Abstract
Introduction
The manuscripts chosen for this yearbook address key controversies in the screening and detection of gestational diabetes mellitus (GDM), glycemic metrics using intermittent and continuous glucose monitoring (CGM) in type 1 diabetes and the theory of beta‐cell regeneration. GDM diagnostic criteria were initially established over 50 years ago to identify women at high risk for the development of diabetes after pregnancy (1). Over the past 20 years, the emphasis has switched more to identifying women and their offspring at increased risk of obstetric and neonatal complications. The 2000–2006 Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study found that oral glucose tolerance tests (OGTT) measures were continuously associated with cord‐blood C‐peptide, large‐for‐gestational‐age, birth weight, cesarean delivery (one standard deviation rise in glucose increased primary cesarean delivery by 8%–11%) and clinically relevant neonatal hypoglycemia (although rates were low at approximately 2%) (2). The observational study design, lack of glycemic thresholds for risk, and the fact that these outcomes are not all clinically relevant has challenged translation into clinical practice. Two studies which sought to inform approaches for GDM screening and diagnosis are included (3, 4). Hillier et al. demonstrated higher rates of GDM diagnosed (16.5 vs 8.5%) for one‐step screening using a 75 g 2‐h OGTT as recommended by the International Association of Diabetes and Pregnancy Study Groups (IADPSG) criteria compared to two‐step screening using a 50 g glucose challenge test, and if applicable, subsequent 100 g 3‐h OGTT using Carpenter‐Coustan diagnostic criteria (3). Overall rates of neonatal morbidity were low, approximately 2% for shoulder dystocia and respiratory distress, 4% for neonatal jaundice, and 8% for neonatal hypoglycemia with no between group differences in large‐for‐gestational‐age neonates, perinatal morbidity, cesarean delivery, or hypertensive disorders. The numbers of women not screened for GDM were low (approximately 6%) in this research study setting, but it was this hard‐to‐reach group who had the highest rates of stillbirth (3). The Comparison of Two Screening Strategies for Gestational Diabetes (GDM2) trial also demonstrated that more women (14.4% vs 4.5%) were diagnosed and treated by IADPSG compared to Carpenter‐Coustan criteria (4). Taken together, the results of these trials suggest a substantial increase in the number of GDM diagnoses using IADPSG criteria without significant difference in neonatal outcomes. It is now time to focus on ensuring that hard‐to‐reach pregnant women are screened for GDM and that the longer‐term maternal and child health outcomes are appropriately addressed in those who are treated.
The UK‐population‐based National Pregnancy in Diabetes (NPID) study describes the outcomes of over 17,000 contemporary pregnancies in those with type 1 and type 2 diabetes (5). It demonstrates the significant impact of maternal characteristics, which were more apparent than variations in clinical care, on adverse pregnancy outcomes and highlights concerning rates of perinatal mortality in type 2 diabetes. Maternal glycemic control and body mass index (BMI) are key modifiable risk factors in both type 1 and 2 diabetes pregnancies. The limited center‐to‐center variation in antenatal glycemia and pregnancy outcomes suggests that interventions to improve maternal glucose are needed across all diabetes maternity clinics (5). The landmark Metformin in women with type 2 diabetes in pregnancy (MiTy) randomized controlled trial provides important data regarding the use of metformin as an adjunct to insulin therapy in type 2 diabetes pregnancies (6). The CONCEPTT and LOIS‐P studies both demonstrate that even for those using insulin pumps and CGM, achieving target glycemic control remains exceptionally challenging for pregnant women with type 1 diabetes (7, 8). Data from LOIS‐P suggest that prandial insulin adjustments are largely inadequate to overcome rising insulin resistance resulting in suboptimal glycaemia from 18 to 30 weeks gestation (8). An interesting application of CGM in mice suggests that large glucose excursions were related to eating episodes (9). CGM profiles showed increased glycemic variability and increased mean amplitude of postprandial glucose excursions in the second half of pregnancy. The importance of psychosocial support for pregnant women is highlighted in a Danish study that shows high levels of anxiety and depression symptoms affecting one in every three to four pregnant women with type 2 diabetes (10). Higher levels of anxiety and/or depressive symptoms were associated with higher HbA1c in late pregnancy.
There are only limited data comparing different glucose monitoring systems in pregnancy. A Danish study comparing intermittently scanned with real‐time CGM found a clinically relevant 5% difference in nocturnal time‐below‐range among Libre users (11). Clinicians should be wary of reducing basal insulin doses based on these data. Another area of controversy in type 1 diabetes pregnancy is whether maternal C‐peptide reflecting endogenous insulin secretion increases in late gestation (12). Data from CONCEPTT trial participants found striking rates of perinatal complications in mothers with first appearance of C‐peptide detected on a highly sensitive immunoassay during late pregnancy. This unexpected finding suggests potential transplacental transfer of C‐peptide from the fetus rather than maternal beta cell hyperplasia.
Key Articles Reviewed for the Article
Hillier TA, Pedula KL, Ogasawara KK, Vesco KK, Oshiro CES, Lubarsky SL, Van Marter J
Davis EM, Abebe KZ, Simhan HN, Catalano P, Costacou T, Comer D, Orris S, Ly K, Decker A, Mendez D, Day N, Scifres CM
Murphy HR, Howgate C, O'Keefe J, Myers J, Morgan M, Coleman MA, Jolly M, Valabhji J, Scott EM, Knighton P, Young B, Lewis‐Barned N; National Pregnancy in Diabetes (NPID) advisory group
Feig DS, Donovan LE, Zinman B, Sanchez JJ, Asztalos E, Ryan EA, Fantus IG, Hutton E, Armson AB, Lipscombe LL, Simmons D, Barrett JFR, Karanicolas PJ, Tobin S, McIntyre HD, Tian SY, Tomlinson G, Murphy KE; MiTy Collaborative Group
O'Malley G, Ozaslan B, Levy C, Castorino K, Desjardins D, Levister C, McCrady‐Spitzer SK, Church MM, Kaur RJ, Reid C, Kremers WK, Doyle FJ, Trinidad MC, Rosenn B, Pinsker JE, Kudva YC, Dassau E
Tundidor D, Meek CL, Yamamoto JM, Martínez‐Bru C, Gich I, Feig DS, Murphy HR, Corcoy R on behalf of the CONCEPTT Collaborative Group
Wuyts C, Simoens C, Pinto S, Philippaert K, Vennekens R
Ásbjörnsdóttir B, Vestgaard M, Do NC, Ringholm L, Andersen LLT, Jensen DM, Damm P, Mathiesen ER
Nørgaard SK, Mathiesen ER, Nørgaard K, Ringholm L
Meek CL, Oram RA, McDonald TJ, Feig DS, Hattersley AT, Murphy HR on behalf of the CONCEPTT Collaborative Group
A Pragmatic, Randomized Clinical Trial of Gestational Diabetes Screening
Hillier TA1,3, Pedula KL1,5, Ogasawara KK4,5, Vesco KK1, Oshiro CES3, Lubarsky SL2, Van Marter J1
1Center for Health Research, Kaiser Permanente Northwest, Portland, Oregon; 2Division of Perinatology, Department of Obstetrics and Gynecology, Northwest Permanente, Kaiser Permanente, Portland, Oregon; 3Center for Integrated Health Care Research and the Division of Perinatology; 4Department of Obstetrics and Gynecology; 5Hawaii Permanente Medical Group, Kaiser Permanente Hawaii, Honolulu
Background
There is no consensus about the best approaches to screen for gestational diabetes mellitus (GDM). This trial aimed to compare one‐step screening with a 75 g 2‐h oral glucose tolerance test (OGTT) as per the International Association of Diabetes and Pregnancy Study Groups (IADPSG) criteria with two‐step screening in which a 3‐h Carpenter‐Coustan 100 g OGTT is performed only as indicated after a positive 50 g glucose challenge test.
Methods
A pragmatic, randomized trial comparing the two screening approaches at two Kaiser Permanente sites in Northwest USA and Hawaii was performed. Clinical teams were aware of the randomization allocations, and GDM management was as per routine clinical care. Primary outcomes were a GDM diagnosis, large‐for‐gestational‐age (LGA) infants, a perinatal morbidity and mortality composite outcome (stillbirth, neonatal death, shoulder dystocia, bone fracture, or nerve palsy), gestational hypertension or preeclampsia, and primary cesarean section.
Results
23,792 participants (age 29.4±5.5 years, BMI 27.4±6.7 kg/m2, 55.5% white ethnicity) were randomized to either one‐step IADPSG (n=11,922) or two‐step GDM screening (n=11,870) with a non‐fasting 50 g glucose challenge test (GCT) followed by 100 g OGTT for those with above target post‐GCT glucose levels (>130 mg/dL or 7.2 mmol/L and >140 mg/dL or 7.8 mmol/L for Northwest and Hawaii, respectively). Adherence was lower for one‐step screening (66 vs 92%) with higher rates of GDM diagnosed (16.5 vs 8.5%) for one‐step versus two‐step screening approaches. There were no differences in rates of LGA (8.9 vs 9.2%), perinatal morbidity, or mortality (3.1 vs 3.0%), gestational hypertension or preeclampsia (13.6 vs 13.5%), or primary cesarean section (24.0 vs 24.6%). Small for gestational age rates (8.5 vs 8.1%) and diabetes medication use (42.6 vs 45.6%) were comparable.
Conclusion
Strict adherence to randomization procedures was difficult with suboptimal adherence to the one‐step screening protocol. However, GDM incidence (16.5 vs 8.5%) and neonatal hypoglycemia rates (9.2 vs 7.5%) were higher using the one‐step IADPSG approach without differences in LGA, perinatal morbidity, cesarean delivery, or hypertensive disorders.
Comment
In women who are screened by either approach for GDM, the pregnancy outcomes and perinatal mortality rates are reassuring (stillbirth 1 in 200 pregnancies, neonatal death 1 in 1,000 pregnancies) with overall low rates of neonatal morbidity (approximately 2% shoulder dystocia, 2% respiratory distress, 4% jaundice, and 8% hypoglycemia). Pregnancy outcomes were less favorable among women who were not screened for GDM, with notably higher stillbirths (33/1450).
Perinatal Outcomes of Two Screening Strategies for Gestational Diabetes Mellitus: A Randomized Controlled Trial
Davis EM, Abebe KZ, Simhan HN, Catalano P, Costacou T, Comer D, Orris S, Ly K, Decker A, Mendez D, Day N, Scifres CM
Department of Medicine, the Department of Obstetrics, Gynecology and Reproductive Science, the Department of Psychiatry, School of Medicine, and the Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, PA; the Department of Obstetrics and Gynecology, Tufts University, Boston, MA; and the Department of Obstetrics and Gynecology, Indiana University School of Medicine, Indianapolis, IN
Background
This trial aimed to compare the prevalence of gestational diabetes (GDM) and its associated obstetric and neonatal complications in women screened by the International Association of Diabetes and Pregnancy Study Groups (IADPSG) and Carpenter‐Coustan diagnostic criteria.
Methods
This was a blinded, randomized, comparative effectiveness trial. All participants with a non‐fasting 50 g glucose challenge test to exclude diabetes were randomized to screening for GDM with either IADPSG or Carpenter‐Coustan criteria. GDM management was as per routine clinical care in 10 obstetric clinics around Pittsburgh, Pennsylvania. The primary outcome was large‐for‐gestational‐age (LGA). Prespecified secondary outcomes were small‐for‐gestational‐age (SGA), cesarean section delivery, and neonatal (clinical hypoglycemia, hyperbilirubinemia, stillbirth, shoulder dystocia, brachial plexus injury) and maternal morbidity (third‐ or fourth‐degree perineal laceration, postpartum hemorrhage, or hypertensive disorder).
Results
921 participants (age 28.7±5.2yr, BMI 26.8±7.0kg/m2, 55.6% white ethnicity) were randomized to either the 2‐h IADPSG 75 g OGTT (n=461) or 3‐h Carpenter‐Coustan 100 g OGTT (n=460) groups. The overall incidence of GDM was 9.7%, with higher incidence (14.4% vs 4.5%, P<0.001) in the IADPSG group. The IADPSG group also had higher rates of diabetes medication use (9.3% vs 2.4%; P<0.001) and higher rates of neonatal morbidity (18.8 vs 13.4%). Rates of LGA (7.7 vs 8.5%), SGA (12.0 vs 12.9%), cesarean section delivery (29.2 vs 27.9%), and maternal morbidity (17.4 vs 17.6%) did not differ between groups. Maternal hypoglycemia (4.3 vs 17.8%) and symptoms (nausea, vomiting, dizziness) following the OGTT were more frequent in the Carpenter‐Coustan group.
Conclusion
The Comparison of Two Screening Strategies for Gestational Diabetes (GDM2) trial demonstrated that more women were diagnosed and treated for GDM by IADPSG criteria but without differences in rates of LGA, cesarean section delivery or maternal morbidity. Neonatal morbidity and healthcare use were higher in the IADPSG group.
Comment
Approximately 1 in 10 women had GDM with a three‐fold increase using IADPSG compared to Carpenter‐Coustan diagnostic criteria. The results of these trials taken together suggest a significant increase in the number of women diagnosed with diabetes but similar neonatal outcomes. However, with LGA rates comparable to the background maternity population, short‐term outcomes for mothers and babies are largely reassuring. We should now focus on ensuring that all women are appropriately screened and on developing evidence‐based interventions to minimize the longer‐term cardiometabolic outcomes for high‐risk women and offspring.
Characteristics and Outcomes of Pregnant Women with Type 1 or Type 2 Diabetes: A 5‐Year National Population‐Based Cohort Study
Murphy HR1, Howgate C2, O'Keefe J2, Myers J3, Morgan M4, Coleman MA5, Jolly M6, Valabhji J7, Scott EM8, Knighton P2, Young B9, Lewis‐Barned N10; National Pregnancy in Diabetes (NPID) advisory group
1Norwich Medical School, University of East Anglia, Norwich, UK; Division of Women's Health, St Thomas' Campus, King's College London, UK; Elsie Bertram Diabetes Centre, Norfolk and Norwich University Hospital, Norwich, UK; 2Clinical Audit and Registries Management Service (CARMS), NHS Digital, Leeds, UK; 3Maternal and Fetal Health Research Centre, St Mary's Hospital, Manchester, UK; 4Department of Obstetrics, Singleton Hospital, Swansea Bay University Health Board, Swansea, UK; 5Princess Anne Hospital, University Hospital Southampton NHS Foundation Trust, Southampton, UK; 6NHS England and NHS Improvement, London, UK; 7NHS England and NHS Improvement, London, UK; Department of Diabetes and Endocrinology, St Mary's Hospital, Imperial College Healthcare NHS Trust, London, UK; Division of Diabetes, Endocrinology and Metabolism, Imperial College London, London, UK; 8Division of Clinical and Population Sciences, Leeds Institute of Cardiovascular and Metabolic Medicine, University of Leeds, Leeds, UK; 9National Diabetes Audit, NHS Digital, Leeds, UK; 10Department of Diabetes and Endocrinology, Northumbria Healthcare NHS Foundation Trust, Northumberland, UK
Background
The higher risk of maternal and neonatal complications in pregnancies with type 1 and 2 diabetes is well established. With increasing numbers of pregnancies complicated by type 1 and 2 diabetes, contemporary population‐based data are needed.
Methods
The was a large population‐based cohort study of clinics participating in the National Pregnancy in Diabetes (NPID) audit from 2014 to 2018. It included 172 maternity clinics in England, Wales, and the Isle of Man. Women with type 1 or 2 diabetes diagnosed prior to pregnancy were included. Differences in characteristics and pregnancy outcomes in women with type 1 vs type 2 diabetes were calculated and compared.
Results
This cohort included 17,156 pregnancy outcomes of which 8,690 (50.0%) and 8,685 (50.0%) were among women with type 1 and type 2 diabetes, respectively. Compared to women with type 1 diabetes, those with type 2 diabetes were more likely to have a higher BMI at their first antenatal visit (BMI ≥30 kg/m2 in 65.0 vs 22.8%; P<0.0001); were more likely to identify as Asian, Black, Mixed, or Other; experienced higher levels of deprivation; were more likely to be treated for hypertension (10.3 vs 3.7%; P<0.0001); and less likely to be on folic acid preconception (22.2 vs 44.1%; P<0.001). However, women with type 2 diabetes were more likely to achieve glycemic targets in early and late pregnancy compared to women with type 1 diabetes. Preterm delivery, large for gestational age neonates, and admission to NICU were more common in pregnancies with type 1 compared to type 2 diabetes. While there was no significant difference between type 1 and 2 diabetes in terms of congenital anomalies (4.5 vs 4.0% for type 1 and 2 diabetes, respectively; P=0.17) or stillbirth (1.0 vs 1.3% for type 1 and 2 diabetes, respectively; P=0.072), neonatal death was significantly higher in type 2 pregnancies compared to type 1 pregnancies (1.1 and 0.7%, respectively; P=0.013).
Conclusions
There are many differences in women with type 1 and 2 diabetes; however, glycemic control and BMI are important modifiable risk factors for adverse pregnancy outcomes in both type 1 and 2 diabetes pregnancies. This large study demonstrates the important impact of maternal characteristics (age, duration of diabetes, BMI, and social disadvantage) on early pregnancy glycemia and all subsequent pregnancy outcomes. It also shows that the impact of these maternal characteristics is more apparent than variations in clinical care, on adverse pregnancy outcomes.
Comment
This large, contemporary cohort highlights that women with type 1 or 2 diabetes continue to experience high levels of pregnancy complications. While women with type 2 diabetes are more likely to achieve target glycemic control compared to women with type 1 diabetes, they are as likely or more likely to experience serious adverse neonatal outcomes (congenital anomaly, stillbirth, or neonatal death). There was surprisingly little center‐to‐center variation in glycemic attainment or obstetric and neonatal outcomes after adjustment for important maternal confounders.
Metformin in Women with Type 2 Diabetes in Pregnancy (MiTy): A Multicentre, International, Randomised, Placebo‐Controlled Trial
Feig DS1,2,3, Donovan LE4,5, Zinman B1,2, Sanchez JJ6, Asztalos E7, Ryan EA8, Fantus IG1,2,3, Hutton E9, Armson AB10, Lipscombe LL1,11, Simmons D1, Barrett JFR6,7, Karanicolas PJ6,7, Tobin S6, McIntyre HD13, Tian SY6, Tomlinson G1,14, Murphy KE1,2,3,15; MiTy Collaborative Group
1Department of Medicine, University of Toronto, Toronto, ON, Canada; 2Lunenfeld‐Tanenbaum Research Institute, Toronto, ON, Canada; 3Sinai Health System, Mount Sinai Hospital, Toronto, ON, Canada; 2Cumming School of Medicine, Department of Medicine, Department of Obstetrics and Gynecology, University of Calgary, Calgary, AB, Canada; 5Alberta Children's Hospital Research Institute, Calgary, AB, Canada; 6Sunnybrook Research Institute, Toronto, ON, Canada; 7Sunnybrook Health Sciences Centre, Toronto, ON, Canada; 8University of Alberta, Edmonton, AB, Canada; 9McMaster University Hamilton, Canada; 10Dalhousie University, Halifax, NS, Canada; 11Women's College Hospital, Toronto, ON, Canada; 12Western Sydney University, Sydney, NSW, Australia; 13Mater Research, University of Queensland, South Brisbane, QLD, Australia; 14Department of Medicine, University Health Network, Toronto, ON, Canada; 15Department of Obstetrics and Gynecology, University of Toronto, Toronto, Canada
Background
Metformin remains an important first‐line treatment for type 2 diabetes outside of pregnancy. However, given that metformin is known to cross the placenta, the risks and benefits of metformin use in the treatment of type 2 diabetes in pregnancy are less clear.
Methods
This was a multicenter, international, masked, placebo‐controlled trial of metformin as an adjunct to insulin therapy in women with type 2 diabetes in pregnancy. Participants were recruited from 29 centers across Canada and Australia. They were randomized to receive either metformin 1 g twice a day or placebo between 6 weeks and 22 weeks plus 6 days gestation. The primary outcome was a composite of fetal and neonatal outcomes including any of the following: pregnancy loss, preterm birth, birth injury, moderate or severe respiratory distress syndrome, neonatal hypoglycemia, or NICU admission >24 hours. Relative risks (RR) and 95% confidence intervals (CI) were calculated for participants randomized to metformin vs placebo using an intention‐to‐treat analysis.
Results
A total of 502 participants were randomized, with 253 randomized to metformin and 249 to placebo. There was no difference in the primary composite outcome between participants randomized to metformin compared to placebo (RR 1.02 [95% CI 0.83, 1.26]). However, neonates of participants randomized to metformin had a lower birthweight (mean weight 3156±742 vs 3375±742 g; P=0.0016), lower adiposity measures, were less likely to be extremely large for gestational age (RR 0.58 [95% CI 0.34, 0.97]) but more likely to be small for gestational age (RR 1.96 [95% CI 1.10, 3.64]). Additionally, participants randomized to metformin had less overall weight gain (7.2 vs 9.0 kg; P<0.0001), a lower HbA1c in late pregnancy (5.9 vs 6.1%; P=0.015), required less insulin in late pregnancy (109.8 vs 155.3 units/day; P<0.0001), and were less likely to have a cesarean section (RR 0.85 [95% CI 0.73, 0.99]).
Conclusions
There was no difference in the primary composite outcome in participants randomized to metformin compared to placebo. However, there were many important differences in secondary outcomes including less maternal gestational weight gain, lower insulin doses and HbA1c in late pregnancy, fewer cesarean sections, lower neonatal adiposity, fewer extreme large for gestational age neonates and more small for gestational age neonates in participants randomized to metformin and their neonates compared to those randomized to placebo.
Comment
While there were many important benefits of metformin, these must be weighed against the increased risk of small for gestational age neonates. Since many participants in both used metformin in the first trimester (mean gestational age at randomization 16.5‐ and 16.4‐weeks gestation in the metformin and placebo groups, respectively), first trimester use of metformin is not evaluated in this trial. Additionally, metformin monotherapy was not assessed since all participants were using metformin as an adjunct to insulin therapy. Future studies are required to understand the etiology of the higher risk of small for gestational age neonates as well as to follow up children of the Mity participants.
Longitudinal Observation of Insulin Use and Glucose Sensor Metrics in Pregnant Women with Type 1 Diabetes Using Continuous Glucose Monitors and Insulin Pumps: The LOIS‐P Study
O'Malley G1, Ozaslan B2, Levy C1, Castorino K3, Desjardins D4, Levister C1, McCrady‐Spitzer SK4, Church MM3, Kaur RJ4, Reid C4, Kremers WK4, Doyle FJ2, Trinidad MC5, Rosenn B6, Pinsker JE3, Kudva YC4, Dassau E2
1Icahn School of Medicine at Mount Sinai, Division of Endocrinology, Diabetes and Bone Diseases, New York, NY; 2Harvard John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA; 3Sansum Diabetes Research Institute, Santa Barbara, CA; 4Mayo Clinic, Division of Endocrinology, Diabetes, Metabolism & Nutrition, Rochester, MN; 5Mayo Clinic, Department of Obstetrics and Gynecology, Rochester, MN; 6Icahn School of Medicine at Mount Sinai, Mount Sinai West Hospital, Division of Obstetrics and Maternal‐Fetal Medicine, New York, NY
Background
Previous studies, mostly retrospective and with small samples, described rising prandial and basal insulin doses until week 9, followed by a nadir week 16, and then progressive increases until delivery around gestational week 37. The aim of this study was to prospectively describe gestational changes in continuous glucose monitoring (CGM) glucose levels, pump settings, and total daily insulin doses in pregnant women with type 1 diabetes (T1D) using insulin pump therapy.
Methods
Twenty‐five pregnant women with T1D (aged 30.2±4.8 years, mean HbA1c 6.6±0.9%, weight 77.7±15.9 kg, BMI 28.6±6.2 kg/m2 at 11.2±3.9 weeks gestation) from three U.S. sites (Mount Sinai, Mayo Clinic, Sansum Diabetes Research Institute) were included. Linear mixed‐effects regression models were used to characterize changes in glycemia (HbA1c, CGM metrics) and insulin doses across pregnancy compared to a reference period at the end of the first trimester (12–14 weeks gestation).
Results
Mean HbA1c was 6.5±0.9%, 6.2±0.8%, 6.3±0.7% during the first, second, and third trimesters. Despite near‐optimal HbA1c, CGM time in range (TIR) (63–140 mg/dL) per participant across pregnancy was 59±14% with no differences for daytime versus nighttime daytime (59±16% vs 61±13% respectively) and higher than recommended CGM time above range (time above range [TAR] >140mg/dl) of 38±15% throughout pregnancy. Between 18 and 30 weeks gestation, there was significant deterioration in maternal glycemia with lower CGM TIR and higher CGM time above range (TAR >140 mg/dl) compared to the reference period at 12–14 weeks. Basal insulin doses were higher before 12 and after 24 weeks gestation. Prandial insulin doses were strengthened from 22 weeks onwards; 54% increase in basal insulin and 60% increase in bolus insulin doses. 7/25 (28%) women achieved the recommended T1D pregnancy TIR >70%. Median birth weight was 3700 g (range 2135–4345 g), and 14/25 (56%) infants were large for gestational age with 10/25 (40%) admitted to the neonatal intensive care unit.
Conclusion
Prandial insulin adjustments were inadequate for gestational changes in insulin sensitivity. Time in range increased significantly only at the end of the third trimester. Hybrid closed loop systems may be needed to improve insulin dosing in women with T1D using insulin pump therapy during pregnancy.
Comment
Clinicians should be aware of the potential for deterioration in glucose levels from mid-gestation and advise women that strict attention to dietary intake and escalation of prandial insulin doses is needed to prevent rising hyperglycemia from 18 to 30 weeks gestation.
Continuous Glucose Monitoring Time‐in‐Range and HbA1c Targets in Pregnant Women with Type 1 Diabetes
Tundidor D1,2, Meek CL3,4, Yamamoto JM5,6, Martínez‐Bru C7, Gich I8,9, Feig DS10, Murphy HR11,12, Corcoy R1,2,13,14, on behalf of the CONCEPTT Collaborative Group
1Institut de Recerca de l'Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 2Department of Medicine, Universitat Autònoma de Barcelona, Barcelona, Spain; 3Institute of Metabolic Science, University of Cambridge, Cambridge, UK; 4Cambridge Universities NHS Foundation Trust, Cambridge, UK; 5Department of Internal Medicine, Faculty of Medicine, University of Manitoba, Winnipeg, Canada; 6Department of Medicine, Cumming School of Medicine, University of Calgary, Calgary, Canada; 7Laboratory Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 8Department of Clinic Epidemiology and Public Health, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 9CIBER Epidemiología y Salud Pública (CIBERESP), Instituto de Salud Carlos III, Madrid, Spain; 10Mount Sinai Hospital, Sinai Health System, Department of Medicine, University of Toronto, Lunenfeld‐Tanenbaum Research Institute, Toronto, Canada; 11Norwich Medical School, University of East Anglia, Norwich, UK; 12School of Life Course Sciences, King's College London, London, UK; 13Servei d'Endocrinologia i Nutrició, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 14CIBER‐BBN, Madrid, Spain
Background
Continuous glucose monitoring (CGM) has been shown to improve maternal glycemic control and reduce neonatal complications when used in real time. The recently published international consensus on time in range for CGM in type 1 diabetes pregnancies recommends >70% time in pregnancy target range (3.5 to 7.8 mmol/L) with <25% above range and <4% time below range. However, little data are published regarding how frequently these targets are obtained.
Methods
This was a secondary analysis of the Continuous Glucose Monitoring in Pregnant Women with Type 1 Diabetes (CONCEPTT) trial. Glycemic control was assessed in early pregnancy and at 24‐ and 34‐weeks gestation with 6‐day CGM profiles. The primary outcome was the percentage of participants who achieved the international consensus time in range targets for type 1 diabetes pregnancies. The association between glycemic control and secondary outcomes was assessed using logistic regression analyses.
Results
The number of participants included in this study was 221 in early pregnancy, 197 at 24 weeks, and 172 at 34 weeks gestation. In early pregnancy, only 7.7% of women achieved the time in range target of >70%. This increased as pregnancy progressed with 10.2% and 35.5% of women at 24 and 34 weeks, respectively, achieving time in range targets. After adjustment, attaining the time below range target in the first trimester was associated with an increased risk of pre‐eclampsia and at 24 weeks was associated with an increased risk of neonatal hypoglycemia.
Conclusion
Only a limited number of participants in the CONCEPTT trial were able to achieve time in range targets recommended by the international consensus guidelines.
Comment
This secondary analysis of CONCEPTT highlights that this is true for both HbA1c and CGM targets. Since this was a secondary analysis of a randomized trial of real‐time CGM vs capillary glucose monitoring alone, half of participants in this analysis were not using real‐time CGM, which has been shown to improve time in range. However, even taking this into account, few women achieved glycemic targets. While the number of women reaching time in range targets increased as pregnancy progressed, even at 34 weeks, only 35.5% of participants achieved time in range targets.
Continuous Glucose Monitoring During Pregnancy in Healthy Mice
Wuyts C1,2, Simoens C1,2,3, Pinto S1,2, Philippaert K1,2, Vennekens R1,2
1Department of Cellular and Molecular Medicine, Laboratory of Ion Channel Research, TRP Research Platform Leuven, KU Leuven, Leuven, Belgium; 2VIB Center for Brain & Disease Research, Leuven, Belgium; 3Department of Chronic Diseases, Metabolism and Ageing, Laboratory of Clinical and Experimental Endocrinology, KU Leuven, Leuven, Belgium
Background
Data on the metabolic adaptations describing changes in insulin sensitivity, secretion, and glucose fluxes in normal healthy pregnancy are limited. The authors applied continuous glucose monitoring (CGM) to healthy, free‐living mice to examine gestational changes in glucose excursions and glycemic variability associated with spontaneous feeding episodes during pregnancy.
Methods
Two groups of female mice with (transmitter group n=14) and without CGM (sham group n=17) and fed ad libitum were compared. Sensors were inserted into the aortic arch and generally well tolerated albeit with some failures related to biting/scratching. Oral glucose tolerance tests (OGTT) were performed using an Aviva Accu‐Chek glucometer with glucose samples measured at 0, 15, 30, 60, and 120 min after glucose loading before pregnancy at embryonic day 8 (E8), 16 (E16), and 2 days postpartum (2pp). CGM and OGTT data were compared as well as CGM metrics before, during, and after pregnancy. Telemetry and video monitoring allowed for detailed analysis of prandial and non‐prandial glycemic excursions.
Results
Fasting glucose and area under the curve (AUC) of the OGTT were higher during pregnancy. There was a good correlation between OGTT and CGM data (R2=0.81) with far more glycemic information provided by CGM. Average glucose levels increased from before pregnancy toward mid‐gestation (E10) and then decreased until delivery. All large glucose excursions were related to eating episodes. Glucose uptake and clearance were both faster before and after pregnancy than during gestation. MAGE reflecting short‐term glucose variability was higher before and after pregnancy, compared to during pregnancy. Hypoglycemic events were sparse, but three mice experienced short hypoglycemic episodes toward the end of pregnancy. Mid‐gestation (E10) postprandial day‐time MAGE and iAUC are significantly lower compared to early or late gestation. Daytime glucose clearance is notably higher postprandially in late compared to early or mid‐gestation. Glucose variability (standard deviation and interquartile range) decreased in early pregnancy and increased thereafter with maximal variability in the 24 h before delivery. Postprandial excursions were larger, with higher amplitude AUC and faster kinetics. They had a longer time to peak and longer duration compared to non‐prandial hyperglycemic excursions. These larger excursions are the ones that seem to fluctuate more during pregnancy.
Conclusion
This study demonstrates distinctive circadian patterns of glucose excursions during normal healthy pregnancy. In mice as in women, mean glycemia increased to mid‐term and decreased toward the end of pregnancy. CGM demonstrates a complex relation between circadian rhythm (opposite in mice who are predominantly nocturnal feeders), glycemic control, and the stage of pregnancy.
Comment
More detailed CGM data are needed to understand the evolution of glycemia across pregnancy in healthy women and in those with GDM.
Prevalence of Anxiety and Depression Symptoms in Pregnant Women with Type 2 Diabetes and the Impact on Glycaemic Control
Ásbjörnsdóttir B1,2, Vestgaard M1,2, Do NC1,2, Ringholm L1,3, Andersen LLT4, Jensen DM4,5,6, Damm P1,2, Mathiesen ER1,2
1Center for Pregnant Women with Diabetes, Departments of Endocrinology and Obstetrics, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark; 2Department of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark; 3Steno Diabetes Center Copenhagen, Gentofte, Denmark; 4Department of Gynaecology and Obstetrics, Odense University Hospital, Odense, Denmark; 5Steno Diabetes Center Odense, Odense University Hospital, Odense, Denmark; 6Department of Clinical Research, University of Southern Denmark, Odense, Denmark
Background
Outside of pregnancy, anxiety and depression are more common in people living with type 2 diabetes compared with the non‐diabetes population. However, less is known about the prevalence of anxiety and depression symptoms in type 2 diabetes during pregnancy.
Methods
This was a pre‐specified analysis of a prospective cohort study of women with type 2 diabetes in pregnancy in Denmark. Participants were asked to complete the Hospital Anxiety and Depression Scale questionnaire in both early and late pregnancy. This questionnaire has two scales each with seven questions, with one measuring anxiety symptoms and the other depressive symptoms. A total score of ≥ 8 was used to represent the presence of anxiety and/or depressive symptoms.
Results
A total of 90 women with type 2 diabetes and 88 without diabetes were included. Compared to women without diabetes, those with type 2 diabetes were more likely to have both anxiety symptoms (36 vs 6%; P<0.001) and depressive symptoms (14 vs 2%; P=0.003) in early pregnancy. Similarly, in late pregnancy, women with type 2 diabetes were more likely to have anxiety symptoms (31 vs 11%; P=0.002) and depressive symptoms (23 vs 4%; P=0.002) compared to women without diabetes. In late pregnancy, women with anxiety and/or depressive symptoms had a higher HbA1c (6.1 vs 5.8%; P=0.04).
Conclusions
Women with type 2 diabetes were more likely to have anxiety and/or depressive symptoms in both early and late pregnancy. In late pregnancy, women with anxiety and/or depressive symptoms had a higher HbA1c compared to those without symptoms.
Comment
Outside of pregnancy, both anxiety and depression have been associated with hyperglycemia (13, 14). This study demonstrated not only a high prevalence of anxiety and/or depressive symptoms in people with type 2 diabetes in pregnancy but also that these symptoms were associated with a higher HbA1c in late pregnancy. The study has some limitations, including the relatively small sample size and low prevalence of anxiety and/or depressive symptoms in women without diabetes. However, it highlights the high prevalence of these symptoms in women with type 2 diabetes in pregnancy and the potential effects on diabetes management in this population.
Comparison of Glycemic Metrics Measured Simultaneously by Intermittently Scanned Continuous Glucose Monitoring and Real‐Time Continuous Glucose Monitoring in Pregnant Women with Type 1 Diabetes
Nørgaard SK1,2, Mathiesen ER1,2,3, Nørgaard K3,4, Ringholm L1,2
1Center for Pregnant Women with Diabetes, Rigshospitalet, University Hospital of Copenhagen, Copenhagen, Denmark; 2Department of Endocrinology, Rigshospitalet, University Hospital of Copenhagen, Copenhagen, Denmark; 3Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark; 4Steno Diabetes Center Copenhagen, Gentofte, Denmark
Background
Real‐time continuous glucose monitoring (RT‐CGM) has been shown to improve pregnancy outcomes in women with type 1 diabetes in pregnancy. However, the role of intermittently scanned glucose monitoring (IS‐CGM) in type 1 diabetes pregnancy remains less clear. Previous observational data have demonstrated that women using IS‐CGM had a higher time below range compared to women using RT‐CGM. However, no studies have compared RT‐ vs IS‐CGM in pregnant women using concomitant RT‐ and IS‐CGM.
Methods
This was a prospective cohort study of women with type 1 diabetes during pregnancy. Women were included if they were using or starting IS‐CGM as part of routine care. In addition to wearing an IS‐CGM, a masked RT‐CGM sensor was placed on the abdomen for 7 days. Each IS‐CGM glucose measurement was matched with the closest glucose measurement from the masked RT‐CGM sensor. The difference in mean sensor glucose between the IS‐ and RT‐CGM over 24 h and overnight (midnight to 6:00) as well as the difference in time below range overnight were compared.
Results
Of the 28 women with type 1 diabetes who agreed to participate, 20 had both IS‐ and RT‐CGM data and were included in the analysis. There was no significant difference in mean sensor glucose over 24 h (median mean sensor glucose 7.3 vs 6.8 mmol/L for IS‐ and RT‐CGM respectively; P=0.50). Overnight, there was no significant difference in mean sensor glucose (median mean sensor glucose 6.7 vs 7.0 mmol/L for IS‐ and RT‐CGM, respectively; P=0.35). However, overnight the time below range was significantly higher using IS‐CGM compared to masked RT‐CGM (5.6 vs 0%; P=0.005).
Conclusions
While mean sensor glucose over 24 h and overnight were similar between IS‐CGM and RT‐CGM, time below range overnight was higher when assessed using IS‐CGM compared to masked RT‐CGM in early pregnancy.
Comment
The results of this study indicate that there are clinically relevant differences between IS‐ and RT‐CGM in time below range and other CGM metrics commonly used to guide insulin adjustments during pregnancy. A recently published randomized controlled trial in people with type 1 diabetes outside of pregnancy found that RT‐CGM was superior to IS‐CGM in improving time in range and reducing hypoglycemia (15). Additional studies are needed to determine the role of IS‐CGM in pregnant women with type 1 diabetes.
Reappearance of C‐Peptide During the Third Trimester of Pregnancy in Type 1 Diabetes: Pancreatic Regeneration or Fetal Hyperinsulinism?
Meek CL1,2,3, Oram RA4, McDonald TJ4,5, Feig DS6,7, Hattersley AT4, Murphy HR2,8,9 on behalf of the CONCEPTT Collaborative Group
1The Wellcome‐MRC Institute of Metabolic Science, University of Cambridge, Cambridge, UK; 2Wolfson Diabetes and Endocrinology Clinic, Cambridge University Hospitals, Cambridge, UK; 3Department of Clinical Biochemistry, Cambridge University Hospitals, Cambridge, UK; 4Department of Diabetes Research, University of Exeter, Royal Devon and Exeter Hospital, Exeter, UK; 5Academic Department of Blood Sciences, Royal Devon and Exeter NHS Foundation Trust, Exeter, UK; 6Leadership Sinai Centre for Diabetes, Mount Sinai Hospital, Department of Medicine, University of Toronto, ON, Canada; 7Lunenfeld‐Tanenbaum Research Institute, Toronto, ON, Canada; 8Department of Women and Children's Health, King's College London, St. Thomas' Hospital, London, UK; 9Norwich Medical School, Faculty of Medicine and Health Sciences, University of East Anglia, Norwich, UK
Background
There is controversy about whether endogenous insulin secretion increases in late gestation in type 1 diabetes pregnancy. This study used samples from the continuous glucose monitoring in type 1 diabetes pregnancy trial (CONCEPTT) participants to assess longitudinal patterns of maternal C‐peptide concentration and associations with glycemia during pregnancy, cord blood C‐peptide, and subsequent neonatal outcomes.
Methods
C‐peptide concentration was measured on serum samples at 12, 24, and 34 weeks from 127 CONCEPTT trial participants. Maternal serum C‐peptide was measured using a highly sensitive direct electrochemiluminescence immunoassay with a mouse monoclonal anti‐C‐peptide antibody. Cord C‐peptide was measured using a standard solid‐phase competitive electrochemiluminescent immunoassay. Maternal glycemia (CGM and HbA1c) and obstetric and neonatal outcomes were obtained from the CONCEPTT database.
Results
Three longitudinal patterns of C‐peptide trajectory were identified. Pattern 1, which included women with undetectable C‐peptide at all time‐points, was the most common (74/127; 58.3%). Pattern 2 included women with detectable C‐peptide at 12 weeks and was less common (22/127; 17.3%). Pattern 3 included women with undetectable C‐peptide at 12 and 24 weeks, with appearance of detectable C‐peptide for the first time at 34 weeks gestation, (31/127; 24.4%). As expected, women with detectable C‐peptide during pregnancy (pattern 2) were approximately three BMI points leaner, older, and had shorter duration of T1D than those with undetectable C‐peptide throughout pregnancy (pattern 1). The group with C‐peptide first detected at 34 weeks (pattern 3) had significantly lower CGM time‐in‐range (TIR) (3.5–7.8 mmol/L); 45% TIR compared to 52 and 55% in groups 1 and 2, significantly higher cord blood C‐peptide concentration (1319 vs 718 pmol/L; P=0.007), and striking rates of neonatal morbidity (large for gestational age 90% vs 59%, neonatal intensive care unit admission, 45% vs 23%, hypoglycemia 42% vs 13%, respiratory distress 13% vs 3%).
Conclusion
This study suggests that the novel appearance of C‐peptide at 34 weeks gestation is most likely of fetal origin due to its associations with maternal hyperglycemia, elevated cord blood C‐peptide, and neonatal complications related to sustained fetal pancreatic hyperinsulinism.
Comment
More data are needed to better understand transfer of insulin, C‐peptide, or related immunoreactive fragments across the placenta. If confirmed in larger studies, the early biochemical identification of hyperinsulinemic offspring could provide new opportunities for detection of the highest risk infants, and may help to inform birth planning decisions in type 1 diabetes pregnancies.
Footnotes
Author Disclosure Statement
HM is on the Medtronic (insulin pump and CGM manufacturer) European scientific advisory board. There are no other competing financial interests.