Studies from Scotland and Canada confirm large increases in the incidence of pregnancies complicated by pregestational type 1 diabetes (T1D). With this increased antenatal workload comes more specialization and staff expertise, which may be important as diabetes technology use increases. While euglycemia remains elusive and obstetrical intervention (earlier delivery, increased operative deliveries) is increasing, there have been some notable successes in the past 5–10 years. These include a decline in the rates of congenital anomaly (Canada) and stillbirths (U.K.) and substantial reductions in both maternal hypoglycemia (both moderate and severe) across many countries. However, pregnant women with T1D still spend ∼30–45% of the time (8–11 h/day) hyperglycemic during the second and third trimesters. The duration of maternal hyperglycemia appears unchanged in routine clinical care over the past decade. This ongoing fetal exposure to maternal hyperglycemia likely explains the persistent rates of large for gestational age (LGA), neonatal hypoglycemia, and neonatal intensive care unit (NICU) admissions in T1D offspring. The Continuous Glucose Monitoring in Women With Type 1 Diabetes in Pregnancy Trial (CONCEPTT) found that pregnant women using real-time continuous glucose monitoring (CGM) spent 5% less time (1.2 h/day) hyperglycemic during the third trimester, with clinically relevant reductions in LGA, neonatal hypoglycemia, and NICU admissions. This article will review the progress in our understanding of the intensive glycemic treatment of T1D pregnancy, focusing in particular on the recent technological advances in CGM and automated insulin delivery. It suggests that even with advanced diabetes technology, optimal maternal dietary intake is needed to minimize the neonatal complications attributed to postprandial hyperglycemia.

Prospective nationwide studies confirm that despite widespread suboptimal glycemic control, the majority of women with pregestational type 1 diabetes (T1D) deliver live-born babies (1,2). While complications attributed to maternal hyperglycemia throughout pregnancy—namely, rates of large for gestational age (LGA), preterm delivery, and neonatal intensive care unit (NICU) admissions—remain high, almost 95% of women with T1D leave hospital with a live-born infant. A large population-based study from the U.K. reported a consistent reduction in stillbirths, from 25.8 to 10.7 per 1,000 births, in women with pregestational diabetes over the past decade (3). The absolute risk of stillbirth in T1D is now 10–13 per 1,000 births, which although higher than the background maternity population risk of <5 per 1,000, is an important success. A Canadian study reported a 23% decline in the rates of congenital anomaly but without improvement in perinatal mortality (4). In contrast, a contemporary study from Scotland found no improvements in stillbirth or perinatal mortality rates for women with diabetes (5). Both the Canadian and Scottish studies describe significant increases in the incidence of pregnancies complicated by pregestational diabetes over the past 15 years (a doubling in Canada, and a 44% increase in T1D and a 90% increase in T2D in Scotland) (4,5).

The reasons for the U.K. reductions in stillbirths are most likely multifactorial, including, for example, improvements in the provision and uptake of prepregnancy care, tighter glycemic targets (HbA1c <6.5% [48 mmol/mol]), and increased specialization of antenatal diabetes care. Over the past 15 years, the number of specialist diabetes maternity clinics has decreased (from 231 to 155), while the incidence of pregnancies complicated by pregestational diabetes has increased. This means that there are now twice as many women with pregestational diabetes per clinic. Balanced with the increased antenatal clinic workload is a more focused concentration of staff expertise, which may be important as diabetes technology becomes more complex. It also means a need for more efficient antenatal diabetes care provision.

Current antenatal diabetes care pathways involve frequent clinic visits with an obstetrician, endocrinologist, diabetes educator, diabetes specialist midwife, or diabetes dietitian typically every 2 weeks from 8 to 36 weeks’ gestation and weekly until delivery (6). This means 15 scheduled face-to-face visits, requiring women to take a morning, afternoon, or full day off work and/or arrange childcare provision. In addition, there are frequent between-visit contacts (face-to-face, telephone, and e-mail) with the diabetes educator for glycemic management. The increased use of technology has the potential to deliver antenatal diabetes care more efficiently to a larger number of women and more effectively in terms of optimizing day-to-day glucose control. In light of increasing demands on limited health care resources, cost-effective diabetes technologies, which enable women to effectively self-manage before, during, and between their pregnancies, are urgently needed. This article will review the progress in our understanding of the intensive glycemic treatment of T1D pregnancy, focusing in particular on the recent technological advances in continuous glucose monitoring (CGM) and automated insulin delivery.

While hyperglycemia at any stage of pregnancy is associated with increased risk of neonatal complications, early pregnancy (the first 6–7 weeks), when organogenesis of major cardiac and neural tube structures occurs, is particularly crucial (7). Hyperglycemia, lack of folic acid supplementation, and taking potentially harmful medications (ACE inhibitors, statins) all contribute to increased rates of cardiac and neural tube anomalies. Tennant et al. (8) demonstrated that even in normally formed offspring without congenital anomaly, the increased risk of fetal and infant death is still largely moderated by maternal glycemic control. Periconception HbA1c levels >6.3% (43 mmol/mol) are associated with increased odds of congenital anomaly, and levels >6.6% (49 mmol/mol) are associated with fetal and neonatal death (8). Therefore, prepregnancy care is universally recommended to optimize maternal glycemia and reduce the most serious adverse pregnancy outcomes.

For women with T1D, the key components of prepregnancy care include preconception folic acid supplementation, the lowest HbA1c level that is safely achievable, and stopping potentially harmful medications (6). Our own work has shown that even with intensive antenatal support, women who do not attend prepregnancy care clinics do not achieve the same glycemic control as those who began before pregnancy (9). A concern is that prepregnancy care clinics benefit educated, advantaged women and fail to engage disadvantaged women who should be prioritized to ensure fairness. This is supported by the U.K. National Pregnancy in Diabetes (NPID) audit data, which show that among women living in the most advantaged areas, 75% take 5 mg preconception folic acid and 25% achieve the National Institute for Health and Care Excellence–recommended HbA1c target of <6.5% (48 mmol/mol) (3). Less than 10% of women from disadvantaged areas achieve the same glycemic target. More innovative community-based approaches engaging primary care physicians and using mobile health technology to raise women’s awareness of and engagement with prepregnancy care should be targeted at disadvantaged groups. Evaluation of a Danish app designed for women attending a diabetes pregnancy clinic reported that 75% of women had downloaded it, with almost half having engaged with it, prior to pregnancy (10).

It is widely accepted that HbA1c levels can be misleading when evaluating individual rather than population-level glucose control, as individuals with the same mean glucose can have different HbA1c values (11). Furthermore, HbA1c does not reflect intra- and interday glycemic excursions or quantify the postprandial hyperglycemia that contributes to fetal and neonatal complications. In women without diabetes, HbA1c is lower during pregnancy due to lower mean glucose, increased erythropoiesis, and shortened red cell life span (12). Extant literature suggests an artifactual lowering of HbA1c (∼0.5%) in pregnancy that is unrelated to maternal glycemia (13,14). However, despite its well-recognized gestational limitations (13,14), HbA1c is routinely used to assess maternal glycemia, potentially providing false reassurance to women and clinicians.

Data from our own clinic population of 102 T1D pregnant women found that the relationship between mean self-monitored blood glucose (SMBG) and HbA1c changed in early pregnancy. We found an even larger fall in HbA1c, of ∼1% (11 mmol/mol), between 12 and 20 weeks’ gestation that was also unrelated to maternal glucose control (Table 1). This means that a mean SMBG of 144 mg/dL (8.0 mmol/L) was associated with an HbA1c of 6.8% (51 mmol/mol) at 12 weeks and with an HbA1c level of 5.9% (41 mmol/mol) at 24 weeks, supporting the view that maternal HbA1c does not adequately reflect antenatal glycemic control (15).

Table 1

Changes in mean glucose, HbA1c, and percentage of capillary glucose levels >8 mmol/L (144 mg/dL) during T1D pregnancy

GestationMean glucose, mmol/LHbA1c, %HbA1c, mmol/mol% Glucose >8 mmol/L (144 mg/dL)
Booking  7.6 ± 1.4 60 ± 15  
12 weeks 7.8 ± 1.5 6.8 ± 1.0 51 ± 11 39.9 ± 15.7 
16 weeks 7.4 ± 1.2 6.3 ± 0.9 45 ± 10 35.2 ± 13.0 
20 weeks 7.3 ± 1.1 5.9 ± 0.8 41 ± 9 35.1 ± 13.0 
24 weeks 7.4 ± 1.0 5.8 ± 0.8 40 ± 9 35.2 ± 13.8 
28 weeks 7.3 ± 1.1 5.9 ± 0.8 41 ± 9 35.4 ± 13.3 
32 weeks 7.1 ± 1.0 5.9 ± 0.7 41 ± 8 33.8 ± 12.3 
GestationMean glucose, mmol/LHbA1c, %HbA1c, mmol/mol% Glucose >8 mmol/L (144 mg/dL)
Booking  7.6 ± 1.4 60 ± 15  
12 weeks 7.8 ± 1.5 6.8 ± 1.0 51 ± 11 39.9 ± 15.7 
16 weeks 7.4 ± 1.2 6.3 ± 0.9 45 ± 10 35.2 ± 13.0 
20 weeks 7.3 ± 1.1 5.9 ± 0.8 41 ± 9 35.1 ± 13.0 
24 weeks 7.4 ± 1.0 5.8 ± 0.8 40 ± 9 35.2 ± 13.8 
28 weeks 7.3 ± 1.1 5.9 ± 0.8 41 ± 9 35.4 ± 13.3 
32 weeks 7.1 ± 1.0 5.9 ± 0.7 41 ± 8 33.8 ± 12.3 

Data are mean ± SD.

‡These data are provided courtesy of Dr. Rosemary Temple and Dr. Katharine Stanley from the Norfolk and Norwich University Hospital NHS Foundation Trust antenatal diabetes clinic. The mean gestational age at the first (booking) visit was 7.2 ± 2.2 weeks, with 80 (78%) women seen at ≤8 weeks and 90 (88%) at ≤10 weeks.

HbA1c can be now calculated according to estimated average glucose (eAG) from CGM measures rather than measured by laboratory assay. Law et al. (16) found that during pregnancy, a 1% (11 mmol/mol) difference in maternal HbA1c is equivalent to 12 mg/dL (0.66 mmol/L) in average glucose levels. Thus, HbA1c is also associated with lower eAG in pregnancy, leading to a recommendation that pregnancy-specific calculations be used and reported. This difference between the pregnancy-specific and nonpregnancy eAG increases with increasing HbA1c values, which means that HbA1c values can be particularly misleading and falsely reassuring in those with suboptimal glycemic control. These authors suggest that patients and clinicians should aim for eAG of 6.4–6.7 mmol/L to minimize risk of LGA (16). This practical solution is applicable to SMBG users (with memory glucometers to calculate a mean glucose over 7–14 days) as well as insulin pump and CGM users.

Novel Markers of Glycemic Control

1,5-Anhydroglucitiol has been approved by the U.S. Food and Drug Administration (FDA) for intermediate assessment of glycemic control and may have a role during pregnancy, reflecting postprandial glycemic excursions (17). More recently, the complement inhibitor glycated CD59 has been suggested as a novel marker of glycemic control (18). It may be useful for identifying pregnancies complicated by hyperglycemia and for identifying mothers at increased risk of delivering LGA newborns. Approximately 75% of LGA infants were born to mothers with a sevenfold increase in median CD59 levels but apparently “normal” results during an oral glucose tolerance test (18). Whether or not these novel markers will prove useful in clinical practice is unclear, especially as direct CGM measures become increasingly used.

CGM Metrics

The vast array of direct CGM metrics facilitates more detailed objective measurement of day-to-day glucose control but complicates our definition of “good” glycemic control. HbA1c lends itself to clear thresholds, with almost all professional organizations suggesting targets of <6.5% (48 mmol/mol) during pregnancy. Similar targets for CGM measures have not yet been established. Outside of pregnancy, there is a move toward standardization of definitions for time spent in the target glucose range (time in range [TiR]) and for both hypo- and hyperglycemic excursions (19). This allows for between-study comparisons and is particularly relevant for T1D pregnancy where CGM data are limited. Suggested CGM metrics for pregnancy in relation to the recent international consensus statement (19) are proposed in Table 2.

Table 2

Consensus statement recommendations with suggestions for T1D pregnancy

NonpregnantPregnant
Data sufficiency 70–80% of possible CGM data 70–80% of possible CGM data 
Data duration Minimum of 2 weeks Not determined 
Time blocks 24 h (midnight to midnight) 24 h (midnight to midnight) 
Nighttime 2400–0600 h  2400–0600 or 2300–0700 h 
Daytime 0600–2400 h 0600–2400 or 0700–2300 h 
Overall control Mean glucose Mean glucose 
% TiR 70–180 mg/dL (3.9–10.0 mmol/L) 70–140 mg/dL (3.9–7.8 mmol/L) 
 70–140 mg/dL (3.9–7.8 mmol/L) 63–140 mg/dL (3.5–7.8 mmol/L) 
Hyperglycemia, level 1 >180 mg/dL (10.0 mmol/L) >140 mg/dL (7.8 mmol/L) 
Hyperglycemia, level 2 >250 mg/dL (13.9 mmol/L) >180 mg/dL (10.0 mmol/L) 
High glucose exposure High blood glucose index High blood glucose index 
Hypoglycemia, level 1 <70–54 mg/dL (3.9–3.0 mmol/L) <70–54 mg/dL (3.9–3.0 mmol/L) or 63–50 mg/dL (3.5–2.8 mmol/L) 
Hypoglycemia, level 2 <54 mg/dL (3.0 mmol/L) <54 mg/dL (3.0 mmol/L) or <50 mg/dL (2.8 mmol/L) 
Hypoglycemia, level 3 Severe hypoglycemia Severe hypoglycemia 
Low glucose exposure Low blood glucose index Low blood glucose index 
Hypoglycemic event 15-min duration 15-min duration 
Prolonged hypoglycemia 120 min 120 min 
Glycemic variability   
 SD Not reported <25 mg/dL suggested 
 Coefficient of variation <36% (stable glycemia) <36% (stable glycemia) 
For research purposes   
 AUC AUC level 1 and 2 hyperglycemia AUC level 1 and 2 hyperglycemia 
 AUC level 1 and 2 hypoglycemia AUC level 1 and 2 hypoglycemia 
 Composite glycemic trial outcomes HbA1c or TiR and level 2 hypoglycemia TiR and level 2 hypoglycemia 
 Broader composite outcomes HbA1c or TiR + hypoglycemia + lipids + BP + weight gain TiR + hypoglycemia + gestational weight gain + obstetric/neonatal outcomes 
NonpregnantPregnant
Data sufficiency 70–80% of possible CGM data 70–80% of possible CGM data 
Data duration Minimum of 2 weeks Not determined 
Time blocks 24 h (midnight to midnight) 24 h (midnight to midnight) 
Nighttime 2400–0600 h  2400–0600 or 2300–0700 h 
Daytime 0600–2400 h 0600–2400 or 0700–2300 h 
Overall control Mean glucose Mean glucose 
% TiR 70–180 mg/dL (3.9–10.0 mmol/L) 70–140 mg/dL (3.9–7.8 mmol/L) 
 70–140 mg/dL (3.9–7.8 mmol/L) 63–140 mg/dL (3.5–7.8 mmol/L) 
Hyperglycemia, level 1 >180 mg/dL (10.0 mmol/L) >140 mg/dL (7.8 mmol/L) 
Hyperglycemia, level 2 >250 mg/dL (13.9 mmol/L) >180 mg/dL (10.0 mmol/L) 
High glucose exposure High blood glucose index High blood glucose index 
Hypoglycemia, level 1 <70–54 mg/dL (3.9–3.0 mmol/L) <70–54 mg/dL (3.9–3.0 mmol/L) or 63–50 mg/dL (3.5–2.8 mmol/L) 
Hypoglycemia, level 2 <54 mg/dL (3.0 mmol/L) <54 mg/dL (3.0 mmol/L) or <50 mg/dL (2.8 mmol/L) 
Hypoglycemia, level 3 Severe hypoglycemia Severe hypoglycemia 
Low glucose exposure Low blood glucose index Low blood glucose index 
Hypoglycemic event 15-min duration 15-min duration 
Prolonged hypoglycemia 120 min 120 min 
Glycemic variability   
 SD Not reported <25 mg/dL suggested 
 Coefficient of variation <36% (stable glycemia) <36% (stable glycemia) 
For research purposes   
 AUC AUC level 1 and 2 hyperglycemia AUC level 1 and 2 hyperglycemia 
 AUC level 1 and 2 hypoglycemia AUC level 1 and 2 hypoglycemia 
 Composite glycemic trial outcomes HbA1c or TiR and level 2 hypoglycemia TiR and level 2 hypoglycemia 
 Broader composite outcomes HbA1c or TiR + hypoglycemia + lipids + BP + weight gain TiR + hypoglycemia + gestational weight gain + obstetric/neonatal outcomes 

Optimal sensor accuracy is considered as MARD ≤10% in pregnant and nonpregnant settings. AUC, area under the curve.

CGM accuracy is most commonly assessed using the mean absolute relative difference (MARD) between CGM and SMBG values. A MARD ≤10% is optimal for research and for clinical decision making and is applicable regardless of pregnancy status. The percentage TiR target range is usually 70–140 mg/dL or 63–140 mg/dL during T1D pregnancy, lower than the 70–180 mg/dL (3.9–10.0 mmol/L) range outside pregnancy. The consensus statement suggests categorizing the level of hypoglycemia into levels of increasing severity from level 1 to level 3. Level 1 is an alert of potential impending hypoglycemia. Level 2 is a glucose level <54 mg/dL (3.0 mmol/L) with or without symptoms. Level 3 is a severe hypoglycemia episode requiring assistance. As fasting glucose is lower during pregnancy and sensor accuracy is lower in the hypoglycemic range, paying attention to the lower threshold is important when quantifying hypoglycemia. Diabetes pregnancy clinicians need to consider whether to adapt the standardized international thresholds or to establish pregnancy-specific ones using the more stringent T1D pregnancy thresholds of 63 mg/dL (3.5 mmol/L) and 50 mg/dL (2.8 mmol/L).

It has been a decade since we first described the longitudinal CGM measures during T1D pregnancy. These data provided some of the first insights into direct fetal exposure to maternal hyperglycemia (20). Our original CGM data indicated that T1D women spent only 43% TiR of 70–140 mg/dL (10.4 h/day) in early pregnancy, rising to 56% (13.5 h/day) in late pregnancy (20). Despite enormous efforts, they still spent 33% of the time (8 h/day) hyperglycemic (>140 mg/dL) during the third trimester. Maternal hypoglycemia (<70 mg/dL) was widespread, ∼13% (3 h/day). Very comparable findings were reported in women with T1D from a Danish CGM trial using SMBG measures: 58% in target (70–144 mg/dL), 14% below 70 mg/dL, and 28% above the slightly higher 144 mg/dL hyperglycemic threshold (21).

Comparing these data with the recent Continuous Glucose Monitoring in Women With Type 1 Diabetes in Pregnancy Trial (CONCEPTT) data, the third trimester TiR (63–140 mg/dL) is largely unchanged: 61% in the SMBG group (22) (Fig. 1A). The most striking difference is the substantial reduction in maternal hypoglycemia (Fig. 1B). The CONCEPTT CGM group spent only 3% time below 63 mg/dL compared with 13–14% time below 70 mg/dL in the earlier CGM studies. This is not only due to differences in sensor accuracy or hypoglycemia thresholds, as clinical hypoglycemia events and severe hypoglycemia episodes were also reduced, most likely due to increasing use of insulin analogs (23,24).

Figure 1

CGM measures of women with T1D during pregnancy in CONCEPTT. A: Time in T1D pregnancy target range 63–140 mg/dL (3.5–7.8 mmol/L). The home glucose monitoring (HGM) control group spent 52% TiR at baseline (12.5 h/day), rising to 61% (14.6 h/day) at 34 weeks’ gestation. The CGM group spent 52% TiR at baseline (12.5 h/day), rising to 68% (16.3 h/day) at 34 weeks’ gestation. B: Time spent hyperglycemic (>140 mg/dL [7.8 mmol/L]). The HGM control group spent 40% time hyperglycemic at baseline (9.6 h/day), reducing to 32% (7.7 h/day) at 34 weeks’ gestation. The CGM group spent 39% time hyperglycemic at baseline (9.4 h/day), reducing to 27% (6.5 h/day) at 34 weeks’ gestation; P = 0.03. C: Time spent hypoglycemic (<63 mg/dL [3.5 mmol/L]). The HGM control group spent 8% time hypoglycemic at baseline (1.9 h/day), reducing to 4% (1.0 h/day) at 34 weeks’ gestation. The CGM group spent 6% time hypoglycemic at baseline (1.4 h/day), reducing to 3% (0.7 h/day) at 34 weeks’ gestation; P = 0.10. This figure is reproduced with permission from Yamamoto and Murphy (48).

Figure 1

CGM measures of women with T1D during pregnancy in CONCEPTT. A: Time in T1D pregnancy target range 63–140 mg/dL (3.5–7.8 mmol/L). The home glucose monitoring (HGM) control group spent 52% TiR at baseline (12.5 h/day), rising to 61% (14.6 h/day) at 34 weeks’ gestation. The CGM group spent 52% TiR at baseline (12.5 h/day), rising to 68% (16.3 h/day) at 34 weeks’ gestation. B: Time spent hyperglycemic (>140 mg/dL [7.8 mmol/L]). The HGM control group spent 40% time hyperglycemic at baseline (9.6 h/day), reducing to 32% (7.7 h/day) at 34 weeks’ gestation. The CGM group spent 39% time hyperglycemic at baseline (9.4 h/day), reducing to 27% (6.5 h/day) at 34 weeks’ gestation; P = 0.03. C: Time spent hypoglycemic (<63 mg/dL [3.5 mmol/L]). The HGM control group spent 8% time hypoglycemic at baseline (1.9 h/day), reducing to 4% (1.0 h/day) at 34 weeks’ gestation. The CGM group spent 6% time hypoglycemic at baseline (1.4 h/day), reducing to 3% (0.7 h/day) at 34 weeks’ gestation; P = 0.10. This figure is reproduced with permission from Yamamoto and Murphy (48).

Close modal

However, the fetal exposure to maternal hyperglycemia is essentially unchanged in routine care. Women using CGM spent 5% (1.2 h/day) less time hyperglycemic (27 vs. 32%) at 34 weeks’ gestation (Fig. 1C) (22). However, pregnant women using only SMBG still spent 45% time (11.5 h/day) hyperglycemic at 24 weeks’ gestation, which likely explains why LGA rates remain persistently high (25).

CGM allows for unprecedented characterization of the day-to-day, within-day, and between-day glycemic variability, with a vast array of potential metrics assessing the amplitude, frequency, and duration of deviations above and below target range. Most of these measures are highly correlated and dependent on mean glucose, making it difficult to accurately assess the independent contribution of glycemic variability beyond overall glucose control. Importantly, the risk of maternal and/or fetal complications increases both with amplitude and with duration of the glycemic excursion, yet most of the traditional glycemic variability metrics ignore the time axis of CGM data. More sophisticated statistical methods (time series, functional data analysis) may provide new insights into which time periods (day vs. night) to target glycemic interventions (26).

Even with specialist prepregnancy care clinics and motivated advantaged attendees, only a third to a half of attendees achieve target HbA1c levels before conception (9,27). The recent CONCEPTT specifically evaluated the effectiveness of CGM for improving glucose control in women planning pregnancy. Previous studies have evaluated the role of retrospective and/or intermittent real-time CGM during pregnancy (21,28) but did not include women planning pregnancy. Full details of the study protocol and results are published elsewhere (22,29). In brief, there were 53 women assigned to CGM and 57 to SMBG (control). They had a long duration of diabetes (18 and 19 years, respectively), with correspondingly high rates of microvascular complications (37%). There was widespread use (∼75%) of insulin pump therapy. Importantly, most women were already overweight or obese, with only 40% having a preconception BMI of ≤25 kg/m2.

The frequency of CGM use (Guardian or MiniMed MiniLink system) was reasonably high over the first 3 months (median 6.7 days), with some attenuation over 6 months (median 6.2 days). HbA1c fell in both groups (7.6% [59 mmol/mol] to 7.1% [54 mmol/mol] and 7.3% [56 mmol/mol] in the CGM and SMBG groups, respectively), with 50% of CGM and 40% of SMBG women reaching target HbA1c levels. However, as both groups improved, the between-group differences were small and not statistically significant. Likewise, although the direct CGM measures favored CGM, with greater reductions in mean glucose (from 158 to 144 mg/dL [8.8 to 8.0 mmol/L]) and more of an increase in TiR of 63–140 mg/dL (from 42 to 48%), these were not statistically significant. Thirty-four women conceived (17 CGM, 17 control), and their glucose control (mean HbA1c at confirmed pregnancy 6.9 vs. 7.0% [52 vs. 53 mmol/mol]) and pregnancy outcomes did not differ. Although the numbers of women who conceived were very small, there was 3-kg less gestational weight gain in the CGM group (10.4 vs. 13.4 kg from confirmed pregnancy to 34 weeks’ gestation), suggesting that CGM users may have been making substantial dietary adjustments.

Despite adjustments to diet and glycemia, 75% of these babies had a composite poor outcome, which included miscarriage, LGA, neonatal hypoglycemia, hyperbilirubinemia, respiratory distress, and NICU admission. These data do not mean that CGM and/or sensor-augmented pump therapy are ineffective, but they highlight that despite increased use of advanced diabetes technology, euglycemia in T1D pregnancy remains elusive before and during pregnancy.

An unexpected finding from CONCEPTT was that the treatment effect of CGM was comparable in women using insulin pumps and multiple daily injections (MDI). This supports recent data supporting CGM use in MDI users outside of pregnancy (30). Indeed, in CONCEPTT, the CGM MDI users had the best overall glucose control, with almost 70% TiR (63–140 mg/dL), 26% time >140 mg/dL, and 3% time <63 mg/dL (Table 3). Furthermore, their glycemic variability was comparable to that of insulin pump users. Some authors have suggested that insulin pump therapy has not yet lived up to the expectations of health care professionals (31). The randomized studies in pregnancy are outdated, with older pumps and MDI regimes and small sample sizes, and are not applicable to current clinical practice (32). More recent descriptions are observational and subject to bias and confounding factors (31,33). Pumps are now in such widespread clinical use among women of reproductive years that an adequately powered, randomized trial would not be feasible in most antenatal diabetes clinics, with appropriate diabetes technology infrastructure and educator expertise.

Table 3

CGM measures among women in CONCEPTT using insulin pump and MDI during pregnancy

Pump users (N = 98)
10–11 weeks’ gestation
34–35 weeks’ gestation
CGM (N = 50)Control (N = 48)CGM (N = 35)Control (N = 37)
Mean glucose, mg/dL 131 ± 22 133 ± 22 121 ± 18 126 ± 16 
% TiR* 53 ± 12 54 ± 14 66 ± 13 62 ± 14 
% Time >140 mg/dL 39 (26–47) 39 (29–49) 27 (20–37) 32 (27–41) 
% Time <63 mg/dL 8 (3–13) 6 (3–10) 3 (1–7) 4 (2–7) 
Hypoglycemia episodes** 0.8 (0.6–1.0) 0.7 (0.5–0.9) 0.5 (0.3–0.8) 0.5 (0.4–0.7) 
Coefficient of variation, % 42 (37–47) 40 (36–46) 31 (28–37) 35 (29–40) 
SD, mmol/L
 
3.0 (2.5–3.4)
 
3.1 (2.5–3.6)
 
2.2 (1.8–2.5)
 
2.4 (2.0–3.0)
 
MDI users (N = 116)
 
 10–11 weeks’ gestation
 
34–35 weeks’ gestation
 
 CGM (N = 57)
 
Control (N = 59)
 
CGM (N = 42)
 
Control (N = 40)
 
Mean glucose, mg/dL 131 ± 22 139 ± 18 121 ± 14 126 ± 7.0 
% TiR* 50 ± 13 50 ± 13 69 ± 13 61 ± 17 
% Time >140 mg/dL 39 (30–49) 41 (34–51) 26 (17–36) 31 (24–39) 
% Time <63 mg/dL 8 (5–17) 6 (2–12) 3 (1–6) 5 (2–9) 
Hypoglycemia episodes** 0.8 (0.6–1.0) 0.7 (0.3–0.9) 0.5 (0.3–0.8) 0.5 (0.3–0.8) 
Coefficient of variation, % 43 (39–48) 43 (36–49) 33 (28–37) 34 (29–38) 
SD, mmol/L 3.2 (2.7–3.6) 3.2 (2.7–3.9) 2.2 (1.8–2.5) 2.3 (2.0–2.8) 
Pump users (N = 98)
10–11 weeks’ gestation
34–35 weeks’ gestation
CGM (N = 50)Control (N = 48)CGM (N = 35)Control (N = 37)
Mean glucose, mg/dL 131 ± 22 133 ± 22 121 ± 18 126 ± 16 
% TiR* 53 ± 12 54 ± 14 66 ± 13 62 ± 14 
% Time >140 mg/dL 39 (26–47) 39 (29–49) 27 (20–37) 32 (27–41) 
% Time <63 mg/dL 8 (3–13) 6 (3–10) 3 (1–7) 4 (2–7) 
Hypoglycemia episodes** 0.8 (0.6–1.0) 0.7 (0.5–0.9) 0.5 (0.3–0.8) 0.5 (0.4–0.7) 
Coefficient of variation, % 42 (37–47) 40 (36–46) 31 (28–37) 35 (29–40) 
SD, mmol/L
 
3.0 (2.5–3.4)
 
3.1 (2.5–3.6)
 
2.2 (1.8–2.5)
 
2.4 (2.0–3.0)
 
MDI users (N = 116)
 
 10–11 weeks’ gestation
 
34–35 weeks’ gestation
 
 CGM (N = 57)
 
Control (N = 59)
 
CGM (N = 42)
 
Control (N = 40)
 
Mean glucose, mg/dL 131 ± 22 139 ± 18 121 ± 14 126 ± 7.0 
% TiR* 50 ± 13 50 ± 13 69 ± 13 61 ± 17 
% Time >140 mg/dL 39 (30–49) 41 (34–51) 26 (17–36) 31 (24–39) 
% Time <63 mg/dL 8 (5–17) 6 (2–12) 3 (1–6) 5 (2–9) 
Hypoglycemia episodes** 0.8 (0.6–1.0) 0.7 (0.3–0.9) 0.5 (0.3–0.8) 0.5 (0.3–0.8) 
Coefficient of variation, % 43 (39–48) 43 (36–49) 33 (28–37) 34 (29–38) 
SD, mmol/L 3.2 (2.7–3.6) 3.2 (2.7–3.9) 2.2 (1.8–2.5) 2.3 (2.0–2.8) 

Data are mean ± SD and median (interquartile range) as appropriate. These data are adapted from the CONCEPTT results publication (22).

*TiR for T1D pregnancy was defined as 63–140 mg/dL (3.5–7.8 mmol/L).

**Hypoglycemia episodes were defined as CGM levels <63 mg/dL for at least 20 min. Distinct episodes were counted only if separated by at least 30 min.

A pragmatic randomized controlled trial evaluating the relative effectiveness of pumps over MDI in nonpregnant U.K. participants found a mean change in HbA1c at 24 months of −0.8% and −0.4% (−9 and −4.5 mmol/mol) for pump and MDI users, respectively (34). After adjustment for confounders, the difference in favor of pump users was smaller (−0.24% [−2.7 mmol/mol]) and not statistically significant. The accompanying psychosocial evaluation found that pump users showed greater improvement in diabetes treatment satisfaction, more dietary freedom, and fewer diabetes hassles (35). This supports an emerging point of view that pump therapy may be more beneficial for psychosocial outcomes than for glycemic outcomes. CGM, with all its alarms and annoyances, may be more effective for focusing the mind on minimizing glycemic excursions.

While there have been incremental improvements in sensor accuracy and usability over the past decade, there have been three key developments in terms of CGM accessibility. These include a growing evidence base regarding the clinical effectiveness in MDI users (30), the FDA approval of CGM measures for insulin dosing, and the introduction of the FreeStyle Libre, the first factory-calibrated intermittent glucose monitoring system (30,36). Data from the T1D Exchange registry showed very little CGM use (<10%) compared with widespread insulin pump use (60%) during 2013–2014 (37). However, the increasing recognition that CGM benefits both pump and MDI users means that CGM is becoming more applicable for day-to-day glycemic management for a wider patient population (22,30).

The FDA endorsement of real-time CGM (specifically the Dexcom G5) for replacing SMBG is also an important step forward. While calibration and some checking of SMBG is still recommended, particularly during hypoglycemia, exercise, and driving, both the clinical and cost-effectiveness of CGM will be greatly enhanced with sensors accurate enough for premeal insulin dosing. The longer duration of newer generation sensors, lasting for up to 10 days, will benefit users and payers and further improve CGM accessibility.

The real game-changer is the introduction of the FreeStyle Libre (Abbott Diabetes Care) intermittent glucose monitoring system (36). The sensors last 14 days without the need for additional SMBG calibration tests. It is so easy and intuitive that it is marketed directly to consumers (without the need for physician recommendation) and neither patient nor staff training is required. When it was first introduced in the U.K., there was such overwhelming demand that the early supply was inadequate, requiring new manufacturing premises to be built. It is also the first CGM to obtain a specific label for use during pregnancy. A study among 74 pregnant women (39 with gestational diabetes mellitus, 24 with T1D, and 11 with T2D) across 13 sites (9 U.K. and 4 Austrian) demonstrated that, as expected, sensor accuracy, assessed over 14 days at various gestational ages, is comparable (MARD of ∼12%) in pregnant and nonpregnant users (38). We found similar agreement between CGM sensor accuracy for pregnant and nonpregnant users of the Navigator sensor when compared with plasma glucose (39). While other sensors may not have specific licenses for use in pregnancy, it seems that accuracy issues are sensor specific and applicable to pregnant and nonpregnant users. The professional version of the FreeStyle Libre (available in the U.S.) will improve the documentation of glycemic profiles in pregnancy. It provides 14-day masked glucose profiles without requiring SMBG, making it an accessible and affordable research tool.

The obvious appeal of the FreeStyle Libre in clinical practice is the lack of alarms and burdens that are associated with real-time CGM. Data from the T1D Exchange indicate that use of real-time CGM wanes over time and that optimal use as seen in the setting of randomized controlled trials is not widespread (37). If real-time CGM is a high-cost, high-maintenance diabetes companion, the FreeStyle Libre is its low-cost, low-maintenance alternative. Although it still provides real-time continuous glucose data on demand (by scanning the reader), it does not alarm to alert users of out-of-range or rapidly changing glucose levels.

Real-world data provided by the device manufacturers suggest that the benefits on glucose control increase with more frequent glucose checking (36). The estimated HbA1c reductions were most marked (from 8 to 6.7% [64 to 50 mmol/mol]) in users with the most frequent glucose checks (increasing from 4 to 48 checks per day). The average user performed 16.3 checks daily, which is clearly higher than an average SMBG user. The role of FreeStyle Libre in T1D pregnancy is yet to be determined, as is, in particular, whether it is as effective as real-time CGM for improving neonatal outcomes (22). The FreeStyle Libre may be an excellent “entry level” technology for patients not wanting the demands of real-time CGM. It may also help clinicians to determine which patients are candidates for more advanced CGM, sensor-augmented/threshold suspend pumps, and automated insulin delivery systems.

Technological advances in CGM have made the promise of automated insulin delivery, also known as artificial pancreas or closed-loop insulin delivery, a potential clinical therapeutic reality. A recent systematic review and meta-analysis of 585 participants from 27 outpatient studies found consistent glycemic improvements with 12.6% increased TiR (3 h/day) across a variety of closed-loop systems (40). However, thus far, most of the improvement is in overnight glycemia and the between-group differences are very dependent on the level of glucose control in the comparator arm. We also found that overnight closed-loop insulin delivery was associated with a 15% higher overnight time in target range (75 vs. 60%) during T1D pregnancy (41). Preliminary data from this first home closed-loop study suggest feasibility of day-and-night closed-loop therapy throughout pregnancy, including in the hospital during labor and delivery. Most women chose to continue closed-loop therapy after the randomized trial, with generally high levels of satisfaction despite frequent alarms and technical glitches (42).

A subsequent randomized evaluation of day-and-night closed-loop therapy in T1D pregnancy found that closed-loop was as effective as, but not superior to, sensor-augmented pump therapy (43). Women spent 60% TiR during both interventions in the second trimester, but closed-loop therapy reduced the extent and duration of hypoglycemia, suggesting that it is potentially safer. We also found that women who entered pregnancy with better glucose control achieved a fairly constant 70–75% TiR (63–140 mg/dL) across pregnancy. Women with suboptimal glucose control (HbA1c >7.5% [53 mmol/mol]) never caught up and only achieved 65–70% TiR after 28 weeks’ gestation. They took longer to become confident using the study devices and never quite reached the same glycemic control, suggesting that diabetes self-management education is still required for optimal use of hybrid closed-loop systems (11). Other investigators have also warned that automated insulin delivery should not be considered a hands-off option and should be accompanied by appropriate high-quality dietary and diabetes education (11). This is particularly important during pregnancy when the tight postprandial glucose targets require meticulous attention to carbohydrate estimation and bolusing at least 15–30 min before eating (44).

As most women with T1D are now entering pregnancy overweight (35%) or obese (25%), the importance of optimal dietary intake cannot be overstated (3,22). Higher prepregnancy BMI and higher gestational weight gain, independent of maternal BMI and glycemic control, is associated with increasing neonatal birth weight (45). We found that almost half of the total daily carbohydrate intake of U.K. participants in CONCEPTT was from ultra-processed, high-sugar (biscuits, chocolate, and confectionary) sources. This means that these well-educated, motivated pregnant women ate ∼90 g per day of fast-acting added sugars, with little if any nutritional value. There are no fast-acting insulin analogs currently available (or on the horizon) that will be able to safely match these dietary intakes during pregnancy (46).

A recent systematic review and meta-analysis found 19 randomized controlled trials of dietary interventions in gestational diabetes mellitus pregnancy (47). While dietary advice is an increasingly recognized important component of clinical management in T1D pregnancy, there are no evidence-based data to guide clinical practice. A high-quality dietary intervention trial (perhaps of Mediterranean diet or low glycemic index carbohydrates or interventions to minimize consumption of ultra-processed foods) is required to guide optimal implementation of CGM, continuous subcutaneous insulin infusion, and closed-loop delivery in T1D pregnancy. While advanced diabetes technology may add an additional 5–10% time in target range (and possibly more for those with suboptimal glucose control) in T1D pregnancy, optimizing maternal dietary intake is imperative to minimize the immediate and longer-term consequences of postprandial hyperglycemia.

See accompanying articles, pp. 1572 and 1579.

Acknowledgments. I acknowledge the invaluable support from the multidisciplinary diabetes teams at Cambridge University Hospitals NHS Foundation Trust, Cambridge, U.K. (Caroline Byrne, Katy Davenport, Jeannie Grisoni, Kerry Stubbington, Ruth Harding, and Charlotte Patient), Ipswich Hospital NHS Trust, Ipswich, U.K. (Sue Mitchell, Anne Scott, Damian Morris, and Duncan Fowler), and Norfolk and Norwich University Hospital, Norwich, U.K. (Rosemary Temple, Katherine Duffield, Katharine Stanley, and Mike Sampson). I also thank Gerry Rayman (Ipswich Hospital NHS Trust, Ipswich, U.K.) and the Ipswich diabetes team for teaching me (from scratch) the principles of diabetes self-management education and Timothy Skinner (Charles Darwin University, Australia) for encouragement to always “ask a better question.” I thank Dr. Rosemary Temple for kitchen table discussions that have guided my entire diabetes pregnancy research. Likewise, I am indebted to Elisabeth Mathiesen (Rigshospitalet, Copenhagen, Denmark) for her mentorship, friendship, support, and support for truly collaborative research. I thank the entire CONCEPTT team, in particular Denice Feig (University of Toronto, Canada) for an effective, fun, and productive partnership. Finally, I thank the inspiring efforts of Dr. Roman Hovorka (University of Cambridge, Cambridge, U.K.) and the University of Cambridge artificial pancreas team (in particular Janet Allen, Daniella Elleri, Gosia Wilinska, Kavita Kumareswaran, Leelarathna Lalantha, Hood Thabit, Sara Hartnell, and Zoe Stewart) and Dr. Eleanor Scott (University of Leeds, Leeds, U.K.) for making automated insulin delivery a potential therapeutic option in T1D pregnancy.

Funding. H.R.M. conducts independent research supported by the National Institute for Health Research (NIHR) (CDF-2013-06-035). She has also received grant income from the U.K. Efficacy Mechanism Evaluation programme (funded by the Medical Research Council and NIHR), Diabetes UK, JDRF, and Diabetes Research & Wellness Foundation.

The views expressed in this publication are those of the author and not necessarily those of the National Health Service, NIHR, or the U.K. Department of Health.

Duality of Interest. H.R.M. serves on the Medtronic European Scientific Advisory Board.

Prior Presentation. Parts of this article were presented at the 78th Scientific Sessions of the American Diabetes Association, Orlando, FL, 23 June 2018.

1.
Persson
M
,
Norman
M
,
Hanson
U
.
Obstetric and perinatal outcomes in type 1 diabetic pregnancies: a large, population-based study
.
Diabetes Care
2009
;
32
:
2005
2009
[PubMed]
2.
Macintosh
MC
,
Fleming
KM
,
Bailey
JA
, et al
.
Perinatal mortality and congenital anomalies in babies of women with type 1 or type 2 diabetes in England, Wales, and Northern Ireland: population based study
.
BMJ
2006
;
333
:
177
[PubMed]
3.
Murphy
HR
,
Bell
R
,
Cartwright
C
, et al
.
Improved pregnancy outcomes in women with type 1 and type 2 diabetes but substantial clinic-to-clinic variations: a prospective nationwide study
.
Diabetologia
2017
;
60
:
1668
1677
[PubMed]
4.
Feig
DS
,
Hwee
J
,
Shah
BR
,
Booth
GL
,
Bierman
AS
,
Lipscombe
LL
.
Trends in incidence of diabetes in pregnancy and serious perinatal outcomes: a large, population-based study in Ontario, Canada, 1996–2010
.
Diabetes Care
2014
;
37
:
1590
1596
[PubMed]
5.
Mackin
ST
,
Nelson
SM
,
Kerssens
JJ
, et al.;
SDRN Epidemiology Group
.
Diabetes and pregnancy: national trends over a 15 year period
.
Diabetologia
2018
;
61
:
1081
1088
[PubMed]
6.
National Institute for Health and Care Excellence. Diabetes in pregnancy: management from preconception to the postnatal period [Internet], February 2015. Available from https://www.nice.org.uk/guidance/ng3. NICE guideline. Accessed 31 December 2017
7.
Mills
JL
,
Baker
L
,
Goldman
AS
.
Malformations in infants of diabetic mothers occur before the seventh gestational week: implications for treatment
.
Diabetes
1979
;
28
:
292
293
[PubMed]
8.
Tennant
PW
,
Glinianaia
SV
,
Bilous
RW
,
Rankin
J
,
Bell
R
.
Pre-existing diabetes, maternal glycated haemoglobin, and the risks of fetal and infant death: a population-based study
.
Diabetologia
2014
;
57
:
285
294
[PubMed]
9.
Murphy
HR
,
Roland
JM
,
Skinner
TC
, et al
.
Effectiveness of a regional prepregnancy care program in women with type 1 and type 2 diabetes: benefits beyond glycemic control
.
Diabetes Care
2010
;
33
:
2514
2520
[PubMed]
10.
Nørgaard
SK
,
Nichum
VL
,
Barfred
C
, et al.
Use of the smartphone application “Pregnant with Diabetes.”
Dan Med J
2017
;
64
:
A5417
11.
Kahkoska
AR
,
Mayer-Davis
EJ
,
Hood
KK
,
Maahs
DM
,
Burger
KS
.
Behavioural implications of traditional treatment and closed-loop automated insulin delivery systems in type 1 diabetes: applying a cognitive restraint theory framework
.
Diabet Med
2017
;
34
:
1500
1507
[PubMed]
12.
Radder
JK
,
van Roosmalen
J
.
HbA1c in healthy, pregnant women
.
Neth J Med
2005
;
63
:
256
259
[PubMed]
13.
Nielsen
LR
,
Ekbom
P
,
Damm
P
, et al
.
HbA1c levels are significantly lower in early and late pregnancy
.
Diabetes Care
2004
;
27
:
1200
1201
[PubMed]
14.
Herranz
L
,
Saez-de-Ibarra
L
,
Grande
C
,
Pallardo
LF
.
Non-glycemic-dependent reduction of late pregnancy A1C levels in women with type 1 diabetes
.
Diabetes Care
2007
;
30
:
1579
1580
[PubMed]
15.
Kerssen
A
,
de Valk
HW
,
Visser
GH
.
Do HbA1c levels and the self-monitoring of blood glucose levels adequately reflect glycaemic control during pregnancy in women with type 1 diabetes mellitus?
Diabetologia
2006
;
49
:
25
28
[PubMed]
16.
Law
GR
,
Gilthorpe
MS
,
Secher
AL
, et al
.
Translating HbA1c measurements into estimated average glucose values in pregnant women with diabetes
.
Diabetologia
2017
;
60
:
618
624
[PubMed]
17.
Dungan
KM
,
Buse
JB
,
Largay
J
, et al
.
1,5-Anhydroglucitol and postprandial hyperglycemia as measured by continuous glucose monitoring system in moderately controlled patients with diabetes
.
Diabetes Care
2006
;
29
:
1214
1219
[PubMed]
18.
Ghosh
P
,
Luque-Fernandez
MA
,
Vaidya
A
, et al
.
Plasma glycated CD59, a novel biomarker for detection of pregnancy-induced glucose intolerance
.
Diabetes Care
2017
;
40
:
981
984
[PubMed]
19.
Danne
T
,
Nimri
R
,
Battelino
T
, et al
.
International consensus on use of continuous glucose monitoring
.
Diabetes Care
2017
;
40
:
1631
1640
[PubMed]
20.
Murphy
HR
,
Rayman
G
,
Duffield
K
, et al
.
Changes in the glycemic profiles of women with type 1 and type 2 diabetes during pregnancy
.
Diabetes Care
2007
;
30
:
2785
2791
[PubMed]
21.
Secher
AL
,
Ringholm
L
,
Andersen
HU
,
Damm
P
,
Mathiesen
ER
.
The effect of real-time continuous glucose monitoring in pregnant women with diabetes: a randomized controlled trial
.
Diabetes Care
2013
;
36
:
1877
1883
[PubMed]
22.
Feig
DS
,
Donovan
LE
,
Corcoy
R
, et al.;
CONCEPTT Collaborative Group
.
Continuous glucose monitoring in pregnant women with type 1 diabetes (CONCEPTT): a multicentre international randomised controlled trial
.
Lancet
2017
;
390
:
2347
2359
[PubMed]
23.
Mathiesen
ER
,
Kinsley
B
,
Amiel
SA
, et al.;
Insulin Aspart Pregnancy Study Group
.
Maternal glycemic control and hypoglycemia in type 1 diabetic pregnancy: a randomized trial of insulin aspart versus human insulin in 322 pregnant women
.
Diabetes Care
2007
;
30
:
771
776
[PubMed]
24.
Mathiesen
ER
,
Hod
M
,
Ivanisevic
M
, et al.;
Detemir in Pregnancy Study Group
.
Maternal efficacy and safety outcomes in a randomized, controlled trial comparing insulin detemir with NPH insulin in 310 pregnant women with type 1 diabetes
.
Diabetes Care
2012
;
35
:
2012
2017
[PubMed]
25.
Kerssen
A
,
de Valk
HW
,
Visser
GH
.
Increased second trimester maternal glucose levels are related to extremely large-for-gestational-age infants in women with type 1 diabetes
.
Diabetes Care
2007
;
30
:
1069
1074
[PubMed]
26.
Law
GR
,
Ellison
GT
,
Secher
AL
, et al
.
Analysis of continuous glucose monitoring in pregnant women with diabetes: distinct temporal patterns of glucose associated with large-for-gestational-age infants
.
Diabetes Care
2015
;
38
:
1319
1325
27.
Egan
AM
,
Danyliv
A
,
Carmody
L
,
Kirwan
B
,
Dunne
FP
.
A prepregnancy care program for women with diabetes: effective and cost saving
.
J Clin Endocrinol Metab
2016
;
101
:
1807
1815
[PubMed]
28.
Murphy
HR
,
Rayman
G
,
Lewis
K
, et al
.
Effectiveness of continuous glucose monitoring in pregnant women with diabetes: randomised clinical trial
.
BMJ
2008
;
337
:
a1680
[PubMed]
29.
Feig
DS
,
Asztalos
E
,
Corcoy
R
, et al.;
CONCEPTT Collaborative Group
.
CONCEPTT: Continuous Glucose Monitoring in Women with Type 1 Diabetes in Pregnancy Trial: a multi-center, multi-national, randomized controlled trial - study protocol
.
BMC Pregnancy Childbirth
2016
;
16
:
167
[PubMed]
30.
Beck
RW
,
Riddlesworth
T
,
Ruedy
K
, et al.;
DIAMOND Study Group
.
Effect of continuous glucose monitoring on glycemic control in adults with type 1 diabetes using insulin injections: the DIAMOND randomized clinical trial
.
JAMA
2017
;
317
:
371
378
[PubMed]
31.
Abell
SK
,
Suen
M
,
Pease
A
, et al
.
Pregnancy outcomes and insulin requirements in women with type 1 diabetes treated with continuous subcutaneous insulin infusion and multiple daily injections: cohort study
.
Diabetes Technol Ther
2017
;
19
:
280
287
[PubMed]
32.
Mukhopadhyay
A
,
Farrell
T
,
Fraser
RB
,
Ola
B
.
Continuous subcutaneous insulin infusion vs intensive conventional insulin therapy in pregnant diabetic women: a systematic review and metaanalysis of randomized, controlled trials
.
Am J Obstet Gynecol
2007
;
197
:
447
456
[PubMed]
33.
Kallas-Koeman
MM
,
Kong
JM
,
Klinke
JA
, et al
.
Insulin pump use in pregnancy is associated with lower HbA1c without increasing the rate of severe hypoglycaemia or diabetic ketoacidosis in women with type 1 diabetes
.
Diabetologia
2014
;
57
:
681
689
[PubMed]
34.
Heller
S
,
White
D
,
Lee
E
, et al
.
A cluster randomized trial, cost-effectiveness analysis and psychosocial evaluation of insulin pump therapy compared with multiple injections during flexible intensive insulin therapy for type 1 diabetes: the REPOSE Trial
.
Health Technol Assess
2017
;
21
:
1
278
[PubMed]
35.
Lawton
J
,
Kirkham
J
,
Rankin
D
, et al.;
REPOSE Group
.
Who gains clinical benefit from using insulin pump therapy? A qualitative study of the perceptions and views of health professionals involved in the Relative Effectiveness of Pumps over MDI and Structured Education (REPOSE) trial
.
Diabet Med
2016
;
33
:
243
251
[PubMed]
36.
Dunn
TC
,
Xu
Y
,
Hayter
G
,
Ajjan
RA.
Real-world flash glucose monitoring patterns and associations between self-monitoring frequency and glycaemic measures: a European analysis of over 60 million glucose tests
.
Diabetes Res Clin Pract
2018
;
137
:
37
46
37.
Miller
KM
,
Foster
NC
,
Beck
RW
, et al.;
T1D Exchange Clinic Network
.
Current state of type 1 diabetes treatment in the U.S.: updated data from the T1D Exchange clinic registry
.
Diabetes Care
2015
;
38
:
971
978
[PubMed]
38.
Scott
E
,
Bilous
RW
,
Kautzky-Willer
A
.
Accuracy, user acceptability, and safety evaluation of FreeStyle Libre Flash Glucose Monitoring System when used by pregnant women with diabetes
.
Diabetes Technol Ther
2018
;
20
:
180
188
39.
Murphy
HR
,
Elleri
D
,
Allen
JM
, et al
.
Closed-loop insulin delivery during pregnancy complicated by type 1 diabetes
.
Diabetes Care
2011
;
34
:
406
411
[PubMed]
40.
Weisman
A
,
Bai
JW
,
Cardinez
M
,
Kramer
CK
,
Perkins
BA
.
Effect of artificial pancreas systems on glycaemic control in patients with type 1 diabetes: a systematic review and meta-analysis of outpatient randomised controlled trials
.
Lancet Diabetes Endocrinol
2017
;
5
:
501
512
[PubMed]
41.
Stewart
ZA
,
Wilinska
ME
,
Hartnell
S
, et al
.
Closed-loop insulin delivery during pregnancy in women with type 1 diabetes
.
N Engl J Med
2016
;
375
:
644
654
[PubMed]
42.
Farrington
C
,
Stewart
ZA
,
Barnard
K
,
Hovorka
R
,
Murphy
HR
.
Experiences of closed-loop insulin delivery among pregnant women with type 1 diabetes
.
Diabet Med
2017
;
34
:
1461
1469
[PubMed]
43.
Stewart
ZA
,
Wilinska
ME
,
Hartnell
S
, et al
.
Day-and-night closed-loop insulin delivery in a broad population of pregnant women with type 1 diabetes: a randomized controlled crossover trial
.
Diabetes Care
. 13 March 2018 [Epub ahead of print]. DOI: 10.2337/dc17-2534
44.
Murphy
HR
,
Elleri
D
,
Allen
JM
, et al
.
Pathophysiology of postprandial hyperglycaemia in women with type 1 diabetes during pregnancy
.
Diabetologia
2012
;
55
:
282
293
[PubMed]
45.
Secher
AL
,
Parellada
CB
,
Ringholm
L
,
Asbjörnsdóttir
B
,
Damm
P
,
Mathiesen
ER
.
Higher gestational weight gain is associated with increasing offspring birth weight independent of maternal glycemic control in women with type 1 diabetes
.
Diabetes Care
2014
;
37
:
2677
2684
[PubMed]
46.
Russell-Jones
D
,
Bode
BW
,
De Block
C
, et al
.
Fast-acting insulin aspart improves glycemic control in basal-bolus treatment for type 1 diabetes: results of a 26-week multicenter, active-controlled, treat-to-target, randomized, parallel-group trial (onset 1)
.
Diabetes Care
2017
;
40
:
943
950
[PubMed]
47.
Han
S
,
Middleton
P
,
Shepherd
E
,
Van Ryswyk
E
,
Crowther
CA
.
Different types of dietary advice for women with gestational diabetes mellitus
.
Cochrane Database Syst Rev
2017
;
2
:
CD009275
[PubMed]
48.
Yamamoto
JM
,
Murphy
HR
.
Emerging technologies for the management of type 1 diabetes in pregnancy
.
Curr Diab Rep
2018
;
18
:
4
[PubMed]
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.