Abstract
Aims
We compared sensor‐derived glycaemic metrics in pregnant women with type 1 diabetes (T1D) randomised to faster acting insulin aspart (faster aspart) or insulin aspart (IAsp).
Methods
A pre‐planned secondary analysis of the CopenFast trial included women with T1D using intermittently scanned continuous glucose monitoring (isCGM) during pregnancy. Glycaemic metrics, including time in range (TIRp, 3.5–7.8 mmol/L) and time below range in pregnancy (TBRp, <3.5 mmol/L), were evaluated in the intervals: from randomisation (median 9.5 weeks, interquartile range 9.0–11.0) to 21 weeks, from 22 to 33 weeks and from 34 to 37 weeks.
Results
In total, 113 (91%) of 124 women using isCGM in the original trial were included. At randomisation, glycaemic metrics were comparable in both groups. Women randomised to faster aspart achieved higher TIRp from 22 to 33 weeks (estimated treatment difference 5.1% [95% confidence interval 0.3; 9.7], p = 0.04) and mean TIRp >70% from randomisation to 21 weeks onwards, while this was achieved after 34 weeks in women randomised to IAsp. TBRp remained stable around 4% throughout pregnancy in both groups. One (2%) versus 5 (9%) experienced ≥1 severe hypoglycaemic event (odds ratio 0.93 [−0.2; −0.01], p = 0.04). Infant birthweight standard deviation score was lower in the faster aspart group (estimated treatment difference −0.5 [−0.9; −0.03], p = 0.04); however, this attenuated when adjusting for parity (p = 0.10).
Conclusions
Women using faster aspart achieved more TIRp and experienced less severe hypoglycaemia compared to women using IAsp. Infant birthweight was lower and thereby more appropriate in the faster aspart group; however, this attenuated when adjusting for parity.
Keywords: continuous glucose monitoring, hypoglycaemia, pregnancy, type 1 diabetes
What is already known?
Faster‐acting insulin as part (faster as part) is an improved formulation of insulin as part (IASP).
Achieving glycaemic metrics in target in early pregnancy is important for normal fetal growth in women with type 1 diabetes (T1D).
What this study has found?
Women using faster as part achieved more time in range and experienced less severe hypoglycaemia compared to women using IAsp.
Infant birthweight was lower in the faster as part group; however, this attenuated when adjusting for parity.
What are the implications of the?
Our findings support the use of faster as part over IAsp during pregnancy in women with T1D.
1. INTRODUCTION
During pregnancy, in women with type 1 diabetes, appropriate insulin treatment is essential to obtain strict glycaemic control, with a focus on postprandial glucose excursions, to reduce the risk of adverse pregnancy outcomes while preventing events of mild and severe hypoglycaemia. 1 Faster acting insulin aspart (faster aspart) is an improved formulation of conventional insulin aspart (IAsp), resulting in a higher early exposure and greater early glucose lowering effect without increasing hypoglycaemia compared to IAsp in non‐pregnant persons with type 1 diabetes. 2 , 3 The effect and safety of faster aspart compared to IAsp during pregnancy were evaluated in the CopenFast trial, a randomised controlled open‐label, single‐centre trial conducted from 2019 to 2023, in 203 women with type 1 and type 2 diabetes. 1 Treatment with faster aspart resulted in similar fetal growth and maternal HbA1c levels with less severe hypoglycaemia and no additional safety issues compared to IAsp. In women with type 1 diabetes, intermittently scanned continuous glucose monitoring (isCGM) data for 7 days, collected from the patient records three times during pregnancy, showed lower mean sensor glucose at 21 weeks in the faster aspart group compared to the IAsp group.
In a subgroup analysis of women with type 1 diabetes, a more favourable effect of faster aspart than IAsp on birthweight standard deviation (SD) score was seen both in women using multiple daily injections (MDI) and insulin pump therapy, although the difference between the faster aspart group and the IAsp group did not reach statistical significance. 1
Normal fetal growth is associated with achieving sensor‐derived glycaemic metrics in target in early pregnancy in women with type 1 diabetes. 4 , 5 , 6 However, women with type 1 diabetes often spend more time above the recommended glycaemic metrics targets in the majority of pregnancy 7 , 8 , 9 and the sensor‐derived glycaemic metrics targets are often not reached until the final weeks of pregnancy, if at all. 4 , 9
In Denmark, all women with type 1 diabetes, not already using continuous glucose monitoring (CGM), are offered an isCGM during pregnancy planning and pregnancy. 10 isCGM measures interstitial glucose concentrations every 60 s and reports a glucose value every 15 min 11 and offers easily accessible glycaemic metrics via smartphone and thereby a unique insight to daily glycaemic patterns. 12 No studies have investigated how the use of faster aspart or IAsp influences continuously measured sensor‐derived glycaemic metrics in women with type 1 diabetes during pregnancy.
We hypothesised that use of faster aspart leads to more favourable glycaemic metrics with less hypoglycaemia compared to IAsp during pregnancy. Therefore, the aim of this study was to compare sensor‐derived glycaemic metrics during pregnancy in women with type 1 diabetes randomised to faster aspart or IAsp.
2. METHODS
2.1. Study design and population
This was a pre‐planned secondary analysis of sensor‐derived glycaemic metrics based on prospectively collected data in pregnant women with type 1 diabetes who participated in the CopenFast trial. 1 , 12
In the original trial, 1 150 women with type 1 diabetes used either isCGM (n = 124) or real‐time CGM (rtCGM) (n = 26) during pregnancy. Data from isCGM were collected prospectively throughout pregnancy from periconception (first day of last menstrual cycle) until delivery, while data from rtCGM were primarily collected at trial visits only. Therefore, data from rtCGM were excluded from this analysis. At randomisation at median 9.5 (interquartile range 9.0–11.0) weeks, 122 women used isCGM initiated prior to pregnancy and two women initiated isCGM shortly after randomisation. Out of these 124 women using isCGM, five women with early fetal loss and six without available CGM data were excluded, leaving a total of 113 (75%) of the 150 women with type 1 diabetes in the original trial for this analysis. 1 Freestyle Libre was used by 93% (n = 105) and 7% (n = 8) used Freestyle Libre 2 in early pregnancy. During pregnancy, seven women changed to Freestyle Libre 2.
Prior to randomisation all 113 women used IAsp except one who used faster aspart and one who used insulin glulisine. Eighty‐eight women used MDI and 25 women used insulin pumps.
2.2. Sensor‐derived glycaemic metrics
All isCGM data were recorded and uploaded using LibreView (Abbott Diabetes Care, Alameda, California, USA). For each woman, the raw isCGM spreadsheets were downloaded from periconception or the day of isCGM application in early pregnancy, respectively, and until 37 completed weeks. In cases of preterm delivery (before 37 weeks), isCGM data were downloaded until delivery. Based on a standard range of summary glycaemic metrics, the following were calculated from each raw isCGM spreadsheets: mean sensor glucose, mean sensor glucose coefficient of variation (CV), time in range (TIRp, 3.5–7.8 mmol/L), time above range (TARp, >7.8 mmol/L) and time below range in pregnancy (TBRp, <3.5 mmol/L). Mean sensor glucose CV (%) was calculated by dividing glucose SD by the mean of the corresponding glucose reading and multiplying by 100.
2.3. Routine diabetes and pregnancy care
All women followed the routine diabetes and pregnancy care program for pregnant women with diabetes consultations by a diabetes specialist approximately every 2 weeks throughout pregnancy. 1 , 12
At each visit, insulin doses were titrated according to isCGM targets of 4.0–5.5 mmol/L preprandially and 4.0–7.0 mmol/L postprandially as well as the mean sensor glucose value of 5.0–6.0 mmol/L, TIRp >70%, TARp <25% and TBRp <4%. 12 , 13 All women, irrespective of the randomisation group, were recommended to inject mealtime insulin 15–30 min before meals and to use sensor‐derived glycaemic metrics to adjust mealtime insulin dose and basal insulin dose, as indicated, every 3–5 days between routine visits to obtain the recommended isCGM targets. 6 HbA1c targets were < 48 mmol/mol (6.5%) before 20 weeks and < 38 mmol/mol (5.6%) after 20 weeks. 1 All women received the same recommendations on medical nutritional therapy, carbohydrate counting and physical activity. 1 , 10 , 12
Women were recommended to aim for a gestational weight gain according to pre‐pregnancy BMI as follows: 10–15 kg if BMI <25 kg/m2, 5–8 kg if BMI 25–29.9 kg/m2 and 0–5 kg if BMI ≥30 kg/m2. 10 The women attended routine obstetric visits at approximately 9, 12, 21, 27, 33 and 36 weeks. Fetal growth was routinely assessed by ultrasound at 27, 33 and 36 weeks.
2.4. Data collection and definitions
At trial visits in the original trial at median 9, 21, 33 and 35 weeks, data on gestational age, HbA1c, weight, blood pressure, insulin dose, presence of albuminuria, the number of self‐reported events of mild hypoglycaemia (events with symptoms familiar to the woman as hypoglycaemia and managed by herself 14 ) the previous week, events of severe hypoglycaemia (requiring third‐party assistance 15 ) and gestational weight gain (calculated from the last weight measured, often at 35 weeks, and self‐reported pre‐pregnancy weight 16 ) were recorded. Diabetic retinopathy was assessed by retinal photo screening in early pregnancy and, when indicated, later in pregnancy. 17 For this analysis, women were categorised as having excessive gestational weight gain if exceeding the recommendations from the National Academy of Medicine (≥16.0 kg if pre‐pregnancy BMI <25 kg/m2, ≥11.5 kg if pre‐pregnancy BMI 25–29.9 kg/m2 and ≥9.0 kg if pre‐pregnancy BMI ≥30 kg/m2. 18
The following pregnancy outcomes were recorded: presence of preeclampsia (office blood pressure ≥ 140/90 mmHg with presence of albuminuria or new onset of symptoms of maternal organ dysfunction after 20 weeks 19 ), gestational hypertension (hypertension after 20 weeks without fulfilling the criteria for preeclampsia 19 ) and mode of delivery. The following neonatal outcomes were recorded: gestational age at delivery given as weeks (+days), birthweight, offspring sex, admission to the neonatal intensive care unit and neonatal hypoglycaemia (plasma glucose <2.2 mmol/L 2 h after birth).
The birthweight SD score was calculated to adjust for gestational age and sex based on growth curves usually used in Scandinavia 20 to reflect how appropriate the infant's weight was compared to infants from the background population taking sex and gestational age into account. The birthweight SD score thus indicated how far from the expected birthweight the newborn was. The closer this value was to zero, the more appropriately grown the newborn was. Large and small for gestational age (LGA and SGA) were defined as birthweight >90th percentile and < 10th percentile, respectively. 20 Early preterm delivery was delivery before 34 completed weeks, and preterm delivery was delivery before 37 completed weeks.
2.5. Statistical analysis
Categorical variables were presented as numbers (%) and numerical variables as mean (SD) or median (interquartile range) as appropriate.
The mean sensor‐derived glycaemic metrics: mean sensor glucose, mean sensor glucose CV, TIRp, TARp and TBRp were calculated for each woman and, based on trial visits in the original trial at randomisation, 21 and 33 weeks where routine fetal ultrasound scans took place, 1 analysed in the pregnancy intervals from randomisation to 21 completed weeks, from 22 to 33 completed weeks and from 34 to 37 completed weeks when comparing women randomised to faster aspart or IAsp.
For each woman and for each week, the mean sensor‐derived glycaemic metrics were calculated across the 24 h day from periconception and until 37 completed weeks and plotted for women randomised to faster aspart and IAsp, respectively.
Continuous outcomes were analysed by linear regression analysis. The assumptions for linear regression were addressed by quantile–quantile plots and histograms of residuals. Categorical variables, including the number of women with severe hypoglycaemia, were analysed using logistic regression analysis. The number of severe hypoglycaemic events was presented as summarised counts and analysed as a continuous outcome using a linear model.
A maximum of 15 min between isCGM data readings was considered as complete isCGM data, as this is the time interval between automatic sensor reports. To calculate the percentage of missing data, time between two isCGM readings exceeding 15 min was considered as missing. Missing isCGM data were reported as a percentage of total time from randomisation to delivery for each woman.
Due to an apparent baseline difference in parity between groups, posthoc adjustment of birthweight SD score for parity was performed.
Correction for multiple testing was not performed.
R version 4.1.0 (R Core Team, 2021, R Foundation for Statistical Computing, Vienna, Austria) was used for all data analyses. 21 A two‐sided p value <0.05 was regarded as statistically significant.
2.5.1. Ethics
Written consent was obtained from all participants. The trial protocol was approved by The Danish Medicines Agency (2018‐004680‐31) and the Regional Ethics Committee (H‐19029966) and published before completed enrolment. 12
This trial is registered with ClinicalTrials.gov, NCT03770767.
3. RESULTS
In total, 59 (52%) and 54 (48%) were randomised to faster aspart and IAsp, respectively. Baseline characteristics are presented in Table 1.
TABLE 1.
Baseline characteristics in 113 women with type 1 diabetes using intermittently scanned continuous glucose monitoring and faster acting insulin aspart or insulin aspart during pregnancy.
| Faster acting insulin aspart (n = 59) | Insulin aspart (n = 54) | |
|---|---|---|
| Age, years | 32 ± 5 | 31 ± 5 |
| Duration of diabetes, years | 14.0 (8.5–20.5) | 15.5 (7.0–22.0) |
| Pre‐pregnancy weight, kg | 68.5 (64.0–77.5) | 68.0 (63.0–80.0) |
| HbA1c, mmol/mol | 46 (43–53) | 48 (44–55) |
| HbA1c, % | 6.4 (6.1–7.0) | 6.5 (6.2–7.2) |
| Pre‐pregnancy BMI, kg/m2 | 24.7 (22.6–27.3) | 24.9 (22.1–26.9) |
| Normal (<25 kg/m2) | 33 (56) | 27 (50) |
| Overweight (25–30 kg/m2) | 19 (32) | 21 (39) |
| Obese (>30 kg/m2) | 7 (12) | 6 (11) |
| Nulliparous | 35 (59) | 19 (35) |
| Nordic origin | 56 (95) | 46 (83) |
| Treatment modality | ||
| Multiple daily injections | 48 (81) | 40 (74) |
| Insulin pump | 11 (19) | 14 (26) |
| Number of women with severe hypoglycaemia the year preceding pregnancy | 2 (2) | 0 (0) |
| Normal hypoglycaemia awareness at 9 weeks a | 26 (58) | 19 (56) |
| Systolic office blood pressure, mmHg | 112 ± 9 | 113 ± 11 |
| Diastolic office blood pressure, mmHg | 73 ± 7 | 74 ± 7 |
| Diabetic retinopathy | 19 (32) | 18 (33) |
| Albuminuria, albumin/creatinine ratio > 30 mg/g | 3 (5) | 1 (2) |
Note: Data are presented as median (IQR), mean (SD) or n (%).
Data on hypoglycaemia awareness were obtained from a questionnaire where 79 (70%) of 113 women responded. Self‐estimated hypoglycaemia awareness was defined as normal when the woman answered ‘always’ to the question ‘Do you recognise symptoms, when you have a hypoglycaemic event?’
Mean sensor glucose and mean sensor glucose CV decreased from randomisation until 37 weeks in both women randomised to faster aspart and IAsp with no difference between the groups (Figure 1a,b).
FIGURE 1.
Sensor‐derived glycaemic metrics throughout pregnancy in 113 women with type 1 diabetes using intermittently scanned continuous glucose monitoring. The sensor‐derived glycaemic metrics were calculated for each woman for each gestational week and plotted according to insulin treatment with faster acting insulin aspart (blue) or insulin aspart (red). The data on sensor‐derived glycaemic metrics presented in the brackets were calculated as the mean value of each woman's average and were analysed in pregnancy intervals: Randomisation to 21 completed weeks, 22–33 completed weeks and 34–37 completed weeks and compared between women randomised to faster acting insulin aspart or insulin aspart. Data were available for 100% (n = 113), 100% (n = 113) and 96% (n = 108) of women in the pregnancy interval randomisation to 21 comleted weeks, 22–33 completed weeks and 34–37 completed weeks. (a) Mean sensor glucose throughout pregnancy (mmol/L). (b) Mean glucose coefficient of variation throughout pregnancy (%). (c) Mean time in range in pregnancy (3.5–7.8 mmol/L) (%). (d) Mean time above range in pregnancy (>7.8 mmol/L) (%). (e) Mean time below range in pregnancy (<3.5 mmol/L) (%).
TIRp increased from randomisation throughout pregnancy in both groups. Women randomised to faster aspart achieved higher TIRp in the pregnancy interval 22–33 weeks compared to women randomised to IAsp (estimated treatment difference 5.1% [95% confidence interval (CI) 0.3; 9.7], p = 0.04). Mean TIRp >70% was achieved from the pregnancy interval randomisation to 21 weeks and onwards, while this was not obtained until the pregnancy interval 34–37 weeks in women randomised to IAsp (Figure 1c).
TARp decreased from randomisation throughout pregnancy in both groups. Women randomised to faster aspart achieved lower TARp in the pregnancy interval 22–33 weeks compared to women randomised to IAsp (estimated treatment difference − 4.9 [95% CI −9.4; −0.3], p = 0.04). Mean TARp <25% was achieved from the pregnancy interval 22–33 weeks onwards, while this was not achieved until the pregnancy interval 34–37 weeks in women randomised to IAsp (Figure 1d).
Mean TBRp remained close to target <4% throughout pregnancy in both groups with no differences between the groups (Figure 1e).
The proportion of missing isCGM data was 7.8% in women randomised to faster aspart and 11.0% in women randomised to IAsp.
HbA1c levels at 35 weeks were similar between women randomised to faster aspart and IAsp (Table 2).
TABLE 2.
Maternal, pregnancy and neonatal outcomes in 113 women with type 1 diabetes using intermittently scanned continuous glucose monitoring and faster acting insulin aspart or insulin aspart during pregnancy.
| Faster acting insulin aspart (n = 59) | Insulin aspart (n = 54) | Estimated treatment difference (95% CI) or odds ratio (95% CI) | p‐value | |
|---|---|---|---|---|
| Maternal and pregnancy outcome | ||||
| HbA1c at 35 weeks, mmol/mol | 42 (39–47) | 43 (40–47) | −1.0 (−3.4; 0.9) | 0.25 |
| HbA1c at 35 weeks, % | 6.0 (5.7–6.5) | 6.2 (5.8–6.5) | −0.2 (−0.3; 0.08) | ‐ |
| Normal hypoglycaemia awareness at 33 weeks a | 17 (45) | 19 (56) | 0.91 (0.7; 2.3) | 0.48 |
| Number of women with severe hypoglycaemia from randomisation to delivery | 1 (2) | 5 (9) | 0.93 (−0.2; −0.01) | 0.04 |
| Preeclampsia | 8 (14) | 5 (10) | 1.04 (0.9; 1.2) | 0.48 |
| Caesarean section | 18 (31) | 22 (41) | 1.10 (0.9; 1.3) | 0.26 |
| Gestational weight gain, kg | 15.0 ± 6 | 14.0 ± 5 | −1.00 (−1.8; 2.6) | 0.72 |
| Excessive gestational weight gain b | 29 (49) | 27 (50) | 0.99 (0.8; 1.2) | 0.93 |
| Gestational age at birth, days | 266 (260–270) | 264 (259–268) | 2.00 (−0.8; 5.1) | 0.16 |
| Preterm delivery before 34 weeks | 0 (0) | 0 (0) | ‐ | ‐ |
| Preterm delivery before 37 weeks | 12 (20) | 10 (19) | 1.12 (0.4; 2.9) | 0.81 |
| Neonatal outcome | ||||
| Birth weight, g | 3534 ± 499 | 3646 ± 511 | −112 (−301; 76) | 0.24 |
| Birth weight SD score | 1.0 ± 1.2 | 1.5 ± 1.2 | −0.5 (−0.9; −0.03) | 0.04 |
| Large for gestational age | 27 (46) | 32 (59) | 0.87 (0.7; 1.1) | 0.15 |
| Small for gestational age | 2 (3) | 0 (0) | 1.03 (1.0; 1.1) | 0.18 |
| Admission to neonatal care unit | 13 (22) | 14 (26) | 1.0 (0.9; 1.2) | 0.97 |
| Neonatal glucose levels 2 h after delivery, mmol/L | 3.3 ± 1.2 | 2.9 ± 1.0 | 0.4 (−0.1; 0.7) | 0.17 |
Note: Data are presented as median (IQR), mean (SD) or n (%). Data were available >95% of the women unless otherwise stated.
Data on hypoglycaemia awareness were obtained from a questionnaire where 72 (64%) of 113 women responded. Self‐estimated hypoglycaemia awareness was defined as normal when the woman answered ‘always’ to the question ‘Do you recognise symptoms, when you have a hypoglycaemic event?’
Excessive gestational weight gain: Exceeding the recommended gestational weight gain according to pre‐pregnancy BMI recommended by the National Academy of Medicine (pre‐pregnancy BMI <25 kg/m2: ≥16 kg, pre‐pregnancy BMI 25–29.9 kg/m2: ≥11.5 kg and pre‐pregnancy BMI of ≥30 kg/m2: >9.0 kg).
The number of mild hypoglycaemic events was similar in both groups from randomisation throughout pregnancy. The number of women experiencing at least one severe hypoglycaemic event from randomisation until delivery differed significantly between women randomised to faster aspart and women randomised to IAsp (1 (2%) and 5 (9%), odds ratio 0.93 [95% CI −0.2; −0.01], p = 0.04). The total number of severe hypoglycaemic events was 1 and 8 in women randomised to faster aspart compared to women randomised to IAsp (OR: −0.13 [−0.26; −0.00], p = 0.05).
The one event with severe hypoglycaemia in the faster aspart group occurred at gestational age 12 weeks. In the IAsp group, the first events with severe hypoglycaemia occurred between 8 and 21 weeks. Three women in the IAsp group each reported one additional event with severe hypoglycaemia, all in the second trimester. A total of 5 (63%) severe hypoglycaemic events occurred during sleep.
Insulin doses, in particular mealtime insulin doses, increased during pregnancy with no differences between groups, except for a higher basal insulin dose at 35 weeks in women randomised to IAsp (Table 3).
TABLE 3.
Insulin dose and mild hypoglycaemia in 113 women with type 1 diabetes using intermittently scanned continuous glucose‐monitoring and faster acting insulin aspart or insulin aspart during pregnancy.
| Faster acting insulin aspart (n = 59) | Insulin aspart (n = 54) | Estimated treatment difference (95% CI) | p‐value | |
|---|---|---|---|---|
| Prior to pregnancy | ||||
| Basal insulin dose, IU/24 h | 20.5 ± 7.4 | 22.3 ± 10.9 | −1.8 (−5.5; 1.8) | 0.32 |
| Mealtime insulin dose, IU/24 h | 17.0 ± 8.5 | 18.3 ± 10.4 | −1.3 (−4.9; 2.5) | 0.51 |
| Total daily insulin dose, IU/24 h | 37.5 ± 13.6 | 40.6 ± 17.0 | −0.05 (−0.2; 0.1) | 0.59 |
| At randomisation at 9 weeks | ||||
| Basal insulin dose, IU/24 h | 17.9 ± 6.7 | 20.1 ± 9.5 | −2.1 (−5.2; 0.1) | 0.18 |
| Mealtime insulin dose, IU/24 h | 17.7 ± 8.5 | 18.6 ± 9.6 | −0.8 (−4.4; 2.8) | 0.66 |
| Total daily insulin dose, IU/24 h | 35.6 ± 13.5 | 38.7 ± 17.9 | −0.05 (−0.2; 0.1) | 0.59 |
| Episodes of mild hypoglycaemia the previous week | 5 (3–9) | 6 (3–8) | 0.03 (−2.2; 2.2) | 0.99 |
| 21 weeks | ||||
| Basal insulin dose, IU/24 h | 18.7 ± 8.3 | 20.6 ± 12.0 | −1.8 (−5.7; 1.9) | 0.34 |
| Mealtime insulin dose, IU/24 h | 22.7 ± 10.2 | 26.1 ± 13.9 | −3.4 (−7.9; 1.2) | 0.14 |
| Total daily insulin dose, IU/24 h | 41.4 ± 16.1 | 46.7 ± 22.8 | −0.1 (−0.6; 0.05) | 0.20 |
| Episodes of mild hypoglycaemia the previous week | 5 (3–9) | 6 (3–8) | 0.6 (−0.7; 1.9) | 0.36 |
| 33 weeks | ||||
| Basal insulin dose, IU/24 h | 26.2 ± 14.5 | 30.9 ± 19.4 | −4.7 (−11.1; 1.6) | 0.14 |
| Mealtime insulin dose, IU/24 h | 42.2 ± 24.5 | 43.9 ± 19.3 | −1.8 (−9.1; 5.6) | 0.66 |
| Total daily insulin dose, IU/24 h | 68.4 ± 27.8 | 74.8 ± 34.6 | −0.08 (−0.2; 0.1) | 0.30 |
| Episodes of mild hypoglycaemia the previous week | 3 (1–5) | 4 (3–5) | −0.7 (−1.7; 0.4) | 0.21 |
| 35 weeks a | ||||
| Basal insulin dose, IU/24 h | 23.7 ± 12.3 | 31.5 ± 22.3 | −7.7 (−14.7; −0.7) | 0.03 |
| Mealtime insulin dose, IU/24 h | 43.7 ± 21.9 | 45.9 ± 23.9 | −2.3 (−11.3; 6.7) | 0.61 |
| Total daily insulin dose, IU/24 h | 67.4 ± 28.9 | 77.4 ± 38.3 | −0.12 (−0.3; 0.1) | 0.17 |
| Episodes of mild hypoglycaemia the previous week | 3 (1–6) | 3 (1–5) | −0.5 (−2.0; 1.1) | 0.54 |
Note: Data are presented as median (IQR) or mean (SD). Data were available >95% of the women unless otherwise stated.
Data were available for 89% (n = 53) of women using faster acting insulin aspart and 93% (n = 50) of women using insulin aspart, respectively.
The birthweight SD score was lower in infants born to women randomised to faster aspart compared to women randomised to IAsp (estimated treatment difference − 0.5 [−0.9; −0.03], p = 0.04). However, this attenuated when adjusting for parity (estimated treatment difference − 0.4 [−0.8; 0.1], p = 0.10). All other pregnancy and neonatal outcomes were similar between the groups (Table 2).
4. DISCUSSION
In this pre‐planned secondary analysis of the CopenFast trial including 113 women with type 1 diabetes with continuous use of isCGM during pregnancy, women randomised to faster aspart achieved TIRp >70% earlier in pregnancy, higher TIRp from 22 to 33 weeks and less severe hypoglycaemia compared to women randomised to IAsp. Infant birthweight was lower and thereby more appropriate in the faster aspart group; however, this attenuated when adjusting for parity.
These data on continuous use of isCGM from randomisation throughout pregnancy provide additional data to the original trial, where isCGM data for 7 days three times during the trial were collected from the patient records. 1 Regardless of treatment allocation, sensor‐derived glycaemic metrics improved throughout pregnancy. However, women randomised to faster aspart achieved the recommendations for glycaemic metrics earlier in pregnancy compared to women randomised to IAsp. With its earlier onset of appearance in the circulation and greater early glucose‐lowering effect, faster aspart provides a more physiological profile, compared to IAsp, and thereby mimics the endogenous prandial insulin secretion with lower postprandial glucose levels as seen in healthy individuals. 2 , 22
As in the original trial, the numbers of severe hypoglycaemic events were low, especially in women randomised to faster aspart. This supports a beneficial effect of faster aspart over IAsp in combination with continuous use of isCGM initiated before pregnancy and used until delivery for minimising severe hypoglycaemia during pregnancy in women with type 1 diabetes.
Women randomised to faster aspart delivered more appropriately sized infants with lower and thereby more appropriate birthweight SD scores compared to women randomised to IAsp, although this attenuated when adjusting for the non‐modifiable variable parity. Lower mean glucose, higher TIRp and early achievement of TIRp >70% are associated with more appropriate fetal size, 4 , 6 , 8 emphasising the importance of appropriate insulin treatment in order to achieve glycaemic metric targets early in pregnancy 7 and for improving fetal outcomes, including birthweight. 4 , 6 , 8 Parity is known to affect infant birthweight in women with type 1 diabetes and multiparity may lead to slightly larger infants. 23 This suggests that adjustment for baseline covariates as parity can be considered in future randomised controlled trials to account for potential imbalances between groups.
Insulin doses, in particular mealtime insulin doses, increased during pregnancy. Pregnancy in women with type 1 diabetes is characterised by alterations in insulin requirements from week to week. From around 16 weeks, insulin requirements increase gradually, and it is typically necessary to mainly increase mealtime insulin dose. The basal insulin dose also increases gradually, but to a smaller extent than the mealtime insulin. 24 , 25 , 26 The higher doses of faster aspart and IAsp seen from 21 to 35 weeks compared to basal insulin doses thus reflect what can be expected during pregnancy. The observed lower basal insulin dose between the groups at 35 weeks may be coincidental since basal insulin doses were similar between the groups at all other trial visits.
In general, isCGM is the most commonly used CGM device worldwide, which is also the case in our centre. Additionally, there might be differences between values obtained by isCGM and rtCGM; thus, to ensure data homogeneity, we decided to analyse data only from women using this device.
This is, to our knowledge, the first report with detailed data on sensor‐derived glycaemic metrics during pregnancy in women with type 1 diabetes randomised to faster aspart or IAsp. Based on the favourable effects of faster aspart on glycaemic metrics and maternal hypoglycaemia, in the CopenFast trial, we suggest that women with type 1 diabetes can be offered faster aspart as part of pregnancy planning, so they can make informed choices regarding insulin treatment in a future pregnancy. To reduce potential temporarily fluctuating glucose values during pregnancy, a change in insulin type should ideally be made before pregnancy or after organogenesis is completed.
Strengths of this secondary analysis include the pre‐planned trial design based on a randomised control trial with prospectively collected data in a well characterised cohort of women with type 1 diabetes. All women were treated according to the same treatment recommendations by the same diabetes team. The isCGM data were continuously collected throughout pregnancy, and large amounts of data were included in this secondary analysis. The amount of missing isCGM data was low in both groups and is not likely to inflict bias.
The cohort was from a well‐defined geographical area, with few exclusion criteria which might benefit the generalisability of this secondary analysis. However, the single‐centre trial design could contribute to less external validity. In the original trial, the intervention with faster aspart or IAsp was not blinded and thus more prone to systematic differences (bias) in other interventions given to the two groups. 1 However, the primary outcome birthweight SD score was an objective outcome with a standardised measurement method, which may be less affected by a lack of blinding. 27 , 28 Irrespective of the randomisation group, all women were given the same recommendations on timing of mealtime insulin dosing and adjustment of insulin dose. The majority of women used an isCGM device (Freestyle Libre) without hypoglycaemia alerts, but now isCGM with alerts (Freestyle Libre 2) is offered to all persons with type 1 diabetes. Adjustment for multiple testing was not performed, and therefore there is a risk of a type 1 statistical error.
In conclusion, this secondary analysis of the CopenFast trial supports the use of faster aspart over IAsp for improved glycaemic metrics and reduction of severe hypoglycaemia during pregnancy in women with type 1 diabetes. Infant birthweight was lower and thereby more appropriate with use of faster aspart; however, this attenuated when adjusting for parity.
AUTHOR CONTRIBUTIONS
The original trial was initiated by the principal investigator L.R, E.R.M. and P.D. J.C.S., S.K.N., E.R.M. and L.R. collected the data. J.C.S. performed all the statistical analyses. J.C.S. wrote the first draft of the article. S.K.N., K.N., T.D.C., P.D., E.R.M. and L.R. critically read and revised the article. All the authors have approved the article and have consented to the publication of the article before submission.
FUNDING INFORMATION
This trial was an Investigator‐Sponsored Study funded by Novo Nordisk A/S, grant number U1111‐1209‐6358. The grant covered full‐time salaries for J.C.S. and S.K.N., as well as part‐time salaries for E.R.M. and L.R.
CONFLICT OF INTEREST STATEMENT
This trial was an Investigator Sponsored Study funded by Novo Nordisk A/S, grant number U1111‐1209‐6358. The grant covered full‐time salaries for J.C.S. and S.K.N., as well as part‐time salaries for E.R.M. and L.R. After the initial submission and before the revision of this manuscript S.K.N. signed an employment contract with Novo Nordisk A/S as part of a position commencing 1st of September 2024. E.R.M. has contracts with Novo Nordisk for the Expect trial and the Evolve trial investigating newer insulin analogues and insulin pump treatment in pregnant women with diabetes, has received fees from Novo Nordisk for lectures, received financial support from Novo Nordisk for travelling to one international scientific meeting in the field of diabetes, and is participating on an advisory board for Novo Nordisk. P.D. has participated in clinical studies on the use of insulin in pregnant women with pre‐existing diabetes in collaboration with Novo Nordisk, but no personal honorarium was involved. K.N. has received funding from Novo Nordisk for an investigator‐initiated trial (grant number U1111‐1209‐6358) with faster acting insulin aspart in insulin pump‐treated adults with T1D, and is an advisory board member, for which an honorarium was given to her institution Steno Diabetes Center Copenhagen. K.N. owns stocks in Novo Nordisk. T.D. declares no conflict of interests.
ACKNOWLEDGEMENTS
The authors kindly thank all participants in the trial and the nurses Ann‐Sofi Lunde, Birgitta Ellingsgaard, Charlotte Barfred, Olga M. Green, Signe Lou‐Møller, and Vibeke Ladefoged at the Center for Pregnant Women with Diabetes. The authors also acknowledge Julie Lyng Forman from the Statistical Advisory Section of Biostatistics at the University of Copenhagen and Rigshospitalet and Mads Albrecht Andersen for statistical support.
Søholm JC, Nørgaard SK, Nørgaard K, et al. Sensor‐derived glycaemic metrics in pregnant women with type 1 diabetes randomised to faster acting insulin aspart or insulin aspart—A secondary analysis of the CopenFast trial. Diabet Med. 2025;42:e15467. doi: 10.1111/dme.15467
Clinical Trials Registry No: NCT03770767.
DATA AVAILABILITY STATEMENT
The data set analysed during the study is available from the corresponding author on reasonable request.
REFERENCES
- 1. Nørgaard SK, Søholm JC, Mathiesen ER, et al. Faster‐acting insulin aspart versus insulin aspart in the treatment of type 1 or type 2 diabetes in pregnancy and post‐delivery (CopenFast): an open‐label, single‐centre, randomised controlled trial. Lancet Diabetes Endocrinol. 2023;11(11):811‐821. doi: 10.1016/S2213-8587(23)00236-X [DOI] [PubMed] [Google Scholar]
- 2. Heise T, Pieber TR, Danne T, Erichsen L, Haahr H. A pooled analysis of clinical pharmacology trials investigating the pharmacokinetic and Pharmacodynamic characteristics of fast‐acting insulin Aspart in adults with type 1 diabetes. Clin Pharmacokinet. 2017;56(5):551‐559. doi: 10.1007/s40262-017-0514-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Heise T, Hövelmann U, Zijlstra E, Stender‐Petersen K, Jacobsen JB, Haahr H. A comparison of pharmacokinetic and Pharmacodynamic properties between faster‐acting insulin Aspart and insulin Aspart in elderly subjects with type 1 diabetes mellitus. Drugs Aging. 2017;34(1):29‐38. doi: 10.1007/s40266-016-0418-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Scott EM, Murphy HR, Kristensen KH, et al. Continuous glucose monitoring metrics and birth weight: informing Management of Type 1 diabetes throughout pregnancy. Diabetes Care. 2022;45(8):1724‐1734. doi: 10.2337/dc22-0078 [DOI] [PubMed] [Google Scholar]
- 5. Meek CL, Tundidor D, Feig DS, et al. Novel biochemical markers of Glycemia to predict pregnancy outcomes in women with type 1 diabetes. Diabetes Care. 2021;44(3):681‐689. doi: 10.2337/DC20-2360 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Søholm JC, Nørgaard SK, Nørgaard K, et al. Mean glucose and gestational weight gain as predictors of large‐for‐gestational‐age infants in pregnant women with type 1 diabetes using continuous glucose monitoring. Diabetes Technol Ther. 2024;26(8):1‐11. doi: 10.1089/dia.2023.0583 [DOI] [PubMed] [Google Scholar]
- 7. Benhalima K, Beunen K, Siegelaar SE, et al. Management of type 1 diabetes in pregnancy: update on lifestyle, pharmacological treatment, and novel technologies for achieving glycaemic targets. Lancet Diabetes Endocrinol. 2023;11(7):490‐508. doi: 10.1016/S2213-8587(23)00116-X [DOI] [PubMed] [Google Scholar]
- 8. Kristensen K, Ögge LE, Sengpiel V, et al. Continuous glucose monitoring in pregnant women with type 1 diabetes: an observational cohort study of 186 pregnancies. Diabetologia. 2019;62:1143‐1153. doi: 10.1007/s00125-019-4850-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Sanusi AA, Xue Y, McIlwraith C, et al. Association of Continuous Glucose Monitoring Metrics with Pregnancy Outcomes in patients with preexisting diabetes. Diabetes Care. 2024;47(1):89‐96. doi: 10.2337/dc23-0636 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Ringholm L, Nørgaard SK, Rytter A, Damm P, Mathiesen ER. Dietary advice to support Glycaemic control and weight Management in Women with type 1 diabetes during pregnancy and breastfeeding. Nutrients. 2022;14(22):1‐16. doi: 10.3390/nu14224867 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Battelino T, Alexander CM, Amiel SA, et al. Continuous glucose monitoring and metrics for clinical trials: an international consensus statement. Lancet Diabetes Endocrinol. 2023;11(1):42‐57. doi: 10.1016/S2213-8587(22)00319-9 [DOI] [PubMed] [Google Scholar]
- 12. Nørgaard SK, Mathiesen ER, Nørgaard K, Clausen TD, Damm P, Ringholm L. CopenFast trial: faster‐acting insulin Fiasp versus insulin NovoRapid in the treatment of women with type 1 or type 2 diabetes during pregnancy and lactation—a randomised controlled trial. BMJ Open. 2021;11(4):1‐8. doi: 10.1136/bmjopen-2020-045650 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Battelino T, Danne T, Bergenstal RM, et al. Clinical targets for continuous glucose monitoring data interpretation: recommendations from the international consensus on time in range. Diabetes Care. 2019;42(8):1593‐1603. doi: 10.2337/dci19-0028 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Nielsen LR, Pedersen‐Bjergaard U, Thorsteinsson B, Johansen M, Damm P, Mathiesen ER. Hypoglycemia in pregnant women with type 1 diabetes: predictors and role of metabolic control. Diabetes Care. 2008;31(1):9‐14. doi: 10.2337/dc07-1066 [DOI] [PubMed] [Google Scholar]
- 15. Seaquist ER, Anderson J, Childs B, et al. Hypoglycemia and diabetes: a report of aworkgroup of the American diabetes association and the endocrine society. Diabetes Care. 2013;36(5):1384‐1395. doi: 10.2337/dc12-2480 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Ásbjörnsdóttir B, Rasmussen SS, Kelstrup L, Damm P, Mathiesen ER. Impact of restricted maternal weight gain on fetal growth and perinatal morbidity in obese women with type 2 diabetes. Diabetes Care. 2013;36(5):1102‐1106. doi: 10.2337/dc12-1232 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Pappot N, Do NC, Vestgaard M, et al. Prevalence and severity of diabetic retinopathy in pregnant women with diabetes—time to individualize photo screening frequency. Diabet Med. 2022;39(7):1‐9. doi: 10.1111/dme.14819 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Institute of Medicine and National Research Council Committee to Reexamine IOM Pregnancy Weight Guidelines . Weight Gain during Pregnancy: Reaxamining the Guidelines. The Natl Acad Press; 2009. doi: 10.17226/12584 [DOI] [Google Scholar]
- 19. American College of Obstetricians and Gynecologists . Hypertension in pregnancy. Obstet Gynecol. 2013;122(5):1122‐1131. doi: 10.1097/01.AOG.0000437382.03963.88 [DOI] [PubMed] [Google Scholar]
- 20. Maršál K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr Int J Paediatr. 1996;85(7):843‐848. doi: 10.1111/j.1651-2227.1996.tb14164.x [DOI] [PubMed] [Google Scholar]
- 21. R Core Team . R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Published online 2021. https://www.r‐project.org/
- 22. Mathieu C, Bode BW, Franek E, et al. Efficacy and safety of fast‐acting insulin aspart in comparison with insulin aspart in type 1 diabetes (onset 1): a 52‐week, randomized, treat‐to‐target, phase III trial. Diabetes Obes Metab. 2018;20(5):1148‐1155. doi: 10.1111/dom.13205 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Glinianaia SV, Tennant PWG, Bilous RW, Rankin J, Bell R. HbA1c and birthweight in women with pre‐conception type 1 and type 2 diabetes: a population‐based cohort study. Diabetologia. 2012;55(12):3193‐3203. doi: 10.1007/s00125-012-2721-z [DOI] [PubMed] [Google Scholar]
- 24. Skajaa GO, Fuglsang J, Kampmann U, Ovesen PG. Parity increases insulin requirements in pregnant women with type 1 diabetes. J Clin Endocrinol Metab. 2018;103(6):2302‐2308. doi: 10.1210/jc.2018-00094 [DOI] [PubMed] [Google Scholar]
- 25. García‐Patterson A, Gich I, Amini SB, Catalano PM, De Leiva A, Corcoy R. Insulin requirements throughout pregnancy in women with type 1 diabetes mellitus: three changes of direction. Diabetologia. 2010;53(3):446‐451. doi: 10.1007/s00125-009-1633-z [DOI] [PubMed] [Google Scholar]
- 26. Dori‐Dayan N, Cukierman‐Yaffe T, Zemet R, et al. Insulin requirements during pregnancy in women with type 1 diabetes treated with insulin pump. Diabetes Metab Res Rev. 2024;40(3):e3771. doi: 10.1002/dmrr.3771 [DOI] [PubMed] [Google Scholar]
- 27. Anand R, Norrie J, Bradley JM, Mcauley DF, Clarke M. Fool's gold? Why blinded trials are not always best. BMJ. 2020;368:l6228. doi: 10.1136/bmj.l6228 [DOI] [PubMed] [Google Scholar]
- 28. Moustgaard H, Clayton GL, Jones HE, et al. Impact of blinding on estimated treatment effects in randomised clinical trials: meta‐epidemiological study. BMJ. 2020;368:1‐4. doi: 10.1136/bmj.l6802 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data set analysed during the study is available from the corresponding author on reasonable request.