Introduction

Necrotizing enterocolitis (NEC) is a common and devastating gastrointestinal emergency of preterm birth that occurs in 7–12% of very low birth weight infants1,2. The mortality rate associated with NEC ranges from 20 to 40%, and survivors are at increased risk for poor long-term growth and neurodevelopmental impairment despite improvements in medical technology and neonatal care over the past several years3,4. The exact etiology of NEC remains unclear, but multiple factors, such as formula feeding, prematurity, low birth weight, intestinal ischemia and abnormal microbial colonization, are considered risk factors5,6,7. Classification as small for gestational age (SGA) is assigned if a newborn has a birth weight < 10th percentile for their gestational age8, suggesting possible intrauterine growth retardation and growth insufficiency. The risk of developing NEC in SGA neonates is more than double that in appropriate for gestational age (AGA) neonates9. However, the risk factors associated with the development of NEC in SGA infants remain unclear. To our knowledge, few studies have focused on the risk factors for NEC in SGA infants. The aim of this study was to identify the potential risk factors for NEC in SGA infants.

Results

Clinical features

During the study period, 2,912 SGA infants were admitted to the Department of Neonatology, Jiangxi Hospital Affiliated to the Children's Hospital of Chongqing Medical University (CHCMU). Among these infants, 150 (5.15%) developed NEC (Bell’s stage ≥ II), 143 of whom were eligible for enrollment; the other 7 patients were excluded because they were discharged from the hospital during the first 24 h of hospitalization (n = 4) and had incomplete information (n = 3). Consequently, 143 matched SGA infants without NEC were included in the control group.

Table 1 shows the comparison of demographic characteristics between the two groups. No differences in neonatal baseline factors or maternal factors were found between the two groups (P > 0.05). Compared with non-NEC infants, infants with NEC required a longer duration of hospitalization (P = 0.046) and had higher overall mortality (P = 0.000). After adjustment, the mortality rate in the NEC group (Table 1).

Table 1 Demographic characteristics of SGA neonates with and non-NEC.
Full size table

was still significantly greater than that in the control group (P = 0.000).

Table 2 shows the risk factors associated with NEC identified with the univariate analysis. A higher incidence of anemia and sepsis and a lower incidence of prophylactic probiotic administration were found in neonates with NEC (P < 0.05). In infants with NEC, sepsis occurred at a mean of 2 (range 1–5) days before the onset of NEC; the time of sepsis onset after birth and the timing of NEC onset are presented in Fig. 1. No differences in the incidence of patent ductus arteriosus, respiratory failure, apnea, respiratory distress syndrome or polycythemia were found between the two groups (P > 0.05).

Table 2 Identification of risk factors associated with NEC by univariate analysis.[n (%)].
Full size table
Figure 1
figure 1

The time of sepsis onset after birth and the timing of NEC onset.

Full size image

The most important risk factors for NEC in SGA infants in the less than P3 and P3–P10 subgroups are shown in Table 3. Specifically, sepsis was an important risk factor for NEC in SGA infants in the less than P3-P10 subgroup (P = 0.012), and anemia was an important risk factor for NEC in the P3 subgroup (P = 0.034). Additionally, prophylactic probiotics appeared to reduce the incidence of NEC in the P3 group (P = 0.005).

Table 3 Analysis of risk factors for NEC in different SGA infants, [n(%)].
Full size table

Table 4 shows the independent risk factors identified by the multivariate logistic regression model. Neonatal anemia (P = 0.015) and sepsis (P = 0.007) were considered independent risk factors for NEC in SGA infants, and SGA infants prophylactically administered probiotics were less likely to have NEC (P = 0.004).

Table 4 Identification of independent risk factors for NEC by multivariate logistics regression model.
Full size table

To further clarify whether the presence of these risk factors affects the prognosis of NEC in infants, we compared subgroups of NEC survivors and nonsurvivors. Table 5 shows that the presence of sepsis significantly increased the mortality of NEC infants (P = 0.016), whereas the prophylactic administration of probiotics might reduce their mortality (P = 0.000). Here, anemia did not increase the mortality of NEC patients (P = 0.732).

Table 5 The impact of risk factors on the prognosis of NEC infants, [n(%)].
Full size table

Discussion

SGA infants are considered a high-risk population for NEC9,10. In this population, the incidence of NEC in SGA infants was 5.15% (150/2912). Other studies have reported that the incidence rate of NEC in SGA infants is between 3.2% and 6.02%9,11. Notably, a multicenter survey of the Chinese population revealed that the incidence rate of NEC in SGA infants may be as high as 20.41%12.

However, the exact risk factors for NEC in SGA infants remain unclear. We found that neonatal anemia and sepsis were risk factors for NEC, and prophylactic administration of probiotics might decrease the incidence of NEC in SGA infants. This study may provide scientific evidence for prevention and treatment strategies for NEC.

We found that SGA infants with sepsis were more vulnerable to NEC. The mechanism by which sepsis causes NEC is thought to be multifactorial. Bacteria from hematogenous and gut-derived infections can directly destroy intestinal epithelial cells, and bacterial products such as endotoxins can cause necrosis of the intestinal tract13,14. Various inflammatory mediators, such as platelet-activating factor, tumor necrosis factor-α, interleukin (IL)-1, IL-6 and IL-10, contribute to the onset and progression of NEC15,16. We also found that anemia was associated with the development of NEC in SGA infants. Anemia can impair splanchnic perfusion, including that in the intestine, resulting in increased anaerobic metabolism and the production of byproducts such as lactic acid17,18. Additionally, anemia can impair the normal maturation of vascular autoregulation in the intestine19. These effects can trigger a cascade of events leading to ischemic-hypoxemic mucosal gut injury, predisposing neonates to NEC17,18.

Finally, we found that prophylactic probiotics were associated with a lower incidence of NEC in SGA infants. Several clinical trials have demonstrated that probiotic administration after birth decreases the incidence of NEC in neonates20,21. Inappropriate bacterial colonization in the gastrointestinal tract plays an key role in the development of NEC. Probiotics may promote the colonization of beneficial microbiota, inhibit the growth of pathogens, improve the function of the gut mucosal barrier, and prevent the incidence of NEC5,22. Therefore, a lack of probiotic use may be associated with a higher incidence of NEC23,24.

In this study, the overall mortality rate of the NEC group was significantly greater than that of the control group, and after adjusting for mortality due to NEC, the mortality rate in the NEC group was still significantly greater than that of the control group. Sepsis, anemia, respiratory failure, and other factors might significantly increase the mortality rate of patients with NEC7,25,26; therefore, the higher mortality rate in the NEC group observed here might be closely related to the presence of multiple comorbidities in the SGA infants themselves.

There are several limitations in this study, including errors and bias inherent to the retrospective nature of the study. Moreover, this was only a single-center study and might not represent the characteristics of the entire Chinese SGA population. Therefore, prospective multicenter studies are needed to clarify the high-risk factors for NEC in the SGA population.

In conclusion, sepsis, anemia and a lack of probiotic use were independent risk factors for NEC in SGA infants in the present study. Thus, more attention should be given to SGA neonates with anemia and sepsis in future medical practices. Additionally, prophylactic probiotic use may reduce the incidence of NEC in SGA neonates.

Methods

Study population

This study was designed as a 1:1 matched case–control study. Consecutive SGA neonates who were admitted to the Department of Neonatology, Jiangxi Hospital Affiliated to Children’s Hospital of Chongqing Medical University (CHCMU) from Jan 2008 to Dec 2022, were included. This retrospective study was approved by the Ethics Committee of Jiangxi Hospital Affiliated to CHCMU (Approval No. 2016–19), and use of the database containing the evaluated data was permitted by the Ethics Committees of Jiangxi Hospital Affiliated to CHCMU. The requirement for informed consent was waived by the Ethics Committee of Jiangxi Hospital Affiliated to CHCMU. All study protocols were carried out in accordance with the ethical standards of the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. SGA was defined as a birth weight < 10th percentile for each newborn’s gestational age according to the growth chart for Chinese neonates8,27. SGA neonates with diagnosed NEC (Bell’s stage ≥ II) were included in the NEC group28. The SGA neonates without NEC admitted during the same period to the same hospital (the admission time of the control group infants did not differ from that of the NEC group by more than 3 months) were screened as possible controls; those matched for gestational age (difference of < 3 days) and birth weight (difference of < 100 g) were selected. When there were multiple candidate infants , one infant was randomly selected for inclusion in the control group by a computer. Neonates whose medical information was incomplete or who were discharged from the hospital during the first 24 h of hospitalization were excluded from the study.

Data collection

The demographic characteristics, including maternal factors such as maternal age, antibiotic exposure during pregnancy, antenatal glucocorticoid exposure, gestational hypertension, intrauterine cholestasis during pregnancy, anemia during pregnancy, gestational diabetes mellitus, premature rupture of the membrane (> 18 h), fetal distress and meconium-stained amniotic fluid, were recorded. The neonatal factors included gender, gestational age, birth weight, mode of delivery, feeding type, and Apgar score at 1 and 5 min. Risk factors prior to the occurrence of NEC, such as neonatal anemia, sepsis, patent ductus arteriosus, respiratory failure, apnea, respiratory distress syndrome and polycythemia, were also recorded. Laboratory examinations and clinical outcomes were collected retrospectively from the hospital’s neonatal database. Neonatal anemia was defined as a hemoglobin or hematocrit concentration greater than 2 standard deviations below the mean for postnatal age29. Sepsis that developed prior to the onset of NEC was diagnosed on the basis of clinical manifestations and the growth of bacteria on blood culture and ancillary tests such as leukopenia (WBC < 5 × 109/L) or leukocytosis (WBC > 25 × 109/L for ≤ 3 days or WBC > 20 × 109/L for > 3 days), a platelet count < 100 × 109/L, an immature-to-total neutrophil ratio (I:T ratio) ≥ 0.16, and a C-reactive protein > 8 mg/L30,31,32. NEC was defined according to the modified Bell’s criteria as Bell Stage II or greater28,33. All infants were treated with suitable and necessary interventions according to their conditions, with the possible interventions including cessation of enteral feeding, nasogastric suction and parenteral nutrition, antibiotic therapy and surgical intervention. The data were collected, reviewed, deidentified, and anonymously analyzed by the authors, and the Ethics Committee of Jiangxi Hospital Affiliated to CHCMU waived the requirement for informed consent because of the anonymized nature of the data and the scientific purpose of the study.

Statistical analysis

All analyses were conducted using SPSS 24.0 (SPSS Inc., Chicago, IL, USA). The Kolmogorov‒Smirnov test was used to assess the normality of continuous variables. Normally distributed variables were analyzed using Student’s t test, and skewed variables were analyzed with the Mann‒Whitney U test. The chi-square test and Fisher’s exact test were used to compare categorical variables between the two groups. All potential risk factors related to NEC incidence were included in the multivariate regression model to identify independent risk factors for NEC. P < 0.05 was considered statistically significant.