Advanced fetal cardiac monitoring in gestational diabetes mellitus: HbA1c remains the relevant predictor of perinatal outcome under optimal metabolic control.

Figures

• 

  Fig. 1. Cohort composition of 242 women who consented, 172 were included: 116 in the control cohort and 56 in the GDM cohort (31 diet-controlled [dGDM], 25 insulin-dependent [iGDM]). Adverse perinatal outcomes (APO; combined endpoint of operative delivery/emergency cesarean due to fetal distress, NICU admission, umbilical cord pH < 7.1, APGAR Score at 5 min < 7) occurred in 25 controls, 7 dGDM, and 5 iGDM cases

• 

  Fig. 2. Composition of adverse perinatal outcome in the total cohort: adverse perinatal outcomes (APO; combined endpoint of operative delivery/emergency cesarean due to fetal distress, NICU admission, umbilical cord pH < 7.1, and/or APGAR Score at 5 min < 7). A total of 25 cases in the control group and 12 in the GDM group experienced one or more of these events, in various combinations

• 

  Fig. 3. Maternal metabolic control in the study cohort: GDM, gestational diabetes mellitus; iGDM, insulin-managed GDM; dGDM, dietary-managed GDM; A insulin at Visit 1 differed between control and GDM (p = 0.023) and between dGDM and iGDM (p = 0.046); B HbA1c at Visit 1 differed between control and GDM (p = 0.032), but not between dGDM and iGDM (n.s.); C HbA1c at birth showed no difference between dGDM and iGDM (n.s.); * significant result < 0.05, n.s., non-significant result (statistical comparison by Mann–Whitney U Test). Displayed below are medians (interquartile range)

• 

  Fig. 4. Maternal metabolic state in dependence on the perinatal outcome in GDM (gestational diabetes mellitus) cohort: APO, adverse perinatal outcomes (combined endpoint of operative delivery/emergency cesarean due to fetal distress, NICU admission, umbilical cord pH < 7.1, APGAR Score at 5 min < 7): A HbA1c at Visit 1 differed between No APO and APO (p = 0.045); B HbA1c at birth differed between No APO and APO (p = 0.0017); * significant result < 0.05 (statistical comparison by Mann–Whitney U Test). Displayed below are medians (interquartile range)

• 

  Fig. 5. cCTG (computerized cardiotocography)-derived fHRV (fetal heart rate variability) parameters in the control and GDM cohort. Data are median (interquartile range), unless otherwise specified; GDM gestational diabetes mellitus, cCTG: computerized cardiotocography, A accelerations (p = 0.022); B short-term variability; C amplitude range (p = 0.037); D RMSSD; E ratio of sdnn to rmssd; F sdnn; G baseline heart rate; H kurtosis; I low-frequency band power; J high-frequency band power; K complexity parameter (p = 0.028); L complexity parameter; M acceleration capacity (p = 0.031); N deceleration capacity; O functional fetal autonomic brain age score * significant result < 0.05, n.s. non-significant result (statistical comparison by Mann–Whitney U Test)

• 

  Fig. 6. fECG (fetal electrocardiogram) parameters in the control and GDM cohort. Data are median (interquartile range), unless otherwise specified; GDM, gestational diabetes mellitus; fECG, fetal electrocardiogram; A gestational age; B PR-time; C QR-time; D RS-time; E QRS-time; F QT-time; G mean heart rate (p = 0.017); H RMSSD; I short-term variability (p = 0.046); J mean minute range; * significant result < 0.05, n.s., non-significant (statistical comparison by Mann–Whitney U Test)

• 

  Fig. 7. Predictive capacity of the selected fECG (fetal electrocardiogram) and cCTG (computerized cardiotocography) parameters individually adjusted for gestational age at delivery, maternal BMI, and HbA1c at study visit and additionally for GDM in the total cohort, each row represents one generalized linear model GDM, gestational diabetes mellitus, fECG, fetal electrocardiogram, cCTG, computerized cardiotocography, FHR, fetal heart rate, STV, short-term variation, numAC15 × 15, acceleration, w5AMP10, amplitude range, w5LF, low-frequency band power, w5HF, high-frequency band power, nFABAS, functional fetal autonomic brain age score