Diabetes, Polycystic Ovary Syndrome, Insufficient Glandular Tissue, and lactation. Dr Kate Rassie video 8_10_24

"PCOS, Diabetes and Breastfeeding" is an excellent presentation by Dr Kate Rassie, delivered to The NDC Institute on 8 October 2024. Dr Rassie is an

• Endocrinologist, Monash Health and Jean Hailes for Women's Health, Melbourne

• PhD candidate, Monash Centre for Health Research and Implementation, Monash University, Melbourne.

Here are the topics covered and how to find them.

Endocrinology + lactation topic	Slide number (pdf of presentation )	Time stamp (video of presentation)
Gestational Diabetes Mellitis (GDM): proposed maternal metabolic benefits of breastfeeding	5-8	06:33mins
GDM + lactation: possible postpartum difficulties, interventions, mechanisms (reset vs preset hypotheses)	9-20	10:49mins
Type 2 Diabetes Mellitus (T2DM) + lactation, delayed onset of lactogenesis II, interventions	21-31	22:46mins
Antenatal expression and the DAME study	33-37	34:57mins
Polycystic Ovary Syndrome (PCOS) + lactation, possible mechanisms	38-43	45:48mins
PCOS + Insufficient Glandular Tissue	44-46	55:43mins
PCOS + lactation: metformin	47-50	59:51mins
Type 1 Diabetes Mellitis (T1DM) + lactation: key management considerations	51-57	1:03:42mins

Transcript

Pam: So it's lovely to see that you've all made it. So it is seven o'clock here in Brisbane. So let's get underway. And I would like to start by acknowledging the traditional custodians of the lands upon which I live and work, the Yugurta and Tubal peoples, and pay my respect to elders past, present, and emerging. And it's my great pleasure to be able to introduce you tonight to Dr. Kate Rassi, who's giving our inaugural talk, really, as part of the NDC Institute's Guest Speaker Series over the next 12 months. And I'll let Kate introduce herself. And Kate has a background in diabetes, polycystic ovary syndrome, and breastfeeding. But as an endocrinologist, she's very kindly agreed to join me for the talk. I'll hand over to you now, Kate, and just so delighted to have you with us and speaking to us.

Kate: Thanks so much for tuning in. As Pam said, this is the first of three lectures that I'm going to do for the NDC Institute over the next few months, and all of them are going to look at endocrine and metabolic aspects of breastfeeding. So as Pam said, I am an endocrinologist. I've got two key areas of clinical interest, really, one of which is diabetes, metabolism, and obesity, and the other of which is women's health and hormones. And I work across those areas in both public and private practice. So in my public role at Monash Health, I do a lot of diabetes in pregnancy and general diabetes, and I also do quite a lot of polycystic ovary syndrome. And then in my private role at Jean Hales here in Melbourne, I do a lot of PCORs, a lot of menopause, a lot of weight management and menstrual mood disorders, so PMS, PMDD. And my PhD is a little bit intersectional between those two areas. So my PhD examines the links between gestational metabolic disease and lactation. And I guess that's what is highly relevant to what we're going to be talking about tonight.

So no conflicts of interest in relation to tonight's material. And I'm gonna start by asking you just to look at this Renaissance painting to get us started. So this is attributable to a French artist called Fragonard, and he was painting in the 1770s. So if you were an affluent European woman at this time, like this one here, you wouldn't breastfeed for very long, if at all, and instead you'd pay to send your infant away to the countryside for one to two years where he or she will be looked after by a dedicated wet nurse. So this painting is believed to be a depiction of one such aristocratic mother taking her infant or visiting her infant who was being looked after rurally in the setting. So this practice had its advantages. The children, of course, got a country upbringing away from the infectious diseases that were rife in the cities at those times. But it also allowed this paid wet nurse to stay at home with her own family whilst generating an income. And for her, that process of extended lactation over time helped her to space out her own children. Of course, this was an era before contraception.

But I think one thing we can be fairly sure about is that neither of the women in this picture would have been concerned about their metabolic risk. We now, though, live in a pretty different world. So we've still got lots of socioeconomic influences on the demographics of breastfeeding, but we've definitely moved into an era in which the main threat to many of our, or the health of us as women, really is that of cardiometabolic disease. So tonight, I'm going to cover off a number of key points, and I'm going to do it really via a series of case studies because I think that helps to bring it to life. So I'm going to talk about the proposed maternal metabolic benefits of breastfeeding, particularly for women who might be emerging from a gestational diabetes pregnancy. I'm then going to talk about some of the caveats we need to have when we look at that research. I'm going to speak about the difficulties that women with gestational diabetes and type 2 face in the postpartum period, and some of the interventions that might prepare them to lactate. I'm going to address a question around whether diabetic milk is indeed full of sugar, and whether insulin enters breast milk. I'm going to think a little bit about the links between PCOS and breastfeeding, and then I'm briefly going to talk about some of the key considerations for type 1 diabetics when they enter lactation. So quite a lot here, but let's step through it via a series of case studies.

Maternal metabolic benefits of breastfeeding & diabetes [00:06:15.29]

So we'll meet our first woman. This is Anusha, and Anusha's 34 years old. She's in her first pregnancy, and she's 34 weeks old, 34 weeks pregnant, and she's been diagnosed with gestational diabetes based on an OGT at 26 weeks. So here in Australia, you'll all be aware, all of our women undergo a full 75-gram oral glucose tolerance test with a baseline, a one-hour and a two-hour glucose, and that's universal at 24 to 28 weeks gestation. Some women with risk factors, of course, will have an earlier OGTT, perhaps in the first trimester, and then that will be repeated at the 24 to 28-week time slot if it's negative.

But Anusha, you'll see, has exceeded the thresholds on her one-hour and two-hour tests on the OGTT, and so she's earned herself a GDM label. We use the IADPSG targets here in Australia, which are used by a number of countries worldwide. Anusha's modified her diet. She's switched across to multigrain bread. She's reduced the portions of rice and pasta, and she's walking for 15 minutes after every meal, but she's still requiring some prandial, or mealtime insulin, so she's having four units of Novarapid with her lunch and her dinner meals.

And she's been correctly, but a little confrontingly, told that her lifetime risk for type 2 diabetes, now that she's had GDM, may be as high as 50 to 60%. So she asks you whether women with GDM like her manage to breastfeed successfully, and I think the answer to this has to be yes. So it's important to acknowledge when you discuss with her that yes, there are some mechanisms that make lactation more challenging for women with diabetes, but overall, and I think particularly for women with GDM rather than pre-gestational diabetes, the differences in their breastfeeding outcomes at a population level are pretty subtle. So they'll vary according to the study setting, but there's one quite good recent systematic review that I looked at with 16 studies, which showed that probably there are no differences in initiation of breastfeeding between women with and without GDM. Some of those studies suggest reduced rates of exclusive breastfeeding at hospital discharge.

Some of them suggest reduced durations of exclusive breastfeeding in women with GDM, but these are pretty subtle in terms of their effect size, and they're also inconsistent across studies. So I think that overall, being positive about Anusha's chances of a successful and rewarding breastfeeding relationship isn't at all unrealistic. And her next question is whether breastfeeding will help to reduce her risk of developing type 2 diabetes, and the answer to this really, if we look at the observational research, has to be yes. So lactation's consistently associated with a significantly reduced maternal risk of developing type 2 diabetes, and that's probably in the order of about a 30% risk reduction. And that's in all comers, but it's a similar magnitude of risk reduction when you apply that specifically to women who've had a GDM pregnancy.

And that, of course, is going to be particularly beneficial for them, given that their absolute risk is so high, as we've just heard. Anusha? And this really just shows you that in pictorial format. So this is from a recent systematic review and meta-analysis of 13 cohort studies, showing that the risk reduction for type 2 diabetes development was 0.66, and this is among women who are breastfeeding after a GDM pregnancy, when compared to those who aren't breastfeeding. And you can see that that confidence interval doesn't cross zero, and that's a statistically significant result.

And so really the observational data linking breastfeeding to improved metabolic outcomes in women long-term is pretty consistent, and it's pretty compelling. And there are lots and lots of similar slides that I could have put in here, because there's really good systematic review level evidence linking breastfeeding to reduced risks of developing hypertension, to reduced risks of developing cardiovascular events. And so Anusha thinks that sounds great, and she's interested in the mechanism for that. So when you start to read around this, you'll pretty quickly come across this idea that can be broadly termed the reset hypothesis. And this is the idea that during pregnancy, visceral fat accumulates, insulin resistance increases, and there are increases in lipid and triglyceride levels.

And all of those changes reverse more quickly and more completely after pregnancy when lactation occurs. So what evidence do we have to support that? What are the possible mechanisms? So there are a few lines of evidence here. And the first of those is the idea that lactation reduces maternal adiposity.

So we can look at animal models here, and animal models are great for lactation studies, really. It's easy to control variables like dietary intake, like breastfeeding exclusivity, and you can feasibly do an RCT, which we know is near impossible in terms of human studies randomizing women to lactate or not. So in rats, fat deposition increases over pregnancy. And when rats lactate, they mobilize those stored lipids from adipose tissue into mammary tissue in the breast and then into milk. And lactating animals, therefore, have smaller adipose cells or fat cells and lower levels of lipoprotein lipase activity compared to non-lactating controls. And those favorable metabolic changes hang around even after lactation. So even 21 days post weaning, rats who've lactated have significantly less fat cells than those who haven't. The effects of lactation on maternal adiposity become a little bit stickier when we try to look at this in humans. So in humans, the direct impact of lactation on postpartum weight change gets a bit difficult to test. There are so many factors at play

There's dietary intake, there's energy expenditure, and these things are notoriously difficult to measure. We all agree, though, that probably there's an extra four to 500 calories of energy expenditure for a woman over the first six months if she's exclusively breastfeeding. We do, though, have quite a number of studies that suggest that if a woman is in a well-nourished context, the environment in which all of us live, where food is amply available, most breastfeeding women within the first three months, at least, postpartum, will increase calories and reduce activity subconsciously to meet the energy demands of lactation. Whereas beyond three months, there's some evidence that maybe lactating women, if they are still exclusively breastfeeding, are more likely to mobilize fat stores a bit later down the track. And there are some quite nice MRI and skinfold studies showing that prolonged lactation, beyond about that three months mark, is associated with mobilization of fat, particularly from the suprailiac and the sort of mid-thigh regions when you compare with women who are no longer lactating.

So beyond the influences of breastfeeding on adiposity, there's this idea that lactation helps to reestablish glucose homeostasis after delivery. So this table's taken from an early study in the 1990s, and it was around this time that we began to realize that women who are consistently lactating or breastfeeding have lower glucose levels on an oral glucose tolerance test compared with their non-breastfeeding counterparts. So this was a study, it was six weeks postpartum, all of these women had had a recent GDM pregnancy. There's 400 women around about in each group, 400 lactating, 400-ish non-lactating. And I've just superimposed the values in millimoles per liter here because it was initially in milligrams per deciliter. But you can see here that there was significantly lower fasting glucose and lower two-hour glucose on an OGTT in the lactating group compared with the non-lactating woman. And that clearly and easily meets statistical significance in both groups. Similarly for the proportion who remain diabetic at that time and for the area under the curve when you integrate under the glucose curve. So it's quite clear that lactation has benefits even as early as six weeks for glycemia in women. And why does that happen?

Well, in really simplistic terms, glucose is simply siphoned off by the process of breast milk production. So in women who breastfeed, I'm always amazed by the statistics. So about 50 grams of glucose per day is diverted to the mammary gland for milk production. That's a lot. You think about what 50 grams of butter looks like when you're baking. And the really key thing here is that glucose uptake is non-insulin mediated. So insulin's not having to push that glucose into the breast tissue. The thought is that lactating women therefore have lower blood glucose and insulin concentrations despite higher rates of glucose production and lipolysis when you compare them to women who aren't lactating. So in simplistic terms, they're making glucose all the time, but their glucose levels are still lower because it's being siphoned off into the breast via a process that doesn't need insulin. And the thought is that maybe this unloads or takes the pressure off the pancreatic beta cells, therefore preserving long-term insulin production in these women.

There's also some really interesting thoughts around the idea that the hormone prolactin, which we all know is essential to breastfeeding, that it might play a direct role in improved glucose homeostasis. So again, a lot of the evidence here is from animal models, but definitely in rodents, prolactin can actually act directly on the beta cells, the insulin-secreting cells of the pancreas. So it binds to prolactin receptors on the cell surface and it acts via a series of secondary messengers, including serotonin, to enhance beta cell function and to enhance insulin production. And indeed, when you culture human pancreatic beta cells in a test tube and you expose them to prolactin, you see an increase in glucosestimulated insulin secretion, which is fascinating. So I think overall, there's certainly mechanistic evidence that breastfeeding lowers glucose levels, that it improves insulin secretion, it improves insulin action.

And maybe to some extent, this helps to explain the association between breastfeeding and reduced risk of diabetes in the long-term. And then the third arm of this reset hypothesis is the idea that lactation has advantages for lipid metabolism. So triglycerides and total cholesterol will go up across pregnancy, they'll fall again after delivery. And that fall seems to be aided by lactation. We've got lots of work documenting lower LDL levels, lower triglycerides, higher HDL levels in lactating women compared to their non-lactating counterparts. And that's particularly when lactation's prolonged. So if you were going to kind of summarize this reset hypothesis, you'd say, look, in pregnancy, insulin resistance increases, lipids go up, visceral fat accumulates. After the pregnancy is complete, there's going to be a gradual reversion back to the non-pregnant baseline metabolic state. And that's very much aided by lactation. So if you don't lactate, those metabolic changes of pregnancy hang around for a longer period.

Caveats in research [00:17:55.04]

And this is used often, this argument, or this hypothesis is used to explain these findings of reduced type 2 diabetes, reduced metabolic syndrome, more favorable lipid profiles, lower rates of hypertension and cardiovascular disease in women who've lactated. So what I've showed you so far is observational data showing an association between breastfeeding and improved metabolic outcomes. And I've talked about some of the physiological mechanisms that might link those two. But I think there's a big but here. And like with everything in medicine, the reality is much more complex.

And I think we've got to think really critically about things. Because all of the human evidence that I've just shown you is observational. And the lactating groups in these studies that I've shown are women who are self-selected by their ability to breastfeed. So when we think critically about studies that show improved metabolic profiles in lactating women, we also have to think about a couple of other things. And the first of those is confounding.

So women who breastfeed in observational trials, we know are typically more likely to engage in other healthful behaviors. So they're typically wealthier, they're better educated, they're less likely to smoke. And we can and do correct for those when we do our statistics, but we're always going to have issues with residual or unmeasured confounding. And then I think the other big issue here is this possibility of reverse causation. And that's the idea that maybe these four more favorable metabolic profiles that we're consistently seeing in our breastfeeding women simply reflect the fact that metabolically healthy women are more able to breastfeed in the first place.

So they appear in those groups by virtue of their pre-pregnancy metabolic health. And so the other way of looking at this would be through the lens of the preset hypothesis. And if you were making this argument, you'd say pre-existing maternal metabolic dysregulation impacts on lactation performance. And all of this observational evidence doesn't necessarily show us that lactation improves maternal metabolism. Rather, it's just showing us that metabolically healthy women are more likely to successfully breastfeed.

And you may go on to speculate that maybe adverse lactation outcomes may be a marker for future underlying maternal cardiometabolic disease risk, a little bit in the same way that we're now beginning to see preeclampsia as a marker of adverse future cardiometabolic risk. Let's explore this a little bit more. And this is the last slide on this theme, but I wanted to just highlight this author here. This is Laurie Nomsen Rivers, and she is a significantly impactful researcher in this area, she's from the States. And she really has, she's the chief author on many of the seminal papers linking metabolism and metabolic disease to lactation. So she's a great place to start if you're diving into the literature here. But this is quite an illustrative scatterplot. And I just thought I'd briefly describe this recent study to you from one of her really good papers, looking at the links between breastfeeding and metabolism. So this was a case control study. So they found 18 women who had severely low milk output.

So this is milk volumes here on the y-axis. So these 18 women had severely low milk output less than 300 mils per 24 hours. And then they have groups of nested controls and external controls who have more normal milk outputs. These women with the severely low milk outputs had been thoroughly evaluated for correctable issues such as infrequent or ineffective breast emptying. And they were really careful to point out in this paper that these women were emptying their breasts as frequently, if not more frequently than the woman with the normal output. And they then profiled these women. They looked at a number of metabolic parameters across the women in both of the groups. They looked at BMI and waist circumference and glucose and insulin and lipids and blood pressure. And they came up with this composite score called a metabolic severity score. And you can see those scores here plotted on the graph. So this is a 24-hour milk output on the y-axis and metabolic health on the x-axis. And so this area represents a zone of metabolic health that's better than average and this side worse than average. And you can see that the women with the very low supply are overwhelmingly overrepresented in this zone here representing adverse metabolic health. The circles with the faint circle around them represent GDM cases. And again, you can see there's a massive overrepresentation of GDM in the low supply woman.

Diabetes and postpartum difficulties [00:22:55.22]

And the similar thing was observed with class two obesity, which again was massively overrepresented in these women with the very low milk supply. So this really is becoming a real, a key area of interest within the literature. We'll move on now to our second case. So Zahra is 36. She has a BMI of 39 and she's a little bit different to Anusha. She was diagnosed with type two diabetes prior to pregnancy. So when she was 32. And outside of pregnancies, Zahra takes metformin, good dose of metformin twice a day and weekly trulicity or delagolatides, it's GLP-1 agonist, she takes that weekly. And since pregnancy has been recognized, she's been on a basal bolus insulin regimen as is typical. So protofan at night and now 22 units of Novarapid three times a day.

So she's got to 34 weeks and she's got a big baby on board, abdominal circumference on the 95th and estimated fetal weight above the 99th and her obstetrician's suggesting a section at 38 weeks and she's really keen to breastfeed. So she asks you what breastfeeding difficulties she might encounter due to her type two diabetes and whether there's anything she can do now to help that. So in order to answer this, I wanted to take you right back just briefly to the beginning of Zahra's pregnancy and think about what's happening in the breasts as the body prepares to lactate for the first time. And I do apologize if this is familiar. So during the first trimester of pregnancy, we have really rapid growth of the ductal lobular alveolus system in the breasts.

And then lactogenesis one or secretory initiation occurs during the second trimester. So from about 16 to 22 weeks of pregnancy, and this is very much under endocrine control. So we've got these rapidly rising levels of these progestational hormones, estrogen and progesterone, but also these lactogen hormones, prolactin and human placental lactogen or HPL, which is placentally derived. And these initiate milk synthesis. They help the alveoli to differentiate into secretory cells. They upregulate the gene expression of all the key genes involved in milk production. And it's now that a woman will begin to produce small amounts of colostrum, but we've still got the placenta in situ and so the really high levels of placental progesterone are keeping a break on copious milk production. Lactogenesis two or secretory activation is again under endocrine control. And the key precipitating factor here is this rapid fall in progesterone after the placenta is delivered. And so for women with uncomplicated deliveries, secretory activation or lactogenesis two would typically occur 48 to 72 hours postpartum.

And it's marked as we know by swelling of the breasts by the onset of copious milk production. And women will describe this as milk coming in. And then lactogenesis three later on is established milk production. This is under autocrine or local control. It's driven by the process of ongoing milk removal. And the main lactational hormone that's around at this time is of course prolactin, but we've got oxytocin two responsible for the let down reflex. So one of the key themes in the literature when you look at the impacts of diabetes on breastfeeding outcomes is this idea of the delayed onset of secretory activation or the delayed onset of lactogenesis two. And that's defined as it occurring more than 72 hours after delivery. Cues of course are breast swelling, milk leakage, a change in the milk's physical appearance from colostrum to transitional milk and breastfulness and pain. We've actually got lots of work showing that a mother's perception of her milk coming in stacks up really well when you compare that to very objective and scientific measures such as test weighing of baby before and after feeds.

So women are actually able to tell you very accurately to within a couple of hours. And so actually maternal perception of lactogenesis onset is used in many of the most robust studies in this area. But you can also define it biochemically. So there are lots of changes in milk composition around this time because it's a time when the lactocyte-type junctions are closing. So you see a rise in milk citrate and lactose with a fall in the sodium chloride and protein. And you may wonder why the timing of lactogenesis onset matters so much. Well, of course, we've got these immediate concerns about meeting our infant's needs for hydration and nutrition. But we also know that infant feeding patterns in these first few days of life are a critical determinant of later feeding practices. So studies show a really strong association between delayed onset of lactogenesis and shorter overall breastfeeding duration. So this has a big impact on women.

The evidence linking delayed onset of lactogenesis in two diabetes is pretty incontrovertible. So two good recent systematic reviews, the top one here by our esteemed colleague, Lisa Amir, who you'll see often on the Breastfeeding Conference Circuit. She had 10 studies in her review in 2015, seven of them type 1 diabetes studies and three GDM. And then this Chinese group summarized the evidence particularly focused on GDM. But the broad findings are really consistent across these reviews.

The first being that type 1 and GDM are clearly and consistently associated with delayed onset of lactogenesis. In the Chinese paper, they suggest that the prevalence of lactogenesis delay is probably about 35% for women with gestational diabetes across all studies. And they also highlight the fact that there are other factors that impact on the risk of delayed lactogenesis. So it's more common in primates and women who are of advanced age who are on insulin treatment for their diabetes, those who are obese and those who have worse metabolic control or worse glycemic control. But Zara, remember, she's got type 2. So she's saying, what about me? And she'd be right to highlight that. I think we've got a key issue in the pregnancy and lactation literature actually around a lack of data in pregnant and breastfeeding women with pre-pregnancy type 2 diabetes. And I think that reflects the fact that it's only been over the last sort of maybe 20 years that large numbers of women at a population level are entering pregnancy with pre-existing type 2 rather than type 1 or GDM. So we've got quite a big gap in the literature here across a number of key outcome areas.

I was really pleased then to see this pilot trial. This is from some colleagues in Brisbane. There are lots of names familiar to you, some big names in the Australian lactation space. But these guys did a beautiful, elegant study. So they took 14 women with type 2 diabetes with a mean BMI of 30.6, so in the obese range. And then they had 10 BMI matched, so also obese but not diabetic controls. And then they had 12 women with normal BMI and no diabetes. And they looked at lactogenesis onset via maternal perception. Plus they also defined this biochemically by measuring breast milk components twice daily for five days. And their findings were pretty clear. So the citrate concentrations as being a key lactogenesis marker, they rose slower and they plateaued at lower values in the women with type 2 compared with the non-diabetic controls, both the lean non-diabetic controls, but also the women with obesity but no diabetes. And this suggested that there was delayed onset of lactogenesis that was associated with the diabetes more so than the BMI. They also found that within the group of type 2 diabetics, the larger insulin doses were associated with later lactogenesis. In the discussion, they talk about the fact that in addition to the biological mechanisms associated with the diabetes, the woman with type 2 diabetes also had lots of other clinical factors that are probably adversely impacting on their lactogenesis and their breastfeeding. So 10 of the 14 women had babies with neonatal hypos and 10 of the 14 women had babies who had formula supplementation as an inpatient.

Interestingly, when you extended this out to sort of four months, the type 2 diabetic woman and the BMI matched nondiabetic controls were less likely to be exclusively breastfeeding at four months compared to the lean non-diabetic woman. So both diabetes and obesity seem to have a bearing on breastfeeding exclusivity at four months, which was interesting. The obesity appeared to have an impact later, but not to have such a bearing on lactogenesis. So when you think about the mechanisms that link delayed onset of lactogenesis to diabetes, I think you've got to think about them in two sort of broad buckets. And the first bucket relates to the diabetes physiology itself.

So diabetes and the often comorbid obesity, it's a state of chronic inflammation which can impair substrate uptake. So glucose and oxygen and long chain fatty acids are more difficult to deliver to the breast. But there's also, and I think this is a really key takeaway point, insulin and the insulin receptor interaction. So the interaction of insulin with the membrane-bound receptor are really important in switching on genes that govern mammary differentiation and milk synthesis. So insulin and its receptor are really important to switch on these genes that are integral to breastfeeding initiation.

And so insulin resistance or absence probably impairs milk production really down at the granular sort of lactocyte level. The second bucket is the clinical consequences of the diabetes. So we know that mothers with diabetes are more likely to have preterm births. Preterm babies, of course, have less developed suckling reflexes and lower muscle tone and smaller mouths and a whole set of factors that makes it more difficult for them to breastfeed. They're also more likely to have instrumental delivery and cesarean sections.

They're more likely to have babies with hypos, mother-baby dyad separations more common, and babies are more likely to be given early formula supplementation. And all of those clinical factors are independent risk factors over and above the diabetes physiology itself, which contribute to delayed lactogenesis. What can we do? I think there's overall, there's a dearth of literature here and specific interventions, but I think it really comes back to simple measures. So optimizing glycemic control during pregnancy and pre-delivery is really important.

We've seen that women with worse metabolic control have more delayed lactogenesis and more inferior breastfeeding outcomes. So a lot of this comes down to good models of care. I know a lot of big tertiary hospitals like the one that I work at will have endocrinologists and diabetes educators co-located at a maternity clinic alongside obstetricians and web wives in the antenatal space. But that's obviously something that's a luxury of a large tertiary hospital, and it may be more difficult to access in more remote settings. Discussing breastfeeding initiation and affirming it and offering support antenatally I think is really important.

There's a beautiful study out of Deakin University. I haven't put it on the slide, but they looked at women with prepregnancy type one and type two across three large metropolitan Melbourne hospitals. And they looked at all of the factors in these women with pre-existing diabetes that predicted successful breastfeeding at three months. And after multiple regression, the only one that was statistically significant was the woman's pre-birth intention to breastfeed. And so I think this is a really key thing to ask about in antenatal reviews and to affirm women on if they say, yes, they are intending to breastfeed. I think it can be difficult because we're so distracted by all the things happening in the antenatal space, but I think it's a really good thing and it's brief and I think can have really significant impacts on postpartum outcomes. Antenatal breastfeeding education can be really beneficial. We know the ABA offers really good group community-based education settings. Antenatal expression of colostrum I'm going to talk about in the next couple of slides because that's a really important intervention. Early skin-to-skin and the opportunity to breastfeed within that golden hour, the avoidance of separation of the mother-baby diet unless that's clearly indicated.

Mothers who are medically unstable in an ICU, we can get them to express breast milk immediately following stabilization. And of course, early and targeted LC involvement and support. So I want to talk a little bit about antenatal expression in the DAME study because this is a really key Australian study that's made international waves in this space. And again, I apologize if it's something that you're all familiar with. But here we're talking about the manual expression of colostrum in the last few weeks of pregnancy.

The goal being to store small amounts of colostrum to be given to the infant prior to the established milk supply. So prior to lactogenesis, it perhaps stayed two or three. And for mothers with diabetes in particular, the stored colostrum can be used to correct mild neonatal hypoglycemia. We know it's as effective at stabilizing blood sugars in this context as formula. Women would typically be advised to check with their antenatal care provider first. And from 36 weeks, they might do sort of three to five minutes of hand expressing each breast two to three times a day. The colostrum can be collected into a medicine cup and aspirated into a syringe, or it can be directly aspirated at the breast. The syringe is then capped and labeled and frozen within 24 hours of expressing. We're really expecting small amounts here, just a few mils. And 25% of women, this is a figure from the DAME study, are unable to express antenatally.

And that is not a harbinger of low supply postpartum, which I think is really important to convey explicitly to women. The Royal Women's Hospital from Melbourne has a really good online instruction handout for patients. And that's usually the PDF that I would print off. The DAME study is important to talk about here because it was the seminal trial proving the safety and the efficacy of this intervention. So the background is that in the early 2000s, this was being done a lot. It was being promoted by breastfeeding counsellors. But in the clinical literature, there really was a dearth of evidence around it. So the goal from this group based out of Latrobe was to couch this practice in some cold hard clinical evidence. And the background was that there were actually some safety considerations. So antenatal manipulation of the breasts comes with a theoretical risk of a surge in oxytocin, prompting uterine activity, prompting contractions, and possibly promoting premature labour. And there were a couple of small, very small UK studies in sort of 2011, 2012, which maybe did show an increased risk of early delivery and NICU admissions, a signal in that direction. And indeed, there were some calls for this practice to be halted awaiting further research. So this was the definitive trial. It was called the DAME study, and it was published ultimately in the Lancet in 2017, which is a huge achievement, I think, for a lactation paper. So conducted across six hospitals in Victoria, all of the women had gestational diabetes or pre-existing diabetes in pregnancy.

They were all low-risk, single-term pregnancies. And the intervention was rolled out at 34 to 37 weeks. The women were randomised to either express twice daily or just to standard care. The expressing group did about a median of 20 episodes, so pretty good compliance with the intervention. But the volumes were small, and I think this is important to bear in mind. This is a total volume over the entire expressing period. The median volume was 5.5 mils. And ultimately, they showed no difference in NICU admissions, which was good. It appeared that this was safe. It's not promoting preterm delivery. And there was similarly no difference in mean gestational age between the groups. So that was great. It proved the safety of the intervention. There was also a moderate association between being in the expressing group and baby receiving exclusive breast milk to 24 hours. That had softened by the time of hospital discharge, and that benefit was not sustained at three months.

Overall, this is now considered a safe and an accepted practice. The study's conclusion was that there's no harm in advising women with diabetes and pregnancy who are at low risk of complications to express from 36 weeks, and there's possibly some evidence of benefit. But there's a couple of caveats here, and I think the first of those is this is not advisable for women with particular obstetric risk factors that might predispose them to preterm birth. So imminent preterm labor, any issues around cervical incompetence or cerclage, any history of antepartum hemorrhage or placenta previa, it's also judicious to say, look, if you're noticing an increase in uterine activity, particularly linked to the expressing episodes, you should probably stop. And I think the other real key thing to note here is that the postnatal use of the colostrum to treat hypos on the ward, for example, needs to be as advised by a lactation consultant or a skilled healthcare professional, because we really don't want that colostrum to take the place of a feed directly at the breast.

It's absolutely that direct feeding that needs to take priority while supply is being established. So Zara, back to our Zara, she's had her baby, she had an elective section and early skin-to-skin was prioritized. Her babe had mild hypoglycemia, and she was able to supplement with her express colostrum that she was pleased about. She's four days postpartum now, her milk's come in, and she's really motivated to breastfeed. She's got back on her metformin, but remembering she was previously on trulicity pre-pregnancy.

Breastmilk & diabetes [00:40:25.18]

Her fasting sugars with just the metformin monotherapy were still in the double digits on day two postpartum. So she's just been put back on a little bit of her basal insulin, just her protofan. And she says to you, will my milk be full of sugar if I'm running my sugars high? And then she's saying, what about my insulin? Does that get into milk? And will it be safe for baby? So the answer to her first question, will my milk be full of sugar? So if we think about the way that glucose gets up into the breast, it's passively transported. And remember I told you earlier, this is an insulin independent mechanism. It's passively transported in by this GLUT1 transporter.

And the majority of that glucose represented here by the orange hexagon is linked up to galactose to synthesize lactose. And lactose is the disaccharide, which is the main carbohydrate component of breast milk. And it's the main determinant of breast milk volume. Once lactation's established, most of our studies suggest that the lactose concentration of breast milk samples from women with diabetes is similar to that of control. So no major difference in lactose.

Breast milk does have in it a little bit of free glucose. And we do have studies suggesting that that amount is likely to be higher and indeed more variable in mothers with type 1 and type 2 diabetes when you compare that to control women. So this graph here shows that yes, there is a statistically significantly more free glucose in the milk of type 1 and type 2 compared with control women. But the really important thing here is the scale on the y-axis. Because you'll see that the typical lactose concentration in breast milk is 75 grams per liter. And these units here, we're sort of looking at 0.8 grams per liter. We also know that experimentally induced hyperglycemia in women with diabetes does elevate breast milk glucose. So if you get diabetic women, you hook them up to a glucose drip. Yes, their sugars go up. And about 40 to 90 minutes later, their breast milk sugar will go up. But again, glucose concentrations in the breast milk are 100 times lower than lactose concentration. And glucose makes up about 0.4% of the energy content of breast milk compared with 40% for lactose. So the clinical significance of that variation is really dubious. The next question is around whether insulin enters the breast milk. And if you look at up-todate, this is the snippet you'll see.

Exogenous or injected insulin use is compatible with breastfeeding. Insulin's thought to be a normal component of human milk and is not expected to be orally bioavailable to the infant. Let's break this down a little bit because it's a little bit more nuanced, I think, than that. We know that insulin's present in both colostrum and mature milk, both in women with diabetes and in non-diabetic control women. And we also know that injected insulin taken by diabetic women does get into breast milk. What's really interesting is that regardless of where it comes from, the concentrations of insulin in breast milk are similar to the concentrations of insulin in serum. And that's really interesting because most other proteins that are the size of insulin move down a concentration gradient into the milk. So they're present in the milk at about 100 times lower concentrations than in the serum. So the fact that the amount of insulin in the milk is equal to that of the serum implies that there's some sort of active transport mechanism for insulin into the breast milk, i.e. that it's actively being pumped into the breast milk. They haven't identified the transporter yet, but it certainly seems that way.

And that's really curious from an evolutionary perspective. We think that insulin in the breast milk has no biological action. The standard understanding has been that insulin's a protein, so it's really rapidly denatured in the acid environment of the stomach, and it wouldn't be biologically or orally available to an infant when it's consumed. But there are a few lines of evidence in recent years that have challenged this. And the first of those is that insulin receptors are present on the intestinal epithelium of other large mammals like cows. And it's thought that milk-borne insulin in baby cows or calves helps with their intestinal maturation. It's possible there's a similar role in humans. So if you have preterm infants infected by things like necrotizing enterocolitis, for example, and you give them insulin orally, there are some studies which shows that that enhances GI function. And there's also some suggestion of a possible role for breast milk insulin in the development of a healthy infant gut microbiome and the development of healthy gut immune tolerance. So if all of those are the case, it is tempting to see insulin as an important bioactive component of human milk.

And if that's the case, I think it's also important to acknowledge that the artificial recombinant insulin that we give to women who have diabetes is only sort of one to three amino acids different to human insulin. So here's a Lantus molecule. There's three amino acids that are different. And the substitutions are well away from the domains of this protein that would interact with the receptor. So it's reasonable to think that any beneficial role that might be filled by the milk-borne insulin of a non-diabetic mom, her endogenous insulin, which has just entered the breast milk, would probably be exactly matched by the exogenous insulin in the breast milk of a diabetic mother. But we definitely need more research in this area. It's a really interesting one.

PCOS [00:45:36.00]

All right, this is a little bit of a diversion now away from diabetes and looking a little bit more into polycystic ovary syndrome or PCOS. So let's move to Martine. Martine's 29. She's just had her first baby and she's two weeks postpartum. And she's got a background of polycystic ovary syndrome. Prior to her pregnancy, she was only having three to four cycles a year. And so she required the assistance of a fertility specialist to conceive and she was given some letrozole for ovulation induction, which would be our first line fertility approach for someone with oligo and ovulatory fertility and PCOS. Martine has a BMI of 37. And at two weeks, she's been referred to you with suspected low supply. She's got a really unsettled baby who's still well below birth weight. And she's saying to you, could this be because of my PCOS? I wanted to recap here a little bit about PCOS in general before we move to the PCOS breastfeeding material. So it's the most common endocrinopathy amongst women of reproductive age.

Depends on what populations you're studying, but the prevalence is probably somewhere between five and 18% of women. I usually say to my woman, one in 10. And the diagnosis is according to the Rotterdam criteria. So for adults with PCOS, you need to have two of the three criteria. The first one of those is oligo or amenorrhea. So cycles that are less than 21 or more than 35 days in length, or a woman who's having less than eight cycles a year. The second of those is hyperandrogenism. So androgen excess. And that might be clinical. It might be based on symptoms such as hirsutism, acne, androgenic alopecia, or it might be biochemical evidence of androgen excess.

So you'd be looking at total testosterone, but also at SHBG and free testosterone. And then the third criteria is polycystic ovarian morphology on an ultrasound. So ideally this is done transvaginally. It's done by an experienced operator with a high-resolution transducer. And you're looking for 20 or more follicles on either ovary or ovarian volumes of 10 mils or above. The latest iteration of the PCOS guidelines came out in 2023, and they made a new comment that if ultrasound is not performed or is not available, then AMH or anti-Mullerian hormone, if that's elevated, can be used as a substitute for the polycystic ovarian morphology. And this will be a typical polycystic ovary appearance. I do quite a lot of PCOS. I like to describe it to women in terms of four key domains or four key groups of features. And we know a lot, and we talk a lot about the endocrine features.

So acne and hirsutism and alopecia, and all of these are driven by endocrine, so by androgen excess. And then the reproductive features are also well-documented and often top of mind for women. So anovulation, which often represents or often presents as menstrual irregularity, but also endometrial hyperplasia, particularly if these women are having lots of anovulatory cycles, and also increased rates of pregnancy complications. What patients often aren't so well-versed in are the metabolic features of the syndrome, so a really clear basis in insulin resistance with a high diabetes risk, high rates of metabolic-associated liver disease, dyslipidemia, cardiovascular disease, and also the psychological features, which is a really interesting area and one we're just beginning to get to grips with. So high rates of anxiety, depression, eating disorders, and also general distress and low self-esteem

I think a few kind of take-home tips for managing PCOS, breastfeeding aside, I think the syndrome has a really confusing name, and indeed, there is actually an international push, perhaps, to rename it. They're not cysts, really, on the ovaries. They're small, immature follicles in arrested development. The pathophysiology of this condition is really, really complex, and it's quite abstract stuff, which is really difficult to describe to patients. There's also lots of myths around this, and it's got a very big social media presence.

So I think that education is a large part of working through this with our patients. It's also a disease where patient priorities vary hugely across the lifespan and also between individuals. So it's a long-term chronic condition. It needs a really good therapeutic relationship. So I see a lot of women in their sort of teens and 20s perhaps predominantly concerned about menstrual irregularity, about hirsutism. Perhaps in their 30s, we move to looking at distress around subfertility and doing lots of preconception planning, optimizing metabolic factors in order to fall pregnant, and perhaps pursuing fertility assistance. And then often in their 40s, these women are looking at weight management and metabolic sequelae. I find that there's a really big overlap, particularly in recent years with obesity and weight management, and also with mental health. Those are probably the two biggest things that I talk to my PCOS woman about, or that I spend most of my time talking to my PCOS woman about weight management and their mental health, because there's big overlaps with those areas. And the Monash 2023 Evidence-Based Guidelines really should be the go-to for PCOS management.

So the full guideline, all of this is easily accessible online. The full guideline extends to more than 200 pages, but there's a really nice plotted summary, which is available as a PDF, which is about 20 pages and has key practice points. There was also a really big translation budget for this work. And my colleagues at MCRI, where I work at Monash University, led by Helena Teed, have developed this Ask PCOS app. So this was co-designed with individuals living with PCOS. It includes evidence-based articles, there's question prompt lists for discussions with health professionals, there's symptom trackers, there's a patient discussion forum on there. So that's quite a good tool to promote to your patients. Let's move now to just think a little bit about PCOS and breastfeeding. So unfortunately, there's not a whole lot of data around this, which is one of the issues, but we've certainly got small case control studies and lots of individual case studies suggesting reduced breastfeeding success in PCOS compared with PCOS, women with PCOS when compared with controls. And it's certainly something that if you have discussions with lactating consultants who've been practicing for years, they will anecdotally say, yes, there is absolutely a link between breastfeeding difficulties and PCOS.

I've recently published some work building on work from my predecessors, looking at women with metabolic disease and PCOS in particular within the Australian Longitudinal Study of Women's Health. So this is an enormous epidemiological study. I looked in particular at a community-based cohort within the study to whom there'd been sort of 13,000 live births. And within that cohort, we found, when we looked at a continuous variable, so self-reported duration of breastfeeding, there were slightly shorter durations in PCOS compared with non-PCOS women, so seven versus nine months, subtle, but it was statistically significant. When we did a logistic regression where we looked at the odds of initiating breastfeeding and the odds of breastfeeding to six months or more, there were actually no differences according to PCOS status in this particular study.

What we did find in this study or this analysis was that the main independent predictor of breastfeeding initiation and duration was maternal BMI, much more so than any specific metabolic condition. It does, though, appear, particularly from the literature, that PCOS does seem to be broadly associated with reduced breastfeeding, with poorer breastfeeding outcomes or with more barriers to breastfeeding. And I think a lot of that probably is mediated by obesity. There's a big overlap there. But there are potentially other disease-specific factors. So we know, and we've talked about the fact that insulin and the interaction between insulin and its receptor is really important for milk synthesis. It probably switches on key genes involved in lactation. There might also be a direct detrimental effect of androgens on milk production. So testosterone, back in the good old days, was given therapeutically to shut down lactation. This is a little paragraph taken from JAMA in 1941, where they say, in the presence of certain constitutional diseases, such as TB or diabetes or heart disease, there may be valid reason to prohibit lactation.

The measures employed for this purpose until recently were empiric, generally consisting of tight breast binders, Camp 4 liniment, belladonna unctions, restriction of fluids, saline purges, and numerous other procedures of questionable value. And then they talk about the way that they've inhibited lactation by administering large slugs of intramuscular to testosterone propionate, which we would never do now. But it's interesting that testosterone directly inhibits lactation. And we've got small observational studies showing that maternal androgens at mid to late pregnancy, both in general populations and in PCOS populations, seem to be associated, seems to be negatively associated with breastfeeding. And then lastly, and I'm just gonna have a couple more slides on this, because I think it's really interesting and important, is this link possibly between PCOS and breast hypoplasia or insufficient glandular tissue, IGT?

If we think about IGT as an entity, we know we've got a lot of research suggesting that there's not a huge correlation between breast size and milk production ability overall. But we do know that successful lactation requires adequate glandular tissue. And that relies on sufficient ductal proliferation in pregnancy, sorry, in puberty, and then again in a woman's pregnancy. Mammary hypoplasia is the general term used to describe an anatomical lack of glandular tissue. And you'd specifically call that insufficient glandular tissue if that then translated to breastfeeding difficulties.

So the lactation textbooks are full of diagrams like this one, and they describe breasts that might be sort of tubular shaped or perhaps droopy, maybe asymmetrical breasts, wide intramammary width, greater than about 3.8 centimeters, breasts which fail to increment across pregnancy in terms of breast size, or perhaps breasts with very minimal breast tissue in the inferior quadrants. And in some women, breasts that look like this do seem to predict primary lactation insufficiency, and sometimes early cessation of exclusive breastfeeding despite a woman's best efforts. But it's murky, there's no agreed definition or diagnostic criteria for IGT, and it's also unclear which specific factors may make breasts that look like this more or less relevant to a woman's milk production ability. The link between PCOS and this condition is well described, but again, it's mainly at the level of case reports and case series, and also in a few now very outdated studies looking at soft tissue radiographs of women with PCOS in their breast tissue. But there are some quite strong theoretical arguments.

So one of them is that women who have PCOS and are having lots of anovulatory cycles or aren't menstruating much at all don't have that lovely cyclical progesterone exposure that many regularly cycling women get every month in their luteal phase. Progesterone, we know, is really important to glandular breast development, so maybe that lack of progesterone exposure leads to impaired glandular tissue development in puberty in these women. We also know that women with PCOS and with adverse metabolic health don't tend to increment their breast size so well across pregnancy. And again, there's this idea around insulin resistance and defective interactions between insulin and the receptor causing difficulty with activating the genes that are essential for breast development. And then furthermore, maybe, again, there's a direct inhibitory effect of androgens on the mammary gland.

What can we do if we're faced with a woman who we think may have IGT possibly in the context of PCOS? Well, I think a lot of it is about anticipating difficulties antenatally before they occur. So really good granular antenatal history taking. Does this woman have known PCOS, known diabetes, known obesity? Is she someone who's struggled with subfertility perhaps for one of those reasons? Has there been a history of surgery or injury to the breasts or chest? Some women may have had surgery for cosmesis to correct hypoplastic breasts. What about breast changes during pregnancy? Were there weight changes during pregnancy? And then the breast exam, of course, looking for anatomy that's analogous to that which I've just shown you.

So particularly widely spaced, tubular or asymmetrical breasts. Acknowledging the limitations here and that this is not an exact science. And again, I think acknowledging breastfeeding priorities in a woman's psychosocial context and supports what things are going to be going on for her in the postpartum period. What setups might she need? If we do get primary insufficiency milk supply postpartum, our strategies really in the context of insufficient glandular tissue are as per usual guidelines, including lactagogue therapy. And there are a number of established protocols here that you'll all be familiar with. Martine though has a specific question about metformin because someone on Reddit has said to her that it might help with supply in PCOS. And she's got a really good thought here. So metformin, as we know, is an insulin sensitizer. We also know that it's safe in lactation unlike many of our diabetes drugs.

And it makes theoretical sense given what we know about how important insulin probably is for lactogenesis. We've only got one trial here and it was done by Laurie Nomsen-Rivers, the researcher who I showed you the photograph of before, was done by her group. And it was a small pilot trial looking at the feasibility of metformin to treat low supply in women with insulin resistance. So in order to be in the trial, you had to have either elevated fasting glucose, a history of GDM, PCOS, or central obesity. And this is only a pilot trial. So tiny, tiny numbers. They assigned 10 women to metformin, five women to placebo, and they started the intervention pretty late. It's sort of 36 days postpartum, which was later than they'd hoped. They also had a really intensive test weighing protocol. So weighing babies before and after feeds in order to assess milk supply. And the participants found it difficult to adhere to that. They also ramped up the dose of metformin pretty quickly. So only seven out of 10 metformin participants completed the course. What they found, perhaps unsurprisingly, was a little bit disappointing. There was a trend in the right direction, but the median peak change in milk supply over a 24-hour period was plus 22 for the metformin completers versus minus 58 for the placebo and the woman who didn't complete metformin. So there was a trend, but it wasn't significant. Ultimately, none of the participants in the trial, even those who completed the metformin, perceived it as having been worthwhile for their supply. And these authors had feasibility concerns about whether this was worth upscaling to a larger trial. What about if we gave it earlier on? What about if we gave it during pregnancy, for example, for women with PCOS?

This, again, this is a little bit different, but it was a follow-up of a big study looking at metformin versus placebo for pregnant women with PCOS. So they weren't just looking at breastfeeding, but this was the breastfeeding paper. And all of these women were randomized to metformin or placebo, and they took it all pregnancy before stopping at delivery. And then they looked at outcomes at one year postpartum. And what they found with respect to breastfeeding was that in the entire cohort, those women whose breasts didn't increment across pregnancy, so it didn't get bigger across pregnancy, had shorter breastfeeding durations, both exclusive and partial breastfeeding, and also had worse metabolic health. So they tended to be more hypertensive, more obese. And when you look back at pregnancy, they had had higher first trimester insulin and glucose levels. But what they didn't find was any differences in breastfeeding outcomes or breast size increment according to whether you'd had metformin or placebo. So overall, worse metabolic health seemed to be associated with a lesser breast size increment, but metformin didn't seem to change that. Overall, when I'm asked about this, I generally tend to say, look, metformin makes mechanistic sense in the setting of insulin resistance, but currently there's insufficient evidence to support its use primarily for augmenting milk supply.

Patients who might be taking it for diabetes or PCOS can safely continue it during lactation. And I generally say, look, theoretically, this may well regulate the hormonal environment in a way that supports lactation, but there's just not enough evidence or not enough evidence yet for us to start initiating particularly for that purpose. Thanks for hanging in there. I've come to my last case, and this is a pretty quick one, really, and this is a type 1 diabetes case, because I think it's really important for completeness to cover this important cohort of patients, and they are particularly vulnerable and difficult to manage in pregnancy.

T1DM [01:04:00.2]

So let's meet Alice. Alice is our last case, and she's 29, and she's got type 1, and she was diagnosed when she was seven. In childhood, under Peds, she was on a pump, but she got a bit sick of that, and she's now on multiple daily injections of insulin. So Optisulin or Lantus is her basal or background insulin, and then Novarapid for her meals. And she's a well-controlled individual who's well-engaged with her endocrinologist and a diabetes educator. Now, preconceptually, she had her HbA1c down at 7.1%. That's great. ADIPS, so the Australasian Diabetes and Pregnancy Society, their guidelines suggest less than 6.5% for pre-gestational diabetes, either type 1 or type 2, if that's achievable without hypos. In practice, that is very, very difficult to get in a type 1 diabetic without significant hypoglycemia. I would consider 7.1% to be excellent. This is changing a little bit with our increasing use of insulin tech, and particularly hybrid closed-loop devices. But Alice is doing pretty well if she's on multiple daily injections and she's got her HbA1c down to 7.1% before conception. She's using a Freestyle Libre device to monitor her sugars. She weighs 60 kilos. And so let's look at her prepregnancy, non-pregnant insulin requirements. So she's on 30 units a day, which just conveniently is 0.5 units per kilo of body weight. And you can see that there's a 50-50 split between her basal insulin and her bolus insulin. So 15 units of her 30 is background or basal insulin, is opto-insulin given in the morning. And then 15 is her mealtime insulin, and that's given as five, five, and five. So five with breakfast, lunch, and dinner. And that's a carb ratio of about one unit of insulin to cover 10 grams of carbs.

Let's catch up with her now 37 weeks pregnant. She's seeing an endocrinologist for the last time in this pregnancy. What's happened to her insulin requirements? So this is a type one diabetic here in the green, and this is what insulin requirements tend to do across pregnancy. So you can see that in the first trimester, women actually become a lot more insulin sensitive. So there's a real dip. And it's at this time often women are off their food because they're feeling pretty unwell. And we get pretty worried about hypoglycemia. We often back off on their insulin doses even below pre-pregnancy levels. We talk to them about driving because we really need to be safe in this time. After that, we get this really steady increase across trimester two and three, and then a little bit of a plateau and sometimes a little bit of a drop off towards the end of pregnancy from about 36 weeks on. But generally by the time women are at the end of pregnancy or in their late third trimester, they're on about 2.5 times their total daily doses from pre-pregnancy. And the other thing that happens is there's a little bit of a difference in the split between basal and bolus insulin. So now that we catch up with Alice, she was on 30 units pre-pregnancy, which was split into 15 and 15 basal and bolus. But she's now on 75 units, so two and a half times the total daily dose.

But you'll see now that it's more like a third basal and two thirds bolus insulin. And this is pretty typical. So whilst both basal and bolus insulin requirements increase across gestation, there's a greater increase in the bolus or the mealtime insulin relative to the basal doses. And you can see that her carbohydrate ratios have become stronger. So she's needing more insulin to cover, sorry, more insulin to cover her carbohydrates. So one unit of insulin is only covering three grams of carbs at breakfast. One unit's only covering four grams at lunch or dinner. And she's now on much bigger and over rapid doses, so 12 to 18 units per meal. Pretty common to be more insulin resistant in the morning. So Alice says to you that she's keen to breastfeed and she wants to know what to look out for in the first few days as she establishes breastfeeding. Her midwife has told her about a previous type one diabetic that she looked after who had lots of problems with hypos at the breast. So Alice is a bit worried about that. This is tricky. So insulin requirements fall immediately following the delivery of the placenta. And they've dropped pretty markedly. So they go down to actually 30 to 50% lower than pre-pregnancy doses. So this is really psychologically challenging for our patients. They go from the most insulin resistant they have ever been in their entire life to this exquisite state of insulin sensitivity where their total daily doses are often lower than their pre-pregnancy state. And the timing of this can be a little bit variable. For some women, it really truly does happen immediately.

For some women, it takes maybe up to sort of six hours, six to 12 hours. Women who are in hospital with type one diabetes and are delivering should have a delivery plan and a postpartum plan provided by an endocrinologist. And typically they would have hourly blood sugars in hospital until the first meal. I usually withhold bolus insulin with the first meal because there's a tendency to hypo. But what's really important is they prepare these women that their blood sugars are gonna be a bit of a moving target for the first three to five days postpartum. Often the endocrinology team at a big hospital would be coming around and seeing these women daily if not more frequently. And what's really important is that it's appropriate for them now to relax their glycemic targets. So control of diabetes during pregnancy is strict. ADIPS would recommend less than 5.3 fasting, less than 6.8 at two hours for someone with pre-gestational diabetes. I would now say, look, five to 10 across the board is perfectly acceptable. You know what? If you are going to 12 or 13 after a meal, I'm not terribly bothered about it. They need to have plenty of hyposnacks on hand. Even in the first six weeks when she goes home, Alice has got a tough ship to steer. And I always say this to women.

So their insulin requirements are continuing to change. It's really important that they maintain their diet with sufficient carbs to support lactation, to maintain their blood glucose levels and prevent them becoming ketoacidotic. They're needing to self-monitor assiduously and they need to do a lot more adjustment than they're used to. And all of this is whilst they're coping with the demands of new motherhood and the broken sleep and the changes in circadian rhythms. And so episodes of mild hypo and hyperglycemia are inevitable. And it really is about not being perfect about it and just getting through. Actually, this is where some of the type one diabetes tech that we've now got, CGM, the hybrid closed loop devices, in some instances, it really comes into its own because it allows these women to outsource some of that mental load to their technology. How does breastfeeding influence on all of this? Look, there aren't a whole lot of studies as with much of this. And a lot of them are quite dated. They're often done back in the era where women had isolated finger stick measurements rather than the much richer and more complete CGM data we're now used to with a lot of our diabetes trials. But on average, the literature suggests that the insulin requirements in exclusively breastfeeding women with type one diabetes probably remain about 20% lower than their pre-pregnancy values from months one to four postpartum. So whilst this sort of perhaps an archetypal woman who wasn't lactating would drop down and then fairly rapidly within a few days be back near her pre-pregnancy requirements, they probably hover at 20% less for women who are breastfeeding, but there's significant individual variation. So the IOM recommends that women who, all women actually, who are breastfeeding take a minimum of 210 grams of carbs per day and at least 1800 calories a day. I think that's probably, it's not a whole lot. I think most women would need a lot more than that. And just as a bit of a kind of index, so this is probably a sandwich with two slices of toast bread. That would be about 40 grams of carbs. So 210 grams is the minimum per day. There's a lot of fear around hypos at the breast, particularly in the immediate postpartum period.

The current understanding is that this probably just reflects the fact that this is a really insulin sensitive time rather than anything specific to the suckling episode. When you look at CGM studies, yep, you do get a little bit of a dip in sugars when the infant's suckling at the breast, but only a very small minority of those will result in clinically significant hypos. And a lot of hypos in the postpartum period are completely unrelated to suckling. So I usually say to my woman, look, routine consumption of carbs during your night feeds, for example, in order to prevent hypos is probably not necessary, particularly in this era in which we live when most people just can swipe their CGM on their phone anytime. So that's on the condition though that these women are eating enough carbs and that their insulin doses have been properly reduced by their team and by themselves in the postpartum period.

So Alice comes back to see us. She's made it to eight weeks postpartum. So let's just recap what's happened to her insulin. Remember, early on before pregnancy, she was on 30 units a day, a 50-50 split, so 15 basal and 15 prandial. By the end of pregnancy, it was 75 units, so it had increased by two and a half times, and it was more like a third basal and two thirds prandial. And now she's breastfeeding. Her overall requirements are about 20% lower than even pre-pregnancy. And again, it's reverted to more of that non-pregnant sort of 50-50 split. When Alice comes to see us, she shows us her Freestyle Libre, and she says, look, my average glucose is sort of eight to nine, and I'm having three to four mild hypos a week. She says, look, the doses that I'm on at the moment are probably a little bit conservative. I could definitely get tighter control, but I'm pretty happy with that, and I would totally agree. So her key priority at this time is to avoid hypos while she's busy attending to her baby. She should be watching her blood sugar on her phone during her long feeds, but I don't think that in the era in which we live, she needs to prophylactically be front-loading carbohydrates in order to stop herself having hypos. And I'd say to her, look, we'd anticipate a really gradual return to pre-pregnancy doses perhaps over the next two to three months. I do think it's important to point out here that this is a highly idealized case with nice round numbers.

And the reality with type one is a lot more complex and a lot more messy. A lot of these women not eating three square meals a day, they're having a much more erratic intake, and they may be giving themselves multiple small boluses. Their insulin sensitivity can change significantly over the course of the day. So some people are very insulin resistant in the morning and not so much at night. Some women gain and then retain a lot of weight over their pregnancy. So they're more insulin resistant postpartum than they were pre-pregnancy just because they're carrying a lot more weight. So it really is about individualizing this to the patient and regularly reviewing them as an endocrinologist. That brings me to the end. I hope that that's been somewhat helpful. I've been through a lot and I appreciate that it's probably felt like a bit of a crash course, but let's just recap some of the kind of key takeaways from each of the cases.

So in the first case, I talked about the fact that lactation is consistently associated with improved maternal metabolic outcomes, but it's unclear whether that reflects the reset hypothesis or the preset hypothesis. Perhaps it's a little bit of both. I think that's probably the most likely scenario. We've talked about the fact that delayed lactogenesis onset is common in women with diabetes and pregnancy, that antenatal colostrum expression is really safe for this group. And it may be helpful, but it's an individual decision and it's certainly not the be-all and end-all. PCOS and breastfeeding is under-researched. There's probably a big overlap with obesity here. And there is a link to insufficient glandular tissue, which is interesting, but mainly that's at the anecdotal case study type level. The role of metformin is unclear and there's certainly not enough sufficient evidence to initiate it primarily for the indication of increasing supply. And then type 1 diabetes in the postpartum is characterized by insulin sensitivity and it's really appropriate to relax those strict glycemic targets that women were sticking to during pregnancy. And of course, with type 1, specialist support is vital. And that doesn't just have to be an endocrinologist. Often it's the diabetes educators who are excellent at supporting in that as well. So that brings me to the end of my material. Really happy to take any questions.

Questions [01:17:16.23]

Participant: I do have a question, if that's all right. Thanks so much, Dr. Rassi. That was an incredible talk. So much information. I was just, I don't know if I heard you correctly with the first case, but you mentioned a 30% risk reduction for women who breastfeed in terms of reducing their risk of type 2 diabetes. And that was across both gestational diabetics and non-diabetics. Is that, do they have, I mean, is that a lifetime risk reduction?

Kate: Yeah, it is. Yeah, that's a lifetime risk reduction. Yeah, so the studies will vary in terms of the duration of follow-up, but often they look at sort of 10, 15, 20 years. So there are really significant, again, observational data, but yeah, really significant apparent metabolic benefits. Yeah. And is that, I presume they have to breastfeed for some amount of months. Yeah, again, that varies according to studies. A lot of the studies do show a dose-response relationship for a lot of those. So the longer you breastfeed, the more benefit you get. But again, we have to think about the possibility of reverse causality there, because maybe women who are fitter and healthier prior to pregnancy and were never going to get diabetes, yes, they managed to breastfeed for a long time because they were metabolically healthy to begin with. So I think that, yes, there's caveats that apply when we're looking at that observational research, but I think there are certainly metabolic benefits when you look at it. You absolutely cannot argue with that evidence.

Participant: Kate, I just have a question about the women who've used a low-carbohydrate or even ketogenic diet pre-pregnancy, and then they embark on their pregnancy. If you've got any specific or some guidelines around what they should be aiming for carbohydrate-wise?

Kate: Yeah, I think that's a really, really good point, because I think that's actually something that we're seeing a lot of. I, we have good or fairly good evidence that babies need carbs, really, that to grow a baby, you need carbs. My usual rule of thumb for pregnancy is a minimum of 175 grams of carbs a day, and there's quite good evidence at the dietetic level to support that. We don't have, we worry theoretically about what long-term systemic ketosis might do to a fetus. I usually really empathize with these women, particularly with type 1 diabetics. A lot of them achieve very good control on very low-carbohydrate diets prior to pregnancy, and it's very difficult to just give that away. Often, I think it's about finding a happy medium. 175 grams of carbs is, you know, isn't a huge amount, and my usual go-to with this is to refer these women to a dietician, because we've got some excellent colleagues in the dietetic space who are well-versed in low-carb diets and are well-versed in pregnancy and can help to work out some compromises that, you know, often it's about spreading carbohydrates across the day, it's about choosing really low-GI carbohydrates. But I work with, I see this a lot. I don't think there's an easy answer, but I would certainly discourage strict ketosis, keto-type diets, those ones that, you know, restrict carbs to less than 30 grams a day. I would strongly discourage them both in pregnancy and in lactation. I don't think we all need to be having heaps and heaps of carbs, but my general rule of thumb would be 175 grams a day for pregnancy and 210 grams a day for breastfeeding. And I think that the help of a dietician can be really, really beneficial there.

Participant: This is less of a clinical question, but more sort of policy and practicalities. I mean, obviously we don't have government funding for IBCLC support for women. I know that in the center that I've been working at recently, there's no particular attention to breastfeeding internationally, like specific to women with preexisting diabetes. And their interaction with midwives is actually quite minimal. So they miss out on a lot of education because they're mostly seeing the doctors in the clinic. Being in this space that you're in, are you aware of any advocacy or policy?

Kate: I think you're completely right. It's exactly the same in my institution. I often slipped into the consult because I'm personally interested in it. But there's absolutely nothing formalized in my institution at Monash either. And I think that's a really big gap. I often, if women seem interested, I often refer them to non-diabetes specifics in general, and the ABA community classes and that sort of thing. But I too am yet to find specific resources that are targeted towards women in these groups.

Participant: I can see a question here in the chat. I was wondering about looking at the differences in breastfeeding rates in GDM women when looked at in cultures where breastfeeding is the norm socially, e.g. in the developing world.

Kate: I think that's really good point. Breastfeeding epidemiology is really interesting. If you look, it's the one sort of public health intervention or one metric of public health where in actual fact the developing world is doing a lot better than many OECD countries. And you look at Sub-Saharan Africa, it's the norm to be exclusively breastfeeding and often to exclusively breastfeed right up until, you know, to be continuing to breastfeed right up until sort of four or five years old because of insecurities in water supply and that sort of thing. And the marked public health benefits in countries where gastrointestinal illnesses can easily claim infant lives. In developed countries, it's the complete opposite. So it's the higher socioeconomic and more educated woman and higher incomes that have the higher breastfeeding rates and then that decrements down through the lower decile. So it's a really interesting point. I think that in the developing world, there is probably less gestational diabetes, but the infrastructure in a lot of the underdeveloped countries is such that GDM often isn't screened for. So I work a lot at Monash Health in interpreter only clinics where we see large numbers of refugee women from Sudan and from Afghanistan, and often they're grand multips and they mightn't be in their eighth or ninth pregnancy. And you ask them whether they've had GDM in previous pregnancies and they all reply by the interpreter, I don't know, it was never tested for. So I think the GDM, I think those underdeveloped countries probably have very good breastfeeding data, but I don't know a lot about their GDM rates. And I suspect that it might be a disease that's falling under the radar.

Participant: Hi, I just, I was the one that asked the question. And so I guess I was just asking, and I added a little added to the end just then, but because we don't really live in a breastfeeding normative culture and I work as an IBCLC midwife and so I, in my experience, in my understanding of everything, so much of successful breastfeeding to have that abundant milk supply, for like the genesis to an ongoing is really influenced by breastfeeding support and the amount of breastfeeding sessions or expressing sessions if they're not feeding in those very early few postpartum days. And so often that's not as well done to put it nicely in that immediate postpartum as it should be. And I feel like that has a big influence on not just GDM moms, ongoing abundance of milk supply, but every mom's milk supply who feels they have insufficient milk. And I'm just one, and I guess that was also prompting my thoughts because we often look at everything through a very Western lens. And if those aspects of the GDM, it would be, maybe I feel it might be interesting, you know, if just to see it, those breastfeeding rates even in other cultures.

Kate: That's right. Yeah, I think a lot of the difficulties that women with diabetes or with any significant maternal complications in the postpartum period, a lot of the difficulties that they face are atrogenic. So they're determined by the institutions in which we practise. So the mother-baby separation and the constant visitors to the room and they're not being able to hold baby and that sort of thing. So it will be interesting to look at contexts which may be a more resource poor, but also maybe don't have those institutional factors. I'm not sure that we've got the data because I guess the other thing with all of this data collection is resource intensive. And so I think there's probably just a big knowledge gap there in terms of those under-resourced settings, but I think you raise a really good point.

Pam: I just want to thank you again for such a superb presentation. You have such a formidable knowledge of the research actually that underpins your specialist knowledge. And also I really value the critical lens that you bring as you're reading the research too. And work that's just so pertinent to those of us listening in tonight in our practices. So I'm really grateful to you, Kate. And I think I can say that I'm speaking on behalf of all of those listening in. So thank you very much. My pleasure. No, it was a pleasure to attend and I'm really looking forward to catching up.