Neonatal Outcomes in Late Preterm Infants of Diabetic Mothers Exposed to Antenatal Steroids: A Cohort Study

Objective: To assess whether exposure to antenatal steroids (ANS) was associated with net neonatal benefits in late preterm infants born to mothers with pre-gestational diabetes mellitus on antidiabetic medications.

Methods: This retrospective cohort study included women with pre-existing diabetes mellitus and nonanomalous singleton gestation who delivered between 34^0/7 and 36^6/7 weeks at Inova Fairfax Women's and Children's Hospital between 2012–2020. Maternal characteristics and neonatal outcomes were determined from electronic health records. The co-primary dichotomous dependent variables were newborn oxygen support use (other than blow-by) and neonatal hypoglycemia requiring neonatal intensive care unit (NICU) admission. The primary independent variable was ANS exposure in the late preterm period from at least 12 hours to 7 days prior to delivery. Correlational analysis was used to identify independent predictors statistically significantly associated with the co-primary outcomes. Categorical techniques were used to identify statistically significant differences between neonate groups for potential confounders requiring adjustment with multivariate analyses. Logistic regression was used to determine the independent variables that best predicted the co-primary outcomes. A p-value of ≤0.05 was considered statistically significant. Analyses were conducted using SAS v9.2 (SAS Institute, Cary, NC).

Results: 99 mother-infant dyads met inclusion criteria. Of those, 39 were exposed to ANS and 60 were not. Infants exposed to ANS were more likely (48.1% versus 28.9%) to develop neonatal hypoglycemia requiring NICU admission (OR=2.29, 95% CI {0.99 to 5.28}, p = 0.053). Although marginally statistically significant, the number needed to harm (NNH) was 5. However, power was low, and we would have required a minimum sample size of 83 per arm to detect a significant difference of 19.2% at 80% power. The NNH of ANS on hypoglycemia was not offset by a favorable number needed to treat (NNT) of ANS to reduce oxygen support use. In fact, 43.6% of exposed infants required oxygen support use, while only 30% of the unexposed group required oxygen use (p = 0.17). Power for this association also was low and would have required a minimum sample of 183 per arm to detect a significant difference of 13.6% at 80% power. Other univariate associations with hypoglycemia were found (see [Table 1]). A logistic regression was performed, but an appropriate model could not be fitted.

Conclusion: Although limited by a small sample size, our findings suggest that any potential benefit of ANS in late preterm infants of diabetic mothers should be weighed against a likely higher risk of neonatal hypoglycemia requiring NICU admission.

Fig. 1 Gestational age at delivery

Table 1 Predictors of neonatal hypoglycemia in univariate analysis

Variables	Neonatal hypoglycemia		p-Value	OR (95% CI)
	Yes N = 54 (%)	No N = 45 (%)
Neonatal
Antenatal Steroids			0.053	2.29 (0.99–5.28)
Yes	26(48.1)	13(28.9)
No	28(51.9)	32(71.1)
Respiratory support			<0.0001	13.8 (4.33–44.1)
Yes	31(57.4)	4(8.9)
No	23(42.6)	41(91.1)
NICU admission			0.0001	262.4 (15.1–4563)
Yes	54(100)	13(28.9)
No	0(0)	32(71.1)
Length of NICU stay (days)	13.6 {10.3, 16.8}	3.56 {1.43, 5.68}	<0.0001	*
Feeding difficulties			0.002	7.43 (2.54–21.7)
Yes	26(48.1)	5(11.1)
No	28(51.9)	40(88.9)
Jaundice requiring phototherapy			0.0009	4.60 (1.87–11.3)
Yes	44(81.5)	22(48.9)
No	10(18.5)	23(51.1)
Transient tachypnea of newborn			0.002	6.52 (2.04–20.9)
Yes	21(38.9)	4(8.9)
No	33(61.1)	41(91.1)
Respiratory distress			0.0004	*
Yes	18(33.3)	2(4.4)
No	36(66.7)	43(95.6)
Gestational age at delivery (weeks)			0.002	*
34 weeks	12(22.2)	3(6.7)
35 weeks	20(37.0)	8(17.8)
36 weeks	22(40.8)	34(75.5)
ANS exposure prior to delivery (hours)	28.6 {18.1, 39.0}	15.9 {5.95, 25.8}	0.04	*
Maternal
Intrapartum glycemia (binary)			0.002	0.28 (0.10–0.79)
≤126 mg/dL	32(62.7)	36(85.7)
>126mg/dL	19(37.3)	6(14.3)
Intra-partum glycemia (interval)	125.7 {112.0, 139.4}	102.8 {94.7, 110.9}	0.05	*
Hemoglobin A1C in 2nd or 3rd trimester(binary)			0.02	0.33 (0.13–0.86)
<7	24(50.0)	27(75.0)
≥7	24(50.0)	9(25.0)
Hemoglobin A1C in 2nd or 3rd trimester (interval)	7.01 {6.68, 7.35}	6.32 {6.02, 6.61}	0.003	*

No conflict of interest has been declared by the author(s).

Publication History

Article published online:
17 September 2021