Maternal Hypertension Disorders and Neonatal Acute Kidney Injury: Results from the AWAKEN Study

Abstract

Objective

This study aimed to examine the association between maternal hypertension (HTN) exposure and neonatal acute kidney injury (AKI).

Study Design

Retrospective cohort study of 2,162 neonates admitted to 24 neonatal intensive care units (NICUs). Neonates were classified into the following exposure groups: any maternal HTN, chronic maternal HTN, preeclampsia/eclampsia, both, or neither. Demographics, clinical characteristics, and AKI status were compared using Chi-square and analysis of variance. General estimating logistic regression was used to estimate adjusted odds ratios and included a stratified analysis for site of delivery.

Result

Neonates exposed to any maternal HTN disorder had a tendency toward less overall and early AKI. When stratified by inborn versus outborn, exposure to both maternal HTN disorders was associated with a significantly reduced odds of early AKI only in the inborn neonates.

Conclusion

Exposure to maternal HTN, especially preeclampsia/eclampsia superimposed on chronic HTN, was associated with less likelihood of early AKI in the inborn group.

Hypertensive disorders of pregnancy affect 6 to 10% of pregnancies in the United States and are known to be associated with poor maternal and neonatal outcomes.^1–3 The definition of maternal hypertension (HTN) disorders has evolved but is generally classified as (1) chronic HTN (blood pressure [BP] ≥ 140/90mm Hg before 20 weeks’ gestation), (2) pregnancy induced or gestational HTN (new onset HTN diagnosed after 20 weeks’ gestation), (3) preeclampsia (new onset HTN diagnosed after 20 weeks’ gestation with proteinuria and/or evidence of end-organ compromise), and/or (4) eclampsia (preeclampsia with new onset generalized seizures).^4–6 Preeclampsia and/or eclampsia superimposed on chronic HTN occurs often and is associated with a higher prevalence of both maternal and neonatal morbidity and mortality and fetal growth restriction (FGR).^2,7 Furthermore, exposure to maternal HTN is associated with long-term effects, including elevated BP and adverse cardiovascular profiles, for both mother and offspring.^8–13

Recent studies have documented a paradoxical association between maternal HTN exposure and reduced neonatal mortality.^14–18 In addition, exposure to maternal HTN has been associated with less-severe brain injury and retinopathy of prematurity while not influencing other preterm morbidities such as bronchopulmonary dysplasia.^14,19–21 Maternal HTN exposure in multiple gestation pregnancies has been consistently noted to be associated with better neonatal outcomes.^22,23 Although the mechanism(s) linking maternal HTN exposure to improved neonatal outcomes is not clear, some have postulated that it could be related to an adaptation to accelerate fetal organ maturation in the stressed fetus or be related to the beneficial effects of antenatal steroids or magnesium sulfate on fetal outcomes.¹⁴ None of these large cohort studies have included data on neonatal kidney outcomes.

In several retrospective cohort studies describing neonatal acute kidney injury (AKI) in preterm infants, maternal HTN disorders have been associated with less neonatal AKI.^24–28 In the Assessment of Worldwide Acute Kidney Epidemiology in Neonates (AWAKEN) cohort, the largest multicenter and international study evaluating neonatal AKI to date, the incidence of neonatal AKI was nearly 30% and was associated with a higher adjusted odds of death.^29–31 Outborn delivery and gestational age (GA) <29 or >36 weeks are among the most highly associated risk factors for neonatal AKI in the AWAKEN cohort.^29,30,32 This secondary analysis of the AWAKEN dataset was designed to investigate whether exposure to maternal HTN disorders was associated with the presence, timing, or severity of neonatal AKI and to identify potential maternal or neonatal factors that modified this association such as the site of delivery and GA.

Materials and Methods

Study Population

AWAKEN is an international, retrospective, multicenter cohort study of neonates admitted to 24 neonatal intensive care units (NICUs) participating in the Neonatal Kidney Collaborative (NKC; Supplementary Appendix A [available in the online version]). Each center received approval from their Human Research Ethics Committee or Institutional Review Board and consent was waived. A detailed description of the methodology and the participating institutions was published.³³ Briefly, data were collected from January 1, 2014, to March 31, 2014, on each participant. For inclusion, neonates had to receive ≥ 48 hours of intravenous fluids. The inclusion criteria were designed to capture sick neonates at significant risk for AKI and those who had an expected hospitalization of at least 48 hours. Infants receiving routine care in the newborn nursery were not included in this study. Exclusion criteria included the following: (1) not admitted to the NICU during the study period, (2) admission at ≥14 days of life, (3) congenital heart disease requiring surgical repair at <7 days of life, (4) lethal chromosomal anomaly, (5) death within 48 hours of NICU admission, and (6) severe congenital kidney and urinary tract abnormalities. Data were collected from admission until discharge from the NICU, death, or 120 days after birth. Maternal and neonatal demographic and clinical information were extracted from the neonatal and/or maternal chart depending on data availability.

Definition of Acute Kidney Injury

AKI was defined by the neonatal modification of the Kidney Disease: Improving Global Outcomes (KDIGO) criteria (1) as an increase in serum creatinine (SCr) by at least 0.3 mg/dL, and (2) increased SCr of at least 50% more than the previous value and/or urine output (UOP) <1mL/kg/hour averaged over a 24-hour period from days 2 to 7.³³ First, we looked at the association between maternal HTN exposure and overall AKI. Then we looked at the association between maternal HTN exposure and early AKI, defined as occurring within the first 2 to 7 days after birth, and late AKI, defined as >7 days after birth.^30,31 AKI severity stages were assigned and are included in Supplemental Table S1 (available in the online version).²⁹ The frequency and methodology for SCr monitoring and measurement were center dependent and variations by site have been previously published.²⁹ One-hundred forty patients did not have sufficient SCr or UOP data to determine AKI status. To determine whether the 140 neonates’ missing SCr/UOP data were different from those who had the data, a Chi-square and t-test analyses comparing the two groups were conducted and found that those with missing data were generally full-term neonates, of normal birth weight, and less sick than their counterparts; therefore, it is likely that they did not have AKI. In addition, there was no difference in the distribution of maternal HTN category by whether the neonate had SCr/UOP data, suggesting that their inclusion in statistical analyses would result in a nondifferential bias toward the statistical null. This factor has also been described in prior AWAKEN publications that also showed no difference in mortality and length of stay outcomes by AKI status, suggesting no selection bias from the inclusion of the 140 neonates without an AKI status.²⁹ All data were stored in a web-based database, MediData Rave, housed at Cincinnati Children’s Hospital Medical Center.

Maternal Hypertension Categories

Data regarding maternal HTN status were collected in the following manner: chronic HTN (HTN occurring prior to the current pregnancy with or without current medications), preeclampsia (with or without current medications including any mother with pregnancy induced HTN), and eclampsia (any seizures associated with maternal HTN or HELLP (hemolysis, elevated liver enzymes, and a low platelet count) syndrome.³³ We then classified the group into five maternal HTN categories based on the presence or absence of chronic HTN and/or preeclampsia/eclampsia (pre/eclampsia). Thus, for the purposes of our secondary analysis, each neonate was categorized as born to a mother with (1) any HTN disorder, (2) chronic HTN, (3) pre/eclampsia (pre/eclampsia), (4) neither, or (5) both (chronic HTN and pre/eclampsia). The “both” category was an independent group, and neonates in this category were not counted in either the chronic HTN or pre/eclampsia categories. Maternal antihypertensive medication use including calcium channel blockers, central α-agonists, β-blockers, vasodilators, and angiotensin converting enzyme inhibitors, was recorded.

Statistical Analyses

Neonatal and maternal characteristics were compared among HTN groups using a Pearson’s Chi-square test and analysis of variance (ANOVA) for categorical and continuous variables, respectively. A Pearson’s Chi-square test was used to compare the staging of AKI stratified by timing (i.e., early vs. late) by HTN group. A generalized estimating equation (GEE) logistic regression was used to account for variations between centers and determine odds ratios (OR) and 95% confidence intervals (95% CIs) for the association between maternal HTN disorders and AKI. A logistic regression with a backward selection process was conducted to adjust for potential confounders. We included all potential covariables from Table 1 and removed them in a stepwise fashion. A variable had to have p-value of <0.2 to remain in the final adjusted model. Ultimately, the final adjusted analysis included the following variables: site of delivery, GA, multiple gestation, and Apgar score at 5 minutes. The remaining covariables were not included given that the p-value was >0.2.

Table 1.

Neonatal and maternal characteristics by maternal HTN type

Demographics	No HTN (n=1,710)	Any HTN (n=452)	p-Valuea	Chronic HTN (n=121)	Pre/eclampsia (n=261)	Both (n=70)	p-Valuea
Male (%)	982 (57.4)	247 (54.6)	0.56	67 (55.4)	140 (53.6)	40 (57.1)	0.89
Ethnicity (%)
Hispanic	223 (13.0)	70 (15.5)	<0.01	10 (8.3)	49 (18.8)	11 (15.7)	<0.01
Non-Hispanic	1,192 (69.7)	333 (73.7)		100 (82.6)	178 (68.2)	55 (78.6)
Unknown	295 (17.3)	49 (10.8)		11 (9.1)	34 (13.0)	4 (5.7)
Race (%)
White	975 (57.0)	237 (52.4)	<0.01	64 (52.9)	141 (54.0)	32 (45.7)	<0.01
Black	281 (16.4)	132 (29.2)		43 (35.5)	63 (24.1)	26 (37.1)
Other	454 (26.6)	83 (18.4)		14 (11.6)	57 (21.8)	12 (17.1)
Gestational age (%)
< 29 weeks	224 (13.1)	52 (11.5)	<0.01	19 (15.7)	24 (9.2)	9 (12.9)	<0.01
29–36 weeks	676 (39.5)	282 (62.4)		57 (47.1)	178 (68.2)	47 (67.1)
36+ weeks	810 (47.4)	118 (26.1)		45 (37.2)	59 (22.6)	14 (20.0)
Gestational Age (weeks)	34.4 ± 4.6	33.0 ± 3.8	<0.01	33.3 ± 4.4	33.0 ± 3.4	32.3 ± 3.8	<0.01
Fetal growth restriction	128 (7.49)	71 (15.7)	<0.01	17 (14.05)	46 (17.62)	8 (11.43)	<0.01
Birth weight (g)	2382 ± 985	1,986 ± 969	<0.01	2,188 ± 1,109	1,955 ± 913	1,784 ± 822	<0.01
Site of delivery (%)
Inborn	922 (53.9)	342 (75.7)	<0.01	91 (75.2)	194 (74.3)	57 (81.4)	<0.01
Outborn	788 (46.1)	110 (24.3)		30 (24.8)	67 (25.7)	13 (18.6)
Resuscitation (%)
PPV	719 (42.1)	234 (51.8)	<0.01	60 (49.6)	137 (52.5)	37 (52.9)	<0.01
Intubation	441 (25.8)	108 (23.9)	0.41	33 (27.3)	59 (22.6)	16 (22.9)	0.64
Chest compression	80 (4.7)	8 (1.8)	0.01	2 (1.7)	4 (1.5)	2 (2.9)	0.46
Apgar score at 5 minutes	7.6 ± 2.0	7.7 ± 1.7	0.61	7.8 ± 1.5	7.6 ± 1.9	7.7 ± 1.7	0.77
Maternal age at delivery (y)	28.3 ± 6.0	29.7 ± 6.5	<0.01	31.0 ± 6.0	28.9 ± 6.6	30.7 ± 6.2	<0.01
Diabetes	181 (10.6)	114 (25.2)	<0.01	43 (35.5)	51 (19.5)	20 (28.6)	<0.01
Kidney disease	10 (0.6)	9 (2.0)	<0.01	5 (4.1)	2 (0.8)	2 (2.9)	<0.01
Delivery type
C-section, scheduled	221 (12.9)	62 (13.7)	<0.01	20 (16.5)	30 (11.5)	12 (17.1)	<0.01
C-section, unscheduled	624 (36.5)	252 (55.8)		60 (49.6)	146 (55.9)	46 (65.7)
Unknown	91 (5.3)	27 (6.0)		9 (7.4)	15 (5.8)	3 (4.3)
Vaginal birth	774 (45.3)	111 (24.6)		32 (26.5)	70 (26.8)	9 (12.9)
Intrapartum Infection	166 (9.7)	35 (7.7)	0.20	13 (10.7)	19 (7.3)	3 (4.3)	0.26
Amniotic fluid volume
Polyhydramnios	66 (3.9)	14 (3.1)	0.51	10 (8.3)	3 (1.2)	1 (1.4)	0.03
Oligohydramnios	77 (4.5)	25 (5.5)		6 (5.0)	15 (5.8)	4 (5.7)
Normal	1,567 (91.6)	413 (91.4)		105 (86.8)	243 (93.1)	65 (92.9)
Vaginal bleeding	81 (4.7)	10 (2.2)	0.02	3 (2.5)	6 (2.3)	1 (1.4)	0.12
Hemorrhage	57 (3.3)	9 (2.0)	0.14	4 (3.3)	5 (1.9)	0 (0.0)	0.28
Multiple gestation	303 (17.7)	74 (16.4)	0.50	26 (21.5)	45 (17.4)	3 (4.3)	<0.01
Assisted conception	120 (7.0)	38 (8.4)	<0.01	19 (15.7)	18 (6.9)	1 (1.4)	<0.01
Maternal steroids	533 (31.2)	226 (50.0)	<0.01	42 (34.7)	137 (52.5)	47 (67.1)	<0.01
NSAIDs	52 (3.0)	12 (2.7)	0.67	9 (7.4)	3 (1.1)	0 (0.0)	<0.01
Antihypertensives	50 (1.2)	179 (39.6)	<0.01	48 (39.7)	90 (34.5)	41 (58.6)	<0.01

Abbreviations: HTN, hypertension; NSAID, nonsteroidal anti-inflammatory drugs; PPV, positive pressure ventilation.

Note: Values are n (%) except plus/minus values represent mean ± standard deviation.

^ap-Values based on Chi-square and analysis of variance for categorical and continuous variables, respectively.

A stratified analysis of site of delivery (inborn vs. outborn) was conducted to determine the interaction with maternal HTN exposure and neonatal AKI. GA categories were defined as <29, 29 to 36, and 36+ weeks as in prior AWAKEN publications. FGR was specified if noted in the neonatal and/or maternal chart. For analyses of AKI events after the first week of life, the analytical population was limited to those who had stayed in the hospital at least 7 days and did not have AKI in the first week of life.

Results

Prevalence of Maternal Hypertension Disorders

The number of neonates included in this secondary analysis was 2,162 (Fig. 1). Table 1 shows neonatal and maternal characteristics and demographic differences by maternal HTN category. Overall, 21% (452/2,162) of the neonates were exposed to a maternal HTN disorder compared with 79% (1,710/2,162) who were not. Of these, 27% (121/452) were exposed to chronic HTN; 58% (261/452) to pre/eclampsia; and 15% (70/452) to both HTN conditions (pre/eclampsia superimposed on chronic HTN).

Fig. 1.

Flow diagram of enrollment of patients into the AWAKEN study. Because exclusion criteria are not mutually exclusive, some potential participants could have been excluded for multiple reasons and are counted in each exclusion category. AKI, acute kidney injury; AWAKEN, the Assessment of Worldwide Acute Kidney Epidemiology in Neonates; Cr, creatinine; IVF, in vitro fertilization; NICU, neonatal intensive care unit; UOP, urine output.

Neonatal Characteristics

The majority of infants exposed to maternal HTN were white of non-Hispanic ethnicity (Table 1). Black neonates were more likely to be exposed to any maternal HTN compared with White or “Other” neonates (32% (132/413) versus 20% (237/ 1212) and 15% (83/537), respectively (Table 1). Most neonates exposed to maternal HTN were born preterm, in the GA category of 29 to 36 weeks (Table 1). Neonates born to mothers with both maternal HTN conditions had the lowest GA(32.3 ± 3.8 weeks). FGR was more prevalent in the maternal HTN exposure categories. Birth weight was lower in neonates exposed to any maternal HTN category, with the lowest birth weight in the neonates exposed to both maternal HTN conditions (1,784 ± 822 g) compared with the neither category (2,382 ± 985 g). Neonates exposed to any maternal HTN were more likely to be inborn compared with neonates not exposed to maternal HTN (i.e., neither group; 342/452 [76%] vs. 922/ 1,710 [54%], respectively [p < 0.01]).

Maternal Characteristics and Intrapartum Complications

Mothers with any HTN were more likely to be older in age, have diabetes, and/or kidney disease than those without HTN (Table 1). Unscheduled cesarean (C-section) was the primary mode of delivery for neonates born exposed to any maternal HTN category. Multiple gestation was lowest in the both maternal HTN category. There was more maternal steroid exposure in all maternal HTN categories compared with the neither group and was the greatest among mothers with pre/eclampsia (52.5%) and both HTN disorders (61.1%). Maternal antihypertensive use among all maternal HTN categories was 179/452 (39.6%). Of note, 50 (2.9%) mothers without HTN (i.e., neither category) were also reported to be receiving antihypertensive medications.

Timing and Severity of Neonatal Acute Kidney Injury by Maternal Hypertension Type

Neonates exposed to any maternal HTN had a lower incidence of overall and early AKI compared with those not exposed (Supplemental Table S2 (available in the online version). Maternal HTN exposure largely did not alter the distribution of severity of AKI. Stage-1 AKI was the most frequent in all the maternal HTN groups. The frequency of late AKI was less comparing no HTN to any HTN but not significantly different after subdividing by maternal HTN category.

Maternal Hypertension Medication Use and Neonatal Acute Kidney Injury

Overall, antihypertensive medication use was reported in 179/452 (39.6%) mothers. Of note, 50 (2.9%) mothers in the neither category were also reported to be receiving antihypertensive medication which could have reflected use for an alternate indication such as tocolysis (Table 1). Use of maternal antihypertensive medication was not associated significantly with neonatal AKI in either the crude or adjusted models (Supplemental Table S3; available in the online version).

Maternal Hypertension and Neonatal Acute Kidney Injury-Stratified Analysis by Site of Delivery

The ORs and associated 95% CIs for the association between maternal HTN and the timing of AKI are shown in Table 2. Any exposure to maternal HTN was associated with a lower odds of overall, early, and late AKI in the unadjusted analysis. However, the adjusted OR, after consideration of multiple confounders including site of delivery and GA, differed from the crude OR by >25%, in favor of AKI (Table 2. Hence, a stratified analysis by site of delivery was conducted and demonstrated a significant difference between the inborn and outborn settings with maternal HTN exposure being associated with a lower odds of AKI only in the inborn group (Table 3).

Table 2.

ORs and associated 95% CIs for the association between maternal HTN exposure and timing of neonatal AKI

	Crude ORa (95% CI)	Adjusted ORa,b (95% CI) including site of delivery	Adjusted ORa,b (95% CI) excluding site of delivery
Overall AKI (n = 605)
Any HTN	0.65 (0.49–0.87)	0.87 (0.63–1.21)	0.77 (0.56–1.05)
No HTN	Referent	Referent	Referent
Chronic HTN	0.64 (0.40–1.02)	0.80 (0.50–1.27)	0.69 (0.44–1.09)
Pre/eclampsia	0.70 (0.49–0.99)	0.98 (0.68–1.39)	0.87 (0.61–1.23)
Both	0.50 (0.26–0.93)	0.67 (0.36–1.24)	0.57 (0.29–1.11)
Early AKI (n = 456)
Any HTN	0.68 (0.51–0.89)	0.93 (0.69–1.25)	0.81 (0.60–1.08)
No HTN	Referent	Referent	Referent
Chronic HTN	0.68 (0.40–1.16)	0.89 (0.50–1.57)	0.76 (0.44–1.30)
Pre/eclampsia	0.76 (0.53–1.07)	1.07 (0.77–1.47)	0.94 (0.67–1.32)
Both	0.39 (0.19–0.81)	0.55 (0.28–1.09)	0.46 (0.22–0.98)
Late AKI (n = 149)
Any HTN	0.61 (0.3–0.97)	0.77 (0.46–1.30)	0.75 (0.48–1.16)
No HTN	Referent	Referent	Referent
Chronic HTN	0.59 (0.27–1.32)	0.64 (0.29–1.41)	0.59 (0.27–1.31)
Pre/eclampsia	0.58 (0.31–1.11)	0.81 (0.41–1.61)	0.58 (0.31–1.11)
Both	0.75 (0.32–1.78)	0.89 (0.38–2.08)	0.75 (0.32–1.78)

Abbreviations: AKI, acute kidney injury; CI, confidence interval; HTN, hypertension; OR, odds ratio; Pre, preeclampsia.

Note: Late AKI excludes those who had both early+late AKI.

^aBased on generalized estimating equation (GEE) logistic regression.

^bAdjusted for gestational age, multiple gestation, and 5-minute Apgar’s score.

Table 3.

Stratified analysis by site of delivery for the association exposure to maternal HTN and timing of neonatal AKI

	Crude ORa (95% CI)	Adjusted ORa,b (95% CI)
Outborn
Overall AKI (n = 349)
Any HTN	1.17 (0.75–1.83)	1.30 (0.82–2.08)
No HTN	Referent	Referent
Chronic HTN	1.97 (0.93–4.16)	2.25 (1.09–4.64)
Pre/eclampsia	1.01 (0.55–1.87)	1.15 (0.63–2.08)
Both	0.74 (0.23–2.39)	0.72 (0.23–2.29)
Early AKI (n = 280)
Any HTN	1.10 (0.73–1.65)	1.21 (0.80–1.83)
No HTN	Referent	Referent
Chronic HTN	1.80 (0.76–4.27)	1.92 (0.79–4.66)
Pre/eclampsia	0.88 (0.57–1.34)	0.99 (0.65–1.51)
Both	1.04 (0.34–3.15)	1.08 (0.37–3.17)
Late AKI (n = 69)
Any HTN	1.42 (0.76–2.65)	1.59 (0.77–3.26)
Inborn
Overall AKI (n = 256)
Any HTN	0.62 (0.40–0.95)	0.68 (0.44–1.07)
No HTN	Referent	Referent
Chronic HTN	0.42 (0.19–0.93)	0.42 (0.19–0.90)
Pre/eclampsia	0.73 (0.44–1.22)	0.87 (0.53–1.44)
Both	0.57 (0.31–1.08)	0.62 (0.31–1.24)
Early AKI (n = 176)
Any HTN	0.70 (0.45–1.08)	0.77 (0.50–1.19)
No HTN	Referent	Referent
Chronic HTN	0.46 (0.18–1.16)	0.48 (0.19–1.18)
Pre/eclampsia	0.94 (0.58–1.53)	1.08 (0.68–1.73)
Both	0.32 (0.12–0.83)	0.34 (0.13–0.91)
Late AKI (n = 80)
Any HTN	0.48 (0.24–0.97)	0.55 (0.27–1.09)
No HTN	Referent	Referent
Chronic HTN	0.39 (0.11–1.31)	0.38 (0.11–1.30)
Pre/eclampsia	0.36 (0.16–0.83)	0.56 (0.25–1.23)
Both	1.03 (0.43–2.43)	1.30 (0.48–3.50)

Abbreviations: AKI, acute kidney injury; CI, confidence interval; HTN, hypertension; OR, odds ratio; Pre, preeclampsia.

Note: Late AKI excludes those who had both early+late AKI.

^aBased on generalized estimating equation (GEE) logistic regression.

^bAdjusted for gestational age, multiple gestation, and 5-minute Apgar’s score.

In the stratified analysis for site of delivery, for those neonates outborn, chronic HTN was associated with a greater odds of overall AKI in the adjusted model OR of 2.25 (95% CI: 1.09–4.64; Table 3). In contrast, in the inborn neonates, chronic HTN was associated with a lower odds of overall AKI in the adjusted model OR of 0.42 (95% CI: 0.19–0.90). Neonates exposed to both maternal HTN disorders demonstrated the lowest odds of early AKI with an OR of 0.34 (95% CI: 0.13–0.91) in the adjusted model. Late AKI was not associated with maternal HTN in the adjusted model in either the inborn or outborn setting.

Association between Maternal Hypertension Disorders and Other Key Covariates and Early Acute Kidney Injury

The relative impact of maternal HTN exposure and category on early AKI was analyzed in comparison to other key neonatal and maternal variables (Supplemental Table S4; available in the online version; Fig 2). Factors favoring higher rates of early AKI were outborn delivery and GA <29 or >36 weeks. Factors favoring lower rates of early AKI were maternal HTN disorders (in particular the both maternal HTN category), inborn delivery, GA of 29 to 36 weeks, multiple gestation, and Apgar score of >7 at 5 minutes.

Fig. 2.

Forest plot showing maternal HTN by category alongside other variables associated with early AKI. Crude odds ratios (OR) are in red and adjusted OR are in blue. AKI, acute kidney injury; FGR, fetal growth restriction; GA, gestational age; HTN, hypertension.

Discussion

This secondary analysis of the AWAKEN cohort represents the largest study to date to examine the association between maternal HTN and neonatal AKI. Here, we show that exposure to maternal HTN was associated with a tendency toward a lower likelihood of AKI for inborn neonates. In addition, exposure to maternal pre/eclampsia superimposed on chronic HTN was associated with the lowest likelihood of early neonatal AKI in the inborn setting.

Large cohort studies aiming to evaluate the effect of maternal HTN exposure on neonatal outcomes have largely centered on infant mortality, various illness severity scores, and specific prevalent preterm conditions such as retinopathy and bronchopulmonary dysplasia.^15,20,34 Few of these studies report any neonatal outcomes specific to the kidney. One study used reduced UOP as part of their neonatal illness severity score (the Score of Neonatal Acute Physiology [SNAP]) but none have incorporated SCr and/or AKI specifically as part of the scoring system.¹⁷ Several retrospective studies aiming to define the risk factors associated with neonatal AKI in preterm and very low birth weight infants have documented that maternal HTN is associated with a reduced risk of neonatal AKI.^24–28 These studies, however, differed in the study population selection and looked at the most vulnerable preterm infant categories unlike the AWAKEN cohort that included high risk neonates across the full spectrum of GA categories. In addition, most of these studies have found a reduced incidence of AKI associated with preeclampsia exposure, with fewer showing this reduction with chronic HTN and none included the superimposed (i.e., both) or neither category for comparison. This study highlights the importance of not only categorizing maternal HTN exposure when evaluating for associations with neonatal kidney outcomes but also exploring the interaction between covariates, such as delivery setting and GA category, in interpreting these associations

Site of delivery has been noted in several studies to impact neonatal outcomes with outborn status being associated with sicker infants.^35,36 In the AWAKEN cohort, both primary and secondary analyses have noted that outborn delivery was associated with a higher likelihood of neonatal AKI.^29,30,32 This finding is postulated to be related to (1) improved access to perinatal and surgical services specializing in maternal fetal medicine that manage high-risk pregnancies such as maternal HTN, and (2) immediate evaluation and treatment in the NICU for vulnerable neonates. In this study, the majority of neonates exposed to maternal HTN were inborn. Stratifying for site of delivery, helped to overcome this bias in the study and look more granularly at the association between maternal HTN exposure and neonatal AKI.

From the AWAKEN cohort and other studies, it has been shown that neonatal AKI affects GA groups differently.^29–31 Most preeclampsia and/or eclampsia pregnancies deliver at late preterm, 34 to 36 weeks’ GA with an increased incidence of FGR.^1,5 This was noted to be true in the AWAKEN cohort as well. However, even in preterm neonates <29 weeks’ GA, less perinatal morbidity has been reported with exposure to maternal HTN.^15,16 In the AWAKEN cohort, the reduced odds of early AKI with any maternal HTN exposure was most prominent in the 22–to 28-week GA group,³⁰ while for late AKI it was noted in the ≥36-week GA group.³¹ In stratifying the analysis by timing of AKI, maternal HTN was associated with a reduced odds of early but not late AKI. Mechanistically, this perhaps can be explained in that the direct and immediate effect of maternal HTN exposure, and its management is most relevant to the fetus and neonate within the pre- and perinatal periods.

The trend toward a reduced odds of neonatal AKI with maternal HTN exposure may indicate that maternal HTN directly or indirectly, via another mediator, lowers the likelihood of neonatal AKI. For example, maternal antihypertensive medication use, and hence control of maternal HTN, could explain the reduction in neonatal AKI seen. Other studies looking at perinatal outcomes in mother’s adherent to antihypertensive therapies have shown mixed results on improved or worsening outcomes for the neonate.³⁷ In this study, maternal antihypertensive medication use was not significantly associated with neonatal AKI. In addition, a small number of mothers who were not classified with HTN (2.9%) were reported to be on antihypertensive medication. This could have been related to use of certain medications such as calcium channel blockers for tocolysis; however, the indication for antihypertensive use was not recorded.³⁸ Antenatal corticosteroids have been shown to be used more widely in maternal HTN disorders and are independently associated with reduced neonatal mortality.^15,18,39 In this study, we also found a higher rate of antenatal steroid use in maternal HTN disorders. Interestingly, in regard to the interaction between site of delivery and improved neonatal outcomes, steroid use was found to account for why inborn preterm neonates had less intraventricular hemorrhage than outborn neonates.³⁶ It is possible that improved uteroplacental perfusion in hypertensive pregnancies^15,18 could lead to improved transfer of nutrients and oxygen to the fetus which may result in improved kidney function.¹⁸ Pre/eclampsia could also lead to intermittent fetoplacental ischemic preconditioning^35,36 that could “prepare” the kidney to sustain an ischemic event around the time of birth. Finally, it is possible that AKI is less common in those who have high risk pregnancies because delivery is often planned to avoid risk to both mother and baby. This is in contrast to a preterm infant delivered precipitously who requires NICU support for other critical neonatal-specific conditions.¹⁸ Understanding the mechanism-driving protective factors for neonatal AKI is beyond the scope of this study; however, it is an important area of future research.

On the most extreme end of this spectrum is pre/eclampsia superimposed on chronic HTN which occurs in 40% of women with chronic HTN and is associated with the most maternal and neonatal mortality.^1,2 Interestingly, in this study, neonates with exposure to both maternal HTN disorders had the lowest odds of overall and early AKI in the univariate and multivariate models. Close prenatal care and expectant management for those mother–infant dyads without severe features (i.e., end-organ damage) up to 37 weeks’ gestation and immediate delivery of those fetuses with any severe features after 34 weeks’ gestation may be related to improved renal outcomes.^34,36 Weintraub et al noted that both maternal preeclampsia and/or use of antenatal magnesium sulfate for preeclampsia or neuroprotection was related to a reduced incidence of AKI in a group of <30-week preterm neonates.²⁸ Neonatal medications, such as caffeine, a methylxanthine, have been shown to reduce the risk of AKI in AWAKEN and other cohorts.⁴⁰ These therapies could also be implicated in the reduction of early AKI seen in the maternal HTN group and needs to be further explored. Neither magnesium therapy nor indication for delivery were collected in the AWAKEN study.

Limitations

Despite the significant strengths of this analysis, we acknowledge, several limitations. The determination of maternal HTN category was limited to reviewing neonatal and maternal charts and was dependent on accuracy of transferred information by each center which could have led to misclassification bias. Ascertainment bias from those outborn could have resulted in an underreporting of maternal HTN given limited access to maternal records. Also, the inclusion of pregnancy-induced HTN within the pre/eclampsia category in the original data collection limited the ability to differentiate a less-severe maternal HTN phenotype (i.e., pregnancy induced/gestational HTN) from pre/eclampsia. Magnesium sulfate exposure use was not recorded as a maternal antihypertensive or other medication used; hence, we could not associate this with neonatal AKI in the current analysis. Finally, significance was lost for certain associations in the fully adjusted model, even after stratification, highlighting the fact that there are likely multiple mediators at play.

In addition, despite the apparent trend toward a reduced odds of neonatal AKI in the short term, the risk of maternal HTN exposure on perinatal programming of chronic kidney and cardiovascular disease in preterm born individuals, with or without neonatal AKI, requires further study as this population has shown long-term sequelae of exposure to adversity in the pre- and perinatal periods.^11,40–45 For example, in the Generation R study, a large prospective Norwegian study investigating the end-organ effects of an adverse intrauterine environment, fetal blood flow redistribution away from the kidneys and to the brain correlated with decreased fetal kidney volume.⁴⁶ The follow-up to this study showed that differences in kidney volume seen in fetal life persisted at 6 years of age with lower estimated glomerular filtration rates in the lower tertiles.⁴⁷

Conclusion

In conclusion, we have shown that exposure to maternal HTN, especially pre/eclampsia superimposed on chronic HTN, was associated with a tendency toward a reduced odds of early neonatal AKI in the inborn delivery setting. Distinguishing the impact of confounders versus effect modifiers related to the association between maternal HTN and neonatal AKI was beyond the scope of this study. Future prospective studies should investigate the role of specific perinatal practices employed in the care of maternal HTN disorders that might alter placental and, hence, fetal renal perfusion and prevent neonatal AKI. In addition, formalizing guidance on timing of transfer to a center with maternal-fetal medicine and NICU services for the high-risk maternal-infant dyads is important to avert adverse outcomes including neonatal AKI. Separately, the long-term consequences of maternal HTN exposure on neonatal kidney development and the risk of chronic kidney and cardiovascular disease needs to be considered.

Key Points.

• Maternal HTN is associated with less neonatal AKI.

• Maternal HTN category is variably associated with AKI.

• Inborn status is an important contributor to this association.