Doppler abnormality predisposes preterm infants with fetal growth restriction to postnatal intestinal disorder.

Tomohiro Ohtani: Conceptualization; methodology; validation; formal analysis; resources; data curation; writing – original draft; visualization; investigation. Mari Ichinose: Conceptualization; methodology; validation; writing – original draft; writing – review and editing; supervision; funding acquisition. Yu Ariyoshi: Validation; investigation; data curation. Miho Irie: Data curation; investigation; validation. Masatake Toshimitsu: Resources; writing – review and editing. Seisuke Sayama: Resources; writing – review and editing. Takahiro Seyama: Resources; writing – review and editing. Hiroshi Muto: Resources; writing – review and editing. Yoshihiko Shitara: Resources; writing – review and editing. Atsushi Ito: Resources; writing – review and editing. Mariko Yoshida: Resources; writing – review and editing. Satsuki Kakiuchi: Resources; writing – review and editing. Akio Ishiguro: Resources; writing – review and editing. Keiichi Kumasawa: Writing – review and editing; supervision. Takayuki Iriyama: Writing – review and editing; supervision; conceptualization. Jun Fujishiro: Supervision; writing – review and editing; project administration. Naoto Takahashi: Supervision; project administration; writing – review and editing. Yasushi Hirota: Writing – review and editing; supervision; project administration. Yutaka Osuga: Supervision; project administration; writing – review and editing.

This study was approved by the ethics committee of the University of Tokyo (No. 20220223NI).

This work was supported by JSPS KAKENHI Grant Number JP22K20935, JP23K15828, and the JAOG Ogyaa Donation Foundation (JODF) research grant.

This study was approved by the ethics committee of our hospital (No. 20220223NI). We retrospectively collected clinical data from the medical records of singleton preterm infants born between 22 weeks 0 days and 28 weeks 6 days of gestation at our hospital between January 1, 2013, and December 31, 2022. We excluded the following cases: intrauterine fetal death, congenital abnormalities, multiple pregnancies, and transfer to another hospital after birth. We defined ID as NEC, MRI, or FIP, and the cases were divided into ID and non‐ID groups. Subsequently, we compared maternal and fetal factors between the ID and non‐ID groups. The maternal factors included maternal age, parity, mode of delivery, and maternal complications, such as hypertensive disorders of pregnancy (HDP), gestational diabetes, clinical chorioamnionitis (CAM), endometriosis, and adenomyosis. Fetal factors included fetal growth restriction (FGR) and oligohydramnios, which were both detected using ultrasonography. The fetal ultrasound data collected most closely to birth was used to determine FGR. The Japanese fetal growth curve was used as a reference. Following the guidelines of the International Society of Ultrasound in Obstetrics and Gynecology, 22 we diagnosed FGR if: (1) abdominal circumference (AC) or estimated fetal weight (EFW) was <3rd percentile, or (2) umbilical artery‐absent end diastolic flow (UA‐AEDF) or AC/EFW were 95th percentile, and/or (3) umbilical artery‐PI (UA‐PI) was >95th percentile. Specifically, in FGR cases, the prevalence of Doppler abnormalities such as the brain sparing effect (cerebroplacental ratio <1), UA‐AEDF, UA‐reverse end‐diastolic flow (UA‐REDF), and ductus venosus (DV) a‐wave reversal were recorded. In addition, the duration from the onset of any Doppler abnormalities to delivery was recorded. A case with uncertain onset of Doppler abnormalities was considered as missing values in the analysis comparing the time from the onset of Doppler abnormalities to delivery between ID and non‐ID groups. The fetal indications for delivery in FGR cases were comprehensively determined by carefully assessing cardiotocographic monitoring (CTG), ultrasound findings including fetal growth, fetal movements, amniotic fluid volume, and Doppler abnormalities. In particular, non‐reassuring fetal status and worsening Doppler abnormalities were considered as the factors determining delivery. With respect to maternal indications, delivery was indicated in the presence of preeclampsia (PE) with severe features or HELLP syndrome. Clinical CAM was diagnosed based on Lencki's criterion. 23 Delivery was indicated when this diagnostic criterion was met. The factors examined at birth included sex, gestational age at birth, birth weight, small for gestational age (defined as birth weight <10th percentile of the birth weight standard values for each gestational age in Japan), 24 umbilical artery pH (UApH), umbilical vein pH (UVpH), Apgar score (1 and 5 min scores), placental weight, fetal weight/placental weight ratio, and histological CAM. We examined the placentas pathologically in all cases of preterm delivery, and histological CAM was defined as either stage II or higher based on Blanc's classification. 25 Univariate comparisons of obstetric risk factors between the ID and non‐ID groups were performed for each variable, and the Benjamini–Hochberg false discovery rate correction was applied to adjust for multiple testing. The Mann–Whitney U test was used for single comparisons without correction. Statistical significance was set at p <0.05 for both nominal and adjusted p ‐values. All statistical analyses were performed using JMP® Pro 17.2.0.

A flowchart illustrating the selection process for the study subjects is shown in Figure  1 . Of the 162 preterm cases, 43 were excluded from the study, including 23 cases of intrauterine fetal death or stillbirth, 11 congenital abnormality cases, seven twin pregnancy cases, and two cases of transfer to other hospitals after birth. Finally, 119 singleton preterm infants were included in the analysis. All cases received essential neonatal resuscitation, and there was no case in which delivery or neonatal resuscitation was withheld due to early gestational age or small fetal size. Flow diagram of the study. Of the 162 preterm infants, 43 were excluded, and 119 singleton preterm infants were included in this study. Of the 119 patients, 22 (18.5%) were postnatally diagnosed with ID. The breakdown of ID cases was as follows: 14 patients underwent MRI, 5 patients had NEC, and 4 patients had FIP. However, MRI and NEC were observed in one case (Figure  2 ). Percentage and the details of intestinal disorder. Of the 119 preterm infants eligible for this study, 22 (18.5%) were diagnosed with ID. Of these, MRI was performed in 14 cases, NEC in 5, and FIP in 4. One patient had both an MRI and NEC diagnosis. FIP, focal intestinal perforation; ID, intestinal disorder; MRI, meconium‐related ileus; NEC, necrotizing enterocolitis. A comparison of risk factors between the ID and non‐ID groups is shown in Table  1 . Maternal characteristics such as age, parity, and mode of delivery did not significantly differ between the groups. No infant in either group was born from mothers diagnosed with gestational hypertension; however, PE and superimposed PE were more frequent in the ID group than in the non‐ID group. Clinical CAM, endometriosis, and adenomyosis also did not significantly differ between the groups. In the ultrasound evaluations, FGR was more common in the ID group than in the non‐ID group (ID group 10/22 [45.5%] vs. non‐ID group 23/97 [23.7%], nominal p  = 0.047, adjusted p  = 0.496). Therefore, FGR was potentially associated with ID. In particular, 21 of the 33 cases of FGR (63.6%) were delivered due to non‐reassuring fetal status. Furthermore, oligohydramnios did not significantly differ between the two groups. In the non‐ID group, many cases (17/23 [73.9%]) were attributed to premature rupture of membranes rather than FGR. Factors observable at birth, such as sex, gestational age, birth weight, umbilical cord blood gas level, and Apgar score, did not differ between the two groups. Notably, histological CAM was less common in the ID group than in the non‐ID group (ID group 4/22, [18.2%] vs. non‐ID group 44/97 [45.4%], nominal p  = 0.014, adjusted p  = 0.301). Comparison of obstetric risk factors between the ID group and the non‐ID group. Note : Results are presented as N (%) or median [interquartile range]. Univariate comparisons were performed for each variable using t ‐test for continuous variables and likelihood‐ratio chi‐square test for categorical variables, and Benjamini–Hochberg false discovery rate correction was applied for multiple testing. Statistical significance was set at p <0.05 for both nominal and adjusted p ‐values. Abbreviations: FGR, fetal growth restriction; ID, intestinal disorder; SGA, small for gestational age; UApH, umbilical artery pH; UVpH, umbilical vein pH. Indicate significance of the test results ( p <0.05). Next, we compared the presence of Doppler abnormalities of infants with FGR between the ID and non‐ID groups. Cases without Doppler abnormalities were observed in 30.0% (3/10) of cases in the ID group and 43.5% (10/23) of cases in the non‐ID group. Conversely, FGR cases with any Doppler abnormalities were more frequent in the ID group than in the non‐ID group (70.0% [7/10] vs. 56.5% [13/23]), although the differences were not statistically significant. The percentages for each Doppler abnormality are shown in Table  2 . Brain‐sparing effect, UA‐AEDF, UA‐REDF, and DV a‐wave reversal were observed in the ID and non‐ID groups. The frequencies of these abnormalities did not significantly differ between the groups. In the analysis that compared the time from the onset of Doppler abnormalities to delivery, one case was excluded because the Doppler abnormalities were already present when the patient was transferred to our hospital; thus, the onset of abnormalities was uncertain. Notably, the time from the onset of any Doppler abnormalities to delivery was significantly longer in the ID group than in the non‐ID group. After one case in each group (672 and 624 h, respectively) was excluded as an outlier (as determined using the Smirnov‐Grubbs test) the median time in the ID group was 180 h (interquartile range: 81–243 h) compared to 24 h (interquartile range: 12–192 h) in the non‐ID group ( p  = 0.049) (Figure  3 ). The prevalence of Doppler abnormalities in cases of fetal growth restriction. Note : Comparison of Doppler parameters among FGR cases between ID group and non‐ID group. Overlapping Doppler findings were counted. Data are presented as n (%). Abbreviations: DV, ductus venosus; FGR, fetal growth restriction; ID, intestinal disorder; UA‐AEDF, umbilical artery absent end‐diastolic flow; UA‐REDF, umbilical artery reverse end‐diastolic flow. Comparison of time from onset of Doppler abnormalities to delivery between the intestinal disorder and non‐intestinal disorder groups in FGR cases. The time interval from the first observed Doppler blood flow abnormality to delivery was compared between the ID and non‐ID groups in FGR cases. In one non‐ID case, data from the onset of Doppler abnormalities were unavailable due to transfer from another facility and was regarded as missing values. Additionally, one case each in the ID and non‐ID groups (672 and 624 h, respectively) was excluded as an outlier, as determined using the Smirnov‐Grubbs test. * p <0.05 (Mann–Whitney U test). FGR, fetal growth restriction; ID, intestinal disorder.

In this study, we comprehensively examined the obstetric risk factors for ID, including NEC, FIP, and MRI, among preterm infants, rather than focusing on the individual conditions. We found that FGR was potentially associated with ID, particularly in cases that had prolonged exposure to Doppler abnormalities. Notably, prenatal risk factors detected on ultrasound showed a stronger association with ID than acute‐phase parameters such as gestational age, birth weight, and UApH. While previous studies have focused mainly on postnatal factors, our findings emphasize the importance of the prenatal environment to the intestinal health of preterm infants. Placental insufficiency is the most frequent cause of FGR because it inhibits uteroplacental blood flow, limits oxygen delivery and nutrients to the fetus, and causes chronic fetal hypoxemia. 16 , 17 , 26 This chronic intrauterine stress, which is a characteristic of placental insufficiency, leads to blood flow redistribution that prioritizes vital organs (such as the brain, myocardium, and adrenal glands). Hence, the flow to other organs, including the gastrointestinal tract and skin, becomes restricted. This redistribution is detectable by ultrasound as a brain‐sparing effect, followed by UA‐AEDF or UA‐REDF and the absence or reverse flow of the DV a‐wave. 16 , 26 These Doppler abnormalities are associated with acute postnatal parameters, such as Apgar score and umbilical cord blood gas, and are widely recognized as parameters for determining the timing of delivery in infants with FGR. 20 , 21 Our findings suggest that patients with FGR and prolonged Doppler abnormalities may have a high risk of ID. The duration of exposure to Doppler abnormalities might not have directly determined the delivery timing; however, the timing could suggest a need for personalized postnatal treatment approaches and provide insights into perinatal management, especially for potential associations with long‐term neonatal complications. Next, we discuss the pathophysiological relationship between FGR and each disease classified as an ID. Studies have shown that NEC may have various causes, including impaired intestinal blood flow, dysregulation of the inflammatory cascade, and microbiota alterations associated with intestinal immaturity. 19 , 27 As such, FGR is recognized as a risk factor for the development of NEC, where intestinal immaturity and ischemia due to restricted blood flow contribute to its development. 19 , 28 , 29 Similarly, FIP, a condition that resembles NEC and is often challenging to distinguish from it during diagnosis, frequently occurs in low‐birthweight infants 2 and involves intestinal ischemia and immature bowel motility. 8 A direct relationship between FIP and FGR has not been confirmed. However, PE may be a risk factor for both NEC and FIP, which suggests a potential link to placental insufficiency. 8 , 30 Additionally, MRI has been frequently observed in low‐birthweight infants, 2 and some reports have indicated that FGR may be a risk factor. 31 Furthermore, immature bowel motility has also been associated with the development of MRI, 10 although further research is needed to elucidate its pathogenesis and its relationship with ischemia. To date, research on the pathophysiology of ID has primarily focused on physiological changes such as impaired blood flow and intestinal motility. However, the effects of chronic intrauterine stress on the histological and molecular development of the intestine remain unclear. Both FGR and PE result from placental insufficiency and are therefore closely interrelated. 32 , 33 Although the overall frequency of HDP did not significantly differ between the groups in this study, six of seven (86%) cases of HDP in the ID group presented with FGR compared with 11 of 20 (55%) in the non‐ID group. These findings suggest that HDP without FGR may not be a risk factor for ID, whereas FGR may reflect the effect of placental insufficiency on ID. However, CAM is known as a risk factor for NEC, 34 , 35 , 36 , 37 the present study showed a significantly lower frequency of histological CAM in the ID group. This result might have occurred because there were very few cases of FGR or PE among those who required termination of pregnancy in the presence of CAM. Consequently, our findings suggest that pregnancy termination related to FGR rather than CAM may be more closely associated with ID risk. This study had some limitations. It was a single‐center retrospective study with a small sample size, which may limit generalizability, reduce statistical power, and introduce selection and information bias. Therefore, in comparisons of obstetric risk factors, no statistically significant differences were observed after false discovery rate correction, raising concern about residual confounding that could not be addressed because multivariable adjustment was not feasible. Consequently, further research with larger cohorts is warranted to verify whether FGR is an independent risk factor for ID. Furthermore, further research is needed to determine whether similar results would apply at later gestational ages because the analysis was restricted to preterm births at or before 28 weeks when the ID incidence was relatively high. Finally, this study focused exclusively on obstetric risk factors without evaluating postnatal treatments or complications other than ID. Notably, anti‐inflammatory strategies such as the routine administration of probiotics and artificial milk containing human milk oligosaccharides have been appropriately implemented in the neonatal management of preterm infants. 38 , 39 Therefore, future studies should determine whether individualized interventions based on specific risk factors may effectively prevent ID in infants exposed to intrauterine stress resulting from placental insufficiency. This study reported prenatal factors in relation to comprehensive ID, including NEC, FIP, and MRI, rather than focusing on individual conditions as previous studies have done. We found a potential association between preterm FGR and ID. Notably, infants with FGR who experience prolonged Doppler abnormalities may face an increased risk of developing ID. It suggests that chronic intrauterine stress beyond mere immaturity may contribute to the increased vulnerability of postnatal intestinal function. Therefore, recognizing prenatal risk factors based on individual risk profiles provides a foundation for optimizing and personalizing postnatal treatments to improve long‐term outcomes.

Focusing on long‐term outcomes in order to provide better perinatal care has become increasingly important as the survival rates for preterm infants improve. For example, establishing enteral nutrition is crucial for the development of preterm and low‐birth‐weight infants. However, intestinal disorders (ID), such as necrotizing enterocolitis (NEC), meconium‐related ileus (MRI), and focal intestinal perforation (FIP), 1 , 2 pose challenges to nutritional intake and developmental outcomes in affected infants. Most studies on the obstetric risk factors for ID have focused extensively on NEC, 3 a severe ID that occurs in approximately 5%–12% of infants with very low birth weight. 4 , 5 Additionally, several systematic reviews have revealed that risk factors such as low birth weight, small for gestational age, and ethnicity are associated with NEC. 6 , 7 However, focusing solely on NEC overlooks the larger issue that this condition may represent only one component of an extensive ID spectrum. Researchers have defined FIP as an isolated intestinal perforation without necrotic or obstructive changes, 8 , 9 while MRI is a functional intestinal obstruction characterized by abdominal distension and delayed passage of meconium. 10 Notably, reports show that FIP and MRI have similar incidences to NEC, 8 , 11 , 12 , 13 and all are associated with high mortality rates, surgical interventions, and challenges in establishing nutritional intake. 5 , 8 , 14 Consequently, findings regarding FIP and MRI are important and should not be overlooked. A more comprehensive approach to ID could potentially benefit larger patient populations. Additionally, postnatal risk factors for ID include acute parameters at birth (e.g., Apgar score), postnatal treatments (e.g., blood transfusion), and other complications such as respiratory distress syndrome and symptomatic patent ductus arteriosus. 2 , 7 , 15 However, the relationship between prenatal factors and ID is poorly understood. Nevertheless, the intrauterine environment may influence fetal and postnatal development, including intestinal function. 16 , 17 , 18 , 19 The validity of prenatal factors, such as ultrasound assessments conducted in utero, has often been evaluated in relation to fetal acute‐phase parameters at birth. 20 , 21 However, perinatal management strategies that factor in long‐term outcomes are increasingly needed. Therefore, in this study, we aimed to comprehensively clarify prenatal risk factors for ID in order to address the gaps in current understanding.

The authors declare no conflicts of interest.