Monitoring fetal well-being and delivery outcomes in pregnancies complicated by gestational and pregestational diabetes

doi:10.21203/rs.3.rs-3610102/v1

Background

diabetes is a very common pregnancy complication. This study aims to compare measurements taken during antenatal monitoring in the third trimester and delivery outcome data in pregnancies complicated by maternal diabetes with respect to healthy controls.

Methods

a prospective observational study included a total of 152 pregnant women. 25 had a diagnosis of pregestational diabetes mellitus (PGDM) and 61 were diagnosed with gestational diabetes mellitus (GDM). The remaining 66 were controls. Data collected antepartum for all pregnancies include ultrasound fetal biometry and amniotic fluid evaluation, Umbilical and Middle Cerebral Artery Pulsatility Index (UA-PI, MCA-PI) and computerized CTG (cCTG) monitoring. The mode of delivery and neonatal data, including umbilical cord gas values, were also obtained. The values collected were compared between groups.

Results

GDM and PGDM groups showed significantly lower values of UA-PI compared to Controls (p < 0.001) and GDMs assumed slightly lower values of MCA-PI compared to Controls (p = 0.028). Some cCTG parameters showed small but significant differences among groups. The analyzed groups presented significant differences in the umbilical artery gas analysis values at birth, which were affected by the mode of delivery. The pH was lower in PGDMs compared to both GDMs (p = 0.0279) and Controls (p < 0.0001), and spontaneous deliveries were associated with lower pH values (p = 0.008). pO2 significantly decreased from Controls to GDMs and PGDMs, respectively (Controls vs GDMs p = 0.0057, Controls vs PGDMs p < 0.001, GDMs vs PGDMs p = 0.0002) while pCO2 followed an opposite trend (Controls vs GDMs p < 0.001, Controls vs PGDMs p < 0.001, GDMs vs PGDMs p = 0.0014). Lactates were higher in PGDMs compared to both other groups (Controls vs PGDMs: p = 0.0128, GDMs vs PGDMs: p = 0.0161) and were higher in vaginal deliveries compared to cesarean sections (p = 0.017).

Conclusions

The results suggest that to date there are no antenatal monitoring methods that can accurately define the true well-being of the fetus in pregnancies complicated by diabetes. In fact, neonatal outcome data show greater differences between groups than those identified antepartum by the analyzed methods. We hypothesize that only a multi-parametric/multimodal approach can help in antepartum management.

antepartum monitoring

umbilical artery

middle cerebral artery

computerized cardiotocography

cesarean section

umbilical cord gas values

Physicians caring for diabetic pregnant women continue to face the problem of determining which methods of fetal surveillance are more effective in identifying pregnancies at risk for adverse neonatal outcomes.

Even though diabetes is essentially a maternal disease, the fetus is directly involved because it is exposed to altered maternal metabolic conditions. This is a particular state that can be difficult to monitor with the most used methodologies for antepartum fetal surveillance. At the same time, changes occur abruptly and may bear life-threatening effects on the fetus and the mother (1,2).

Currently, obstetricians can use three major fetal surveillance tests:

Fetal biometry and amniotic fluid evaluation.
Doppler velocimetry of the Umbilical Artery (UA) and Middle Cerebral Artery (MCA);
Cardiotocography (CTG).

a) Fetal biometry is useful because diabetic pregnancies are commonly associated with macrosomia and polyhydramnios, with variable prevalence. The etiology of macrosomia, which causes the newborn to be larger than normal, is that fetal hyperglycemia stimulates a greater release of insulin with an increased intake of glucose in the fetal cells. Excessive neonatal fat mass is associated with the development of childhood obesity and diabetes (3). Another common consequence of maternal diabetes is fetal polyuria, which in turn can cause polyhydramnios. Diabetic pregnancies complicated by macrosomia and polyhydramnios have been associated with worse neonatal outcomes (1,2).

b) International guidelines recommend the use of Doppler velocimetry of the UA in the monitoring of diabetic pregnancies only if there is a known risk of intrauterine growth restriction (IUGR) (2). However, in several countries, fetal Doppler examinations are routinely conducted in pregnancies complicated by diabetes, regardless of the presence of IUGR (4).

c) Antepartum CTG is recommended for monitoring pregnancies of women with pregestational diabetes mellitus (PGDM) (2) and it is considered beneficial for women with gestational diabetes mellitus (GDM) with poor glycemic control or treated with insulin or oral agents (1). Moreover, in the last two decades, much has been done to improve methods of fetal heart rate (FHR) analysis and monitoring to allow for better evaluation of fetal well-being in relation to antepartum pathophysiological control mechanisms (5–7).

In general, however, the optimal timing and frequency of fetal monitoring in pregnancies complicated by diabetes has not been established (2) and it is unclear which one of these prenatal tests is the most accurate predictor of adverse perinatal outcomes.

This work aims to evaluate a comprehensive set of antepartum measurements from all the methods available for fetal surveillance and delivery outcomes in order to characterize differences in the evolution of pregnancy complicated by maternal diabetes compared with healthy controls.

This paper presents an observational prospective study carried out on a population of pregnant women monitored by means of the three major fetal surveillance tests reported in the Background section. Data were collected at the Department of Mother and Child at the University Hospital of Naples Federico II, in collaboration with the Polytechnic of Milan and the University of Pavia. The population consisted of a total of 152 pregnant women: 25 had a diagnosis of pregestational diabetes mellitus (PGDM), 61 were diagnosed with gestational diabetes mellitus (GDM) and 66 were controls.

All pregnant women were admitted, followed up and gave birth in the same Obstetric Unit. Pregnant women were enrolled starting from the third trimester of gestation until the end of gestation. All enrolled subjects read and signed informed consent for participation in the study.

Inclusion criteria for the entire population were defined as follows:

singleton pregnancy;
certain determination of pregnancy dating calculated from the first day of the last menstrual period and confirmed by ultrasound measurements according to the population nomograms (8);
gestational age at or beyond 26 weeks of gestation.
no known pathologies which are not related to diabetes.

Healthy Controls were selected among those who matched the previous three inclusion criteria by adding the condition that they were defined as euglycemic during the screening examinations, i.e., 2-hour oral glucose tolerance test (OGTT) and had normal fasting blood glucose levels for all three pregnancy trimesters.

GDMs were selected on the basis of the OGTT test generally performed at 24-28 weeks of gestation. In a few cases, it was conducted earlier (i.e., at 16-18 weeks of gestation) based on additional risk factors for gestational diabetes. The 2-hour diagnostic OGTT consisted of a three-step approach, with the determination of fasting, 1-hour and 2-hour glucose levels after administration of a 75 g oral glucose solution (9).

PGDM was in most cases a known condition that preceded pregnancy. In a few cases only, PGDM was diagnosed in the first trimester or early second trimester with standard diagnostic criteria, defined as hemoglobin A1C (HbA1C) of 6.5% or higher, fasting plasma glucose of 126 mg/dL or higher, or 2-hour glucose of 200 mg/dL or greater during a 75-g oral glucose tolerance test (10).

Fasting glucose levels at the first and second trimester and pre-pregnancy BMI were retrospectively obtained from hospital records. Population characteristics are summarized in Table 1.

Table 1 Maternal characteristics of the study population.

	Controls	GDMs	PGDMs
N	66	61	25
Pre-pregnancy BMI (Kg/m²)	25.4±4.9	29.8±7.7	27.9±5.5
Age (years)	31.7±5.2	35.0±5.3	31.6±5.8
Number of pregnancies	2 (1-4)	2 (1-9)	2 (1-4)
Fasting glucose 1^sttrim	74 (69-82)	86 (76-92)	123 (114.7-143.7)
Fasting glucose 2^nd trim	78 (72-85)	89 (82.5-99)	133 (122.5-154.5)
Fasting glucose 3^rd trim*	83 (76-88)	90 (82.0-99.2)	142 (114.0-174.2)
1-h Plasma glucose	142 (132-153)	184.5 (169-201)	-
2-h Plasma glucose	123 (132-118)	141 (118.5-159)	-
urinary proteinuria*	0 (0%)	5 (8.2%)	0 (0%)
urinary glycosuria*	0 (0%)	5 (8.2%)	11 (44%)
gestational hypertension*	0 (0%)	11 (18%)	5 (20%)
chronic hypertension	0 (0%)	4 (6.6%)	1 (4%)
preeclampsia*	0 (0%)	3 (4.9%)	0 (0%)
diabetic retinopathy	0 (0%)	0 (0%)	1 (4%)
chronic renal failure	0 (0%)	0 (0%)	1 (4%)

Asterisks (*) indicate values or diagnoses obtained during the course of the study

2.1 Experimental Protocol

Fetal monitoring included fetal biometry and amniotic fluid evaluation, Doppler velocimetry of the UA and MCA and antepartum cCTG monitoring. Moreover, to assess maternal glycemic control, fasting blood glucose levels were measured during the third trimester of pregnancy in all groups.

2.1.1 Amniotic fluid

Concerning amniotic fluid evaluation, the deepest vertical amniotic fluid pocket (DVP) was calculated in all pregnancies through the identification and measurement of the largest vertical diameter within an amniotic fluid pocket. Values lower than 2 cm identify oligohydramnios, while values over 8 cm are considered polyhydramnios (11).

2.1.2 Doppler velocimetry

The antepartum Doppler velocimetry was performed using a Voluson E8 (General Electric Healthcare Technologies, Milwaukee, WI, USA) ultrasound machine equipped with a 2–8MHz transabdominal transducer and color Doppler imaging was used to identify the UA and MCA vessels. Doppler flow spectra were obtained from the UA at the midsection of the umbilical cord and the middle portion of the MCA, as described in (12). The indices considered for the analyses are UA and MCA Pulsatility indices (UA-PI, MCA-PI) and the analysis was performed at the 26^th, 32^nd, and 36^th weeks of gestation.

2.1.3 Antepartum CTG

The antepartum computerized CTG (cCTG) monitoring was performed in a controlled clinical environment with the patient lying on an armchair. The cCTG records were obtained using Philips Avalon FM30, which is equipped with an ultrasound transducer and a transabdominal tocodynamometer. Recordings lasted at least 40 minutes.

In our study, the FHR signal was collected simultaneously employing two computerized systems, namely the Philips IntelliSpace Perinatal (www.philips.com/IntelliSpacePerinatal) and the 2CTG2 system (13).

Although in pregnancies complicated by PGDM CTG monitoring is considered appropriate starting from the 32^nd week of gestation, it can be performed at earlier gestational ages in case of additional complications. In our study, we decided to start CTG monitoring from the 31^st week of gestation with approximately weekly frequency. Recordings include the fetal heart rate (FHR) and uterine contractions signals, and the fetal movements series (FM), which is set to 1 when the mother presses a button to indicate a perceived movement and 0 when she does not. A collection of parameters, both linear and nonlinear, were calculated from the CTG traces stored by the 2CTG2 software.

They include the average of the FHR baseline (MeanFHR) calculated using the Mantel algorithm (14), the number of accelerations per hour (#Acc), and the Short Time Variation (STV) (15). All these are commonly used CTG time-domain parameters (6,15).

Moreover, the distribution of power in LF (0.03-0.15 Hz), MF (0.15- 0.5 Hz), and HF (0.5-1 Hz) was calculated and expressed as the percentage of the total signal power (5). These features were computed on windows of three minutes which were then averaged within the same tracing.

Other parameters which are not used in clinical practice were also examined to provide a more comprehensive analysis of the FHR. In particular, two features that estimate signal complexity were included: the Sample Entropy (SampEn), which was computed setting x=1, m=0.1, and r=1 (16), and the ternary Lempel-Ziv complexity (LZ3), which was computed as reported in (17) setting p to 0.01. To compute this parameter the FHR signal is re-codified to a trace composed of 3 symbols which represent signal increase, decrease or stationary state. LZ3 quantifies the emergence of new patterns in this trace. Values close to 1 indicate high irregularity of the signal while lower values indicate the presence of repetitive patterns. Both SampEn and LZ3 were computed over three-minute windows, which were then averaged within the same recording.

Finally, we applied the Phase Rectified Signal Averaging (PRSA) technique to highlight hidden oscillations in the signal (18). The value of L was set to 40 FHR samples, T to 1 and s to 2. From the PRSA curve we computed the Deceleration Capacity (DC), the Deceleration Reserve (DR) (19) and, by applying the CWT, we estimated the power in MF and HF at the anchor point (20). The values obtained (MFprsa and HFprsa) were then normalized by the total power of the PRSA signal.

Fetal movements (MovF%) were estimated by measuring the percentage of time mothers press the dedicated button over the entire CTG recording.

These measurements were done for routine follow-up in cases and only for study purposes in the controls.

2.1.4 Pregnancy outcomes

Neonatal data were collected for all newborns. They include week and modality of delivery, birthweight, Apgar score at 1 and 5 minutes, UA blood gas values at birth, and access to the neonatal intensive care unit (NICU). UA blood gas values collected are: pH, O2 and CO2 partial pressure (pO2, pCO2), cord arterial base excess (BECF), lactates and bicarbonate (HCO3).

2.2 Statistical Analysis

UA-PI and MCA-PI, which were assessed at three different time points during the last trimester, were compared among groups using repeated measure analysis of variance (ANOVA). In particular, we looked for between-subjects effects in a repeated measures model. When performing pairwise comparisons, p-values were corrected with Bonferroni correction. Concerning the occurrence of oligohydramnios and polyhydramnios, differences between groups were evaluated using Fisher’s exact test.

The comparison of cCTG parameters between the three groups was made through the non-parametric Kruskal-Wallis test since normality was not respected for several of the analyzed features. The test was applied independently for each group of weeks. Additionally, Linear Mixed Effect models (LME) were also employed to make use of all the available recordings and account for repeated measures. For each of the CTG features discussed, we fitted an LME model. The fixed independent variables are the group (i.e., Controls, GDMs, PGDMs) and gestational age, while the subject is modeled as the random factor to account for repeated measures. Before fitting the models, variables were log-transformed when necessary to ensure that they were approximately normally distributed, and all were normalized using the standard scaler.

Umbilical cord gas values were compared through 2-way ANOVA, considering the mode of delivery (spontaneous/cesarean) and the groups as factors. Post-hoc comparisons between groups were corrected using the Bonferroni procedure. The prevalence of Apgar scores less than 7, neonatal intensive care admission, small for gestational age and large for gestational age were compared across groups using Fisher's exact test.

Analyses were conducted in MATLAB R2022b (The Math Works, Inc.).

3.1 Amniotic fluid evaluation

As shown in Table 2, almost all alterations in amniotic fluid values were identified within the diabetic populations. In particular, it was observed a substantial increase in the prevalence of polyhydramnios.

Table 2 Prevalence of polyhydramnios and oligohydramnios

		Controls	GDMS	PGDMs	p-vals
Week 26	Poly.	0%	11.5%	8%	0.009
Week 26	Oligo.	0%	1.6%	0%	0.565
Week 32	Poly.	3%	14.8%	48%	2*10^-6
Week 32	Oligo.	0%	6.6%	0%	0.038
Week 36	Poly.	0%	25.9%	57.1%	4.7*10^-10
Week 36	Oligo.	0%	8.6%	0 %	0.022

3.2 Doppler velocimetry

UA-PI and MCA-PI values are depicted in Figure 1. Both the GDM and PGDM groups showed significantly lower values (p<0.001) of UA-PI compared to Controls, while no significant differences were observed between GDMs and PGDMs during the weeks of gestation examined. Consistently with results already reported in the literature (21), UA-PI decreased with gestational age. As shown in Figure 1, and confirmed by the significant interaction between time and group (p<0.001), in GDMs and PDGMs this decrease is more pronounced than in Controls. Moreover, this decrease during pregnancy is more evident in GDMs compared to PGDMs.

Concerning MCA-PI values, a slight difference between groups was noticed (p=0.027). In particular, GDMs assumed lower values compared to Controls (p=0.028). MCA-PI values have been reported to increase with the gestational age starting from the 20th week of gestation and reach a peak at around 32-34 weeks (21). In our data, however, we observed a significant reduction over time (p<0.001) from week 26 to 36. We did not observe differences in the evolution with time across groups (p=0.338).

For completeness, we underline that 7 participants (3 GDMs and 4 PGDMs) gave birth before the 36th week of gestation, and consequently could not be included in the repeated measures models discussed.

3.3 CTG parameters

CTG parameters at weeks 31 to 34+6, 35 to 36+6 and from 37 to delivery are reported in Table 3. When multiple recordings were available for the same subject in the same group of weeks, the average was considered.

Table 3 Values of FHR parameters from cCTG recordings (median and quartiles)

	31w + 0d – 34w + 6d			35 w + 0d - 36 w + 6d			>37 w + 0d
	Controls	GDMs	PGDMs	Controls	GDMs	PGDMs	Controls	GDMs	PGDMs
MeanFHR [bpm]	140.1 (133.2 - 142.2)	138.8 (136.8 – 142.7)	139.9 (135.0 – 144.1)	139.8 (135.6 – 143.3)	140.0 (135.6 – 144.0)	137.1 (131.9 – 142.8)	138.0 (131.1 – 142.0)	139.2 (134.5 – 146.0)	139.7 (134.2 – 145.5)
STV [ms]	6.45 (4.62 – 7.95)	6.32 (4.65 – 7.38)	5.86 (5.30 – 6.30)	7.09 (5.53 – 9.00)	6.52 (5.44 – 7.30)	5.79 (5.44 – 6.77)	7.84 (5.47 – 9.20)	6.54 (5.30 – 7.61)	6.20 (4.72 – 6.40)
LF [%]	35.8 (30.9 – 37.8)	33.7 (26.4 – 36.5)	35.3 (28.6 – 37.9)	33.3 (30.4 – 40.4)	32.8 (29.4 – 36.6)	32.3 (26.8 – 36.7)	33. 9 (28.7 – 38.2)	33.5 (29.3 – 36.4)	29.7 (25.9 – 34.2)
MF [%]	5.1 (4.4 – 6.4)	5.5 (4.3 – 7.2)	6.5 (4.7 – 8.1)	4.9 (4.1 – 6.4)	5.7 (4.5 – 6.7)	6.0 (4.3 – 6.5)	5.1 (4.1 – 6.2)	4.8 (4.1 – 5.7)	4.3 (3.3 – 5.2)
HF [%]	1.4 (1.3 – 1.8)	1.6 (1.1 – 2.0)	2.0 (1.4 – 2.5)	1.3 (1.0 – 1.8)	1.4 (1.1 – 2.0)	1.8 (1.2 – 2.1)	1.4 (1.1 – 1.7)	1.2 (1.0 -1.5)	1.1 (0.8 – 1.5)
MFprsa [%]	31.1 (27.7 – 37.4)	37.9 (31.5 – 41.0)	36.8 (32.3 – 38.4)	29.4 (20.8 – 37.0)	34.2 (29.1 – 39.5)	33.9 (29.8 – 38.1)	30.8 (24.4 – 34.1)	33.2 (28.4 – 36.6)	30.6 (24.8 – 35.7)
HFprsa [%]	36.3 (22.4 – 40.0)	26.6 (21.9 – 36.0)	31.9 (26.4 – 40.5)	27.0 (17.0 – 35.6)	28.0 (22.1 - 32.7)	29.9 (27.9 - 34.1)	23.6 (17.9 – 31.9)	20.9 (15.7 – 29.0)	27.3 (17.6 – 34.9)
LZ3	0.96 (0.94 – 0.99)	0.97 (0.96 – 0.99)	0.97 (0.97 – 0.99)	0.94 (0.91 – 0.96)	0.96 (0.94 – 0.98)	0.97 (0.95 – 0.99)	0.93 (0.90 – 0.96)	0.96 (0.94 – 0.98)	0.97 (0.95 – 0.98)
SampEn	0.95 (0.84 – 1.14)	0.94 (0.86 – 1.00)	1.02 (0.99 – 1.15)	0.99 (0.75 – 1.14)	1.03 (0.84 – 1.13)	1.00 (0.86 – 1.10)	0.95 (0.79 – 1.05)	0.91 (0.79 – 1.02)	0.96 (0.77 – 1.06)
DC [ms]	2.34 (1.73 – 2.77)	2.58 (2.10 – 2.91)	2.34 (2.15 – 2.48)	2.48 (1.96 – 2.78)	2.57 (2.22 – 3.02)	2.51 (2.21 2.70)	2.74 (2.01 – 3.27)	2.58 (2.17 – 3.07)	2.31 (1.89 – 2.80)
DR [ms]	0.315 (0.212 – 0.400)	0.309 (0.213 – 0.466)	0.308 (0.221 – 0.360)	0.341 (0.199 – 0.457)	0.349 (0.281 – 0.563)	0.323 (0.241 – 0.428)	0.408 (0.276 – 0.578)	0.392 (0.292 – 0.543)	0.376 (0.178 – 0.507)
#Acc. [no./hour]	11.4 (7.2 – 15.6)	12.7 (10.3 – 16.7)	10.1 (7.2 – 16.0)	16.3 (11.0 – 19.2)	15.8 (11.1 – 18.5)	12.8 (8.9 – 18.2)	16.9 (13.9 – 22.0)	15.3 (11.7 – 20.3)	16.1 (13.5 – 22.2)
MovF [%]	18.5 (15.6 – 24.1)	23.9 (17.0 - 27.1)	19.3 (12.7 – 28.4)	22.2 (12.5 – 30.0)	19.5 (14.1 -28.3)	20.4 (14.0 – 24.0)	14.7 (12.2 – 23.6)	18.4 (12.3 – 26.0)	14.6 (9.6 – 25.8)

Concerning differences between the two diabetic groups and Controls, we identified a statistical significance only for MFprsa, LZ3 and STV. In particular, MFprsa significantly increased for GDMs at weeks 35-36 (p=0.03). Similarly, MF% seems to be greater in GDMs and PGDMs in this gestational age (p=0.06). Both diabetic groups showed significantly higher LZ3 at weeks 35-36 (p=0.008 and p=0.006 for GDMs and PGDMs, respectively) and at weeks 37+ (p=0.005 and p=0.011). A borderline-significant reduction in STV (p=0.049) was identified at weeks 37+ for PGDMs.

The results obtained with LME models confirm those obtained using independent non-parametric tests: GDMs and PGDMs are associated with significantly higher values than MFprsa and LZ3. LME models also identified a significant reduction in STV for both diabetic groups. The inclusion of fasting glucose level in the third trimester as a fixed effect did not affect the results for MFprsa and LZ3, while it reduced the relation between STV and the group, making it non-significant. This result suggests that the decrease in STV may be mediated by the higher fasting glucose levels in the diabetic groups.

In general, no significant differences in CTG parameters were identified between the two diabetic groups.

3.4 Pregnancy outcomes

We did not identify significant differences among groups concerning the increase in maternal BMI (ANOVA test, p=0.069), which was on average 3.52±2.12 Kg/m².

As reported in Table 4 and shown in Figure 2, the overall incidence of cesarean section in both diabetic groups was more than double compared to Controls. This result is motivated by the need to anticipate delivery in patients with gestational diabetes (within 40 weeks + 6 days) and pre-gestational diabetes (within 39 weeks + 6 days) according to international guidelines (1,2). In 24% of PGDMs and 6.5% of GDMs delivery happened before 37 weeks due to maternal or fetal complications, such as maternal hyperglycemia or IUGR. These results are in agreement with (22).

Table 4 Maternal delivery outcomes

	Controls	GDMs	PGDMs
Weight gain [Kg/m²]	3.9±2.3	3.1±2.0	3.6±1.9
Cesarean Section	33.3%	68.8%	72%
Pre-Term Birth	0%	6.5%	24%

Substantial differences among groups in the umbilical cord blood gas analysis were identified and are reported in Table 5. The pH of the umbilical cord at birth assumed lower values in PGDMs compared to both GDMs and Controls, and spontaneous deliveries were associated with lower pH. pO2 decreased progressively from Controls to GDMs and PGDMs and was not significantly affected by mode of delivery, while pCO2 followed an opposite trend. LACTATES were higher in PGDMs compared to both other groups and were lower in cesarean sections compared to vaginal deliveries.

PGDMs were associated with higher birth weight compared to Controls and GDMS. In particular, in the PGDM group, 28% of neonates presented a weight over the 95th percentile, calculated using the values reported in (23). Differences would have likely been more marked if delivery had not been anticipated in the diabetic groups.

Despite most of the Apgar scores at one minute (Apgar1) being good or very good (8 or 9) in all the groups, both GDMs and PGDMs were associated with lower Apgar1. An Apgar1 lower than seven was assigned to 1 newborn in the Control group, 5 in GDMs and 2 in PGDMs. At 5 minutes, only one newborn in the PGDMs group had an Apgar score (Apgar5) lower than 7. The percentage of PGDMs that required admission to NICU was about twice that of GDMs, and was higher for vaginal deliveries.

Table 5 Comparison of fetal Outcomes at delivery. Values are reported as mean±std or percentages where appropriate. On the left are reported average values without stratifying by mode of delivery. On the right, values have been divided between vaginal delivery (gray) and cesarean sections (white).

	Controls (Group 1)		GDMs (Group 2)		PGDMs (Group 3)		p-vals
pH	7.26±0.08	7.25±0.08	7.23±0.11	7.21±0.10	7.18±0.09	7.12±0.08	0.1245^1-2 <0.001^1-3 0.0340^2-3 0.008^DEL
		7.28±0.09		7.25±0.12		7.20±0.09
PO_{2 [}_mmHg]	22.12±6.48	21.4±6.38	17.94±7.24	20.17±6.53	11.99±4.65	10.53±4.28	0.0057^1-2 <0.001^1-3 <0.001^2-3 0.8293 ^DEL
		23.5±6.60		16.93±7.39		12.56±4.78
PCO_{2 [}_mmHg]	39.00±8.94	40.62±9.77	49.68±12.97	45.11±10.93	57.67±15.45	63.17±15.36	<0.001^1-2 <0.001^1-3 0.0014^2-3 0.3927^DEL
		35.76±5.95		51.76±13.40		55.53±15.38
BECF _[_meq/L]	-4.84±3.63	-4.99±3.82	-5.26±4.49	-5.45±3.50	-5.92±4.07	-6.69±3.43	0.999^1-2 0.5723^1-3 0.999^2-3 0.4554^DEL
		-4.54±3.31		-5.18±4.91		-5.62±4.35
Lactates _[_mmol/L]	3.38±1.15	3.53±1.13	3.28±1.61	3.40±1.10	4.10±2.02	5. 15±2.25	0.999^1-2 0.0128^1-3 0.0161^2-3 0.0171^DEL
		3.07±1.16		3.22±1.81		3.69±1.82
HCO_{3 [}_mmol/L]	24.47±3.46	24.62±3.26	20.90±4.78	20.94±3.89	20.24±5.39	19.75±4.05	<0.001^1-2 <0.001^1-3 0.999^2-3 0.947^DEL
		24.17±3.90		20.88±5.18		20.44±5.93
apgar1<7 [%]	1.5%	2.3%	8.2%	0%	8%	14.3%	0.1043^1-2 0.1819^1-3 0.999^2-3
		0%		11.9%		5.6%
apgar5<7 [%]	0%	0%	0%	0%	4%	0%	0.999^1-2 0.1819^1-3 0.2907^2-3
		0%		0%		5.6%
weight < 5pc [%]	9.1%	13.6%	13.1%	0%	4%	0%	0.5747^1-2 0.6688^1-3 0.2740^2-3
		0%		19.0%		5.6%
weight > 95pc [%]	0%	0%	4.9%	0%	28%	14.3%	0.1080^1-2 <0.001^1-3 0.0055^2-3
		0%		7.1%		33.3%
NICU	4.55%	4.55%	14.8%	21.1%	28.0%	42.9%	0.0685^1-2 0.0038^1-3 0.2208^2-3
		4.55%		11.9%		22.2%

More pregnancies with polyhydramnios were delivered by Cesarean section, both in PGDMs (92.3%) and GDMs (66.7%). The decision was made because of maternal hyperglycemia and suspected macrosomia. Among pregnancies with polyhydramnios, 3 newborns showed a pH<7.05. This happened both among newborns delivered by cesarean (1 in GDM and 1 PGDM) and vaginal delivery (1 PGDM). Concerning oligohydramnios, 3 cases were diagnosed with IUGR defined as birth weight <3rd percentile, and were delivered by cesarean section. Two newborns required NICU admission, one of whom had a birth pH<7.

The main goal of this work was to compare antepartum measurements including amniotic fluid alterations, fetal Doppler indices, cCTG parameters, and outcome data of the diabetic populations with respect to healthy controls. In the following sections, the main results of this study are summarized and discussed.

4.1 Amniotic fluid evaluation

PGDM and GDM are the most frequent etiological causes of polyhydramnios with a variable prevalence (24). In our study, virtually all cases of polyhydramnios occurred in the diabetic group. In particular, we observed a more than double incidence of polyhydramnios in PGDMs compared to GDMs at the 32nd and 36th week of gestation. Polyhydramnios was associated with worse maternal and fetal outcomes, in particular, a higher percentage of Cesarean sections and fetal pH lower than 7.05 at birth.

However, the significance of polyhydramnios in diabetic pregnancies remains unclear. Shoham et al. (25) did not find an increased risk of perinatal morbidity and mortality in GDMs complicated with polyhydramnios when compared to those with a normal volume of amniotic fluid. In PDGMs, polyhydramnios has been directly related to maternal hyperglycemia and was associated with worse maternal and neonatal outcomes (26). Idris et al. (27) pointed out that a large study on the effect of polyhydramnios in PDGMs on maternal and perinatal outcomes was still missing.

In this study, we observed a very low incidence of oligohydramnios, which was associated with the highest incidence of IUGR, need for cesarean section and admission to NICU.

Indeed, oligohydramnios has been related in the literature to increased risk of IUGR, cesarean delivery, low Apgar score, NICU admission and respiratory distress syndrome (28). However, previous studies examining diabetes and oligohydramnios have come to disparate conclusions.

4.2 Doppler velocimetry

Findings in doppler evaluation show a reduction in the UA-PI and MCA-PI values in both the diabetic populations compared to controls, especially regarding UA-PI. These results suggest the existence of a similar pathophysiological mechanism occurring in GDM and PGDM, that induces a decreased impedance to flow in UA and MCA. These may be generated by changes in the placental structure and vascularization (29). In previous studies (4, 30) it was found that the umbilical cord was significantly larger in fetuses of women with GDM compared to the normal population and UA diameter is much larger in the case of fetal macrosomia (31). This explains why classical reference ranges of UA-PI and MCA-PI do not show adequate efficacy to predict adverse neonatal outcomes in pregnancies complicated by PGDM and GDM. Our results are in agreement with the current literature, which suggests that Doppler evaluation of the UA is probably not a good tool for evaluating the fetal well-being in pregnancies with GDM and PGDM (1, 2, 32). Despite lower values of UA-PI and MCA-PI are associated with better neonatal outcomes in the general population, our results suggest that in the diabetic population it might be useful to introduce a cut-off for lower values of UA-PI and MCA-PI to identify fetuses at higher risk of adverse neonatal outcome. This hypothesis warrants further investigation.

Fetuses of women with GDM and PGDM could be normoxic (33), and so might not show any anomalies in MCA-PI. On the other hand, fetal tissue may not receive sufficient oxygen, even if oxygen is at normal levels in the circulation with normal MCA-PI values (34). Therefore, as shown also in (1, 2, 32, 35–37) MCA-PI values routinely used for monitoring high-risk pregnancies are not able to predict fetal well-being among women with GDM and PGDM.

4.3 cCTG parameters

With regard to cCTG parameters, very few studies investigated their application in the management of diabetes and found weak associations with fetal well-being. Notably, some authors found a positive association between maternal hyperglycemia and elevated FHR baseline and lower STV values analyzed by cCTG in the third trimester in diabetic women (38, 39). Findings were linked with a delay in the physiological maturation of the autonomic nervous system (ANS). Our results did not show higher FHR baseline values but confirmed a small reduction in STV values in the PGDM group compared to Controls at term of pregnancy, which can be linked to maternal hyperglycemia.

In this study, we also investigated a comprehensive set of non-standard cCTG parameters and found some associations with LZC and MFprsa. Concerning the potential application of LZC to FHR monitoring, both diabetic groups showed higher LZ3 values compared to Controls from the 35th week of gestation. To the best of our knowledge, the only study evaluating LZC values in fetuses from diabetic pregnancies is Steyde et al. (40). The results obtained in this study are consistent with those already reported, i.e., an increase in LZ3 in GDMs. Moreover, we also identified an increase in LZ3 in PGDMs. Previous studies applied LZC to monitor IUGR fetuses (41) and showed that they were characterized by higher values, very close to 1, supporting the hypothesis that the FHR belonging to the severe IUGRs is less regular than in healthy fetuses.

Among frequency parameters, MFprsa and MF have shown some discriminative capacity between GDMs and Controls, in agreement with (20, 40). The physiological meaning of this finding and possible clinical implications remain unclear. However, this frequency band has been associated with fetal movements, which may suggest a state of hyperactivity in GDMs.

FHR patterns in combination with fetal movements, as an expression of activity, are the cornerstones to assess fetal well-being. However, it is impossible to attribute to cCTG the same reliability in diabetic pregnancies as we do for normal pregnancies, even when diabetic women present in euglycemic conditions. This is due to the fact that fetal acidemia can occur abruptly as a consequence of sudden maternal hyperglycemia.

4.4 Pregnancy outcomes

Umbilical cord blood gas analysis has increasingly been recognized as the most reliable indication of fetal oxygenation and acid-base condition at birth. Indeed, since blood flows directly from the fetus, umbilical arterial blood accurately reflects its status. However, few studies analyzed umbilical cord blood gas values in neonates of diabetic mothers stratified by mode of delivery. In this study, we detected a significant decrease in pH and pO2 and an increase in pCO2 values in neonates of diabetic mothers. Moreover, values in PGDMs deviated more than GDMs compared to healthy pregnancies. These results suggest the presence of a trend towards chronic acidosis in diabetic pregnancies. Indeed, hyperglycemia causes a chronic deficit in the transport of oxygen and fetal metabolites, which coupled with osmotically induced placental villous edema can lead to impaired fetal oxygen transport in diabetic pregnancies (42, 43). This trend towards chronic acidosis is further confirmed by higher lactates and lower HCO3 levels in PGDMs compared to GDMs and Controls. Moreover, this trend is more evident after spontaneous delivery compared to elective cesarean section. This is probably because labor involves a condition of physiological fetal stress. Therefore, when a fetus has an increased susceptibility to hypoxia/acidosis before labor, it is more frequent to observe altered umbilical cord blood gas values following spontaneous delivery. Reduced oxygen levels and increased lactate levels at birth have also been found in GDMs with normal maternal glucose levels during pregnancy (44). Indeed, multiple reviews (22, 45) confirm that both PGDM and GDM are associated with pregnancy complications and adverse neonatal outcomes, and the majority show that adverse neonatal outcomes are more common in PGDMs than in GDMs.

The present work aimed at representing a valuable aid for managing pregnancies complicated by maternal diabetes. Various methodologies have been applied and the extracted indices were compared in cases and controls. Although some significant variations were observed, the main conclusion is that none of the methods by itself can provide a reliable evaluation of fetal well-being. Therefore, it is important to strengthen the few existing analysis tools in order to extract as much information as possible. Fetal Doppler analysis has shown the importance of an in-depth study on fetal placental hemodynamics in pregnancies complicated by diabetes. Concerning cCTG, the only parameters which appear to have some discrimination abilities are not conventional, which suggests that standard methods of analysis do not capture all the information from the FHR signal. We hypothesize that a quantitative multi-parameter approach could improve the management of diabetic pregnancies.

Our present work can be considered a preliminary investigation of a very complex maternal-fetal condition, and its results need future confirmation in a larger and more heterogeneous population.

Umbilical Artery (UA); Middle Cerebral Artery (MCA); Cardiotocography (CTG); intrauterine growth restriction (IUGR); pregestational diabetes mellitus (PGDM); gestational diabetes mellitus (GDM); fetal heart rate (FHR); oral glucose tolerance test (OGTT); hemoglobin A1C (HbA1C); fetal movements series (FM); neonatal intensive care unit (NICU); computerized CTG (cCTG); average of the FHR baseline (MeanFHR); number of accelerations per hour (#Acc); Short Time Variation (STV); Sample Entropy (SampEn); ternary Lempel-Ziv complexity (LZ3); Phase Rectified Signal Averaging (PRSA); Deceleration Capacity (DC); Deceleration Reserve (DR); Fetal movements (MovF%); UA Pulsatility Index (UA-PI); MCA Pulsatility Index (MCA-PI); Linear Mixed Effect models (LME); Cord arterial base excess (BECF); autonomic nervous system (ANS).

Ethics approval and consent to participate: This research involving human data was performed in accordance with the Declaration of Helsinki. All patients/participants provided their written informed consent to participate in this study. The dataset was completely anonymized before data analysis. The study was reviewed and approved by the ethics committee CAMPANIA 3, which comprises “Ethics Committee A.O.U. Federico II – A.O.R.N. Cardarelli”

Consent for publication: Not applicable

Availability of data and materials: the datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests: the authors declare that they have no competing interests.

Funding: this work was supported by PRIN 2017 MIUR, grant No. ICT4MOMs #2017RR5EW3.

Authors' contributions: MGS, GM, and GMM contributed to the conception and design of the study. MGT and ST carried out the recording of the data. GS, ST, MGS, ES and GM planned and carried out the simulations and analysis of data and took the lead in writing the manuscript. GS, ST, MGS and GM contributed to the interpretation of the results. All authors provided critical feedback and helped shape the research, analysis, and manuscript.

Caughey A, Turrentine M. ACOG Practice Bulletin: Gestational Diabetes Mellitus. Obstetrics and gynecology. 2017;130(1):e17–31.
Caughey A, Kaimal A, Gabbe S. ACOG Practice Bulletin No. 201: Pregestational Diabetes Mellitus. Obstetrics & Gynecology. 2018 Dec;132(6):e228–48.
Durnwald CP, Mele L, Spong CY, Ramin SM, Varner MW, Rouse DJ, et al. Glycemic characteristics and neonatal outcomes of women treated for mild gestational diabetes. Obstetrics and Gynecology. 2011 Apr;117(4):819–27.
Sirico A, Rizzo G, Maruotti GM, Aiello E, Morlando M, Arduini D, et al. Does fetal macrosomia affect umbilical artery Doppler velocity waveforms in pregnancies complicated by gestational diabetes? Journal of Maternal-Fetal and Neonatal Medicine. 2016 Oct 17;29(20):3266–70.
Signorini MG, Magenes G, Cerutti S, Arduini D. Linear and nonlinear parameters for the analysis of fetal heart rate signal from cardiotocographic recordings. IEEE Trans Biomed Eng. 2003;50(3):365–74.
Pardey J, Moulden M, Redman CWG. A computer system for the numerical analysis of nonstress tests. Am J Obstet Gynecol. 2002;186(5):1095–103.
Signorini MG, Fanelli A, Magenes G. Monitoring fetal heart rate during pregnancy: Contributions from advanced signal processing and wearable technology. Comput Math Methods Med. 2014;2014.
Hadlock FP, Harrist RB, Sharman RS, Deter RL, Park SK. Estimation of fetal weight with the use of head, body, and femur measurements—A prospective study. Am J Obstet Gynecol. 1985 Feb;151(3):333–7.
Metzger B, Gabbe S, Persson B, Buchanan T, Catalano P, Damn P. International Association of Diabetes and Pregnancy Study Groups Recommendations on the Diagnosis and Classification of Hyperglycemia in Pregnancy. Diabetes Care. 2010 Mar 1;33(3):676–82.
American Diabetes Association. 2. Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes—2018. Diabetes Care. 2018 Jan 1;41(Supplement_1):S13–27.
Magann EF, Chauhan SP, Doherty DA, Lutgendorf MA, Magann MI, Morrison JC. A Review of Idiopathic Hydramnios and Pregnancy Outcomes. Obstet Gynecol Surv. 2007 Dec;62(12):795–802.
Oros D, Figueras F, Cruz-Martinez R, Meler E, Munmany M, Gratacos E. Longitudinal changes in uterine, umbilical and fetal cerebral Doppler indices in late-onset small-for-gestational age fetuses. Ultrasound in Obstetrics and Gynecology. 2011 Feb;37(2):191–5.
Magenes G, Signorini MG, Ferrario M, Lunghi F. 2CTG2: A new system for the antepartum analysis of fetal heart rate. In: IFMBE Proceedings. Springer Verlag; 2007. p. 781–4.
Mantel R, Van Geijn HP, Caron FJM, Swartjes JM, Van Woerden EE, Jongsma HW. Computer Analysis of Antepartum fetal heart rate. Int J Biomed Comput. 1990;25(4):261–86.
Arduini D, Rizzo G, Piana A, Brambilla BP, Romanini C. Computerized analysis of fetal heart rate: I. description of the system (2CTG). Journal of Maternal-Fetal Investigation. 1993;3(3):159–63.
Moorman JR. Physiological time-series analysis using approximate entropy and sample entropy. American Journal of Physiology Heart and Circulatory Physiolog. 2000;2039–49.
Ferrario M, Signorini MG, Magenes G. Comparison between fetal heart rate standard parameters and complexity indexes for the identification of severe intrauterine growth restriction. Methods Inf Med. 2007;46(2):186–90.
Bauer A, Kantelhardt JW, Bunde A, Barthel P, Schneider R, Malik M, et al. Phase-rectified signal averaging detects quasi-periodicities in non-stationary data. Physica A: Statistical Mechanics and its Applications. 2006;364:423–34.
Rivolta MW, Stampalija T, Frasch MG, Sassi R. Theoretical Value of Deceleration Capacity Points to Deceleration Reserve of Fetal Heart Rate. IEEE Trans Biomed Eng. 2020;67(4):1176–85.
Steyde G, Spairani E, Magenes G, Signorini MG. Fetal heart rate spectral analysis in raw signals and PRSA-derived curve: normal and pathological fetuses discrimination. Med Biol Eng Comput. 2023;
Ciobanu A, Wright A, Syngelaki A, Wright D, Akolekar R, Nicolaides KH. Fetal Medicine Foundation reference ranges for umbilical artery and middle cerebral artery pulsatility index and cerebroplacental ratio. Ultrasound in Obstetrics and Gynecology. 2019 Apr 1;53(4):465–72.
Ye W, Luo C, Huang J, Li C, Liu Z, Liu F. Gestational diabetes mellitus and adverse pregnancy outcomes: systematic review and meta-analysis. The BMJ. 2022;
Kiserud T, Piaggio G, Carroli G, Widmer M, Carvalho J, Neerup Jensen L, et al. The World Health Organization Fetal Growth Charts: A Multinational Longitudinal Study of Ultrasound Biometric Measurements and Estimated Fetal Weight. PLoS Med. 2017 Jan 1;14(1).
Frank Wolf M, Peleg D, Stahl-Rosenzweig T, Kurzweil Y, Yogev Y. Isolated polyhydramnios in the third trimester: is a gestational diabetes evaluation of value? Gynecological Endocrinology. 2017 Nov 2;33(11):849–52.
Shoham I, Wiznitzer A, Silberstein T, Fraser D, Holcberg G, Katz M, et al. Gestational diabetes complicated by hydramnios was not associated with increased risk of perinatal morbidity and mortality. European Journal of Obstetrics & Gynecology and Reproductive Biology. 2001 Dec;100(1):46–9.
Hamza A, Herr D, Solomayer EF, Meyberg-Solomayer G. Polyhydramnios: Causes, diagnosis and therapy. Vol. 73, Geburtshilfe und Frauenheilkunde. Georg Thieme Verlag; 2013. p. 1241–6.
Idris N, Wong SF, Thomae M, Gardener G, Mcintyre DH. Influence of polyhydramnios on perinatal outcome in pregestational diabetic pregnancies. Ultrasound in Obstetrics and Gynecology. 2010 Sep;36(3):338–43.
Rabie N, Magann E, Steelman S, Ounpraseuth S. Oligohydramnios in complicated and uncomplicated pregnancy: a systematic review and meta-analysis. Ultrasound in Obstetrics & Gynecology. 2017 Apr;49(4):442–9.
Vambergue A, Fajardy I. Consequences of gestational and pregestational diabetes on placental function and birth weight. World J Diabetes. 2011;2(11):196.
Weissman A, Jakobi P. Sonographic measurements of the umbilical cord in pregnancies complicated by gestational diabetes. Journal of Ultrasound in Medicine. 1997;16(10):691–4.
Cromi A, Ghezzi F, Naro EDI, Siesto G, Bergamini V, Raio L. Large cross-sectional area of the umbilical cord as a predictor of fetal macrosomia. Ultrasound in Obstetrics and Gynecology. 2007 Nov;30(6):861–6.
Shabani Zanjani M, Nasirzadeh R, Fereshtehnejad SM, Yoonesi Asl L, Alemzadeh SAP, Askari S. Fetal cerebral hemodynamic in gestational diabetic versus normal pregnancies: A Doppler velocimetry of middle cerebral and umbilical arteries. Acta Neurol Belg. 2014;114(1):15–23.
Bradley RJ, Brudenell JM, Nicolaides KH. Fetal Acidosis and Hyperlacticaemia Diagnosed by Cordocentesis in Pregnancies Complicated by Maternal Diabetes Mellitus. Diabetic Medicine. 1991 Jun;8(5):464–8.
Salvesen DR, Brudenell MJ, Nicolaides KH. Fetal polycythemia and thrombocytopenia in pregnancies complicated by maternal diabetes mellitus. Am J Obstet Gynecol. 1992 Apr;166(4):1287–93.
Kutuk MS, Dolanbay M, Gokmen Karasu AF, Ozgun MT. Relationship between fetal peak systolic velocity in Middle cerebral artery and umbilical blood gas values and hemoglobin levels in diabetic pregnant women. Journal of Clinical Ultrasound. 2018 Jul 1;46(6):391–6.
Salvesen DR, Higueras MT, Mansur CA, Freeman J, Brudenell JM, Nicolaides KH. Placental and fetal Doppler velocimetry in pregnancies complicated by maternal diabetes mellitus. Am J Obstet Gynecol. 1993 Feb;168(2):645–52.
Leung WC, Lam H, Lee CP, Lao TT. Doppler study of the umbilical and fetal middle cerebral arteries in women with gestational diabetes mellitus. Ultrasound in Obstetrics and Gynecology. 2004 Oct;24(5):534–7.
Costa VN, Nomura RMY, Reynolds KS, Miyadahira S, Zugaib M. Effects of maternal glycemia on fetal heart rate in pregnancies complicated by pregestational diabetes mellitus. European Journal of Obstetrics & Gynecology and Reproductive Biology. 2009 Mar;143(1):14–7.
Gragnano E, Saccone G, Tagliaferri S, Esposito G, Paino JA, Campanile M, et al. Computerized cardiotocography in singleton gestations with diabetes mellitus: a retrospective study. Minerva Obstetrics and Gynecology. 2022 Apr;74(2).
Steyde G, Daniele B, Spairani E, Magenes G, Signorini MG. Influence of Gestational Diabetes on Fetal Heart Rate in antepartum Cardiotocographic Recordings. In: Computing in Cardiology . 2022.
Ferrario M, Maria G S, Magenes G. Complexity analysis of the fetal heart rate variability: early identification of severe intrauterine growth-restricted fetuses. Med Biol Eng Comput. 2009;47:911–9.
Cunningham F, Leveno KJ, Bloom SL, Dashe JS, Hoffman BL, Casey BM, et al. Diabetes Mellitus. In: Williams Obstetrics. 2010. p. 1104–25.
Namavar Jahromi B, Ahmadi N, Cohan N, Jahromi MRN. Comparison of the umbilical artery blood gas, nucleated red blood cell, C-reactive protein, and white blood cell differential counts between neonates of diabetic and nondiabetic mothers. Taiwan J Obstet Gynecol. 2011 Sep;50(3):301–5.
Taricco E, Radaelli T, Rossi G, Nobile De Santis MS, Bulfamante GP, Avagliano L, et al. Effects of gestational diabetes on fetal oxygen and glucose levels in vivo. BJOG. 2009 Dec;116(13):1729–35.
Malaza N, Masete M, Adam S, Dias S, Nyawo T, Pheiffer C. A Systematic Review to Compare Adverse Pregnancy Outcomes in Women with Pregestational Diabetes and Gestational Diabetes. Vol. 19, International Journal of Environmental Research and Public Health. MDPI; 2022.

No competing interests reported.

Monitoring fetal well-being and delivery outcomes in pregnancies complicated by gestational and pregestational diabetes

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

1. Background

2. Methods

2.2 Statistical Analysis

3. Results

4. Discussion

4.1 Amniotic fluid evaluation

4.2 Doppler velocimetry

4.3 cCTG parameters

4.4 Pregnancy outcomes

5. Conclusions

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1