donor milk didn't. And then lastly, infants who fail enteral measures require IV dextrose. And the approach in the past was that all infants who failed enteral measures would receive a bolus of IV dextrose and would be started on the same rate of dextrose-containing fluids. However, given the concern about blood glucose variability and the impact on neurodevelopmental outcomes, we instituted a protocol where a graded approach to IV fluids was instituted. And as you can see here, we found that after implementing this protocol in the post-period, the variability or flux was lower in babies' blood glucoses. And this was associated with a decreased length of NICU stay in the top line graph and a commensurate decrease in NICU cost of care in the bottom line graph. So in summary, neonatal hypoglycemia is common, and it may be more common than previously estimated. Established risk factors should be reexamined in our contemporary context. Neonatal hypoglycemia, particularly when severe, prolonged, or recurrent, is associated with specific neurodevelopmental deficits, and blood glucose lability may also play a role. Treatment thresholds vary widely, and integration of risk factors and functional neurodevelopmental outcomes should drive determination of thresholds. And trials with long-term outcomes should inform clinical care. And just to highlight our gaps, well-powered, unbiased, diverse contemporary studies are needed to determine more specific hypoglycemia risk factors and improve risk stratification paradigms. So factors such as trajectories of maternal glycemia, feeding, fasting versus postprandial glycemia, and the role of other maternal metabolic factors should also play into our risk stratification paradigms. Risk factors and maternal metabolic factors need to be considered in the association between hypoglycemia and long-term child outcome. And neonatal tools to improve evaluation and approach to hypoglycemia, integrating glucose, and other metabolic factors are urgently needed. So I want to thank my collaborators. Thank you, and thank you for your attention. All right, thank you, Dr. Sen, that was fabulous. Our last talk of the session before our question and answer is Dr. Jamie Jofson, who is a pediatric endocrinologist and physician scientist at the Ann and Robert Lurie Children's Hospital of Chicago, an associate professor of pediatrics at Northwestern. She will be presenting to us a talk entitled The Lasting Impact of a Sweet Start to Life, Impact of GDM on Offspring. Hi, good morning. Thank you to the organizers for inviting me to speak today. I have no disclosures. So here's an overview of the adverse outcomes I'm gonna review today. And a very important caveat when we look at data on maternal GDM exposure and offspring outcomes are that many studies do not distinguish the maternal diabetes type or treatment. And for the sake of time today, I'm gonna focus on the metabolic outcomes. And while there's emerging data on other outcomes that's becoming available. So before we dive into some data, I'm gonna provide just a little bit of a personal perspective. So this is a photo of my dad and his youngest brother back in the 70s at my parents' wedding. And around the time I started my research career into the developmental origins of health and disease, my uncle was diagnosed with type two diabetes. And him being a pharmacist and me being an endocrinologist, we spent a lot of time talking about diabetes medications. But now I'm fairly convinced that my grandmother had gestational diabetes when she was pregnant with my uncle. She was older at the time of her pregnancy. It was her fourth pregnancy. And at the time she was pregnant, there was no universal GDM screening. My uncle was born large for gestational age. He was over nine pounds. And my grandmother went on to develop type two diabetes herself around 10 years after her pregnancy with my uncle. So this is an N of one. It's not gonna pass peer review. Let's talk about some other studies that looked at siblings exposed and unexposed to diabetes in pregnancy. So this is data from the Pima Indian population by Donna D'Abelia. And the black bars are children that were born to mothers with diabetes in pregnancy, with the white bars being children, the older children who were born prior to their mother developed diabetes in pregnancy. And if you notice, it's around the age of 12 to 15, so in the peripubertal years that children had an increase, we see increased rates of BMI and obesity. And data such as these sibling studies has also been evaluated in a very large cohort in Sweden. So men that were undergoing conscription had an examination BMI. And the exposure was maternal diabetes in pregnancy. And in a subgroup of this population, they also had data on maternal early pregnancy BMI. At the time the data was collected, majority of the women were lean. And again, a very large study. And what was found was that a BMI difference of nearly 1.0 kilogram per meter square higher in the men born to mothers with diabetes in pregnancy compared to their older brothers whose mothers did not have diabetes when they were pregnant. And early maternal BMI did not alter this association. So the authors concluded there was an intercausal relationship between maternal GDM and offspring obesity. So GDM rates are increasing. And in this data in the United States it's from 2011 until 2019. And it's divided by various racial ethnic groups. But in all groups, rates of GDM are increasing. And these authors further looked at the data based on Asian subgroups and Hispanic Latina subgroups. So in the field of developmental programming, I studied the maternal fuels and how it affects the fetus during critical periods of development. And I'm particularly interested in the metabolic diseases of these children. And we need to talk about plasticity, which is developmental adaptation. And in this data, in this figure from Norbert Frankel who gave his famous Banting lecture at the ADA meeting in 1980, where he expanded upon the Peterson hypothesis. And so we have maternal fuels, the glucose, lipids, amino acids that cross the placenta. As you've seen in our prior speakers today have talked about this pathophysiology. And when it crosses the placenta, the baby, the fetus responds by increasing its insulin production. And insulin in utero is a growth factor. And so this is the mechanism we think behind the pathophysiology of why these babies can be born large for gestational age. And at the Northwestern Diabetes and Pregnancy Program Studies, they evaluated children longitudinally and they identified children to develop increased rates of obesity in the peripueral years and increased adiposity as measured by anthropometrics. And some of these children went on to develop impaired glucose tolerance and type two diabetes. This is data from the Kaiser Health population. And what's interesting about this data is these were five to seven year old children. They were looking at overweight and obese outcomes in the children at five to seven years of age. And these children were born to mothers who all had an abnormal glucose challenge test. So Camille talked all about the glucose challenge test so Camille talked all about the types of screening that's done. And so the women that went on to have a full OGTT who had a normal fasting glucose, the children didn't look much different from the women who had a totally normal OGTT. But it was the fasting glucose when that was elevated above 95 that we see increased rates of overweight and obesity in the children. And so this identifies fasting glucose as important in hyperglycemia in pregnancy. And this is a meta-analysis here evaluating risk of overweight among 34 different cohort studies in nearly 460,000 pregnancies. And what I like about this meta-analysis is that the studies are grouped by when the child BMI was studied. And so at the young age, so when BMI is measured before five years of age, the increased rate of obesity, or overweight, excuse me, is not very much increased. However, between the ages of five to 10, this does increase. And after age 10 and into adulthood, there's a two-fold increased risk of overweight in these offspring born to mothers with gestational diabetes. And the pooled analysis from all of these studies was 1.45. So we're gonna talk a little bit about the HAPO follow-up study. This study's very near and dear to my heart because I'm one of the HAPO follow-up, I was one of the HAPO follow-up investigators. And in the original HAPO study, which it's really important to recognize that women with ovarian diabetes from HAPO were excluded from the study. So these were all healthy pregnant women who happened to undergo a 75-gram fasting OGTT. And both the mothers and their providers were blinded to what their glucose levels were. And the newborn outcomes and the pregnancy outcomes were related to gestational diabetes very much across the continuum of maternal hyperglycemia. So in HAPO follow-up, we studied nearly 4,500 children, and this was a multi-ethnic, multi-racial cohort. And again, we were asking the question, does hyperglycemia in pregnancy less severe than ovarian diabetes? Is it independently associated with increased risk of metabolic outcomes in the children 10 to 14 years later? And so in this data here, we're looking at additive associations of maternal BMI category and GDM with childhood overweight and obesity, and also separately with obesity. And the rates of overweight obesity in the HAPO cohort were about 30%. The rate of obese BMI was 11%. And the reference group, as you can see here, is normal weight BMI women without gestational diabetes. And when these women have gestational diabetes, the odds ratio is 1.5 for a child having overweight BMI. And the highest risk category are the children who are born to mothers with both an obese BMI and gestational diabetes. And that risk is nearly an odds ratio of seven for an obese BMI in those children. So we're gonna talk a little bit about adiposity outcomes. And adiposity is considered to be, in many research opinions of many researchers, kind of a better measure of risk of metabolic disease compared to just the crude measure of BMI. And so body composition using a variety of different types of methods has been demonstrated to be associated between maternal GDM and increased offspring adiposity. So anthropometry is using calipers and skin folds and, say, waist circumference. Bioimpedance is another method. In some studies using DEXA scans, there appears to be about a 3% difference between offspring of mothers with GDM compared to not with GDM. And using MRI to evaluate different fat compartments, there seems to be a more centralized fat distribution in the children born to mothers with GDM. In many studies, maternal BMI does attenuate the results. And some studies have reported sex-specific findings. Interestingly, some studies report adiposity differences yet absent differences in child BMI, which kind of goes back to our thought that BMI is really just a crude measure of adiposity. So again, this is data from the HAPO follow-up study where we measure total body fat with the BODPOD, or air displacement plethysmography. And in the panel on the left, using the reference category of a mother with a normal BMI, and if you look, if you look over here, so these are children born to obese mothers. So on average, these children had 9.3 grams, 9.3 kilograms higher in total weight, and 6% of that was increased percent body fat. And then when we look at the panel on the right of GDM, that confers an increased mean of one kilogram higher in total weight, and 1% higher in body fat. So what this data is really telling us is that maternal BMI definitely has a very large impact on child adiposity, and GDM does somewhat increase that risk. And so then we also wanted to do, we did a mediation analysis in this population because we wanted to ask the question, does newborn adiposity mediate the relationship between maternal predictors, GDM, and BMI, and childhood adiposity? And what I find really interesting about these findings was that maternal glucose had a 15% mediation in the pathway from neonatal adiposity to child adiposity, whereas maternal BMI had just a 3% mediation. And what we can summarize from this finding is that perhaps improving maternal glucose in pregnancy can have an impact on reducing newborn adiposity and child adiposity. And some potential mechanisms. So there's emerging data on genetics and epigenetics, and this was a study from Denmark where young adults underwent a fat biopsy. They had subcutaneous adipose tissue in which methylation studies were performed, and they looked at gene expression in the adipokines in offspring of GDM mothers, offspring of type 1, and then offspring of type 2. And then offspring of the background population. And there were some differences in these gene expressions. So this is a potential mechanism, but I think a lot more data to come in the next few years with many active studies trying to identify potential mechanisms. So let's move on and talk about disorders of glucose metabolism. So this is data from the SEARCH type 2 diabetes in youth study, which looked at the percentage of youth that developed type 2 diabetes. And there was a much higher percentage of youth who had in utero exposure to maternal diabetes. And this is data from Yale, which evaluated obese adolescents. And it was a longitudinal study where these adolescents underwent frequently sampled OGTTs, and they were studying the oral disposition index, which is a measure of beta cell response in the setting of insulin resistance. And what these researchers found was that offspring of mothers with GDM had an odds ratio of greater than five of developing impaired glucose tolerance over a three-year timeframe. And so this decrease in insulin sensitivity with an inability to increase insulin secretion is what led to their reduced oral disposition index. And DI is a really important measure of progression to type 2 diabetes. Not all studies have reported positive findings. So in the Project VIVA study, they did look at this question, and they reported negative findings of GDM and risk of impaired glucose metabolism. There was a large meta-analysis, which provided some evidence, but judged a lot of the evidence to be of low quality. And some investigators cite evidence that associations are due to confounding variables, such as parental obesity. And there's many other potential confounding variables. So the GDM screening type within these studies, whether it was a fasting OGTT or a glucose challenge test, the GDM severity and the treatment that the mother was provided may confound the findings. The timing of offspring follow-up, that's also an important time point to evaluate. So if we look at HAPO data and disorders of glucose metabolism, so all the children in the HAPO follow-up study underwent a 75-gram, two-hour fasting OGTT. And so from that data, we were able to evaluate a number of outcomes of disorders of glucose metabolism. And if you recall, none of these women had gestational diabetes. So on the x-axis is the maternal glucose category, and this hopefully looks similar to some data that Camille had shown earlier on HAPO, the original HAPO study in newborn outcomes. So category five represents the most severe maternal glucose category. And as you can see here, this is the frequency of child impaired fasting glucose. So as maternal glucose category increases, the frequency of child IFG goes up. And similarly, the frequency of impaired glucose tolerance in these children really doesn't start to increase until the highest category of maternal hyperglycemia in pregnancy. And in these graphs here, you can see that both child Mitsuda index and disposition index go down as the maternal category of hyperglycemia increases. So we're gonna move on to talk about hypertension outcomes. There've been some studies in the Pima Indian population which have shown elevated systolic blood pressure among offspring exposed to diabetes in pregnancy, and that finding was independent of adiposity. And higher blood pressure has been demonstrated in a Hong Kong study. And this is data from a study in Portugal which was a prospective birth cohort. And GDM was associated with a higher systolic blood pressure although it was mediated by the child's BMI. There have been a number of conflicting results. So in Project Viva, higher systolic blood pressure at age three was attributed to adiposity. A study in Germany did not identify blood pressure differences. And in a meta-analysis, there was a mean difference in systolic blood pressure in the pooled analysis down here. Although it's really just slightly above the zero line. We'll look at some evidence of dyslipidemia in children exposed to gestational diabetes. So in the EPA cohort, GDM was associated with higher total cholesterol and LDL cholesterol but only in girls. And there were a few finished studies which found lower HDL among offspring born to mothers with gestational diabetes. In the maternal fetal medicine networks studies, there were no differences in triglycerides among the offspring. But these offspring were really just studied till a mean age of seven years. In HAPO data, we found that maternal glucose was associated with a slightly elevated risk of total cholesterol in the children. In a large study in China, they did find that GDM had two times the odds ratio of low HDL and 1.36 odds ratio of high triglycerides among offspring born to mothers with gestational diabetes. There's been just some data in NAFLD among offspring of mothers with gestational diabetes. So the ALSPAC cohort, which was a large cohort in the UK, they assessed fatty liver in adolescents at the ages of 17 to 18. Now the caveat being here is that they didn't really have blood glucose levels to measure, to assess GDM. They identified women as having GDM if they had glycosylurea, so the bad glucose in the urine. But they did find that offspring of mothers with what they termed GDM had an increased risk of fatty liver. Now in the EPOC cohort, they found different results. So they used MRI to measure hepatic fat fraction. And they did find an association between maternal obesity and increased offspring hepatic fat fraction, but not when they isolated out maternal GDM as the exposure. There have been a few studies that have evaluated whether GDM exposure increases risk of precocious puberty. So we do know that population-wide we are seeing an earlier age of pubertal onset, especially in girls. And the EPOC study in Colorado reported a three-month earlier onset of puberty and a faster speed of pubertal growth among offspring exposed to diabetes in utero. And the increased BMI in early childhood mediated the association with early pubertal timing. So increased adiposity and BMI in childhood are well-described associations with earlier pubertal onset. This was data from Denmark. And they did see differences in offspring pubertal status, but it just, once they adjusted for BMI of the child, these differences were no longer statistically significant. And so what's the mechanism going on? Is it the in utero environment that's impacting precocious puberty or early pubertal onset? The postnatal environment? Is it a combination? So let's briefly talk about does treatment of mild GDM reduce child obesity risk? So there have been two large, very similar randomized control trials of an intervention to treat GDM. So this was a trial in Australia. And the trial did reduce macrosomia. Now, just a portion of the offspring were followed up at age four to five. And they did not see a difference in BMI Z-score between the two groups. And that data's not overly surprising to me because I do think that these children were too young to necessarily see a difference in BMI. The maternal fetal networks, so similarly looked at after their randomized trial during pregnancy, they did a follow-up both around four to five, and again, between five to 10. And they just had participation of about 50%. And similarly, their BMI Z-score did not differ between the two groups. Now, I showed you some data earlier that they did find some adiposity differences when they looked at the children between five and 10 years. So let's review. I would say that there's good evidence that gestational diabetes in pregnancy increases childhood offspring risk of obesity, adiposity, and disorders of glucose metabolism. I would say the evidence is moderate for some of these other outcomes. I am very concerned about diabetes, Baguette's diabetes. So mothers who have gestational diabetes and their children are at risk for youth-onset diabetes, and then this can lead to the vicious cycle where these young individuals go on to become pregnant themselves and enter their pregnancy with a diagnosis of type 2 diabetes. What alleviates childhood obesity risk? Well, these are some potential factors. And specifically, breastfeeding and postnatal diet have been evaluated. So the SWIFT study is a study led by Erica Gunderson in Kaiser, Northern California, where they evaluated fetal and early postnatal behaviors associated with infant growth and toddler obesity risk. Now, they did not have a control group. These were all offspring of mothers with gestational diabetes. And what they found was that reduced breastfeeding and increased intake of sugar-sweetened drinks and 100% fruit juice increased the odds of childhood obesity at two to five years. Now, we have some just very recent data that we presented at the ADA meeting last week, and this is not published data, where we identified that long breastfeeding durations, so for at least six months, seem to be protective against obesity risk. So I think that this is some exciting data because at this time, we really do not know of interventions to reduce risk of obesity in offspring of mothers with gestational diabetes. And some breastfeeding data has also been evaluated in a cohort in Australia, and they also identified it to be protective. So in summary, the lasting impact of a sweet start to life impact on offspring of mothers with gestational diabetes. There's good evidence that obesity, adiposity, insulin resistance, reduced beta cell function are adverse outcomes. And some opportunities for intervention, so breastfeeding, diet, avoiding intake of sugar-sweetened drinks, and lifestyle, of course. And we really do need more perspective studies. So I just wanna identify my funding and acknowledge my mentors, collaborators, and my colleagues, both in research and clinical pediatric endocrinology at Lurie Children's Hospital. Thank you. Thank you so much. So now we'll invite all of the speakers up here to the table and do a question and answer session. Let's go ahead and get started with you at the microphone down there. Hi, Suchitra Nenshala from Northern Virginia. I have a question for the last speaker. I'm sorry, I didn't get your name. So thank you for the data you presented. As an adult endocrinologist, it helps me counsel moms during pregnancy. So my first question is just to clarify something. You showed one slide about childhood obesity and the effect of GDM, where there was a 9.3 kilogram weight increase with maternal obesity, and then just one kilogram weight increase beyond that with GDM. And then the subsequent slide said 15% increase in risk with GDM versus 3% with maternal obesity. So I'm a little confused. Oh, sure. So the first slide was really just looking at the haplophala population, so the children in general, when we looked at body composition differences. And so we really do know that maternal obesity probably has the largest impact on rates of child obesity. But the point of the next slide in showing the data, that was a mediation analysis, so that was looking at the pathway from maternal risk factors, so both BMI and glucose tolerance, through newborn adiposity, and then how, in that pathway from newborn adiposity to child adiposity. And so what the point of that slide was saying is that a lot of newborn adiposity can be attributed to maternal glucose in pregnancy, and that's where there's an opportunity for intervention. So really, it's reducing maternal glucose, normalizing maternal glucose in pregnancy, reducing newborn adiposity, and therefore reducing child adiposity. So the reason why I would say that mediation analysis was showing such a small amount of maternal BMI is because we know so much of the relationship between maternal BMI and child BMI is a lot of genetics, a lot of the home environment as well, maybe not as much the inter-unit environment. Okay, and your definition of hyperglycemia versus OBERT-GDM, is that just one abnormal reading on the GTD that you're classifying as hyperglycemia? So the criteria that were used for the original HAPO studies, this was back in 1999 to 2005 when HAPO was conducted, was the traditional, at that time, it was using the OGT levels that, at that time, were diagnostic of gestational diabetes. So they were somewhat higher than they are now on the current WHO and IADPSG criteria for OGTTs. Has anyone looked at just one abnormal reading? Because when we- That's really, Camille's research is looking at one abnormal reading, and so if you have questions about that, please go ahead and ask Dr. Paul. Okay, yeah, I'm sorry I missed your lecture. So just one abnormal glucose reading and the effect on children, because we know that macrosomia in the infant, even with one abnormal reading, you can see that on the fetus, but on the child, just with one abnormal reading, do you see all this morbidity? Yeah, that's a good question. So in our data, in our large US-based cohort, or Massachusetts-based cohort, we have seen an increased risk of large for gestational age, as you mentioned. We have not yet looked at the childhood outcomes, but I'll say, in a review of the literature, there's mixed data on that in smaller studies, but I saw some data presented last week at ADA that there was an increased risk in some of these intermediate group. Glucose tolerance categories and future risk of childhood increased body weight. Okay, can I ask one last sub-question, or? We actually have a bunch of questions in the chat, and there's a line. So hopefully we can come back to it and finish it up. Next question. Thank you for that excellent talk. I'm Afreen Sharif, Duke Endocrinology. I practice adult oncoendocrine, but this is a topic personal to me, so I have a question for Dr. Paul, especially with the sub-phenotypic subtype of insulin deficiency going in. So a couple of questions, one leading into the other. The first one is, what timeline do you recommend that this testing be done? For example, pre-conceptional, post-conceptional, postpartum? And if you identify insulin deficiency with a stimulated C-peptide that's low and an insulin level that's low, how do you recommend follow-up for those mothers? Yeah, so this is a great question, and we're still doing a lot of work on our gestational diabetes physiologic subtypes. So the data that I presented was doing subtyping at the diagnosis of gestational diabetes. So these were cross-sectional studies, at least in our original publication in diabetes care. We looked at insulin levels during the oral glucose tolerance test that was used to diagnose diabetes. And so we really haven't done the full longitudinal look at how these gestational diabetes physiologic subtypes change over time, although we do have a longitudinal cohort and we're analyzing that data now. So for now, while this type of physiologic subtyping hasn't been brought, it's not ready to be brought clinically yet because we don't fully understand all of the implications, the data we have right now is based on at the time of gestational diabetes diagnosis or a little before in some of the other cohorts at 16 to 20 weeks gestation. As far as follow-up, our data and both the data from Dr. Ratnakaran from Toronto show that regardless of which type of physiologic subtype of gestational diabetes that you have, there does seem to be an increased risk of future glycemia. And so it could be that different ways of being hyperglycemic or different physiology underlying that hyperglycemia is also present sort of later in life. So it doesn't seem to be distinct, while it's distinguishing the adverse perinatal outcomes with the more insulin resistant individuals having worse perinatal outcomes maybe because of other fuels other than glucose affecting the fetus, we don't see the same differences in future glucose intolerance risk among those individuals. Thank you so much. Great talks, Rami Alamzadeh, Pediatric Endocrinologist at the University of Tennessee. Dr. Poe, great talk. My question is that as we know, the rate of type one is increasing, over 50% are diagnosed after age 21. How come we are not doing autoantibodies? And I have two other questions, but we should be including islet antibodies once we have this glycemia. Yeah, great point. So when we do this universal screening during pregnancy, as I mentioned, we're just looking for hyperglycemia. A lot of times, I would say 99% of the time, the cause of the hyperglycemia is not looked into, like it's sort of put in this bucket of gestational. And within gestational, all of the causes of hyperglycemia in people of childbearing age are actually present. So there's a lot of, I like that you're emphasizing that there's a lot of sort of misdiagnosis of other types of diabetes as gestational. And so we can pick up type one on that screening, we can pick up monogenic diabetes on that screening. And so I think the key is after the pregnancy ends, because a lot of times, because hyperglycemia is such a problem for the fetus, we end up treating the hyperglycemia similarly, no matter sort of what the underlying disease leading to it is. You know, the key is doing the postpartum testing to see if the hyperglycemia has resolved or not. And if there is persistent, I totally agree with you, if there is persistent hyperglycemia after delivery, auto antibodies can be checked in the right clinical scenario, genetic testing for MODY can be pursued in the right clinical scenario, et cetera. So your point, I think, is very well taken. As far as checking auto antibodies during pregnancy, I would caution against doing that in terms of potential false reassurance, because we don't have good longitudinal studies of type one diabetes associated auto antibodies across gestation. And there's some indication that pregnancy may decrease the levels of some auto antibodies. So I wouldn't wanna miss type one, but certainly like if they're present, then that would be, I agree with you, that that would be diagnostic for type one diabetes. And then briefly, I know we have a long, Dr. Sen mentioned about hyperglycemia. I mean, we see that often consulted on neonatal hyperglycemia. They're still using alcohol on heel sticks, which really over-diagnose hyperglycemia because alcohol interferes. In our older kids, we get them to wash their hands with soap and water. We don't use alcohol and falsely lowers high glucose reading on a meter. I don't know. Unless we can do a continuous glucose monitoring. Yeah, we definitely are interested in investigating other avenues by which to more accurately measure newborn glucose. I think our handheld glucometers are fraught and we're kind of extrapolating them from the adult population. We certainly have best practices to try and get as accurate a reading of newborn heel sampling blood glucose as we can, but you're correct in that there's a lot of challenges in applying those point of care whole blood glucometers to neonates and it's an area, I think, that we would love to kind of see some innovation. Great. And Dr. Olson, the other question is that we treat BMI equally among all ethnic groups. And a big UK study several years ago showed the threshold for type 2 diabetes. It differs for Caucasians versus Afro-Caribbeans and Southeast Asians. We're still treating BMF, looking at the same, you know. So actually, and maybe this is what you're alluding to in your question, the data from HAPO follow-up used what's called the IOTF, the International, like the specific BMI cutoffs based on, like for South Asian subgroups. So those BMI thresholds were lower to identify the different categories of overweight and obesity. And we see quite a difference also. I mean, a study with First Nation study in Canada showed a tenfold higher risk of subsequent development of type 2 diabetes in the offspring of women with GDM. So there is definitely a different epigenetic effect. Yeah, absolutely. Thank you all for attending the session. We are out of time, but I'm sure the speakers would be happy to, excuse me, answer more questions up at the front. And we also have some additional questions in the Q&A that some of you may have asked that they can discuss with you. Thank you.

gestational diabetes mellitus

offspring

diagnosis

personalized treatment strategies

physiological subtypes

neonatal hypoglycemia

underlying mechanisms

neurodevelopmental outcomes

understanding

management

long-term follow-up