Abstract
The main objective of this article is to provide a comprehensive review of continuous glucose monitor (CGM) use in pregnant women with type 1 diabetes (T1D) from the CONCEPTT study including subanalyses. Literature search was accessed through MEDLINE (1966–September 2023) using the key terms: CONCEPTT, pregnancy, women, T1D, and CGM with limitations set to distinguish human subjects written in English. A total of 17 publications including one main clinical trial and 15 subanalyses have been published to date regarding the use of CGM in pregnant women with T1D which were conducted by a research group identified as the CONCEPTT Collaborative Group. While advances in maternal care have resulted in safer pregnancy for both the mother and child, women with preexisting T1D and pregnancy still experience higher rates of complications both in the short and long term. The use of CGM in pregnancy has not been studied extensively until more recently. The CONCEPTT clinical trial was a landmark study that involved several subanalyses. The main trial proved that CGM use in T1D pregnancy resulted in less hyperglycemia in the third trimester, reduced large for gestational age (LGA, >90th percentile), reduced neonatal intensive care unit admissions lasting longer than 24 h, and reduced neonatal hypoglycemia. Although subanalyses showed a variety of results including ‘inconclusive’ due to lack of prespecification, it is believed that CGM in T1D during pregnancy is to be recommended and used for overall improved outcomes.
Introduction
A continuous glucose monitor (CGM) is a sophisticated medical system that provides a continuous stream of glucose data, allowing individuals with diabetes to visualize their glucose levels in real-time and on demand. A CGM is a significant improvement over traditional self-monitoring blood glucose (SMBG) testing, which only provides a single glucose value at a given time, and has proven substantially better accuracy over standard glucometers (1). A CGM system consists of three parts: a sensor, a transmitter, and a monitoring device. The sensor is inserted subcutaneously to measure glucose concentrations in the interstitial fluid (typically every 1–5 min), generating between 288 and 1440 glucose readings daily, and must be replaced every 7–14 days. The transmitter, connected to the sensor, sometimes even built into one unit, allows the sensor to send real-time glucose readings wirelessly (via Bluetooth) to a monitoring device or smartphone (2). Since the introduction of CGM in type 1 diabetes (T1D) in 1999, CGMs have improved convenience and quality of life for patients with diabetes. CGM has also branched out into optimizing care in patients with type 2 diabetes (T2D) and gestational diabetes (GDM). The increase in awareness, availability, affordability, and ease of use of CGM devices have exponentially improved the management of diabetes by capturing real-time glucose data. A CGM can provide a wealth of information to both the patient and healthcare provider, allowing for a comprehensive review of how diet, exercise, medication, insulin regimen, stress, hydration, consistent sleep, and other factors affect blood glucose levels. It will also notify the patient and the people the patient shares data with regarding hypoglycemia and hyperglycemia with alerts and alarms (3, 4). Given the great advances in maternal care, most women can now expect to have a healthy and uncomplicated pregnancy, and deliver a healthy baby at the end of nine months (1). However, pregnancy still presents significant challenges, associated with maternal and fetal risk in women with preexisting T1D and T2D, and gestational diabetes. It is well documented that pregnant women with T1D have increased glucose levels, glucose variability, time in hypoglycemia, impaired hypoglycemia awareness, and, as a result, spend less time in target glucose range for pregnancy than pregnant women with T2D according to the tighter glucose control range targets in pregnancy (3). Hence, pregnant women with T1D continue to be a high-risk population and maternal euglycemia remains difficult to achieve which has detrimental effects on both neonatal and maternal obstetric outcomes (5). T1D complicates pregnancy and requires meticulous management to minimize the risk of complications. Complications include but are not limited to miscarriage, preeclampsia, preterm birth, congenital anomalies, LGA, and NICU admission (6). The CONCEPTT study was a landmark prospective, multicenter, international, randomized controlled trial to evaluate CGM use in pregnant women with T1D (5). The results from CONCEPTT, which was completed in 2016, continue to impact guidelines and clinical practice today, most importantly developing consensus guideline recommendations for time in range (TIR) targets for pregnancy (7). Prior to CONCEPTT it was unclear whether continuous glucose monitoring in pregnant women with T1D improved maternal and neonatal outcomes. At the time the trial began in 2013, there were only a few widely available CGM systems, including the Medtronic Enlite 2 sensor with Guardian REAL-Time or MiniMed MiniLink system (MARD 13.6%) (8) and the Dexcom G4 CGM (MARD 13%) (9) and there were no guidelines on how to interpret CGM data nor what CGM metrics are most important. More and more women were using CGM in pregnancy but there lacked clinical evidence supporting CGM use to improve clinical outcomes. The aim of CONCEPTT was to evaluate the effectiveness of CGM used both preconception and during pregnancy in improving maternal glucose control and subsequently obstetric and neonatal health outcomes (5). The CONCEPTT collaborative group published the main article in 2017 (5); however, there continues to be a plethora of subanalyses (10–25) to glean more insight into the optimal care of this special population that is underrepresented in research. Our objective is to provide a comprehensive review of the findings of the CONCEPTT collaborative group’s primary analysis (5) and subanalyses (10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25) so as to provide a summary of all of the group’s published findings presented in one publication. We searched PubMed for articles published with restrictions to English language with the search terms ‘type 1 diabetes’ and ‘pregnancy’ and ‘continuous glucose monitor’ and ‘women’ and ‘CONCEPTT’. We identified 17 articles including a protocol summary (10), the main article (5) and 15 subanalyses (11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25). The subanalyses reviewed seven general concepts: (i) are insulin pumps better than MDI in neonatal and maternal outcomes? (ii) were dietary patterns different in the treatment arms of the CONCEPTT? (iii) are there associations between maternal glucose metrics and maternal and neonatal outcomes? (iv) is maternal fear of hypoglycemia a factor in glucose control? (v) What tools are best for assessing neonatal growth and fetal size? (vi) what are the economic implications of CGM in pregnancy with type 1 diabetes, and (vii) are there alternative biomarkers that affect maternal and neonatal outcomes? A summary of the CONCEPTT publications is shown in Table 1.
Summary of CONCEPTT publications including secondary and prespecified subanalyses (5, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25).
Publication | Article Type | Aim | Results Summary |
---|---|---|---|
1. Feig et al. BMC Pregnancy Childbirth 2016 (10) | Study protocol (n = 324, 110 planning pregnancy and 214 pregnant women) | Provide detail on study design and outcome measures | Study protocol: multicenter, open-label, randomized, controlled trial of women with T1D either planning pregnancy with HbA1c 7.0–10.0% or early pregnancy (<13 weeks 6 days) with HbA1c 6.5–10.0%. 324 women will be recruited (110 planning pregnancy, 214 pregnant). Analyses performed according to intention to treat. Primary outcome is change in HbA1c from baseline to 24 weeks or conception in women planning pregnancy, baseline to 34 weeks gestation during pregnancy. Secondary outcomes include maternal hypoglycemia, CGM time in, above and below target, glucose variability measures, and maternal and neonatal outcomes. |
2. Feig et al. Lancet 2017 (5) | Main article | Publish main results in primary and secondary outcome | The recommended percentages of glucose readings in the target range for pregnant women with T1D were >70% time for TIR of 63–140 mg/dL, <25% time for TAR of>140 mg/dL, and <4% time TBR of <63 mg/dL. (These values were based on the CONCEPTT trial and a cohort study (Kristensen et al., Diabetologia 2019 (32)) CGM group 5% less hyperglycemia in third trimester, reduced LGA, less NICU admissions and less neonatal hypoglycemia |
3. Feig et al. Diabetes Care (2018) (11) | Prespecified analysis of secondary outcome measure | Compare pumps vs MDI outcomes including glycemic control, QOL, and pregnancy outcomes in CONCEPTT trial | Results demonstrated mixed findings. Neither the basal insulin type nor the type of insulin pump was reported. Authors concluded that future studies designed primarily to compare pump versus MDI would be needed to better determine superiority of one method over the other. |
4. Yamamoto et al. Diabet Med 2019 (12) | Secondary analysis of neonatal hypoglycemia (n = 225; 33 continued CGM until delivery) | (i) Analyze incidence of neonatal hypoglycemia and its associated outcomes at birth and (ii) examine the association between maternal glycemic control and neonatal outcomes (C-peptide level in cord blood), birthweight percentile, and anthropometry measures) | Neonatal hypoglycemia was associated with fetal hyperinsulinemia in the 24 h prior to delivery, birthweight >97th percentile, and infant adiposity. During the second and third trimesters, maternal hyperglycemia (HbA1C ~0.4% higher and TIR 6–7% lower) was associated with neonatal hypoglycemia. |
5. Murphy et al. Diabet Med 2019 (13) | Secondary analysis to model economic outcome (n = 1441 pregnant women with T1DM in England per year) | Model potential cost savings from use of RT-CGM in pregnant women with T1DM | Real-time continuous glucose monitoring (RT-CGM) improves neonatal health outcomes, with fewer LGA infants, fewer NICU admissions and a shorter neonatal length of hospital stay. The approximately three-fold higher costs of RT-CGM use, compared with SMBG, are offset by substantial cost savings, mainly through reductions in NICU admissions and a shorter duration of NICU stay. |
6. Neoh et al. Diabet Med 2020 (14) | Prespecified secondary subgroup analysis (n = 156; 93 pregnant and 63 planning pregnancy) | Describe the dietary intake of women who were pregnant or planning (UK population only) | Overall caloric and carbohydrate intake was similar in both groups. Fat and nonrecommended carbohydrate sources (sugars, preserves, biscuits, cakes) were high in both groups. Fiber intake as well as fruit and vegetable intakes were inadequate in both groups. Micronutrient intake in this UK population was insufficient in 1 of 4 participants. |
7. Neoh et al. Diabetes Care 2020 (15) | Secondary subgroup analysis of 113 CONCEPTT trial participants (n = 93 pregnant women; 38 using an insulin pump and 55 using MDI) | Evaluate the dietary patterns of those using an insulin pump versus MDI | There were no significant differences between groups in total caloric intake, macronutrient intake (carbohydrate, protein, fat, or fiber), or snacking behaviors. Dietary patterns were suboptimal in both groups. |
8. Scott et al. Diabetes Care 2020 (16) | Secondary analysis of CGM profiles at certain time periods and their associated outcomes (n = 100 RT-CGM and n = 100 SMBG) | Determine if temporal glucose profiles differed in: RT-CGM vs SMBG, pump vs MDI, LGA vs normal weight, via functional data analysis | (i) Women using RT-CGM were able to achieve better daytime glucose control than SMBG group, (ii) women with LGA infant had significantly higher glucose at all times, (iii) functional data analyses identified glucose patterns that are otherwise undetectable by standard glucose metrics, (iv) pump users had higher daytime glucose at week 24, but not at baseline or 34 weeks gestation |
9. Meek et al. Diabetes Care 2021 (17) | Prespecified secondary analysis of alternative biomarkers for measuring glycemic control (n = 157) | Assess the predictive performance of traditional biomarkers (HbA1C), CGM, and alternative biomarkers in predicting maternal and neonatal outcomes | Neither laboratory markers nor CGM metrics were able to provide a strong prediction of any pregnancy outcome. In pregnant women with T1D, the use of alternative laboratory markers did not appreciably increase the AUROC for prediction of suboptimal pregnancy outcomes beyond HbA1c, which is already widely available, or CGM metrics, such as TIR and TAR. HbA1c was consistently associated with pregnancy outcomes, suggesting that despite the known limitations of HbA1c for assessing antenatal glycemia, it is still a critically important biomarker for obstetric and neonatal health outcomes. |
10. Meek et al. BMC Pregnancy Childbirth 2021 (18) | Prespecified analysis of the rates of LGA and SGA offspring in the CONCEPTT study population (n = 225 pregnant women and liveborn infants) | Evaluate what growth standard was most appropriate to predict LGA in women with T1D in pregnancy. | GROW and INTERGROWTH birth percentiles were more strongly associated with neonatal complications, WHO standards may not be appropriate for pregnancies complicated by T1D, as they were designed to identify SGA in a mixed population. |
11. Bacon et al. Diabetologia 2021 (19) | Secondary analysis of biorepository samples of placental function (P1GF levels and sFlt-1/PlGF ratio) (n = 157) | To evaluate whether alterations in placental angiogenic factors contribute to variations in infant birth weight among women with T1DM, and whether this was altered by maternal glycemic status. | Infant birth weight in women with T1DM is influenced by both glycemic status and placental function, as measured by P1GF levels and sFlt-1/PlGF ratio. In women with suboptimal control, infant birth weight was higher when placentas were healthy (high P1GF levels or low sFlt-1/P1GF ratio), and birthweight was lower when placentas were unhealthy (low P1GF or high sFlt-1/P1GF ratio) |
12. Tundidor et al. Diabetes Technol Ther 2021 (20) | Secondary analysis CGM and HbA1c data in first (n = 221), second (n = 197) and third (n = 172) trimesters using standardized CGM and HbA1c target metrics | To examine target glucose attainment and associations with pregnancy outcomes. | CONCEPTT trial participants had a low rate of CGM and of HbA1c target attainment especially for the trimester-specific ADA HbA1c targets. Attainment of CGM and NICE HbA1c targets increased throughout gestation and all targets (both NICE/ADA HbA1c and CGM) were more likely to be achieved by RT-CGM users, at 34 weeks’ gestation. ADA HbA1c target achievement was independently associated with better perinatal outcomes. CGM target attainment was low but increased during pregnancy (7.7/10.2/35.5% in first/second/third trimesters) and with RT-CGM use. In the adjusted analyses, achieving TBR target was associated with a higher risk of preeclampsia and neonatal hypoglycemia. ADA HbA1c target attainment <6.0%was low and unchanged during pregnancy (23.5/27.9/23.8%) but increased with RT-CGM use. |
13. Ahmed et al. CMAJ Open 2021 (21) | Post hoc analysis to determine cost implications of CGM use in pregnant women with T1DM in Canada (n = 215) | Determine whether CGM is more expensive than standard SMBG, from the perspective of a third-party payer (government). | The mean cost of care was not different between CGM and SMBG groups. If governments pay for up to $1372 for CGMS, they incur no additional overall costs compared with SMBG, and improve outcomes for pregnant women with T1DM and their babies. |
14. Meek et al. Diabetes Care 2021 (22) | Secondary analysis of biorepository samples for maternal serum and cord blood C-peptide levels in pregnancy (n = 127) | To analyze C-peptide levels and its trajectory throughout pregnancy in maternal serum and cord blood, to examine the hypothesis of B-cell regeneration in pregnant women with T1DM | First maternal C-peptide appearance at 34 weeks was associated with third trimester hyperglycemia, elevated cord blood C-peptide, and high rates of neonatal complications. This suggests transfer of C-peptide from fetal to maternal serum and is inconsistent with pregnancy-related β-cell regeneration. |
15. Abolbaghaei et al. Biomark Res 2021 (23) | Secondary analysis of biorepository samples to identify biomarkers that may predict care associated clinical with outcomes (n = 163) | To quantify extracellular vesicle levels in maternal blood and determine and compare its association with pregnancy clinical outcomes | High levels of platelet extracellular vesicles in early pregnancy were associated with adverse neonatal outcomes (NICU admission rate, hyperbilirubinemia). Assessment of extracellular vesicles may represent a novel approach to personalized care in pregnant women with T1DM. |
16. Bahrami et al. Diabet Med 2022 (24) | Prespecified subanalysis to characterize maternal fear of hypoglycemia (n = 214) | To examine maternal fear of hypoglycemia, glycemia, and pregnancy outcomes in women with impaired and normal awareness of hypoglycemia. | In pregnant women with T1DM, impaired awareness of hypoglycemia (accounting for 30% of CONCEPTT trial participants) is associated with more episodes of severe maternal hypoglycemia and more fear of hypoglycemia. Reassuringly, there was no increase in adverse neonatal outcomes. |
17. Meek et al. Diabetelogia 2023 (25) | Metabolomic insights into maternal and neonatal complications in pregnancies affected by type 1 diabetes (n = 174) | To identify maternal and cord blood metabolic biomarkers for early risk identification for LGA, adiposity, neonatal hypoglycemia, hyperinsulinism, and preeclampsia. | Biorepositories from 174 women with T1D at various time points (93 paired cord blood biorepositories) in the CONCEPTT trial allowed metabolic assays to find correlative relationships with maternal hyperglycemia and suboptimal maternal and neonatal outcomes. Early maternal blood gestational biomarkers positively associated with neonatal hypoglycemia and hyperinsulinemia include maternal triacylglycerol and dietary phenols, while preeclampsia demonstrated association with increased second trimester maternal levels of phosphatidylethanolamines. Maternal hyperglycemia also showed increased carnitine levels in cord blood, linked to LGA. Understanding correlative biomarkers to suboptimal neonatal outcomes opens the possibility of reducing risk factors by early recognition and intervention. |
Does the addition of a CGM to SMBG prior to conception and during pregnancy in women with T1D optimize glycemic control and lead to improved maternal and neonatal outcomes?
The CONCEPTT Collaborative Group was an international compilation of experts who conducted a large, multicenter, open-label, randomized controlled clinical trial in pregnant women with T1D (5). The trial included 31 centers in Canada, England, Ireland, Italy, Scotland, Spain, and the USA. Women with T1D were eligible if they were between 18 and 40 years of age, had a diagnosis of T1D for at least 12 months, and used intensive insulin therapy with multiple daily injections (MDI) or an insulin pump. Those who were planning to become pregnant qualified if they had an HbA1c level between 7.0 and 10%.
From March 2013 to March 2016, 325 women with T1D were recruited who were either planning pregnancy (n = 110, with 34 who conceived during the 24-week planning arm of the trial) or were in the first trimester of pregnancy (n = 215, defined as <14 weeks gestation). The participants primarily identified as European or Mediterranean, were college educated, nonsmokers, average duration of diabetes 17 years, and nearly half re either overweight (BMI 25–29.9) or obese (BMI ≥30). They were assigned to one of two parallel arms – a pregnant arm and a prepregnancy planning arm based on their status at enrollment, then equally randomized to either CGM and SMBG or SMBG alone, and study visits occurred every 4 weeks. Both arms were given detailed treatment algorithms to standardize insulin dose adjustment based on either CGM or SMBG values. The hypothesis was that preconception or gestational glycemic control may improve maternal and neonatal outcomes. Those randomized to the CGM arm received either the MiniMed Guardian CGM, MiniMed Paradigm Veo CGM, or MiniMed 640G system with the Guardian 2 Link Transmitter and Enlite Sensor and would continue their home glucose monitoring regimen via SMBG in combination. Those randomized to the control arm were instructed to continue their home glucose monitoring regimen. Randomization was stratified by mode of insulin delivery (insulin pump or MDI) as well as baseline HbA1c.
Primary analysis of CONCEPTT trial
The primary outcome measure for both arms included change in HbA1c (analyzed at a central lab) from baseline to the prespecified endpoint (34 weeks’ gestation in the pregnancy arm, and at conception or 24 weeks after enrollment in the prepregnancy planning arm). Secondary outcomes for both arms included prespecified glycemic metrics from CGM downloads, such as duration and frequency of hypoglycemia, TIR for pregnancy (63–140 mg/dL), area under the curve, glycemic variability, as well as maternal and neonatal outcomes. For those in the pregnancy arm, prespecified health outcomes included gestational weight gain, gestational hypertension, preeclampsia, mode of delivery, and length of hospital stay. Prespecified neonatal outcomes included preterm delivery, neonatal hypoglycemia, birthweight, neonatal intensive care unit admission for more than 24 h, and duration of neonate hospital stay. In the primary analysis of the pregnancy arm there was a very small difference in participants attaining HbA1c target of ≤6.5% favoring the CGM arm versus the control arm (66% vs 52%, P = 0.0601. In the pregnancy planning arm, there was no statistically significant difference in HbA1c between the CGM and the control arms. With regard to secondary outcomes, the CONCEPTT trial demonstrated significant improvements in TIR in the CGM group (68% TIR) versus the control group (61% TIR, P = 0.0034). The CGM group also demonstrated significantly less time in hyperglycemia. Through a composite of glucose variability metrics, the CGM group demonstrated less glycemic variability with no significant difference in hypoglycemia between groups.
Overall there were no significant differences between the characteristics of the two groups except the proportion of completed study visits was slightly higher in the CGM group (mean 7.2 vs 6.8 visits completed). Analysis of clinical outcomes showed trends toward improved maternal outcomes; however. none were found to be statistically significant. The CONCEPTT group demonstrated improved neonatal outcomes in the CGM arm, including reduction in incidence of LGA infants, reduction in neonatal hypoglycemia requiring i.v. dextrose, and fewer NICU admissions. In the pregnancy planning group CGM use during the pregnancy planning arm was underpowered and thus inconclusive and further analysis was not pursued.
The overwhelming strengths of this study include its RCT design as CGM was rapidly becoming ubiquitous in women with T1D during pregnancy. In fact, during the enrollment period there was some discussion among experts that in countries where CGM was widely available in pregnancy, it was unethical to have a control arm that did not offer real time CGM. Since CONCEPTT, CGM technology has rapidly improved in both accuracy and form factor leading experts to presume that the benefit of CGM in pregnancy complicated by T1D is more robust than CONCEPTT results. Further, automated insulin delivery systems which rely on CGM to automatically adjust insulin delivery are now available and are often used in pregnancy.
The CONCEPTT Trial did present a weakness which was illustrated by multiple countries refusing to participate in the landmark study. In 2016, while the study was enrolling subjects, sensors had entered the international market after federal regulatory approvals in their corresponding countries that were markedly superior as measured by MARD comparison including the Abbott Freestyle Libre (MARD 10%) (26) Dexcom G5 (MARD 9%) (26) both receiving FDA approval in 2016. The inferior accuracy of the sensors acted as a data integrity and clinical care detriment to the study.
Are Insulin pumps better than MDI at achieving glycemic control and providing positive neonatal and maternal outcomes in pregnancies complicated by T1D?
Pregnancy poses unique challenges for women with T1D, necessitating optimal glycemic control to mitigate adverse maternal and fetal outcomes. The data from CONCEPTT was further studied by Feig et al. to determine if there were any differences when using insulin pump therapy versus MDI therapy (11). This prespecified secondary analysis included 248 women. The primary outcome was changes in HbA1c from the time of baseline to 34 weeks of gestation. Other secondary glucose outcomes included CGM TIR (63–140 mg/dL), TAR, TBR, episodes of hypoglycemia (<63 mg/dL; ≥20 min), and glucose coefficient of variation (CV) showcasing glycemic variability. HbA1c and TIR was measured at 12, 24, and 34 weeks’ gestation. The purpose was to compare obstetric and neonatal outcome and patient-related outcome measures in pregnant subjects using insulin pumps versus MDI. At baseline, pump users seemed ‘healthier’, more often in stable relationships (P = 0.003), more likely to take preconception vitamins (P = 0.03), and less likely to smoke (P = 0.02). Pump and MDI users had similar baseline glycemic control: HbA1c (6.84 ± 0.71% vs 6.95 ± 0.58%; P = 0.31) and TIR (51 ± 14% vs 50 ± 13%; P = 0.40). More MDI users achieved target HbA1c levels below 6.5% at 24 weeks (72.1 vs 63.1%; P = 0.009) as well as 34 weeks (65.1 vs 52.0%; P = 0.001). A higher percentage of pump users had hypertensive disorders of pregnancy (30.6% vs 15.5%; P = 0.011). It was also found that infants of maternal insulin pump users had more NICU admissions (44.5% vs 29.6%; P = 0.02). The authors concluded that glycemic outcomes were better in MDI users and MDI users had less gestational hypertension, neonatal hypoglycemia, and NICU admission rates compared to insulin pump users yet cited several limitations to this analysis. Since this study was a secondary analysis of the CONCEPTT study, subjects were not randomized to insulin pumps or MDI, thus introducing a sampling bias. Further, there were no data collected for diabetes management such as insulin dose adjustments, frequency of glucose checks, insulin pump download, or use of bolus calculators. Furthermore, Sacha et al. (27) point out that neither the type of basal insulin used in the MDI treatment group nor the insulin pumps used and pump behavior in the pump group were identified or analyzed. Further studies would be necessary to observe the true effect of insulin pump versus MDI in patients with T1D during pregnancy.
Were dietary patterns and macronutrient intakes different between the treatment arms of the CONCEPTT trial?
Two subgroup analyses were identified that investigated the differences in dietary intake between the pregnancy and pregnancy planning women, and between insulin pump and MDI users, in women with Type 1 diabetes. The former was a prespecified secondary subgroup analysis of macro- and micronutrient intake among the UK subgroup, which included 156 participants (93 pregnant and 63 planning pregnancy) from 14 sites across England, Northern Ireland, and Scotland.
Baseline characteristics were not significantly different between the groups. The mean energy intake was not significantly different between the groups: 1588 kcal/day in women planning pregnancy and 1673 kcal/day in pregnant women. The major sources of energy were similar between groups, consisting of cereals and cereal products (30–33%), meat and meat products (15–18%), vegetables and potatoes (12%), and milk and milk products (10%). The total carbohydrate intake was consistent with the American Diabetes Association’s and UK’s recommendations, with 180 g of carbohydrate/day in women planning pregnancy (54% from recommended sources) and 198 g of carbohydrate/day in pregnant women (56% from recommended sources). Fat consumption exceeded guideline recommendations in both groups: 70 g/day in the women planning pregnancy, and 72 g/day in the pregnant women. Daily fiber intakes were below guideline recommendations in both groups: 15.5 g/day in women planning pregnancy, and 15.4 g/day in pregnant women. The average consumption of fruit and vegetables was inadequate in both groups. Twelve women planning pregnancy (19%) and 24 pregnant women (26%) did not meet micronutrient requirements. This was the first multicenter study that described the dietary habits of women with T1D before and during pregnancy, specifically identifying total energy and micronutrient intake. The findings from this study provide a comprehensive analysis of the dietary habits of women with T1D before and during pregnancy and demonstrate that nutritional guidelines are not being met and their diets are characterized as being high in fat, low in fiber, fruits, and vegetables, and high in nonrecommended carbohydrates, compared with current nutritional guidelines. Approximately one in five met the UK-recommended fruit and vegetable target, and one in four were at risk of micronutrient deficiencies.
Clearly, it is difficult to maintain tight glycemic targets with healthy dietary choices, as this population often substitutes fat for carbohydrates, and often treat hypoglycemia with nonrecommended carbohydrate choices. Limitations from this study include the high likelihood of underreporting calorie intake, lack of verifications of consumed diet, lack of portion sizes, and combined analyses across all trimesters. Generalizability is limited, since results are only representative of the UK population (majority from England), from women with long durations of diabetes (mean 17 years) and a high proportion of participants were well educated.
The second subanalysis, also conducted by Neoh and colleagues, evaluated the dietary patterns of women using an insulin pump and those using MDI in pregnant women with T1D. A total of 93 (82.3%) pregnant women were included in the analysis, 55 using MDI and 38 using an insulin pump. Baseline characteristics were not significantly different between the groups. There were no significant differences between groups in total caloric intake or macronutrient intake. Unfortunately, dietary patterns were suboptimal in both groups, but no significant differences were found in neither total calorie intake nor snacking behaviors. Limitations of this study include the risk of type II error due to small sample size and missing data; furthermore, the food diary methodology does not allow an evaluation of glycemic index.
Are there associations between maternal glucose metrics and maternal and neonatal outcomes?
Neonatal hypoglycemia is a common complication of T1D in pregnancy, and is driven by maternal antenatal glucose control. In a secondary analysis of 225 pregnant women and their infants from the CONCEPTT trial, clinically relevant neonatal hypoglycemia occurred in 15% of term and 40% of preterm infants, and were more likely to be delivered via C-section (83%, P < 0.0001) and admitted to the NICU (90%, P < 0.0001) (12).
The primary objective was to analyze the incidence of neonatal hypoglycemia (defined as glucose concentration <46 mg/dL requiring treatment with i.v. dextrose). The secondary objective was to explore the associations between maternal glycemia and birthweight percentile, neonatal anthropometry measures, and fetal hyperinsulinemia. Findings from this analysis identified high C-peptide levels in cord blood, birthweight >97th percentile (extreme LGA), and skinfold thickness in those neonates with hypoglycemia. During the second and third trimesters, mothers of neonates with hypoglycemia demonstrated modest differences in maternal HbA1c’s (6.6% vs 6.2% ± 0.6 and 6.7% vs 6.3% ± 0.6, respectively), and TIR (46% vs 53% and 60% vs 66%, respectively). Although the statistical power in this study was limited by the small number of women who continued using CGM until delivery (33 CONCEPTT trial participants), there were no differences in intrapartum maternal glycemic control in neonates with and without hypoglycemia.
This secondary analysis suggests that modest improvements in maternal HbA1c (0.4% decrease), and CGM TIR (6–7% increase) during the second and third trimesters are associated with a lower risk of neonatal hypoglycemia. The mechanism appears to be linked with fetal insulin secretion in the 24 h prior to delivery in neonates with hypoglycemia, as demonstrated by the high concentration of cord blood C-peptide, birth weight percentile, and infant adiposity. Achieving tight maternal glycemic control throughout pregnancy, with more focus on clinical intervention during the second and third trimesters, may be more effective in reducing the risk of neonatal hypoglycemia.
Secondary analysis of temporal glucose profiles and their association with outcomes
In 2020, Scott and colleagues published an in-depth analysis of glucose data from the CONCEPTT trial, taking a deeper dive into the standard CGM metrics by time of day. Investigators gathered CGM metrics in daytime (06:01–00:00) and overnight (00:01–06:00) to evaluate (i) RT-CGM vs SMBG (ii) pump vs MDI, and (iii) LGA outcomes vs normal weight. A polynomial regression function (functional data analysis), similar to an ambulatory glucose profile, provided comparisons on temporal differences between glycemic control factors. A total of 200 complete data sets were available (RT-CGM n = 100 and SMBG n = 100), with over 96 h of continuous data at baseline, 24 weeks’, and 34 weeks’ gestation. LGA was defined as >90th percentile using Gestational Related Optimal Weight (GROW) (Gardosi Am J Obstet Gynecol 2018). Results showed that women using RT-CGM had a 7–14 mg/dL lower glucose compared to SMBG and pump users had a 7–16 mg/dL higher glucose at 24 weeks, with no difference at 34 weeks compared to MDI. Women who had an LGA infant had higher glucose at all time points evaluated during the pregnancy. The authors concluded that women using RT-CGM achieved better daytime glucose control. Though interesting, we still hope to identify which times of the day most affect infant outcomes.
In 2019, International Consensus Guidelines on Clinical Targets for CGM were published, including pregnancy-specific targets (7). This provided a much needed framework for evaluating CGM data. Not surprisingly, CONCEPTT Collaborative Group reevaluated participants’ CGM data using the new TIR guidelines, as shown here:
Secondary analysis of CONCEPTT CGM and A1C data using standardized metrics
In a secondary subanalysis of pregnant CONCEPTT trial participants (20), real-time CGM (RT-CGM) and HbA1c measures in the first, second, and third trimesters were reviewed, aiming to examine the associations of target glucose attainment and pregnancy outcomes. Standardized targets for CGM metrics in pregnant women with T1D include: >70% TIR (63–140 mg/dL), <25% TAR (>140 mg/dL), and <4% TBR (<63mg/dL) (7). According to the National Institute for Health and Care Excellence (NICE), the target HbA1c for women with T1D in pregnancy is <6.5% throughout pregnancy while the American Diabetes Association (ADA) target is <6.5% during the first trimester and <6.0% during the second and third trimesters (28, 29). The TIR target attainment was low but increased during pregnancy and increased in the second and third trimesters with RT-CGM use. There were nonsignificant differences in ADA trimester-specific HbA1c target attainment, which was low and unchanged during pregnancy but increased with RT-CGM use.
In conclusion, this secondary analysis identified a low rate of RT-CGM and HbA1c target attainment, especially for the trimester-specific ADA HbA1c targets. Attainment of CGM and NICE HbA1c targets increased throughout gestation and all targets (both NICE/ADA CGM and HbA1c) were more likely to be achieved by RT-CGM users, at 34 weeks’ gestation. Although not statistically significant, the trimester-specific ADA HbA1c target achievement was independently associated with better perinatal outcomes, while the perinatal risks of TBR warrants further study.
Is maternal fear of hypoglycemia a factor in achieving target glucose control metrics?
A prespecified subanalysis of 214 pregnant CONCEPTT trial participants examined maternal fear of hypoglycemia, glycemia, and pregnancy outcomes in women with impaired and normal awareness of hypoglycemia (24). Participants completed hypoglycemia fear surveys (HFS-II) at baseline. Logistic regression and Poisson regression analyses were used to obtain an adjusted estimate for the rate ratio relating awareness to the number of severe hypoglycemic episodes and for several neonatal outcomes in relation to the total HFS-II score. Overall, 30% of participants reported impaired awareness of hypoglycemia (64 participants). Women with impaired awareness of hypoglycemia between 12 and 34 weeks’ gestation had more episodes of severe hypoglycemia (mean 0.44 vs 0.08, P < 0.001) and scored higher on the HFS-II scale (43.7 vs 36.0, P < 0.008). They had more TBR and exhibited more glycemic variability at 12 weeks’ gestation. Higher overall HFS-II scores were associated with a higher risk of severe hypoglycemia episodes (RR 1.78, 95% CI 1.39–2.27). Women with impaired awareness of hypoglycemia had less maternal weight gain, but no differences in neonatal outcomes between women with impaired awareness of hypoglycemia and normal hypoglycemia awareness were found. Higher HFS-II scores were positively associated with nephropathy (OR 1.91, 95% CI 1.06–3.4). CGM use after 12 weeks was not associated with the number of episodes of severe hypoglycemia (RR 0.75, 95% CI 0.49–1.15). In pregnant women with T1D, impaired awareness of hypoglycemia is associated with more maternal severe hypoglycemia episodes and more fear of hypoglycemia but not with adverse neonatal outcomes.
What tools are best for assessing neonatal growth and fetal size in pregnancy with T1D?
This secondary prespecified analysis (18) involved CONCEPTT participants including 225 pregnant women with T1D whose Infants were weighed immediately at birth with GROW, INTERGROWTH, and WHO percentile calculations. Relative risk ratios, sensitivity and specificity were used to assess the different growth standards with respect to neonatal complications, including neonatal hypoglycemia, hyperbilirubinemia, respiratory distress, NICU admission. Accelerated fetal growth was common, with mean birthweight percentiles of 82.1, 85.7, and 63.9 and LGA rates of 62%, 67%, and 30% using GROW, INTERGROWTH, and WHO standards, respectively. Corresponding rates of SGA were 2.2, 1.3, and 8.9%, respectively. LGA defined according to GROW percentiles showed stronger associations with preterm delivery, neonatal hypoglycemia, hyperbilirubinemia, and NICU admission. Infants born >97.7th percentile were at highest risk of complications. SGA defined according to INTERGROWTH percentiles showed slightly stronger associations with neonatal complications. GROW and INTERGROWTH standards performed similarly and identified similar numbers of neonates with LGA and SGA. GROW-defined LGA and INTERGROWTH-defined SGA had slightly stronger associations with neonatal complications. WHO standards underestimated size in preterm infants and are less applicable for use in T1D.
What are the economic impacts of implementing CGM use in pregnancy complicated by T1D?
In the first economic analysis (13) using data from the CONCEPTT study, the authors aim to estimate the costs and potential cost savings associated with the introduction of real-time continuous glucose monitoring (RT-CGM) in pregnant women with T1D. The economic impact model was developed from the perspective of National Health Service (NHS) England and assumes that RT-CGM is used for ~28 weeks (10–38 weeks’ gestation) and that neonates who did not require NICU admission had a normal postpartum stay. Pregnancy and neonatal complication rates and related costs were derived from published literature, national tariffs, and device manufacturers. The device cost of SMBG alone was £588 (~$729) and RT-CGM was £1820 (~$2256). The total annual costs of managing pregnancy and delivery in women with T1D were £23,725,648 (~$29,407,703) with SMBG alone, and £14,165,187 (~$17,557,608) with SMBG and RT-CGM; indicating potential cost savings of approximately £9,560,461 (~$11,850,096) per year. Shorter and reduced NICU stays (mean 6.6 vs 9.1 days, respectively) at daily costs of NICU admissions (£3743 per day in England, i.e. ~$4639) accounted for this difference. The authors conclude that routine use of RT-CGM by pregnant women with T1D would result in substantial savings, and they subsequently published a review aimed at facilitating CGM utilization and access, serving as an update and summary of the benefits of RT-CGM in T1D pregnancy (3).
Subsequently a post hoc analysis (21) was published to evaluate the cost implications of CGM in pregnant women with T1D in three Canadian provinces. Health-care resource data were analyzed in 100 mother–infant pairs in the CGM group and 102 mother–infant pairs in the SMBG control group. The resource costs were based on prices from hospitals in three Canadian provinces. The primary outcome was the difference in the mean total health-care costs for the CGM group compared to the SMBG group.
Costs were determined for CGM devices and sensors from Medtronic Canada and for glucometers and testing strips from Ascensia Diabetes Care Canada, according to 2013 private market rates. The trial protocol indicated the use of one CGM sensor every 6 days, and seven glucose testing strips per day from randomization to delivery in both study groups.
When patients pay for both technologies, CGM is less expensive than SMBG to government health-care payers ($13,270 vs $18,465, respectively) and confers substantial clinical benefit on mothers and their infants. The overall mean cost savings for the CGM group of $5195 is largely driven by reduced NICU admissions. If governments paid for CGM and SMBG devices and supplies at the 2013 private market rates, the mean cost of care was not different between the groups. Mothers and their infants can reap the important clinical benefits of CGM technology at no extra cost to the government.
Based on the two studies reviewed, the overall benefits appear to outweigh the cost, making the economic impacts of CGM use in pregnancy complicated by T1D positive. As CGM technology continues to improve and as more data showcase its value and safety, government and private payers will make a continuous stride to include them on their formulary. SMBG, although useful, will inevitably be a technology of the past.
The CONCEPTT study included collecting samples from participants to form a central biorepository for both prespecified and yet-to-be-specified analysis of biomarkers. The following four subanalyses examine the utility of a variety of biomarkers.
Are there alternative biomarkers associated with maternal glycemic control, maternal outcomes, or neonatal outcomes?
The first subanalysis evaluated the predictive performance of HbA1c, CGM metrics, and alternative biochemical markers of glycemia in maternal and neonatal outcomes including preeclampsia, preterm delivery, LGA, neonatal hypoglycemia, and NICU admission (17). Alternative biomarkers included glycated CD59, 1,5-anhydroglucitol (1,5-AG), fructosamine, and glycated albumin. The results of the evaluation at 12, 24, and 34 weeks’ gestation can be found in Table 2. Overall, CGM TIR and TAR showed the most consistent associations with neonatal outcomes. ROC curves were used to compare the ability of the laboratory markers of glycemia with CGM metrics to predict pregnancy outcomes. However, as expected for interrelated glycemic markers, CIs for ORs and area under the ROC curve (AUROC) were often overlapping. The strongest predictors (defined by the highest AUROC) have been laid out by neonatal outcome and time point in Table 2.
Neonatal outcomes have been laid out by associated CGM findings, strongest predictors (AUROC), and their corresponding time points. Only the highest AUROCs were shown as the strongest predictors.
Outcome | CGM metrics | Time points |
---|---|---|
LGA | TIR, TAR, SD | All |
Mean glucose | 24 and 34 weeks | |
Neonatal hypoglycemia | Mean glucose, TIR, TAR | 24 and 34 weeks |
TAR | First trimester | |
NICU admission | Mean glucose, TIR, TAR, SD | 24 weeks |
TIR | First trimester | |
No CGM metric associations | 34 weeks | |
Outcome | Strongest predictors (AUROC) | Time points |
LGA | 1, 5-AG, TIR (0.64) | First trimester |
TIR, fructosamine, HbA1c (0.64) | 24 weeks | |
TAR (0.67) | 34 weeks | |
Neonatal hypoglycemia | gCD59 (0.61) | First trimester |
gCD59 (0.72) | 24 weeks | |
HbA1c (0.68) | 34 weeks | |
NICU admission | gCD59 (0.73) | 24 weeks |
HbA1c (0.66) | 34 weeks | |
Preeclampsia | Mean CGM glucose (0.65) | First trimester |
Mean CGM glucose (0.72) | 24 weeks | |
Fructosamine (0.76) | 34 weeks | |
Preterm birth | Mean CGM glucose, TAR (0.61) | First trimester |
Mean CGM glucose, TAR, gCD59 (0.64) | 24 weeks | |
HbA1c (0.65) | 34 weeks |
The authors concluded that neither laboratory markers nor CGM metrics were able to provide a strong prediction of any pregnancy outcome (AUROCs mostly 0.70). In pregnant women with T1D, the use of alternative laboratory markers did not appreciably increase the AUROC for prediction of suboptimal pregnancy outcomes beyond HbA1c, which is already widely available, or CGM metrics, such as TIR and TAR. HbA1c was consistently associated with pregnancy outcomes, suggesting that despite the known limitations of HbA1c for assessing antenatal glycemia, it is still a critically important biomarker for obstetric and neonatal health outcomes. While other laboratory biomarkers demonstrated some promise, none were able to significantly increase the AUROC, showing, at best, comparable prediction to HbA1c alone.
Now we turn our attention specifically toward biomarkers of birth weight with Bacon and colleagues in a subanalysis aimed to evaluate whether alterations in placental angiogenic factors contribute to variations in infant birth weight among women with T1D, and whether this was altered by maternal glycemic status (19). It is well known that maternal glycemia alone is not the only factor contributing to birth weight variation among infants of women with T1D. The fetoplacental circulation also impacts fetal size and is influenced by placental angiogenic factors (placental growth factor (P1GF) and soluble fms-like tyrosine kinase (sFlt-1)). A healthy placenta is represented by a high P1GF or a low sFlt-1/P1GF ratio, whereas an unhealthy placenta is represented by a low P1GF or a high sFlt-1/P1GF ratio. Thus far, studies of placental factors have mainly been made in the context of preeclampsia, where low P1GF (<100 pg/mL) and/or a high sFlt-1/P1GF ratio (>85) are useful predictors of impending preeclampsia (a result of persistent maternal placental malperfusion) and tend to be associated with growth-restricted offspring. Linear regression was used to assess the relationship between birth weight Z-score (the primary outcome) and placental health (as measured by P1GF and sFlt-1/P1GF ratio) stratified by maternal glycemic control (CGM and HbA1c measures) and adjusted for potential confounders of maternal BMI, smoking, and weight gain. The rates of preeclampsia, duration of diabetes, HbA1c, and microvascular complications were similar across all groups. In the setting of optimal maternal glycemia (HbA1c <6.5% or CGM time above range (TAR) ≤30%), birthweight was normal with a healthy placenta (increasing P1GF), and birthweight was higher with an unhealthy placenta (decreasing P1GF). However, when maternal glycemia was suboptimal, those with a healthy placenta (higher P1GF) had heavier infants than those with an unhealthy placenta (lower P1GF).
In summary, this analysis highlights the complex interplay between maternal glycemic control, placental health, and infant birth weight in women with T1D. It underscores the clinical importance of monitoring placental function (P1GF levels and sFlt-1/PlGF ratio) in pregnancies with suboptimal glycemic control and normal fetal growth. Measuring P1GF and/or sFlt-1/PlGF ratios in pregnant women with T1D may help to predict compromised placental function and reduce perinatal complications. However, more studies and larger sample sizes are necessary to understand why women with optimal glycemic control still had unhealthy placentas and larger babies.
In another secondary analysis, the CONCEPTT collaborative group further examined the hypothesis of β-cell regeneration in pregnancy with T1D (22). Many experts in T1D and pregnancy have observed the resumption of C-peptide production during pregnancy which is a state of mild immunosuppression. This observation has led scientists to postulate both the drivers of this mechanism as well as the potential therapeutic benefit of using mild immunosuppression to preserve or stimulate β-cell function. Meek and colleagues (22) found three discrete phenotypes of maternal C-peptide production: (i) undetectable throughout pregnancy, n = 74 (58%; <3 pmol/L); (ii) detectable at baseline, n = 22 (17%; 7–272 pmol/L at baseline); and (iii), undetectable in first and second trimesters, then became detectable at 34 weeks, n = 31 (24%; 4–26 pmol/L at 34 weeks). Baseline characteristics and end of pregnancy glucose profiles were similar in phenotypes 1 and 3, yet the pattern 3 group had suboptimal glucose profiles and more neonatal complications. The authors concluded that C-peptide reemergence in the third trimester was a consequence of maternal hyperglycemia and postulated that it is neonatal C-peptide crossing the placenta into maternal serum. In counterpoint, authors Ivanisevic and Djelmis published a rebuttal citing evidence refuting the hypothesis of fetal transfer of C-peptide and maintaining the β-cell regeneration hypothesis (30). In rebuttal, Meek and colleagues provided evidence and possible mechanisms for immunological tolerance augmenting β-cell function in pregnancy. Meek and colleagues reviewed the levels of GAD, islet cell, IA2, and ZnT8 autoantibodies at 12, 24, and 34 weeks in their cohort. Participants with detectable C-peptide (phenotype 2) were more likely to have detectable antibodies than women in phenotypes 1 and 3. However, autoantibody positivity was unchanged for all groups throughout pregnancy, with no evidence of new immune phenomena, which would presumably occur if quiescent β-cells resume function (31). The authors conclude that further research is needed to inform the question of pancreatic regeneration or fetal hyperinsulinism as the mechanism behind the reappearance of C-peptide during the third trimester of pregnancy in T1D.
In an additional secondary analysis (23), the authors quantify the number of circulating EVs in the serum of a subset of CONCEPTT participants, with the aim of evaluating their association with maternal glycemic metrics, maternal blood pressure, and pregnancy outcomes. Secondarily, the authors aimed to evaluate whether RT-CGM as compared to SMBG alter levels of EVs in plasma. Maternal vascular health is critical in maternal and fetal outcomes in pregnancy including vascular endothelial dysfunction being observed in preeclampsia. Biomarkers for maternal vascular health may help identify women with greater risk of pregnancy complications and may even allow for intervention strategies that improve vascular health. In nonpregnant populations with T1D and T2D, elevated levels of circulating extracellular vesicles (EVs), primarily originating from platelets, endothelial cells, and leukocytes, have been shown to predict risk of cardiovascular events independent of traditional risk factors. Results showed there were no significant differences in levels of endothelial, platelet, leukocyte, or total circulating EVs between trimesters. Platelet EV levels were inversely associated with maternal hyperglycemia and glycemic variability, while higher endothelial EV levels were associated with increased risk of NICU admission, neonatal respiratory distress, and overall poor composite fetal outcome. Additionally, baseline platelet EVs were inversely associated with maternal complications, possibly conferring a protective or compensatory effect. In summary, high levels of circulating EVs in early pregnancy were associated with adverse neonatal outcomes and may offer a new approach to predicting and intervening in women at higher risk for vascular impairment in pregnancy complicated by T1D.
As notable by topics of previous substudies, current clinical approaches to assess risk of pregnant women with T1D for neonatal complications including LGA, neonatal hypoglycemia, hyperinsulinemia, adiposity, and preeclampsia focus on metabolic monitoring and glycemic control during the second and third trimesters. These outcomes pose the challenge of identifying metabolic markers earlier in pregnancy for early risk assessment. The most recent substudy by Meek et al. utilized the biorepository generated by the CONCEPTT trial by means of maternal and cord blood untargeted metabolite analysis (ultrahigh-performance liquid chromatography–mass spectrometry) (25). The primary findings include a positive association between maternal TAR with maternal blood lipids and cord blood carnitines. Carnitines present in cord blood demonstrated an association with neonatal adiposity. Carnitines are an essential component in β-oxidation, a process believed to be absent until 3 days of life; however, no widespread positive association was demonstrated between TAR and maternal blood carnitines. A pattern of evidence indicated LGA being linked to increased steroid hormone production, activity of the tricarboxylic acid cycle (citrate) and β-oxidation (carnitines) in maternal blood beginning at week 24.
Adiposity and LGA demonstrated remarkably different metabolite associations, indicating possibly different pathophysiology. Neonatal hyperinsulinemia, assessed by neonatal hypoglycemia and cord blood C-peptide demonstrated positive associations with maternal blood triacylglycerol concentrations at 12 weeks. Preeclampsia demonstrated association with increased maternal phosphatidylethanolamine through the first and second trimester. The biorepository consisting of nonfasting samples presents a significant weakness of this subanalysis. The introduction of exploratory biomarkers indicating increased risk for suboptimal gestational outcomes with T1D prompts an opportunity for further investigation and development of clinical standards for early screening and intervention.
The investigation of biomarkers associated with negative maternal or neonatal outcomes complicated by T1D presents a new set of essential tools the practitioner may utilize during pregnancy to enhance gestational care. Disproportionate prevalence of suboptimal outcomes resulting from diabetes in pregnancy has long faced the medical community, and the first step to properly addressing this phenomenon is understanding effective screening measures. Screening measures, like those outlined by the aforementioned substudy, are capable of being integrated into standard-of-care measures, but first-trimester measures are still underwhelming. Future studies are necessary to continue the work began by these critical substudies, all resulting from the CONCEPTT Trial.
Conclusion
The CONCEPTT clinical trial was a landmark study with multiple subanalyses, undoubtedly with more to be published in the future. With the copious amount of data generated over the past several years, our aim was to create a summary of the findings as well as a timeline of the publications. The CONCEPTT findings have informed both the 2019 Consensus Guidelines for Time In Range as well as the American Diabetes Association’s Standards of Care for Diabetes in Pregnancy (7, 29). RT-CGM is now the standard of care for T1D in pregnancy; thus, it would be unethical to repeat CONCEPTT in a T1D population. However, study design utilized in the CONCEPTT study could inform future studies evaluating the best practice for other types of diabetes in pregnancy, including GDM and T1D. Further analysis is also needed on the benefit of CGM as part of automated insulin delivery systems in pregnancies with diabetes. In the meantime, clinicians must strive to keep abreast of the rapid progress in CGM technology and ever-changing landscape, and the available evidence to support its use in pregnancies complicated by diabetes so as to continue to improve the complications associated with diabetes in pregnancy.
Declaration of interest
Dr Pham is a paid consultant and speaker for Abbott Laboratories. Dr Thorsell receives supplies and research support provided to her institution from Tandem Diabetes Care, Altimmune, Insulet, Dexcom, Medtronic, Eli Lilly, MannKind, Roche, and Abbott Diabetes Care. Dr Castorino receives research support provided to her institution from Dexcom, Abbott, Medtronic, Eli Lilly, MannKind, and Insulet and receives consulting fees from Dexcom. Brandon Cobb has no conflicts to declare.
Funding
This work did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.
References
- 1↑
Meek CL. Monitoring motherhood: monitoring and optimizing glycaemia in women with pre-existing diabetes in pregnancy. Annals of Clinical Biochemistry 2022 59 37–45. (https://doi.org/10.1177/00045632211035815)
- 2↑
Yamamoto JM, & Murphy HR. Emerging technologies for the management of T1D in pregnancy. Current Diabetes Reports 2018 18 4. (https://doi.org/10.1007/s11892-018-0973-9)
- 3↑
Yamamoto JM, & Murphy HR. Benefits of real-time continuous glucose monitoring in pregnancy. Diabetes Technology and Therapeutics 2021 23(Supplement1) S8–S14. (https://doi.org/10.1089/dia.2020.0667)
- 4↑
Murphy HR. Intensive glycemic treatment during T1D pregnancy: a story of (mostly) sweet success! Diabetes Care 2018 41 1563–1571. (https://doi.org/10.2337/dci18-0001)
- 5↑
Feig DS, Donovan LE, Corcoy R, Murphy KE, Amiel SA, Hunt KF, Asztalos E, Barrett JFR, Sanchez JJ, Leiva A, et al.Continuous glucose monitoring in pregnant women with T1D (CONCEPTT): a multicentre international randomised controlled trial. Lancet 2017 390 2347–2359. (https://doi.org/10.1016/S0140-6736(1732400-5)
- 6↑
Farrar D, & Campbell MD. Does continuous glucose monitoring during pregnancy improve glycaemic and health outcomes in women with T1D?-what the CONCEPTT trial adds. Annals of Translational Medicine 2018 6 188. (https://doi.org/10.21037/atm.2018.03.08)
- 7↑
Battelino T, Danne T, Bergenstal RM, Amiel SA, Beck R, Biester T, Bosi E, Buckingham BA, Cefalu WT, Close KL, et al.Clinical targets for continuous glucose monitoring data interpretation: recommendations from the international consensus on time in range. Diabetes Care 2019 42 1593–1603. (https://doi.org/10.2337/dci19-0028)
- 8↑
Chamberlain JJ. Continuous glucose monitoring systems: categories and features. ADA Clinical Compendia 2018 2018 8–10. (https://doi.org/10.2337/db20181-8)
- 9↑
Nakamura K, & Balo A. The accuracy and efficacy of the dexcom G4 platinum continuous glucose monitoring system. Journal of Diabetes Science and Technology 2015 9 1021–1026. (https://doi.org/10.1177/1932296815577812)
- 10↑
Feig DS, Asztalos E, Corcoy R, Leiva AD, Donovan L, Hod M, Javanovic L, Keely E, Kollman C, McManus R, et al.CONCEPTT: continuous glucose monitoring in women with T1D in pregnancy trial: a multi-center, multi-national, randomized controlled trial-study protocol. BMC Pregnancy and Childbirth 2016 167 1–8.
- 11↑
Feig DS, Corcoy R, Donovan LE, Murphy KE, Barrett JFR, Sanchez JJ, Wysocki T, Ruedy K, Kollman C, Tomlinson G, et al.Pumps or multiple daily injections in pregnancy involving T1D: a prespecified analysis of the CONCEPTT randomized trial. Diabetes Care 2018 41 2471–2479. (https://doi.org/10.2337/dc18-1437)
- 12↑
Yamamoto JM, Corcoy R, Donovan LE, Stewart ZA, Tomlinson G, Beardsall K, Feig DS, Murphy HR & CONCEPTT Collaborative Group. Maternal glycaemic control and risk of neonatal hypoglycaemia in T1D pregnancy: a secondary analysis of the CONCEPTT trial. Diabetic Medicine 2019 36 1046–1053.
- 13↑
Murphy HR, Feig DS, Sanchez JJ, de Portu S, Sale A & CONCEPTT Collaborative Group. Modelling potential cost savings from use of real-time continuous glucose monitoring in pregnant women with T1D. Diabetic Medicine 2019 36 1652–1658. (https://doi.org/10.1111/dme.14046)
- 14↑
Neoh SL, Grisoni JA, Feig DS, Murphy HR & CONCEPTT Collaborative Group. Dietary intakes of women with T1D before and during pregnancy: a pre-specified secondary subgroup analysis among CONCEPTT participants. Diabetic Medicine 2020 37 1841–1848. (https://doi.org/10.1111/dme.13937)
- 15↑
Neoh SL, Yamamoto JM, Feig DS, Murphy HR & CONCEPTT Collaborative Group. Dietary patterns of insulin pump and multiple daily injection users during type 1 diabetes pregnancy. Diabetes Care 2020 43 e5–e7. (https://doi.org/10.2337/dc19-1908)
- 16↑
Scott EM, Feig DS, Murphy HR, Law GR & CONCEPTT Collaborative Group. Continuous glucose monitoring in pregnancy: importance of analyzing temporal profiles to understand clinical outcomes. Diabetes Care 2020 43 1178–1184. (https://doi.org/10.2337/dc19-2527)
- 17↑
Meek CL, Tundidor D, Feig DS, Yamamoto JM, Scott EM, Ma DD, Halperin JA, Murphy HR, Corcoy R & CONCEPTT Collaborative Group. Novel biochemical markers of glycemia to predict pregnancy outcomes in women with type 1 diabetes. Diabetes Care 2021 44 681–689. (https://doi.org/10.2337/dc20-2360)
- 18↑
Meek CL, Corcoy R, Asztalos E, Kusinski LC, Lopez E, Feig DS, Murphy HR & CONCEPTT Collaborative Group. Which growth standards should be used to identify large- and small-for-gestational age infants of mothers with type 1 diabetes? A pre-specified analysis of the CONCEPTT trial. BMC Pregnancy and Childbirth 2021 21 96. (https://doi.org/10.1186/s12884-021-03554-6)
- 19↑
Bacon S, Burger D, Tailor M, Sanchez JJ, Tomlinson G, Murphy HR, Feig DS & CONCEPTT Collaborative Group. Can placental growth factors explain birthweight variation in offspring of women with type 1 diabetes? Diabetologia 2021 64 1527–1537. (https://doi.org/10.1007/s00125-021-05438-y)
- 20↑
Tundidor D, Meek CL, Yamamoto J, Martinez-Bru C, Gich I, Feig DS, Murphy HR, Corcoy R & CONCEPTT Collaborative Group. Continuous glucose monitoring time-in-range and HbA1c targets in pregnant women with type 1 diabetes. Diabetes Technology and Therapeutics 2021 23 710–714. (https://doi.org/10.1089/dia.2021.0073)
- 21↑
Ahmed RJ, Gafni A, Hutton EK, Hu ZJ, Sanchez JJ, Murphy HR, Feig DS & CONCEPTT Collaborative Group. The cost implications of continuous glucose monitoring in pregnant women with type 1 diabetes iin 3 Canadian provinces: a posthoc cost analysis of the CONCEPTT trial. CMAJ Open 2021 9 E627–E634. (https://doi.org/10.9778/cmajo.20200128)
- 22↑
Meek CL, Oram RA, McDonald TJ, Feig DS, Hattersley AT, Murphy HR & CONCEPTT Collaborative Group. Reappearance of C-Peptide during the third trimester of pregnancy in type 1 diabetes: pancreatic regeneration or fetal hyperinsulinism? Diabetes Care 2021 44 1826–1834. (https://doi.org/10.2337/dc21-0028)
- 23↑
Abolbaghaei A, Langlois MA, Murphy HR, Feig DS, Burger D & CONCEPTT Collaborative Group. Circulating extracellular vesicles during pregnancy in women with type 1 diabetes: a secondary analysis of the CONCEPTT trial. Biomarker Research 2021 9 67. (https://doi.org/10.1186/s40364-021-00322-8)
- 24↑
Bahrami J, Tomlinson G, Murphy HR, Feig DS & CONCEPTT Collaborative Group. Impaired awareness of hypoglycaemia in women with type 1 diabetes in pregnancy: hypoglycaemia fear, glycaemic and pregnancy outcomes. Diabetic Medicine 2022 39 e14789. (https://doi.org/10.1111/dme.14789)
- 25↑
Meek CL, Stewart ZA, Feig DS, Furse S, Neoh SL, Koulman A, Murphy HR & CONCEPTT Collaborative Group. Metabolomic insights into maternal and neonatal complications in pregnancies affected by type 1 diabetes. Diabetologia 2023 66 2101–2116. (https://doi.org/10.1007/s00125-023-05989-2)
- 26↑
Zimmerman C, Albanese-O'Neill A, & Haller MJ. Advances in type 1 diabetes technology over the last decade. European Endocrinology 2019 15 70–76. (https://doi.org/10.17925/EE.2019.15.2.70)
- 27↑
Sacha JM, & Lane WS. Comment on Feig, et al. Pumps or multiple daily injections in pregnancy involving type 1 diabetes: a prespecified analysis of the CONCEPTT randomized trial. Diabetes Care 2018 42 2471–2479.
- 28↑
NICE. Diabetes in pregnancy: management from preconception to the postnatal period. NICE Guideline [NG3]. London, UK: NICE, 2015. (available at: https://www.nice.org.uk/guidance/ng3)
- 29↑
Nuha A, Sayed E, Aleppo G, Aroda VR, Bannuru RR, Brown FM, Bruemmer D, Collins BS, Hilliard ME, Isaacs D, et al.Management of diabetes in pregnancy: Standards of care in diabetes—2023. Diabetes Care 2023 46 S254–S266. (https://doi.org/10.2337/dc23-S015)
- 30↑
Ivanisevic M, & Djelmis J. Comment on Meek, et al. Reappearance of C-Peptide during the third trimester of pregnancy in type 1 diabetes: Pancreatic regeneration or fetal hyperinsulinism? Diabetes Care 2021 45 1826–1834.
- 31↑
Meek CL, Oram RA, McDonald TJ, Feig DS, Hattersley AT, & Murphy HR. Response to Comment on Meek, et al. Reappearance of C-Peptide during the third trimester in type 1 diabetes pregnancy: Pancreatic regeneration or fetal hyperinsulinism? Diabetes Care 2021 45 1826–1834.
- 32↑
Kristensen K, Ögge LE, Sengpiel V, Kjölhede K, Dotevall A, Elfvin A, Knop FK, Wiberg N, Katsarou A, Shaat Net al. Continuous glucose monitoring in pregnant women with type 1 diabetes: an observational cohort study of 186 pregnancies. Diabetologia 2019 62 1143–1153. (https://doi.org/10.1007/s00125-019-4850-0)