Early cerebral hemodynamic assessment using resistive index and its correlation with neonatal outcomes

Abstract

Background

Cerebral blood flow (CBF) is critical for neonatal brain health, and disruption of autoregulation may predispose infants to brain injury. The resistive index (RI) of the anterior cerebral artery (ACA), measured via Doppler ultrasonography, is a widely used non-invasive marker of CBF, but its relationship with vital parameters in critically ill neonates remains uncertain.

Methods

This retrospective study at Taipei Veterans General Hospital from January 2022 to December 2024 included neonates admitted to the NICU with ACA Doppler ultrasonography within 24 h of birth. RI values were stratified into three groups: low (<0.6), normal (0.6–0.8), and high (>0.8). Clinical characteristics, resuscitation needs, Apgar scores, blood gas analyses, and short- and long-term outcomes were compared across groups.

Results

Among 898 admissions, 386 neonates met the inclusion criteria (mean gestational age 35 ± 4 weeks, mean birth weight 2382 ± 847 g), abnormal RI was identified in 25.1% (low RI: 4.1%, high RI: 21.0%). Abnormal RI correlated with significantly lower pH, higher pCO₂, lower 5-min Apgar scores, and greater need for delivery room intubation (p < 0.01). After adjusting for gestational age, birth weight, and evaluation timing, low RI remained independently associated with increased risks of systemic hypotension and fluid resuscitation (p < 0.05).

Conclusion

Abnormal ACA RI within 24 h of birth is associated with cerebral hemodynamic instability in critically ill neonates. Early RI assessment may provide insights into neonatal hemodynamic status and serve as a supplemental clinical indicator for identifying high-risk infants in the NICU.

Keywords

neonate cerebral blood flow resistive index (RI)anterior cerebral artery neonatal outcomes

Introduction

Cerebral blood flow (CBF) plays a pivotal role in maintaining neonatal brain health, particularly during the vulnerable perinatal and early postnatal periods. Disruptions in cerebral circulation and autoregulation can arise from various pathological conditions, including prematurity, intraventricular hemorrhage (IVH), patent ductus arteriosus (PDA), shock, and hypoxic-ischemic encephalopathy (HIE), and are associated with adverse neurodevelopmental outcomes. These conditions can alter cerebrovascular resistance and compromise oxygen and nutrient delivery to the developing brain.¹

Several non-invasive techniques have been developed to assess CBF, with Doppler ultrasonography being widely utilized in neonatal intensive care. The resistive index (RI), calculated as RI = (peak systolic velocity – end-diastolic velocity)/peak systolic velocity, is one of the most frequently adopted Doppler-derived parameters.² While the normal RI range varies among studies, the present work employs the Johns Hopkins reference range of 0.6–0.8.³ Anterior cerebral artery (ACA) is a preferred vessel for Doppler assessment due to its accessibility via the anterior fontanelle and its sensitivity to systemic circulatory changes. Despite extensive research, the relationship between RI and various clinical parameters, such as Apgar scores and blood gas analysis, remains inconclusive. CBF is thought to be highly related to cerebral metabolism and arterial blood gas levels.⁴ Some studies suggest that alterations in pCO₂ play a key role in determining CBF velocity, raising concerns about its potential contribution to neonatal brain injury.¹ However, previous studies have reported no significant association between RI and 1- or 5-min Apgar scores, umbilical cord pH, or base excess levels.³

Given these uncertainties, it remains unclear how early RI measurements relate to vital perinatal parameters and subsequent morbidity in critically ill neonates. Clarifying these associations could enhance early risk stratification and inform individualized management strategies in the NICU. Therefore, the objective of this study was to evaluate the correlation between ACA RI measured within the first 24 h of life and clinical vital parameters, perinatal conditions, and outcomes in neonates requiring intensive care.

Methods

Study design and setting

This was a single-center, retrospective cohort study conducted at Taipei Veterans General Hospital from January 2022 to December 2024. Eligible participants were neonates admitted to the neonatal intensive care unit (NICU) who underwent ACA Doppler ultrasonography within the first 24 h after birth. Neonates were excluded based on the following criteria: (1) admission to the NICU beyond 24 h after birth; (2) initial ACA Doppler ultrasonography performed more than 24 h after birth (e.g., due to logistical delays during weekends or other administrative factors); (3) ACA RI measurements were not performed or not documented in the specified time frame.

Patient population and clinical management

Admission to the NICU was determined by neonatologists based on clinical indications, including but not limited to respiratory distress, meconium aspiration syndrome (MAS), prematurity or very low birth weight, and other perinatal complications. A majority of patients required respiratory support during hospitalization.

Ultrasound measurement protocol

Routine cranial ultrasonography, including ACA Doppler assessment, was performed for all admitted neonates. Doppler measurements were acquired using a Philips CX50 vascular ultrasound system with a C8-5 curved array transducer (4.4 MHz). Imaging was performed via the anterior fontanelle in the sagittal plane, visualizing the ACA anterior to the corpus callosum. Cranial Doppler examinations were performed by neonatal fellows and subsequently reviewed by attending neonatologists for quality assurance; representative images are shown in Figure 1. RI values were categorized as follows:

• Normal RI (Group N): 0.60–0.80

• Low RI (Group L): <0.60

• High RI (Group H): >0.80

Figure 1.

Representative cerebral Doppler ultrasonography images of the ACA. The Doppler waveforms and RI measurements shown here are representative examples obtained at our institution. The images illustrate the hemodynamic patterns of the three study groups: (a) Low RI group (RI = 0.585); (b) Normal RI group (RI = 0.702); and (c) High RI group (RI = 0.814). Each panel displays the color Doppler flow signal (top) and the corresponding pulse-wave Doppler spectral waveform (bottom) used for clinical assessment. Abbreviations: ACA: anterior cerebral artery; EDV: end-diastolic velocity; PSV: peak systolic velocity; RI: resistive index.

Data collection

Demographic and perinatal data included gestational age, birth weight, sex, delivery mode, use of positive pressure ventilation (PPV), and endotracheal intubation during neonatal resuscitation in the delivery room. Clinical variables collected included 1- and 5-minute Apgar scores and blood gas parameters obtained within the first hour after birth (pH, pCO₂, HCO₃^-, base excess).

Outcome measures

Primary short-term outcomes were interventions required within 7 days of life, including respiratory support, fluid resuscitation, inotropic support, and therapeutic hypothermia. Secondary outcomes included incidence and severity of intraventricular hemorrhage (IVH, graded by Papile classification), severe IVH (grade 3 or 4), bronchopulmonary dysplasia (BPD, per 2019 Jensen criteria), PDA treatment status, NICU and total hospital length of stay, and in-hospital mortality. Subgroup analyses compared preterm (<37 weeks gestation) versus term (≥37 weeks) neonates.

Statistical analysis

Data analyses were performed using Microsoft Excel and SPSS version 25.0. Continuous variables were analyzed using one-way ANOVA with Tukey post-hoc testing, or Kruskal–Wallis test for non-normally distributed data. Multivariate logistic regression analysis was performed to evaluate the independent association between RI groups and clinical outcomes, adjusting for variables that showed significant baseline differences or were clinically relevant. Categorical variables were compared using the Chi-square test. A two-tailed p value <0.05 was considered statistically significant.

Results

Patient characteristic

Between January 2022 and December 2024, 898 patients were admitted to the NICU. Among them, 512 were excluded because no ACA RI Doppler examination records were available within the first 24 h of life Figure 2. The clinical characteristics of the excluded infants are detailed in Table S1. Overall, 386 neonates entered the study, with a mean gestational age of 35 ± 4 weeks and mean birth weight of 2382 ± 847 g. Based on anterior cerebral artery (ACA) resistive index (RI), 289 infants (74.9%) were classified as normal (RI 0.6–0.8), 16 (4.1%) as low RI (<0.6), and 81 (21.0%) as high RI (>0.8). Baseline characteristics across groups are summarized in Table 1. Neonates with low RI had significantly lower gestational age (32.1 ± 6.7 weeks) and birth weight (1872 ± 1121 g) compared with the normal and high RI groups (p < 0.05). The timing of first evaluation was later in the low RI group (10.4 ± 6.8 h vs 7.7 ± 7.1 h and 4.1 ± 4.7 h, p < 0.001). Intubation during delivery room resuscitation was more frequent in the low RI group (37.5%) compared to the other groups (6.6% and 7.4%, p < 0.001). Significant differences were also observed in 5-min Apgar scores, arterial pH, pCO₂, and base excess among the three groups (all p < 0.05).

Figure 2.

Flowchart of patient selection. A total of 898 neonates were admitted to the NICU during the study period. Of these, 512 were excluded. Reasons for exclusion included: age at admission >24 h (n = 199), ACA RI performed ≥24 h after birth (n = 32), and ACA RI not performed (n = 281). The remaining 386 neonates were eligible and included in the final analysis. Abbreviations: ACA: anterior cerebral artery; NICU: neonatal intensive care unit; RI: resistive index.

Table 1.

Baseline characteristics of neonates according to RI classification.

	Low RI (<0.6) n = 16	Normal RI (0.6–0.8), n = 289	High RI (>0.8), n = 81	p value
GA (weeks)	32.1 ± 6.7	35.6 ± 4.0	35.7 ± 3.7	0.005
BW (g)	1872 ± 1121	2386 ± 825	2469 ± 847	0.035
Female, n (%)	8 (50.0)	125 (43.3)	43 (53.1)	0.273
CS, n (%)	8 (50.0)	160 (55.4)	54 (66.7)	0.158
Age at first evaluation (h)	10.4 ± 6.8	7.7 ± 7.1	4.1 ± 4.7	<0.001
PPV in delivery room, n (%)	11 (68.8)	112 (38.8)	31 (38.3)	0.055
Intubation in delivery room, n (%)	6 (37.5)	19 (6.6)	6 (7.4)	<0.001
1-min Apgar	6 [3–7]	7 [5.5–7.5]	7 [5–8]	0.116
5-min Apgar	7.5 [5–8]	8 [8–9]	8 [7–9]	0.012
pH	7.18 ± 0.11	7.24 ± 0.07	7.21 ± 0.09	<0.001
pCO₂ (mmHg)	59.1 ± 16.6	51.6 ± 9.5	53.9 ± 11.1	0.005
HCO₃^- (mmol/L)	21.0 ± 3.2	21.6 ± 4.1	20.9 ± 3.0	0.250
Base excess (mmol/L)	−8.1 ± 4.4	−6.3 ± 3.4	−7.5 ± 3.5	0.004
RI	0.55 ± 0.04	0.71 ± 0.05	0.84 ± 0.04	<0.001

Abbreviations: BW: birth weight; CS: caesarean section; GA: gestational age; PPV: positive pressure ventilation; RI: resistive index.

Data are expressed as mean ± SD, median [interquartile range], or frequency (percent).

Preterm versus term subgroup analysis

When stratified by gestational age (Table 2), preterm infants with low RI demonstrated markedly poorer clinical profiles, including lower gestational age, reduced birth weight, higher rates of delivery room intubation (60.0%), and worse blood gas parameters compared to those with normal or high RI (all p < 0.01). In contrast, among term infants, differences in Apgar scores and respiratory interventions were less pronounced, although significant differences in pH and pCO₂ persisted between RI groups.

Table 2.

Patients’ characteristics by gestational age (preterm vs full-term).

Group	GA (weeks)	BW (g)	Intubation in DR, n (%)	5-min Apgar	pH	pCO₂ (mmHg)	Base excess (mmol/L)	RI
Preterm
Low RI (n = 10)	28.0 ± 4.9	1158 ± 704	6 (60.0)	6.5 [5–8]	7.17 ± 0.13	61.7 ± 20.3	−7.7 ± 4.7	0.55 ± 0.04
Normal RI (n = 149)	32.5 ± 3.4	1806 ± 618	17 (11.4)	8 [7–9]	7.25 ± 0.06	52.2 ± 9.6	−5.8 ± 2.6	0.71 ± 0.05
High RI (n = 40)	32.6 ± 2.9	1887 ± 620	5 (12.5)	8 [7–9]	7.24 ± 0.07	52.1 ± 10.7	−7.1 ± 2.8	0.84 ± 0.04
p value	<0.001	0.004	<0.001	0.037	0.037	0.037	0.007	<0.001
Full-term
Low RI (n = 6)	39.0 ± 0.8	3062 ± 400	0 (0.0)	8 [7–8]	7.18 ± 0.08	54.9 ± 7.5	−8.8 ± 4.0	0.56 ± 0.02
Normal RI (n = 140)	38.8 ± 0.7	3005 ± 509	2 (1.4)	8 [8–9]	7.23 ± 0.07	50.9 ± 9.4	−6.8 ± 4.0	0.72 ± 0.05
High RI (n = 41)	38.6 ± 1.1	3036 ± 609	1 (2.4)	8 [7.5–9]	7.19 ± 0.09	55.7 ± 11.2	−7.9 ± 4.1	0.84 ± 0.05
p value	0.524	0.920	1.000	0.292	0.005	0.020	0.166	0.166

Abbreviations: BW: birth weight; DR: delivery room; GA: gestational age; RI: resistive index.

Data are expressed as mean ± SD, median [interquartile range], or frequency (percent).

Short-term and long-term outcomes

Primary outcomes are presented in Table 3. Compared with the normal RI group, neonates with abnormal RI exhibited higher needs for endotracheal intubation (37.5% in low RI, 22.2% in high RI, vs 13.1% in normal RI; p = 0.011), hypotension management (50.0% and 22.2% vs 13.1%; p < 0.001), fluid resuscitation (43.8% and 22.2% vs 13.1%; p = 0.002), and inotropic support (37.5% and 14.8% vs 8.3%; p = 0.001). Hydrocortisone use was also significantly higher in abnormal RI groups (p = 0.004).

Table 3.

Short- and long-term outcomes by RI classification.

Outcome	Low RI, n = 16	Normal RI, n = 289	High RI, n = 81	p value
Short-term
Endotracheal tube, n (%)	6 (37.5)	38 (13.1)	18 (22.2)	0.011
Hypotension, n (%)	8 (50.0)	38 (13.1)	18 (22.2)	<0.001
Fluid resuscitation, n (%)	7 (43.8)	38 (13.1)	18 (22.2)	0.002
Inotropic support, n (%)	6 (37.5)	24 (8.3)	12 (14.8)	0.001
Hydrocortisone, n (%)	5 (31.3)	20 (6.9)	8 (9.9)	0.004
Therapeutic hypothermia, n (%)	0 (0.0)	1 (0.3)	3 (3.7)	0.055
Long-term
Any IVH, n (%)	4 (25.0)	36 (12.5)	10 (12.3)	0.341
Severe IVH (III–IV), n (%)	2 (12.4)	4 (1.4)	1 (1.2)	0.038
BPD at 36 weeks, n (%)	5 (31.3)	43 (14.9)	14 (17.3)	0.082
PDA treatment, n (%)	5 (31.3)	29 (10.0)	10 (12.3)	0.033
PDA intervention, n (%)	4 (25.0)	6 (2.1)	3 (3.7)	0.001
NICU LOS (days)	12 [4–112]	6 [3–18]	6 [3–21]	0.110
Hospital LOS (days)	21 [9–143]	12 [7–34]	12 [7–35]	0.265
Mortality, n (%)	3 (19.0)	5 (2.0)	1 (1.0)	0.006

Abbreviations: BPD: bronchopulmonary dysplasia; IVH: intraventricular hemorrhage; LOS: length of stay; NICU: neonatal intensive care unit; PDA: patent ductus arteriosus; RI: resistive index.

Data are expressed as frequency (percent) or median [interquartile range].

Regarding long-term outcomes, severe IVH (IVH grade III–IV) was more frequent in the low RI group (12.4%) compared with the normal (1.4%) and high RI (1.2%) groups (p = 0.038). Treatment for PDA was also more common among infants with abnormal RI (p = 0.033), with a notably higher rate of surgical or catheter-based intervention in the low RI group (25.0% vs 2.1% in normal RI; p = 0.001). Mortality was significantly higher in neonates with low RI (19%) compared to those with normal (2%) or high RI (1%) (p = 0.006).

Preterm and term outcome comparisons

As shown in Table 4, preterm infants with low RI experienced the most severe short-term morbidities. They had significantly higher rates of endotracheal intubation (60.0%), hypotension requiring intervention (70.0%), fluid resuscitation (60.0%), and inotropic support (60.0%) compared with preterm infants with normal or high RI (all p < 0.01). Hydrocortisone therapy was also more frequently administered in this group (40.0%, p = 0.012). Severe IVH (grade III–IV) occurred in 20.0% of preterm infants with low RI, significantly higher than those with normal or high RI (p = 0.045). PDA requiring intervention was also notably elevated (40.0% vs 4.0% and 7.5%, p = 0.001). Mortality among preterm infants with low RI reached 30%, compared with 3% in the other groups (p = 0.006).

Table 4.

Outcomes in preterm and full-term infants according to RI classification.

Primary outcomes (short-term)	Preterm				Full-term
Primary outcomes (short-term)	L (n = 10)	N (n = 149)	H (n = 40)	p value	L (n = 6)	N (n = 140)	H (n = 41)	p value
Endotracheal intubation	6 (60.0)	33 (22.1)	13 (32.5)	0.015	0 (0.0)	5 (3.6)	5 (12.2)	0.076
Hypotension^a	7 (70.0)	21 (14.1)	9 (22.5)	<0.001	1 (16.7)	17 (12.1)	9 (22.0)	0.251
Fluid resuscitation^b	6 (60.0)	21 (14.1)	9 (22.5)	0.001	1 (16.7)	17 (12.1)	9 (22.0)	0.251
Inotropic support^c	6 (60.0)	18 (12.1)	6 (15.0)	0.001	0 (0)	6 (4.3)	6 (14.6)	0.070
Hydrocortisone	4 (40.0)	13 (8.7)	3 (7.5)	0.012	1 (16.7)	7 (5.0)	5 (12.2)	0.145
Therapeutic hypothermia	0 (0)	0 (0)	0 (0)	-	0 (0)	1 (0.7)	3 (7.3)	0.107
Secondary outcomes (long-term)
All grades of IVH	4 (40.0)	33 (22.1)	9 (22.5)	0.461	0 (0.0)	3 (2.1)	1 (2.4)	1.000
Severe IVH^d	2 (20.0)	4 (2.7)	1 (2.5)	0.045	0 (0.0)	0 (0.0)	0 (0.0)	-
BPD (at 36 weeks)^e	5 (50.0)	43 (28.9)	14 (35.0)	0.058	-	-	-	-
Grade II	5 (50.0)	42 (28.2)	13 (32.5)	0.061	-	-	-	-
Grade III	0 (0.0)	0 (0.0)	1 (2.5)	0.242	-	-	-	-
PDA receiving treatment	5 (50.0)	28 (18.8)	9 (22.5)	0.064	0 (0.0)	1 (0.7)	1 (2.4)	0.435
PDA medical management^f	4 (40.0)	28 (18.8)	9 (22.5)	0.295	0 (0.0)	1 (0.7)	1 (2.4)	0.435
PDA intervention^g	4 (40.0)	6 (4.0)	3 (7.5)	0.001	0 (0.0)	0 (0.0)	0 (0.0)	-
NICU LOS	42 [12, 158]	16 [7, 34]	18 [6, 39]	0.152	6 [4, 7]	4 [2, 5]	4 [2, 6]	0.182
Hospital LOS	70 [19, 224]	33 [17, 54]	32 [13, 63]	0.328	9 [8, 12]	7 [6, 10]	7 [6, 12]	0.319
Mortality	3 (30)	5 (3)	1 (3)	0.006	0	0	0	-

Abbreviations: BPD: bronchopulmonary dysplasia; H: high; IVH: intraventricular hemorrhage; L: low; LOS: length of stay; N: normal; NICU: neonatal intensive care unit; PDA: patent ductus arteriosus; RI: resistive index.

Data are expressed as frequency (percent) or median [interquartile range].

^aSystolic blood pressure < gestational age or mean arterial pressure < gestation age – 10.

^bRequired fluid bolus >10 mL/kg.

^cInotropic support included Dopamine or Epinephrine.

^dGrade III or IV.

^eClassified according to the 2019 Jensen criteria. Grade II: non-invasive support; Grade III: mechanical ventilation. BPD definition does not apply to full-term infants.

^fMedical treatment included Acetaminophen, Ibuprofen, or Indomethacin.

^gIntervention included surgical ligation or percutaneous catheter-based closure.

In contrast, among term infants, the differences between RI groups were less striking. Although some variation in pH and pCO₂ persisted, short-term outcomes such as intubation, hypotension, and PDA treatment did not differ significantly across RI categories. No deaths occurred among term infants during the study period.

Independent predictors of clinical outcomes

To determine whether RI acts as an independent predictor beyond the confounding effects of prematurity and clinical timing, a multivariate logistic regression was performed (Table 5). After adjusting for gestational age, birth weight, and age at first RI evaluation, low RI remained significantly associated with an increased risk of hypotension (aOR 4.904; 95% CI 1.509–15.935; p = 0.008) and the need for fluid resuscitation (aOR 3.600; 95% CI 1.109–11.687; p = 0.033). However, the associations with severe IVH and mortality were no longer significant after adjustment, suggesting these outcomes are more closely linked to the degree of prematurity than to RI alone.

Non-survivor characteristics

Table 6 summarizes the clinical and laboratory profiles of nine non-surviving neonates. These infants had a mean gestational age of 24.6 ± 0.9 weeks and mean birth weight of 610 ± 133 g, reflecting extreme prematurity. Most presented with low Apgar scores (median 1-min: 3; 5-min: 5) and severe metabolic acidosis (mean pH 7.21 ± 0.07, BE –7.3 ± 3.7). RI values were markedly abnormal (mean 0.67 ± 0.12), with several infants demonstrating both high and low extremes. All non-survivors required endotracheal intubation and intensive cardiovascular support, and the majority had severe IVH (grade III–IV). Mortality occurred despite maximal interventions, underscoring the prognostic significance of abnormal cerebral hemodynamics in extremely preterm neonates.

Table 5.

Univariate and multivariable logistic regression analysis of RI groups for clinical outcomes.

Outcome	RI group	Univariate			Multivariate^a
Outcome	RI group	OR	95% CI	p value	OR	95% CI	p value
Short-term
Endotracheal intubation	Low vs normal	3.963	1.362-11.532	0.012	1.512	0.285-8.030	0.627
Endotracheal intubation	High vs normal	1.887	1.010-3.526	0.046	2.120	0.999-4.496	0.050
Hypotension	Low vs normal	6.605	2.340-18.644	<0.001	4.904	1.509-15.935	0.008
Hypotension	High vs normal	1.887	1.010-3.526	0.046	1.822	0.938-3.542	0.077
Fluid resuscitation	Low vs normal	5.137	1.807-14.608	0.002	3.600	1.109-11.687	0.033
Fluid resuscitation	High vs normal	1.887	1.010-3.526	0.046	1.825	0.941-3.541	0.075
Inotropic support	Low vs normal	6.625	2.216-19.803	0.001	3.183	0.868-11.676	0.081
Inotropic support	High vs normal	1.920	0.914-4.032	0.085	1.658	0.662-4.153	0.281
Hydrocortisone	Low vs normal	6.114	1.935-19.320	0.002	3.420	0.814-14.360	0.093
Hydrocortisone	High vs normal	1.474	0.624-3.482	0.376	2.041	0.892-4.672	0.091
Long-term
Severe IVH (III–IV)	Low vs normal	10.179	1.716-60.367	0.011	0.627	0.046-8.461	0.725
Severe IVH (III–IV)	High vs normal	0.891	0.098-8.081	0.918	5.748	0.207-159.325	0.302
BPD at 36 weeks	Low vs normal	3.503	1.094-11.213	0.035	6.420	0.036-1137.078	0.481
BPD at 36 weeks	High vs normal	1.189	0.613-2.304	0.608	2.651	0.736-9.544	0.136
PDA treatment	Low vs normal	4.075	1.324-12.548	0.014	0.649	0.103-4.097	0.645
PDA treatment	High vs normal	1.263	0.587-2.714	0.550	1.678	0.619-4.552	0.309
Mortality	Low vs normal	0.076	0.016-0.354	0.001	0.436	0.031-6.233	0.541
Mortality	High vs normal	1.408	0.162-12.229	0.756	0.050	0.001-2.453	0.131

Abbreviations: BPD: bronchopulmonary dysplasia; CI: confidence interval; IVH: intraventricular hemorrhage; OR: odds ratio; PDA: patent ductus arteriosus; RI: resistive index.

p values were calculated using univariate and multivariable logistic regression analyses.

^aAdjusted for gestational age, birth weight, and age at first RI evaluation.

Neonates treated with therapeutic hypothermia

Four term neonates with hypoxic-ischemic encephalopathy underwent therapeutic hypothermia (Table 7). These infants exhibited profoundly abnormal blood gases within the first hour of life (mean pH 7.004 ± 0.018, BE –16.5 ± 2.5) and elevated RI values (mean 0.84 ± 0.12). Following hypothermia, RI values showed partial normalization (mean 0.67 ± 0.09). Despite requiring mechanical ventilation and cardiovascular support, all four infants survived to discharge, with variable NICU and hospital lengths of stay. These findings suggest that RI may reflect cerebral hemodynamic disturbances in perinatal asphyxia and could potentially serve as a marker of treatment response during hypothermia.

Table 6.

Clinical and laboratory characteristics of non-surviving neonates (n = 9).

Case	GA (weeks)	BW (g)	Apgar 1 min	Apgar 5 min	pH	pCO₂ (mmHg)	HCO₃^- (mmol/L)	BE (mmol/L)	RI	Age at death (days)	ETT	Hypotension tx	IVH grade	PDA tx
1	23	550	2	5	7.274	52.1	23.6	−3.8	0.641	5	✓	✓	4	Medical
2	23	600	3	7	7.200	66.0	25.2	−4.3	0.455	3	✓	✓	4	Surgical
3	24	730	3	6	7.253	57.2	24.7	−3.5	0.571	314	✓	✓	2	Medical
4	25	740	3	7	7.204	60.3	23.2	−5.8	0.773	6	✓	✓	4	Surgical
5	25	809	4	6	7.241	51.7	21.7	−6.1	0.823	5	✓	✓	4	Surgical
6	23	610	2	4	7.168	45.1	16.0	−12.3	0.615	21	✓	✓	3	Medical
7	25	520	3	5	7.329	34.1	17.5	−7.4	0.763	50	✓	✓	2	Medical
8	24	560	4	5	7.075	59.5	17.0	−13.6	0.577	22	✓	✓	0	Medical
9	26	375	0	2	7.166	56.2	19.9	−9.3	0.785	4	✓	✓	2	Surgical

Summary (mean ± SD or median [IQR]): GA 24.6 ± 0.9 weeks; BW 610 ± 133 g; Apgar 1 min: 3 [2–3.5], 5 min: 5 [4.5–6.5]; pH 7.21 ± 0.07; pCO₂ 53.6 ± 9.4 mmHg; HCO₃^- 21.0 ± 3.5 mmol/L; BE –7.3 ± 3.7 mmol/L; RI 0.67 ± 0.12.

Abbreviations: BE: base excess; BW: birth weight; ETT: endotracheal tube; GA: gestational age; IVH: intraventricular hemorrhage; PDA: patent ductus arteriosus; RI: resistive index; tx: treatment.

✓ = received intervention.

Discussion

In this study, we evaluated the cerebral RI of the ACA within the first 24 h of life in a large cohort of NICU-admitted neonates and investigated its association with perinatal characteristics and clinical outcomes. We found that approximately one-quarter of neonates exhibited abnormal RI values (<0.6 or >0.8). Both low and high RI were associated with adverse vital signs, metabolic derangements, and increased risks of short- and long-term complications compared with neonates with normal RI. Recent reviews have emphasized that impairment of cerebral autoregulation places neonates at heightened risk for brain injury because both hypoperfusion and hyperperfusion can occur when the normal vasoconstrictive and vasodilatory responses to arterial blood pressure fluctuations fail.^5,6 These disturbances have been described across a range of neonatal conditions, including prematurity, IVH, HIE, and systemic hemodynamic instability.^6,7

Our results demonstrate that abnormal RI is closely linked with impaired transition at birth, as evidenced by lower Apgar scores, higher need for delivery room intubation, and acidosis with elevated pCO₂. These findings are consistent with the prior study reporting that cerebral hemodynamics are highly sensitive to perinatal hypoxia, ventilation status, and systemic circulation disturbances.⁸ The neonatal cerebral vasculature is exquisitely sensitive to pCO₂, a reactivity that further complicates hemodynamic stability in this vulnerable population.⁵ Comparative findings from other studies also support that cerebrovascular autoregulation is uniquely vulnerable during the immediate postnatal transition, when fluctuations in perfusion pressure and gas exchange can readily disrupt cerebral blood flow control.⁸

Importantly, our study demonstrates that low RI is a potent and independent predictor of early hemodynamic instability. After adjusting for gestational age, birth weight, and age at first RI evaluation, infants with low RI remained at a significantly higher risk for systemic hypotension and the requirement for fluid resuscitation. These findings support the hypothesis that low RI reflects cerebral hyperperfusion.

In preterm infants, impaired autoregulation of fragile vasculature predisposes the brain to hemorrhagic complications.⁹ The autonomic nervous system in preterm infants is inherently immature, significantly elevating the risk of low CBF during episodes of systemic hypotension.¹⁰ Consistent with previous studies that found no significant link between brain RI and IVH,¹¹ our cohort also showed that low RI may not be the sole independent driver of IVH. However, its strong association with systemic hypotension—which we found to be independent of GA—suggests that low RI serves as a crucial clinical “red flag” for an imminent breakdown in circulatory homeostasis. This systemic instability, as evidenced by the increased need for inotropic support and fluid boluses in the low RI group, likely creates the fluctuating cerebral blood flow patterns that ultimately exceed the tolerance of the fragile preterm brain.

Conversely, high RI was associated with increased cardiopulmonary support and adverse outcomes, albeit with lower mortality than the low RI group. Elevated RI may reflect increased cerebrovascular resistance due to hypoxia, hypoperfusion, or raised intracranial pressure. This pattern has been described in neonates with perinatal asphyxia and correlates with poor neurodevelopmental prognosis. The four neonates with hypoxic-ischemic encephalopathy in our study demonstrated markedly abnormal RI prior to therapeutic hypothermia, with partial normalization following treatment. These observations suggest that RI may also serve as a non-invasive biomarker of treatment response.

While a low RI (<0.6) is traditionally associated with poor prognosis due to loss of autoregulation in HIE,¹² early high RI may serve as a precursor or an alternative marker of severe injury.¹³ Our observation is consistent with the findings of Kumar et al. (2021), who demonstrated that neonates with low 5-min Apgar scores exhibit high RI in the early postnatal phase (mean 3.6 h).¹³ This phenomenon has been previously characterized as a primary insult stage, reflecting a transient period of intense vasoconstriction or reduced diastolic flow before the onset of secondary energy failure.¹³

The timing of measurement is critical for interpreting these findings, as neonatal cerebral hemodynamics are highly dynamic during the first 24 h of life. On a general population level, the High RI group in our study was evaluated at a significantly earlier time point (mean 4.1 h) than other groups. This aligns with the physiological transition of immediate postnatal life, where cerebral vascular resistance is typically at its peak during the initial adaptation to extrauterine circulation, followed by a gradual decline as systemic and pulmonary circulations stabilize.¹⁴ The clinical relevance of this temporal evolution in specific disease states is further demonstrated by the neonates in our cohort who received therapeutic hypothermia (Table 7). In these HIE cases, initial RI values within the first 24 h were notably high (mean 0.84 ± 0.12), but subsequently declined to lower levels (mean 0.67 ± 0.09) following clinical stabilization. This transition explains the disparity between our data and reports of low RI, as earlier assessments are more likely to capture the transient high-resistance phase.

Regarding the hemodynamic influence of a hemodynamically significant PDA (hsPDA), our results present an interesting contrast to established literature. Typically, hsPDA is associated with increased RI due to the “ductal steal” phenomenon, which reduces end-diastolic flow in the cerebral arterioles.^15,16 In our initial univariate analysis (Tables 3 and 4), we observed a paradoxical trend where infants requiring PDA treatment exhibited lower RI values. However, after adjusting for potential confounders—including gestational age and systemic hypotension—in the multivariable logistic regression (Table 5), this association did not reach statistical significance (OR 0.649; 95% CI 0.103–4.097; p = 0.645). This loss of significance suggests that the observed lower RI in the PDA group may not be a direct hemodynamic consequence of the ductal shunt itself, but rather a reflection of the overall clinical instability or extreme prematurity common in these infants. In the immediate transitional phase (the first 24 h of life), the “ductal steal” effect may not yet be the dominant factor influencing cerebral resistance. Instead, other factors that frequently coexist with PDA, such as systemic hypotension or a compensatory loss of cerebral autoregulation (vasodilation), might initially override the resistive effects of the shunt, leading to the observed low RI in the univariate model. Further investigation with serial Doppler and echocardiography is warranted to elucidate better the complex interplay between ductal shunting and cerebral hemodynamics.

Table 7.

Clinical and laboratory characteristics of neonates treated with therapeutic hypothermia (n = 4).

Case	GA (weeks)	BW (g)	Apgar 1 min	Apgar 5 min	pH	pCO₂ (mmHg)	HCO₃^- (mmol/L)	BE (mmol/L)	RI within 24h	RI after TTM	ETT	Hypotension	NICU LOS (days)	Hospital LOS (days)
1	38	2450	3	5	6.984	63.3	14.7	−17.6	0.984	0.791	✓	✓	91	106
2	39	2580	2	4	7.023	43.4	11.0	−19.5	0.814	0.571	✓	✓	30	50
3	39	2900	3	7	6.994	79.2	18.9	−14.0	0.878	0.691	✓	✓	9	13
4	40	3130	1	2	7.016	68.9	17.2	−14.8	0.689	0.628	✓	✓	9	16

Summary (mean ± SD or median [IQR]): GA 39.0 ± 0.8 weeks; BW 2765 ± 308 g; Apgar 1 min: 2.5 [1–3], 5 min: 4.5 [2.5–6.5]; pH 7.004 ± 0.018; pCO₂ 63.7 ± 15.1 mmHg; HCO₃^- 15.5 ± 3.4 mmol/L; BE –16.5 ± 2.5 mmol/L; RI within 24 h: 0.84 ± 0.12; RI after TTM: 0.67 ± 0.09.

Abbreviations: BE: base excess; BW: birth weight; ETT: endotracheal tube; GA: gestational age; LOS: length of stay; NICU: neonatal intensive care unit; pCO₂: partial pressure of carbon dioxide; RI: resistive index; TTM: targeted temperature management (therapeutic hypothermia).

✓ = received intervention.

Our findings suggest the potential clinical value of early cerebral Doppler assessment as a supplemental marker for the assessment of critically ill neonates. The identification of high-risk infants shortly after birth may support closer monitoring for hemodynamic stability. While ACA RI is a simple and non-invasive observational tool, it should be interpreted in the context of gestational age, hemodynamic status, and concomitant morbidities.

Looking forward, future research should prioritize establishing standardized clinical criteria for cerebral Doppler measurements in critically ill neonates, specifically tailored to gestational age and underlying pathologies. The substantial proportion of infants excluded from this retrospective cohort underscores the necessity of moving toward a comprehensive institutional protocol. Transitioning from a single assessment within the first 24 h of life to serial ACA RI monitoring through a prospective approach would ensure higher data completeness. Such longitudinal analysis would allow for a more robust understanding of the dynamic correlations between cerebral hemodynamics and clinical parameters throughout the neonatal course.

This study has several limitations. First, the single-center design and relatively small sample size may limit the generalizability of our findings. Specifically, the limited number of neonates in the low RI group (n = 16) may reduce the power of our multivariate analysis to detect independent associations for less frequent outcomes, such as severe IVH or mortality, after accounting for gestational age, birth weight, and age at first RI evaluation. Second, although the Doppler measurements were performed by neonatal fellows, who underwent standardized training prior to their participation, the involvement of multiple operators may still introduce inter-observer variability.

Third, the retrospective nature of the study led to a high exclusion rate, as early ACA Doppler screening was not a universal mandate during the initial phase of our study period. As shown in Table S1, infants excluded due to the lack of early ACA Doppler data had higher GA and BW. Clinically, it is likely that more stable near-term or term neonates did not undergo cerebral hemodynamic assessment immediately after birth, as the clinical urgency for such monitoring may be lower in this population. Therefore, their initial ACA Doppler scans are more likely to be performed beyond the first 24 h of life. While this ensures our study cohort strictly represents neonates requiring early-life monitoring, the findings may not be fully generalizable to the entire NICU population, particularly those who are more clinically stable at birth. Furthermore, while we performed subgroup and multivariate analyses to address the heterogeneity between preterm and term infants, the differing underlying pathophysiology in these groups remains a potential confounding factor that warrants further investigation in larger, prospective cohorts. Finally, long-term neurodevelopmental outcomes were not assessed, and thus the predictive value of RI for later impairment remains uncertain.

Conclusions

In conclusion, abnormal ACA RI within the first 24 h of life is common in NICU-admitted neonates and is independently associated with early hemodynamic disturbances. Specifically, after adjusting for potential confounders, abnormal RI values—particularly low RI—exhibited a significant independent association with systemic hypotension and the requirement for fluid resuscitation. Both low and high RI values reflect disturbed cerebral hemodynamics with distinct clinical implications. These findings suggest that early assessment of cerebral RI may serve as a supplemental clinical indicator to support risk assessment and hemodynamic monitoring in neonatal intensive care.

Supplemental material

Supplemental material - Early cerebral hemodynamic assessment using resistive index and its correlation with neonatal outcomes

Supplemental Material for Early cerebral hemodynamic assessment using resistive index and its correlation with neonatal outcomes by Ting-An Hsia, Chia-Sui Chou, Wei-Yu Chen, Pei-Chen Tsao, Yu-Sheng Lee, Mei-Jy Jeng in Journal of Neonatal-Perinatal Medicine

Footnotes

ORCID iDs

Ting-An Hsia

Chia-Sui Chou

Wei-Yu Chen

Pei-Chen Tsao

Yu-Sheng Lee

Mei-Jy Jeng

Ethical considerations

This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Institutional Review Board of Taipei Veterans General Hospital (Approval No. 2025-04-003BC).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a grant from Taipei Veterans General Hospital (Grant number: V115C-198).

Declaration of conflicting interests

The authors declare that they have no conflicts of interest related to the subject matter or materials discussed in this article.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.*

Supplemental material

Supplemental material for this article is available online.

References

Kumar

Patodia

Padhan

, et al. Transcranial Doppler: a new stethoscope–voiceover tool for neonatal. Brain. Newborn 2023; 2: 279–290.

Brackley

Kilby

Evans

. Doppler assessment of the fetal and neonatal brain. In: Fetal and Neonatal Neurology and Neurosurgery. 3rd ed. Churchill Livingstone, 2001, pp. 121–137.

Snyder

Perin

Chavez-Valdez

, et al. Head ultrasound resistive indices are associated with brain injury on diffusion tensor imaging magnetic resonance imaging in neonates with hypoxic-ischemic encephalopathy. J Comput Assist Tomogr 2020; 44(5): 687–691. https://doi.org/10.1097/RCT.0000000000001069

Leon

Ortigoza

Ali

, et al. Cerebral blood flow monitoring in high-risk fetal and neonatal populations. Front Pediatr 2022; 9: 748345. https://doi.org/10.3389/fped.2021.748345

Vutskits

. Cerebral blood flow in the neonate. Paediatr Anaesth 2014; 24(1): 22–29. https://doi.org/10.1111/pan.12307

Rhee

da Costa

Austin

, et al. Neonatal cerebrovascular autoregulation. Pediatr Res 2018; 84(5): 602–610. https://doi.org/10.1038/s41390-018-0141-6

Boylan

Young

Panerai

, et al. Dynamic cerebral autoregulation in sick newborn infants. Pediatr Res 2000; 48(1): 12–17. https://doi.org/10.1203/00006450-200007000-00005

Martini

Frabboni

Rucci

, et al. Cardiovascular and cerebrovascular responses to cardio‐respiratory events in preterm infants during the transitional period. J Physiol 2020; 598(18): 4107–4119. https://doi.org/10.1113/JP279730

Brew

Walker

Wong

. Cerebral vascular regulation and brain injury in preterm infants. Am J Physiol Regul Integr Comp Physiol 2014; 306(11): R773–R786. https://doi.org/10.1152/ajpregu.00487.2013

10.

Fenton

Woods

Evans

, et al. Cerebrovascular carbon dioxide reactivity and failure of autoregulation in preterm infants. Arch Dis Child 1992; 67(7 Spec No): 835–839. https://doi.org/10.1136/adc.67.7_spec_no.835

11.

Imanishi

Sumiya

Kanno

, et al. Relationship of cerebral blood volume with arterial and venous flow velocities in extremely low-birth-weight infants. Eur J Pediatr 2023; 182(6): 2821–2832. https://doi.org/10.1007/s00431-023-04969-3

12.

Gerner

Burton

Poretti

, et al. Transfontanellar duplex brain ultrasonography resistive indices as a prognostic tool in neonatal hypoxic-ischemic encephalopathy before and after treatment with therapeutic hypothermia. J Perinatol 2016; 36(3): 202–206. https://doi.org/10.1038/jp.2015.169

13.

Kumar

Singh

Kumar

, et al. Early postnatal color Doppler changes in neonates receiving delivery room resuscitation with low 5 min apgar score—a pilot study. J Perinatol 2021; 41(3): 486–493. https://doi.org/10.1038/s41372-020-00882-x

14.

Connors

Hunse

Gagnon

, et al. Perinatal assessment of cerebral flow velocity wave forms in the human fetus and neonate. Pediatr Res 1992; 31: 649–652. https://doi.org/10.1203/00006450-199206000-00022

15.

Camfferman

de Goederen

Govaert

, et al. Diagnostic and predictive value of Doppler ultrasound for evaluation of the brain circulation in preterm infants: a systematic review. Pediatr Res 2020; 87(Suppl 1): 31–46. https://doi.org/10.1038/s41390-020-0777-x

16.

Al Kanjo

McNamara

Czech

, et al. Systemic hypotension and patterns of cerebral blood perfusion in newborns. J Perinatol 2025; 45: 935–943. https://doi.org/10.1038/s41372-025-02357-3

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.01 MB