A Feasibility Study of Paired Continuous Glucose Monitoring Intrapartum and in the Newborn in Pregnancies Complicated by Type 1 Diabetes

Abstract

Aim:

To describe the continuous glucose monitoring (CGM) profiles of type 1 diabetes (T1D) offspring in the early neonatal period and its association with maternal intrapartum glucose control.

Methods:

A prospective observational study of T1D pregnant women and their neonatal offspring. Women had a CGM sensor inserted 2–3 days before delivery. Infants had a masked CGM sensor inserted as soon as possible after delivery. Maternal glycemic outcomes were time-in-target (70–140 mg/dL [3.9–7.8 mmol/L]), hyperglycemia >140 mg/dL (7.8 mmol/L), and mean CGM glucose during the 24 h preceding delivery. Neonatal outcomes included lowest recorded blood glucose concentration, and CGM measures (glucose <47 mg/dL [2.6 mmol/L], time-in-target (47–144 mg/dL [2.6–8.0 mmol/L]), glucose standard deviation [SD]) during the first 72 h of life.

Results:

Data were available for 16 mother–infant pairs. Mothers had a mean age (SD) 32.3 (4.3) years, T1D duration 17.6 (6.8) years, first antenatal glycated hemoglobin 7.4 (0.8)% (57 [8.5] mmol/mol). In the 24 h preceding delivery, mothers spent mean (SD) 72 (20)% time-in-target (70–140 mg/dL [3.9–7.8 mmol/L]), 19 (15)% time >140 mg/dL (7.8 mmol/L), and 9 (9)% time <70 mg/dL (3.9 mmol/L) with mean (SD) CGM glucose 113 (9) mg/dL (6.3 [0.7] mmol/L). Fifteen infants (93.8%) had ≥1 blood glucose concentration <47 mg/dL (2.6 mmol/L), and five had ≥1 blood glucose concentration <18 mg/dL (1.0 mmol/L). The mean infant CGM glucose on days 1, 2, and 3 of life was 63 (14), 67 (13), 76 (11) mg/dL (3.5 [0.8], 3.7 [0.7], and 4.2 [0.6] mmol/L). Four infants (25%) spent >50% time with CGM glucose levels <47 mg/dL (2.6 mmol/L) on day 1.

Conclusions:

CGM detected widespread neonatal hypoglycemia, even among mothers with good intrapartum glucose control.

Introduction

Infants of mothers with type 1 diabetes (T1D) are at increased risk of complications, of which neonatal hypoglycemia is the most common neonatal concern. Approximately two in three infants born after a pregnancy complicated by T1D have neonatal hypoglycemia, and 30%–40% require admission to neonatal intensive care units (NICU).^1,2 National audit data from England and Wales have highlighted the high prevalence of hypoglycemia as a cause of term admissions and made it a priority target in trying to reduce term admissions and keep mothers and their babies together.³

Although neonatal hypoglycemia is often asymptomatic, it can also have long-term consequences including neurocognitive impairment and lower academic achievement.^4
–6 A population-based study of 1395 children found that even transient newborn hypoglycemia (a single blood glucose measurement <34, 40, or 44 mg/dL [1.9, 2.2, or 2.5 mmol/L]) was associated with lower probability of literacy and mathematics proficiency in fourth-grade tests.⁶ In addition, a recent study of infants identified to be at risk of neonatal hypoglycemia showed that those with hypoglycemia, detected clinically with routine glucose monitoring and/or asymptomatic episodes detected with continuous glucose monitoring (CGM), had a 3- to 4-fold increased risk of low executive function at 4.5 years of age.⁷

Although neonatal hypoglycemia is reported to be more common in infants of T1D mothers with poorer antenatal and intrapartum glycemic control,^8,9 this relationship is inconsistently observed.^10

–16 The timing of the relationship is also poorly understood. Randomized trial data confirm that CGM improves maternal antenatal glycemic control and neonatal health outcomes, with the CONCEPTT trial having described a halving in the odds ratio for neonatal hypoglycemia.¹ However, it is unclear whether the reduction in neonatal hypoglycemia and in NICU admissions was solely attributed to maternal antenatal glycemia or whether intrapartum glycemia during labor and delivery also contributed.

There remains controversy as to the optimal screening strategy and diagnostic thresholds for detection and management of neonatal hypoglycemia.^17

–20 CGM has the potential to inform optimal management of glucose dysregulation in the neonatal period²¹ but CGM data in T1D offspring are limited, particularly in the first 24–72 h of life. The aim of this study was to explore the feasibility of using CGM to describe the CGM glucose profiles of T1D offspring in the early neonatal period. A secondary objective was to investigate the relationship between maternal intrapartum and neonatal postpartum glucose control in T1D pregnancy.

Research Design and Methods

Maternal participants

Pregnant CGM users were approached about potential participation in the study during their third trimester. Those who wished to participate provided informed consent for themselves and assent for their neonate, with parental written consent for neonates confirmed within 24 h of delivery. Inclusion criteria included maternal familiarity with masked or real-time CGM, intensive insulin therapy using either multiple daily injections of insulin (MDI) or insulin pump therapy, and singleton pregnancy. Exclusion criteria included known congenital anomaly or neonates with severe respiratory distress.

Pregnant women with T1D had a CGM sensor inserted 2–3 days before anticipated delivery. Participants already using the Guardian^® REAL-Time or MiniMed Minilink^® CGM (both Medtronic, Northridge, CA) continued their usual CGM. Participants using the Freestyle Navigator II (Abbott Diabetes Care) were fitted with an additional masked CGM sensor for measurements used in this study (iPro™2 Professional CGM; Medtronic). Participants were asked to measure capillary glucose concentration using their routine glucometer and to record at least four daily readings.

Neonatal participants

As soon as possible after delivery (aiming for within 4 h), an appropriately trained member of the research team fitted the newborn infant with a masked CGM device (iPro™2 Professional CGM; Medtronic). The sensors were inserted into the lateral aspect of the infant's thigh. While infants had the sensor in situ, their routine blood glucose monitoring samples were used to calibrate the CGM. Concurrent sampling was performed for measurement of infant blood ketones. There were no additional blood glucose samples taken from infants as a result of this study.

Study design

This is a prospective observational study of pregnant women with T1D and their neonatal offspring. All infant CGM data were masked and therefore not available to parents, clinicians, or researchers. Infants otherwise received standard neonatal clinical care. Standard neonatal care required that all T1D offspring were screened for hypoglycemia, clinically defined as blood glucose concentration <47 mg/dL (2.6 mmol/L). A standard treatment protocol was used for blood glucose concentration <47 mg/dL (2.6 mmol/L). Buccal dextrose gel was not used during this study. All infants required three prefeed blood glucose levels >47 mg/dL (2.6 mmol/L) before hospital discharge. The CGM sensor remained in situ until the infant was ready for hospital discharge, for up to a maximum of 7 days duration.

Outcomes

Maternal and neonatal demographics details were obtained from the medical records, including gestational age at delivery, birthweight, need for supplemental feeds, intravenous (IV) dextrose infusion, or NICU admission. Maternal glycemic outcomes included percentage time-in-target range (70–140 mg/dL [3.9–7.8 mmol/L]), hyperglycemia >140 mg/dL (7.8 mmol/L), and mean CGM glucose concentration during the 24 h preceding delivery. Neonatal outcomes included the lowest recorded blood glucose concentration (measured on either blood gas [Cobas b 221; Roche Diagnostics Limited, United Kingdom], or point-of-care meters [Nova StatStrip^® meters; Novabiomedical, Waltham, MA]), percentage time with sensor glucose concentration <47 mg/dL (2.6 mmol/L), percentage time-in-target range (sensor glucose 47–144 mg/dL [2.6–8.0 mmol/L]), and standard deviation (SD) of sensor glucose within the first 48 h of life. The clinical study protocol was approved by the Health Research Authority, East of England Regional Ethics Committee (14/EE/0001).

Statistical analysis

Glucose outcomes were calculated with GStat Version 2.2 software (University of Cambridge, Cambridge, United Kingdom). Statistical analyses were conducted using SPSS, R and Stata version 14.1. The percentage of time-in-target range, time hypoglycemia, and time hyperglycemia were summarized as mean (SD) or median (interquartile range or IQR) where appropriate. Student's t-tests and Mann–Whitney Wilcoxon tests were performed as appropriate and presented with P-values and/or 95% confidence intervals. A P-value of <0.05 was considered significant.

Results

Twenty-two pregnant women with T1D were recruited, 21 participants had both maternal and infant CGM sensors applied (one pair could not be included because there was insufficient CGM equipment available at the time of delivery). The CGM sensor failed to collect any data from five infants. CGM data were therefore available for 16 mother–infant pairs.

Mothers had a mean age (SD) of 32.3 (4.3) years, T1D duration of 17.6 (6.8) years, first antenatal glycated hemoglobin (HbA_1c) 7.4 (0.8)% (57 [8.5] mmol/mol), and body mass index of 26.1 (4.1) kg/m². Eight mothers (50%) used insulin pump therapy, 4 (25%) used MDI and 4 (25%) used hybrid closed-loop (CL) insulin delivery systems. Full details of the baseline maternal characteristics are provided (Supplementary Table S1; Supplementary Data are available at https://www.liebertpub.com/suppl/doi/10.1089/dia.2018.0021).

Maternal glycemia

In the 24 h before infant delivery, mothers spent a mean (SD) of 72 (20)% time (17.3 h/day) within target range (70–140 mg/dL [3.9–7.8 mmol/L]), 19 (15)% time (4.6 h/day) >140 mg/dL (7.8 mmol/L), and 9 (9)% time (2.2 h/day) <70 mg/dL (3.9 mmol/L) (Table 1). Their mean (SD) sensor glucose was 113 (13) mg/dL (6.3 [0.7] mmol/L). There were no significant differences in percentage time-in-target or mean CGM glucose between mothers treated with insulin pump, MDI, or CL: (mean [SD] time-in-target 66.0 [22.1]% pump, 77.0 [20.1]% MDI, 78.5 [13.4]% CL, P = 0.51); mean [SD] glucose 113 [16] mg/dL (6.3 [0.9] mmol/L) pump, 112 [16] mg/dL (6.2 [0.9] mmol/L) MDI, 115 [2] mg/dL (6.4 [0.1] mmol/L) CL, P = 0.94). There was, however, substantial intraindividual variability in maternal glucose control (Table 1).

Table 1.

Maternal Glucose Control as Measured by Continuous Glucose Monitoring in the 24 h Before Delivery, by Participant

Participant no.	Percentage time in target range ^a	Percentage time above target	Percentage time below target	Mean glucose (mg/dL [mmol/L])
1	100	0	0	94 [5.2]
2	65	26	9	122 [6.8]
3	53	40	7	126 [7.0]
4	50	50	0	139 [7.7]
5	57	12	31	95 [5.3]
6	35	40	25	121 [6.7]
7	68	15	17	122 [6.8]
8	91	0	9	92 [5.1]
9	94	3	3	97 [5.4]
10	65	20	15	113 [6.3]
11	54	31	15	126 [7.0]
12	85	15	0	113 [6.3]
13	96	4	0	108 [6.0]
14	73	16	11	108 [6.0]
15	94	6	0	119 [6.6]
16	70	24	6	115 [6.4]
Mean (SD)	72 (20)	19 (15)	9 (9)	113 (13) [6.3 (0.7)]

Maternal participants used a Medtronic Enlite sensor (either as part of an iPro, Guardian RT, or MiniLink system).

The maternal CGM glucose target range was 70–140 mg/dL [3.9–7.8 mmol/L].

CGM, continuous glucose monitoring; SD, standard deviation.

Obstetric and neonatal outcomes

Infants were delivered at a mean of 37 + 2 weeks' gestation. Two had vaginal deliveries and 14 delivered by cesarean section (9 elective, 5 emergency). Infant birthweight ranged from 2810 to 4675 g (mean [SD] 3887 [519] g). Six infants (37.5%) weighed >4000 g, and 2 (12.5%) weighed >4500 g (Supplementary Table S2). Six infants (37.5%) had a sex and gestational age-specific birthweight centile >90th percentile.

Neonatal glycemia

Fifteen newborns (93.8%) had at least one blood glucose concentration <47 mg/dL (2.6 mmol/L). The lowest recorded blood glucose concentrations ranged from 13 mg/dL (0.7 mmol/L) to 61 mg/dL (3.4 mmol/L), with a mean (SD) of 31 (14) mg/dL (1.7 [0.8] mmol/L) (Table 2). There was no correlation between lowest recorded blood glucose concentration and maternal mean glucose, time-in-target, or hyperglycemia in the 24 h before delivery (R ² = 0.007, P = 0.77 for mean CGM glucose, R ² = 0.03, P = 0.52 for time-in-target, R ² = 0.003, P = 0.85 for hyperglycemia).

Table 2.

Neonatal Delivery Characteristics and Glycemic Control by Conventional and Continuous Glucose Monitoring Measures, by Participant

						First 24 h of life			24–48 h of life
Participant no.	Gestation at delivery (weeks + days)	Birthweight (g)	Admission to NICU	IV dextrose	Lowest recorded blood glucose (mg/dL [mmol/L])	CGM glucose <47 mg/dL [2.6 mmol/L] (%)	Mean [SD] CGM glucose (mg/dL)	Mean [SD] CGM glucose (mmol/L)	CGM glucose <47 mg/dL [2.6 mmol/L] (%)	Mean [SD] CGM glucose (mg/dL)	Mean [SD] CGM glucose (mmol/L)
1	36 + 1	4490	Yes	Yes	18 [1.0]	0	77 [13]	4.3 [0.7]	0	74 [4]	4.1 [0.2]
2	37 + 3	3280	Yes	Yes	31 [1.7]	0	59 [5]	3.3 [0.3]	0	58 [4]	3.2 [0.2]
3	37 + 1	4095	Yes	Yes	61 [3.4]	0	88 [5]	4.9 [0.3]	0	92 [9]	5.1 [0.5]
4	38 + 3	4280	Yes	Yes	34 [1.9]	51	47 [2]	2.6 [0.1]	57	47 [5]	2.6 [0.3]
5	37 + 1	3370	No	No	41 [2.3]	0	54 [2]	3.0 [0.1]	0	54 [2]	3.0 [0.1]
6	37 + 0	4675	No	Yes	13 [0.7]	0	63 [5]	3.5 [0.3]	0	72 [13]	4.0 [0.7]
7	34 + 4	3815	Yes	Yes	16 [0.9]	0	67 [5]	3.7 [0.3]	0	70 [7]	3.9 [0.4]
8	38 + 1	3830	Yes	Yes	41 [2.3]	0	76 [5]	4.2 [0.3]	0	76 [4]	4.2 [0.2]
9	38 + 0	2810	No	No	40 [2.2]	60	47 [5]	2.6 [0.3]	46	49 [13]	2.7 [0.7]
10	37 + 0	3700	No	No	13 [0.7]	54	49 [5]	2.7 [0.3]	2	54 [9]	3.0 [0.5]
11	36 + 4	4035	Yes	Yes	18 [1.0]	0	86 [13]	4.8 [0.7]	0	85 [4]	4.7 [0.2]
12	38 + 4	3455	No	No	43 [2.4]	0	67 [7]	3.7 [0.4]	0	68 [7]	3.8 [0.5]
13	38 + 0	3880	No	No	31 [1.7]	0	56 [2]	3.1 [0.1]	0	61 [4]	3.4 [0.2]
14	38 + 1	3885	No	No	31 [1.7]	0	74 [9]	4.1 [0.5]	0	77 [9]	4.3 [0.5]
15	37 + 5	4635	No	No	41 [2.3]	100	40 [0]	2.2 [0.0]	50	58 [23]	3.2 [1.3]
16	38 + 5	3990	Yes	Yes	20 [1.1]	0	61 [4]	3.4 [0.2]	0	74 [7]	4.1 [0.4]

All infants used the iPro CGM system.

IV, intravenous; NICU, neonatal intensive care units.

For the newborn infants on days 1, 2, and 3 of life the mean (SD) CGM glucose was 63 (14) mg/dL (3.5 [0.8] mmol/L), 67 [13] mg/dL (3.7 [0.7] mmol/L), and 76 [11] mg/dL (4.2 [0.6] mmol/L), respectively. The percentage of time infants spent with CGM glucose levels <47 mg/dL (2.6 mmol/L) varied from 0% to 100% on day 1, 0%–57% on day 2, and 0%–21% on day 3 of life (median [IQR] for day 1 = 0 [0–12.8]%, day 2 = 0 [0–0.5], day 3 = 0 [0–6.0]). Four infants (25%) spent >50% of the time with CGM glucose levels <47 mg/dL (2.6 mmol/L) in the first 24 h. Three of these four infants were not treated with IV dextrose. Two of the three infants who had persistent hypoglycemia on day 3 were also not treated with IV dextrose but remained only enterally fed.

All infants were given supplementary feeds with either expressed breast milk or infant formula. In addition, 10 infants (62.5%) received treatment with IV dextrose. Of the 10 infants who had blood ketones measured, only one had a detectable ketone concentration at any time (4 mg/dL [0.2 mmol/L]). All other ketone measurements were <12 mg/dL (0.1 mmol/L).

The median (IQR) mean absolute relative difference (MARD) between CGM and glucometer readings was 9.7 (7.4, 12.6)%.

Across the 16 infants, there were 51 time-matched sensor and blood glucose results <63 mg/dL mmol/L and 21 time-matched values <50 mg/dL. The MARD was 9.1 (3.6, 16.4)% for values <63 mg/dL, and 9.1 (4.5 16.7)% for values <50 mg/dL. The mean absolute difference (MAD) was 5 mg/dL for values <63 and 4 mg/dL for values <50 mg/dL.

Infants treated with IV dextrose

Although the mothers of infants not treated with IV dextrose had lower mean glucose, higher time-in-target and less hyperglycemia, these differences did not reach statistical significance (Table 3). Infants treated with IV dextrose had a mean CGM glucose level 18 mg/dL (1.0 mmol/L) higher in the first 24 h of life than those who did not receive treatment with IV dextrose (P = 0.006; Table 3). The mean CGM glucose in IV dextrose-treated infants also remained higher on day 2 compared with infants not treated with dextrose (mean [SD] sensor glucose 72 [13] mg/dL [4.0 (0.7) mmol/L] vs. 58 [7] mg/dL [3.2 (0.4) mmol/L], P = 0.03). There was no difference in rates of IV dextrose administration according to mode of delivery (6 of 9 who delivered by elective cesarean section had IV treatment vs. 3 of 7 in those who delivered by emergency cesarean section or vaginal delivery; P = 0.61).

Table 3.

Characteristics of Infants Treated Versus Not Treated with Intravenous Dextrose

	Infant treated with IV dextrose (n = 10)	Infant not treated with IV dextrose (n = 6)	P
Maternal mean CGM glucose (mg/dL)^b	122 [115–126]	111 [101–113]	0.11
Maternal CGM time in target (%)^a	65.0 [53.0–70.0]	89.5 [70.0–94.0]	0.16
Maternal CGM time hyperglycemic (%)^b	25.0 [15.3–37.8]	9.0 [4.5–14.3]	0.12
Maternal CGM glucose SD (mg/dL)	28 [27–31]	27 [26–29]	0.62
Maternal first trimester HbA_1c (mmol/mol)^d	57.3 [9.2]	56.3 [8.1]	0.83
Maternal second trimester HbA_1c (mmol/mol)^d	49.6 [7.7]	50.2 [8.6]	0.89
Maternal third trimester HbA_1c (mmol/mol)^d	51.1 [9.1]	50.7 [6.9]	0.92
Gestation at delivery (weeks)^d	36.9 [1.2]	37.7 [0.6]	0.13
Infant birthweight (g)^b	4035 [3830–4280]	3642 [3391–3835]	0.14
Lowest recorded infant blood glucose (mg/dL)^b	25 [18–36]	34 [32–41]	0.31
Infant mean CGM glucose in first 24 h of life (mg/dL)^d	70 [13]	52 [11]	0.006
Mode of delivery^c			0.61
Elective cesarean section	6	3
Emergency cesarean section or vaginal delivery	3	4

Maternal participants used a Medtronic Enlite sensor (either as part of an iPro, Guardian RT, or MiniLink system). Maternal glucose measures are in the 24 h before delivery. All infants used the iPro CGM system.

The maternal CGM glucose target range was 70–140 mg/dL [3.9–7.8 mmol/L].

Presented as median [IQR].

Presented as number of infants.

Presented as mean [SD].

HbA_1c, glycated hemoglobin; IQR, interquartile range.

Infants with blood glucose concentration <18 mg/dL (1.0 mmol/L)

Although infants whose blood glucose did not drop to ≤18 mg/dL (1.0 mmol/L) seemed to have mothers with better glucose control (lower mean glucose, higher time-in-target, less glucose variability, lower HbA_1c in all trimesters), these differences did not reach statistical significance. Infants with at least one recorded blood glucose ≤18 mg/dL (1.0 mmol/L) were born an average of 1.5 weeks earlier than those who did not drop to ≤18 mg/dL (1.0 mmol/L) (P = 0.01; Table 4). Infants delivered by elective section were more likely to have at least one glucose value ≤1.0 mmol/L than those delivered by emergency section or vaginally (five of nine in those who delivered by elective cesarean section vs. zero of seven who delivered by emergency cesarean section or vaginal delivery; P = 0.03).

Table 4.

Characteristics of Infants With and Without One Blood Glucose Measurement ≤18 mg/dL [1.0 mmol/L]

	Infants with ≥1 blood glucose ≤18 mg/dL [1.0 mmol/L]	Infants who did not have a blood glucose ≤18 mg/dL [1.0 mmol/L]	P
Maternal mean glucose in 24 h before delivery (mmol/L)^b	121 [113–122]	115 [101–122]	0.71
Maternal time-in-target glucose during 24 h before delivery (%)^a	65.0 [54.0–68.0]	77.5 [59.0–93.3]	0.53
Maternal CGM glucose standard deviation during the 24 h before delivery (mg/dL)	29 [26–31]	26 [24–27]	0.71
First trimester maternal HbA_1c (mmol/mol)^d	63.2 [4.3]	53.8 [8.5]	0.12
Second trimester maternal HbA_1c (mmol/mol)^d	54.2 [4.8]	47.6 [8.2]	0.39
Third trimester maternal HbA_1c (mmol/mol)^d	53.8 [7.8]	49.5 [8.1]	0.58
Gestation at delivery (weeks)^d	36.3 [1.0]	37.8 [0.6]	0.01
Infant birthweight (g)^b	4035 [3815–4460]	13855 [3391–4068]	0.31
Mode of delivery^c			0.03
Elective cesarean section	5	4
Emergency cesarean section or vaginal delivery	0	7

The maternal CGM glucose target range was 70–140 mg/dL [3.9–7.8 mmol/L].

Presented as median [IQR].

Presented as number of infants.

Presented as mean [SD].

Discussion

In our study, neonatal hypoglycemia was near universal with 15 of 16 infants having at least one blood glucose measurement <47 mg/dL (2.6 mmol/L) and five having at least one measurement ≤18 mg/dL (1.0 mmol/L). Ten neonates were treated with IV dextrose, and they had a mean glucose that was 18 mg/dL (1.0 mmol/L) higher than untreated infants. A surprising and unexpected finding was that the infants with the most prolonged exposure to hypoglycemia were those who were managed on the routine postnatal ward.

Although some studies have found neonatal hypoglycemia to be associated with maternal intrapartum glucose control,^{13,14,22
–24} these data are inconsistent,²⁵ and the pathogenesis of neonatal hypoglycemia remains poorly understood. It is well established that preterm delivery and large for gestational age are important contributors to hypoglycemia in T1D offspring, and indeed we found that infants who had at least one glucose measurement <18 mg/dL (1.0 mmol/L) were larger and born earlier. Although prematurity itself is a risk factor for hypoglycemia, in our study the explanation for this finding is less clear. The group of infants with glucose values ≤18 mg/dL (1.0 mmol/L) may represent a cohort whose mothers had poorer glycemic control (as suggested by trends in HbA_1c throughout pregnancy) and these women are more likely to be induced early because of obstetric complications. Infants born earlier also tended to be larger (median birthweight centile >99 for infants born at or before 37 weeks vs. 93 for those born after 37 weeks gestation), perhaps because of poorer maternal glycemic control, which may mean that these infants were more likely to be hyperinsulinemic and therefore have lower blood glucose levels.

Our results are consistent with data suggesting that increasing maternal hyperglycemia is associated with neonatal hypoglycemia, although with small numbers the differences in the proportion of CGM time spent above and within target range were not statistically significant. Larger adequately powered studies of maternal intrapartum glucose control are required to understand the contribution of maternal intrapartum hyperglycemia to neonatal hypoglycemia.

Our study also demonstrated that four infants spent more than half of their first day of life with CGM glucose levels <47 mg/dL (2.6 mmol/L). This may be clinically important because even a single low glucose measurement has been associated with lower test proficiency achievement at 10 years of age.⁶ Furthermore, CGM-detected hypoglycemia is associated with lower executive functioning in preschool-aged children.⁷ Although routine clinical blood glucose monitoring did demonstrate some level of hypoglycemia, it did not reveal the severity or duration of hypoglycemia exposure and three of these were treated only with supplementary feeds. These data are consistent with those of Harris et al.,²⁶ who also found that CGM detected greater exposure to hypoglycemia than standard newborn heel-prick monitoring.

Little is known about what represents normoglycemia in the early neonatal period either in terms of target glucose levels or glucose fluctuations, in infants of women with diabetes or healthy newborns.^17,18 In utero, glucose concentrations are normally maintained between 72 and 108 mg/dL (4–6 mmol/L).²⁷ It is reported that neonatal glucose concentrations fall after birth, reaching a nadir at approximately 1-to-2 h after delivery, depending on gestational age and other factors.^28,29 Recent studies in healthy term infants suggest that mean glucose remains steady at ∼54 mg/dL (3.0 mmol/L) during the first 2 days of life, before gradually increasing to ∼72 mg/dL (4.0 mmol/L) thereafter.^30
–32 The mean glucose of 52 mg/dL (2.9 mmol/L) in infants treated with supplementary feeds is similar to that described in other newborn cohorts having glucose levels measured because of an anticipated risk of perinatal hypoglycemia.^32,33 However, clinical concerns remain in T1D offspring because of the potential impact of higher neonatal insulin levels suppressing the availability of alternative fuels.

Those infants treated with IV dextrose had a mean CGM glucose concentration ∼18 mg/dL (1.0 mmol/L) higher than the infants in our study who did not receive IV dextrose. Whether this is beneficial or represents overtreatment remains unclear. Further studies are required to better understand the role of alternative fuels, as well as glycemic exposure on longer term outcomes. The landmark trial of neonatal hypoglycemia and neurocognitive outcomes by McKinlay et al.⁷ included 28 mothers with pregestational diabetes, among whom 20 of 28 (71%) newborns experienced hypoglycemia. The data from this trial suggest a U-shaped relationship between neonatal glucose and subsequent poor neurocognitive outcomes. They also highlighted increased risk of neurosensory impairment in infants with neonatal hypoglycemia followed by higher glucose levels in the first 48 h.

Other neonatal CGM studies were performed on preterm infants without maternal CGM,²¹ or which included offspring of mothers with diabetes, either did not specify the diabetes type³⁴ or included only small numbers of T1D offspring (n = 3).³⁵ These studies did not include details of maternal glucose control and did not specify the timing of CGM initiation.

To our knowledge, ours is the first study to perform CGM both on women with T1D and on their infants. The masked CGM in the early neonatal period provided data under standard clinical conditions. We included a varied population of women, including mothers with long duration of T1D with a range of glycemic control and using a variety of insulin delivery methods. We made every effort to insert the CGM sensor as soon as possible in the first 4 h of life. We also included infants on routine postnatal wards in addition to those in the NICU.

This study also has limitations. Standard neonatal care required screening for hypoglycemia and clinical interventions will have impacted on neonatal glucose control. Other limitations include accuracy of CGM at low glucose levels and the fact that CGM is not designed for use in the early neonatal period. The sensors were inserted manually and use calibration algorithms designed for older children with higher glucose concentrations.³⁶ Our study was not designed to assess accuracy of the CGM system used, and blood glucose measurements were performed as clinically indicated rather than at specific times or glucose thresholds (given CGM data were masked to investigators and clinicians) and so accuracy results must be interpreted with caution. However, we found a median MAD of 9.7% between the CGM and glucometer readings overall, and 9.1% at values <63 mg/dL, suggesting acceptable accuracy.³⁷ This is consistent with previous studies of similar systems in infant populations, which also demonstrated acceptable accuracy using the same sensor as our study with a different monitor,³⁸ and using an older Medtronic sensor.³⁹ The latter study found that this older CGM system was less accurate at lower glucose levels, when >50% of readings had an error in excess of 10%. The study suggested that for readings <72 mg/dL, there was a slight bias for that sensor to overread.

It is accepted that CGM may also fail to calibrate if initiated during hypoglycemia. Within our study, the very early manual sensor insertion in this high-risk study population of T1D newborns may have contributed to the withdrawal of five infants for whom no CGM data were available. The rate of sensor failure in this study was higher than has been observed in other published cohorts,^21,34,35,40 which is likely a result of the earlier insertion, higher risk cohort, and newer sensors (with more difficult manual insertion) than were used in other studies. In some infants with low blood glucose levels, initiation of sensor recording was delayed until stable glycemia was obtained. This has been noted by others^26,35 and may result in an underestimation of exposure to neonatal hypoglycemia in early life. Finally, it is a small single-center feasibility study at a tertiary level academic hospital with a specialized diabetes pregnancy and full neonatal services.

We found a high burden of neonatal hypoglycemia, with 15 of the 16 infants having at least one glucose concentration <47 mg/dL (2.6 mmol/L), and five having at least one glucose <18 mg/dL (1.0 mmol/L). Our study adds to the body of evidence suggesting that CGM can detect hypoglycemia that goes undetected with routine heel-prick glucose monitoring and may be useful in maintaining euglycemia in neonates. We do not yet have an absolute threshold for what constitutes risk from neonatal hypoglycemia. Length of exposure to hypoglycemia is likely to be more important, and CGM is a useful tool to assess overall exposure in a way that single heel-prick tests cannot.

Our study also provides insight into the impact of interventions to optimize glucose levels that will help in determining the potential balance between under- and overtreatment. Larger cohorts using early CGM, but with longer term targeted neurocognitive assessments, are needed to understand the complexities and risk of impaired metabolic transition and our clinical interventions in these infants.

Footnotes

Acknowledgments

The authors thank all the pregnant women with type 1 diabetes who participated, along with their partners and families. They also acknowledge the invaluable support from the diabetes antenatal and postnatal care teams. The neonatal CGM was funded by a Diabetes Wellness and Research Foundation (DRWF) project grant. The CONCEPTT trial was funded by Juvenile Diabetes Research Foundation, Canadian Clinical Trials Network, and the National Institute for Health Research (HRM Career Development Fellowship, CDF-2013-06-035).

Authors' Contributions

H.R.M. and K.B. designed the study protocol. Z.A.S., L.T. screened, enrolled and consented trial maternal (Z.A.S. and H.R.M.) and neonatal (Z.A.S., L.T., and K.B.) participants. L.T., K.B., and Z.A.S. inserted CGM sensors. Z.A.S. analyzed and Z.A.S., H.R.M., K.B. interpretated data. Z.A.S., H.R.M., and K.B. wrote the article, which all authors critically reviewed. Z.A.S., K.B., and H.R.M. had full access to all the data and take responsibility for the integrity of the data, and accuracy of the analyses.

Author Disclosure Statement

H.R.M sits on the Medtronic Scientific Advisory Group. No competing financial interests exist.

References

Feig

, Donovan

, Corcoy

, et al.: Continuous glucose monitoring in pregnant women with type 1 diabetes (CONCEPTT): a multicentre international randomised controlled trial. Lancet, 2017; 390:2347–2359.

Murphy

, Bell

, Cartwright

, 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.

NHS Digital: National Pregnancy in Diabetes Annual Report 2016. 2017. https://digital.nhs.uk/data-and-information/publications/statistical/national-pregnancy-in-diabetes-audit/national-pregnancy-in-diabetes-annual-report-2016 (accessed March 12, 2018 ).

Burns

, Rutherford

, Boardman

, Cowan

: Patterns of cerebral injury and neurodevelopmental outcomes after symptomatic neonatal hypoglycemia. Pediatrics, 2008; 122:65–74.

Lucas

, Morley

, Cole

: Adverse neurodevelopmental outcome of moderate neonatal hypoglycaemia. BMJ, 1988; 297:1304–1308.

Kaiser

, Bai

, Gibson

, et al.: Association between transient newborn hypoglycemia and fourth-grade achievement test proficiency: a populatio-based study. JAMA Pediatr, 2015; 169:913–921.

McKinlay

, Alsweiler

, Anstice

, et al.: Association of neonatal glycemia with neurodevelopmental outcomes at 4.5 years. JAMA, 2017; 171:972.

Maresh

, Holmes

, Patterson

, et al.: Glycemic targets in the second and third trimester of pregnancy for women with type 1 diabetes. Diabetes Care, 2015; 38:34–42.

Inkster

, Fahey

, Donnan

, et al.: Poor glycated haemoglobin control and adverse pregnancy outcomes in type 1 and type 2 diabetes mellitus: systematic review of observational studies. BMC Pregnancy Childbirth, 2006; 6:30.

10.

Cordua

, Secher

, Ringholm

, et al.: Real-time continuous glucose monitoring during labour and delivery in women with type 1 diabetes. Diabet Med, 2013; 30:1374–1381.

11.

Agrawal

, Lui

, Gupta

: Neonatal hypoglycaemia in infants of diabetic mothers. J Pediatr Child Heal, 2000; 36:354–356.

12.

Sargent

, Roeder

, Ward

, et al.: Continuous subcutaneous insulin infusion versus multiple daily injections of insulin for the management of type 1 diabetes mellitus in pregnancy: association with neonatal chemical hypoglycemia. Am J Perinatol, 2015; 32:1324–1330.

13.

Stenninger

, Lindqvist

, Aman

, et al.: Continuous subcutaneous glucose monitoring system in diabetic mothers during labour and postnatal glucose adaptation in their infants. Diabet Med, 2008; 25:450–454.

14.

Curet

, Izquierdo

, Gilson

, et al.: Relative effects of antepartum and intrapartum maternal blood glucose levels on incidence of neonatal hypoglycemia. J Perinatol, 1997; 17:113–115.

15.

Drever

, Tomlinson

, Bai

, Feig

: Insulin pump use compared with intravenous insulin during labour and delivery: the INSPIRED observational cohort study. Diabet Med, 2016; 33:1253–1259.

16.

Kline

, Edwards

: Antepartum and intrapartum insulin management of type 1 and type 2 diabetic women: impact on clinically significant neonatal hypoglycemia. Diabetes Res Clin Pract, 2007; 77:223–230.

17.

Cornblath

, Hawdon

, Williams

, et al.: Controversies regarding definition of neonatal hypoglycemia: suggested operational thresholds. Pediatrics, 2000; 105:1141–1145.

18.

Rozance

, Hay

: Describing hypoglycemia—definition or operational threshold?. Early Hum Dev, 2010; 86:275–280.

19.

Aziz

, Dancey

, Canadian Paediatric Society Fetus and Newborn Committee: Screening guidelines for newborns at risk for low blood glucose. Paediatr Child Heal, 2004; 9:723–729.

20.

Harding

, Harris

, Hegarty

, et al.: An emerging evidence base for the management of neonatal hypoglycaemia. Early Hum Dev, 2017; 104:51–56.

21.

Galderisi

, Facchinetti

, Steil

, et al.: Continuous glucose monitoring in very preterm infants: a randomized controlled trial. Pediatrics, 2017; 140:e20171162.

22.

Pedersen

: Diabetes and pregnancy: blood sugar of newborn infants [PhD thesis]. Copenhagen: Danish Science Press, 1952.

23.

Carron Brown

, Kyne-Grzebalski

, Mwangi

, Taylor

: Effect of management policy upon 120 type 1 diabetic pregnancies: policy decisions in practice. Diabet Med, 1999; 16:573–578.

24.

Taylor

, Lee

, Kyne-Grzebalski

, et al.: Clinical outcomes of pregnancy in women with type 1 diabetes. Obstet Gynecol, 2002; 99:537–541.

25.

Yamamoto

, Benham

, Mohammad

, et al.: Intrapartum glycaemic control and neonatal hypoglycaemia in pregnancies complicated by diabetes mellitus: a systematic review. Diabet Med, 2018; 35:173–183.

26.

Harris

, Battin

, Weston

, Harding

: Continuous glucose monitoring in newborn babies at risk of hypoglycemia. J Pediatr, 2010; 157:198.e1–202.e1.

27.

Economides

, Nicolaides

: Blood glucose and oxygen tension levels in small-for-gestational-age fetuses. Am J Obstet Gynecol, 1989; 160:385–389.

28.

Srinivasan

, Pildes

, Cattamanchi

, et al.: Plasma glucose values in normal neonates: a new look. J Pediatr, 1986; 109:114–117.

29.

Kaiser

, Shasha

, Rozance

: Newborn plasma glucose concentration nadirs by gestational-age group. Neonatology, 2018; 113:353–359.

30.

Diwakar

, Sasidhar

: Plasma glucose levels in term infants who are appropriate size for gestation and exclusively breast fed. Arch Dis Child Fetal Neonatal Ed, 2002; 87:F46–F48.

31.

Hoseth

, Joergensen

, Ebbesen

, Moeller

: Blood glucose levels in a population of healthy, breast fed, term infants of appropriate size for gestational age. Arch Dis Child Fetal Neonatal Ed, 2000; 83:F117–F119.

32.

Guemes

, Rahman

, Hussain

: What is normal blood glucose?. Arch Dis Child, 2016; 101:569–574.

33.

McKinlay

, Alsweiler

, Ansell

, et al.: Neonatal glycemia and neurodevelopmental outcomes at 2 years. N Engl J Med, 2015; 373:1507–1518.

34.

Saw

H-P

, Yao

N-W

, Chiu

C-D

, Chen

: The value of real-time continuous glucose monitoring in premature infants of diabetic mothers. PLoS One, 2017; 12:e0186486.

35.

Wackernagel

, Dube

, Blennow

, Tindberg

: Continuous subcutaneous glucose monitoring is accurate in term and near-term infants at risk of hypoglycaemia. Acta Paediatr, 2016; 105:917–923.

36.

McKinlay

, Chase

, Dickson

, et al.: Continuous glucose monitoring in neonates: a review. Matern Heal Neonatol Perinatol, 2017; 3:18.

37.

Danne

, Nimri

, Battelino

, et al.: International consensus on use of continuous glucose monitoring. Diabetes Care, 2017; 40:1631–1640.

38.

Thomson

, Elleri

, Bond

, et al.: Targeting glucose control in preterm infants: pilot studies of continuous glucose monitoring. Arch Dis Child Fetal Neonatal Ed, 2018 [Epub ahead of print]; DOI:10.1136/archdischild-2018-314814.

39.

Beardsall

, Vanhaesebrouck

, Ogilvy-Stuart

, et al.: Validation of the continuous glucose monitoring sensor in preterm infants. Arch Dis Child Fetal Neonatal Ed, 2013; 98:F136–F140.

40.

Beardsall

, Vanhaesebrouck

, Ogilvy-Stuart

, et al.: Early insulin therapy in very-low-birth-weight-infants. N Engl J Med, 2008; 359:1873–1884.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB