Blood Biomarkers in the Fetally Growth Restricted and Small for Gestational Age Neonate: Associations with Brain Injury

Abstract Fetal growth restriction (FGR) and small for gestational age (SGA) infants have increased risk of mortality and morbidity. Although both FGR and SGA infants have low birthweights for gestational age, a diagnosis of FGR also requires assessments of umbilical artery Doppler, physiological determinants, neonatal features of malnutrition, and in utero growth retardation. Both FGR and SGA are associated with adverse neurodevelopmental outcomes ranging from learning and behavioral difficulties to cerebral palsy. Up to 50% of FGR, newborns are not diagnosed until around the time of birth, yet this diagnosis lacks further indication of the risk of brain injury or adverse neurodevelopmental outcomes. Blood biomarkers may be a promising tool. Defining blood biomarkers indicating an infant’s risk of brain injury would provide the opportunity for early detection and therefore earlier support. The aim of this review was to summarize the current literature to assist in guiding the future direction for the early detection of adverse brain outcomes in FGR and SGA neonates. The studies investigated potential diagnostic blood biomarkers from cord and neonatal blood or serum from FGR and SGA human neonates. Results were often conflicting with heterogeneity common in the biomarkers examined, timepoints, gestational age, and definitions of FGR and SGA used. Due to these variations, it was difficult to draw strong conclusions from the results. The search for blood biomarkers of brain injury in FGR and SGA neonates should continue as early detection and intervention is critical to improve outcomes for these neonates.


Introduction
Fetal growth restriction (FGR) contributes to increased mortality and morbidity in neonates [1].The incidence of FGR is at its highest in the last 20 years [2], accounting for approximately 10% of all pregnancies [2,3].Both FGR and small for gestational age (SGA) have been associated with neurodevelopmental delay [4,5] but these are distinct and different diagnoses.SGA is solely defined by birth weight for gestational age (GA), which ranges from <10th to <3rd percentile, while FGR diagnosis relies on serial measures to establish in utero growth retardation often along with other features including umbilical artery Doppler flow pattern [6,7].Maternal, fetal, placental, and genetic factors may all contribute to FGR [7] but the majority arises from placental insufficiency, whereby a poorly functioning placenta restricts nutrient and oxygen supply to the developing fetus impacting normal development.Furthermore, compromised placental function can result in not only chronic fetal hypoxia but abnormal transfer of maternal hormones to fetal circulation, having significant implications on fetal programming, growth, and development [1,8,9].The brain is particularly affected by this chronic hypoxic environment, with neuronal and white matter injury being major pathophysiological features of FGR [10].Preterm FGR infants have lower gray and white matter complexity compared to preterm and term normally grown infants, with these structural abnormalities persisting at 1 year of age and associated with neurodevelopmental disabilities [10].Neurodevelopmental delays are evident in up to 50% of FGR infants [11,12], yet there is no sensitive method to detect which FGR or SGA neonate may or may not have brain injury.Current methods include cranial ultrasound, magnetic resonance imaging, and electroencephalography [13,14].Yet these methods rely on the proficiency of the operator and lack sensitivity when detecting subtle brain injury.Blood biomarkers, however, are not largely operator dependent and have been used to form highly sensitive tests for many diseases.Once identified, blood biomarkers are an efficient and effective diagnosis tool.Investigating potential biomarkers of FGR brain injury will assist in early diagnosis and identify potential targets for future therapies to reduce or prevent long-term adverse outcomes in FGR and SGA.This paper reviews studies reporting potential biomarkers in cord and neonatal blood or serum of FGR and SGA neonates, up to 28 postnatal days.The purpose of this review is to assist in determining which blood biomarkers may be suitable candidates to pursue in future studies for the early detection of adverse brain outcomes in FGR and SGA neonates.

Methods
A search of the literature on the PubMed, Embase, and Web of Science databases was conducted from conception to September 2022 using search terms including "fetal growth restriction," "intrauterine growth restriction," "small for gestational age," "biomarkers," "proteins," and "blood."Hand searching of these databases was also conducted to ensure all relevant articles were included.Studies were included if (1) the cohorts studied included FGR/SGA neonates where blood samples were taken within the 28 days of life to detect blood biomarker levels and (2) the study was written in English.A total of 18 studies met these inclusion criteria as determined by two reviewers (J.W. and H.M.).Metaanalysis was not conducted due to the heterogeneity of timepoints and specific blood biomarkers measured across the included studies.

Potential Biomarkers of Interest
Biomarkers have been developed as accurate and reproducible indications of health and disease.For example, biomarkers have been used to indicate risk of a specific outcome, diagnostic or predictive tools, monitoring disease progress, and measuring response to therapies [15].Depending on the intended use of the biomarker, different sample types may be used.Blood biomarkers have the benefit of the relative ease of sample collection.With FGR and SGA exerting a negative effect on the brain, it has been hypothesized that blood biomarkers might prove to be an effective tool in detecting brain injury in FGR and SGA neonates.Studies investigating potential blood biomarkers in FGR and SGA neonates have considered markers with roles in inflammation, neurogenesis, growth, development, vasculature, and remodelling, as well as structural proteins.Many of these markers target mechanisms potentially involved in brain injury associated with FGR and SGA.Animal studies have shown that development of the blood-brain barrier (BBB) is initiated early on in embryonic development; however, it is not until after birth that mature cell types, including astrocytes begin to appear [16].Perinatal insults, such as chronic hypoxia, impact the neurovascular unit [17].As demonstrated in large animal models of FGR [18][19][20][21][22][23], FGR infants may be at an increased risk of BBB disruption resulting in a "leaky" BBB.T-cells, red blood cells, and endogenous plasma proteins infiltrate the FGR piglet and lamb brain [18][19][20][21][22][23], making it susceptible to further insult.The leaky nature of the BBB in FGR animal models indicates that blood biomarkers could be used to detect brain injury in FGR neonates.

Inflammatory Biomarkers
Neuroinflammation is observed in FGR animal models in the days following birth [20] whereby activation of microglia, increased levels of pro-inflammatory cytokines and astrogliosis are observed in injured regions of the brain [18,21,23,24].These studies indicate neuroinflammation plays a crucial role in white matter and neuronal disruption that forms part of the pathophysiology of FGR [25].Furthermore, there is evidence inflammation is associated with the disruption of the BBB [20].The chronic hypoxic environment FGR fetuses are exposed to increases the risk of BBB disruption, resulting in infiltration of peripheral proteins to the brain, which in turn increases the susceptibility of the FGR brain to further injury [20].Blood is an appropriate and clinically relevant sample type to screen for biomarkers of inflammation which may be associated with brain injury in FGR/SGA.Four studies to date have explored inflammatory blood biomarkers in FGR neonates (see Table 1; [26][27][28][29]).Yue et al. [26] examined interleukin (IL)-6, IL-8, and IL-10 at birth to postnatal day (P) 5 in a mixed cohort of preterm and term very low birth weight FGR and non-FGR neonates.Boutsikou et al. [29] examined C-reactive protein (CRP) levels in maternal, umbilical cord, and neonatal samples at birth, P1 and 4 in term FGR and AGA (appropriate for GA) neonates.McElrath et al. [27] and Leviton et al. [28] examined numerous cytokines in preterm FGR neonates, grouped as severely growth restricted (birth weight Z-score <−2) and growth limited (birth weight Z-score ≥−2 and <−1), on P1, 7, 14, 21, and 28.
Pro-inflammatory cytokines play a crucial role in inflammatory responses in the FGR brain [25,30].One such pro-inflammatory cytokine is IL-6 which is expressed in response to infection and tissue injury [31].Yue et al. [26] observed that FGR neonates have significantly lower IL-6 levels in cord blood at birth compared to non-FGR neonates.No significant differences were observed in neonatal serum on P1 to 3, yet significant increases in levels of IL-6 in FGR neonates compared to non-FGR neonates on P4 and 5 were observed [26].Although McElrath et al. [27] also investigated IL-6 expression on P1 and 7, increases in severely growth restricted and growth limited neonates were not observed until P14.Leviton et al. [28] also investigated IL-6 concentration on P21 and 28, observing no differences between FGR and AGA at either of these timepoints.McElrath et al. [27] and Leviton et al. [28] also examined concentration of the IL-6 receptor, IL-6-R, but observed no alterations between FGR and AGA at any timepoint investigated in these studies.IL-8 is a member of the CXC chemokine family with a key role in mediating inflammatory responses [32] and is thought to play a role in FGR.Increases in IL-8 levels in FGR neonatal blood were observed on P2, 3, and 4 but not on P1 or 5 or umbilical cord blood at birth compared to non-FGR [26].McElrath et al. [27] and Leviton et al. [28] also observed increases in IL-8 levels in preterm FGR neonates on P14, 21, and 28 although not on P1 or 7.
IL-10 is an anti-inflammatory cytokine.Only one study has investigated IL-10 levels in FGR neonates where increased levels were observed on P2 to 4 compared to non-FGR, which, however, were not present on P5 [26].
CRP is synthesized in response to pro-inflammatory cytokines [33] and therefore an important biomarker to consider in inflammatory responses in FGR.No differences were observed in CRP levels between FGR and AGA from maternal blood at the first sign of labor, umbilical cord blood at birth or neonatal blood at P1 and 4 [29].However, increases in CRP levels were observed in blood from severely growth restricted neonates on P7, 14, 21, and 28 and in growth limited neonates on P7 compared to AGA neonates [27,28].

Summary
Although numerous inflammatory markers have been investigated, only a few have consistently shown differences in FGR and SGA cohorts.McElrath et al. [27] and Leviton et al. [28] investigated similar biomarkers (IL-6, IL-6-R, IL-8, CRP, serum amyloid A, IL-1β, RANTES, ICAM-1, TNFα, and TNF-R2); however, these studies examined different timepoints and therefore provide little supporting evidence for the different findings from each.IL-6, IL-8, and CRP have been considered in three studies; however, there were few overlapping timepoints.Despite this, the findings from these studies do indicate that the levels of CRP and IL-8 are elevated in the weeks following birth in FGR/SGA neonates [26][27][28][29].The levels of IL-6 are increased in neonatal blood after birth but for a shorter period of time and are decreased in umbilical cord blood at birth for FGR/SGA neonates [26][27][28].These results demonstrate the dynamic and changing inflammatory response potentially involved in the pathogenesis of FGR and SGA.Leviton et al. [28] demonstrated that little of this increased inflammatory profile was attributable with secondary matters such as delivery indication or bacteremia.Therefore, these results have important implications for not only detection but for the inflammatory mechanisms potentially involved in FGR and SGA associated brain injury.

Growth and Developmental Biomarkers
Proteins involved in brain growth and development have also been of particular interest as biomarkers in FGR and SGA.Neurogranin has important roles in dendritic spine formation, synaptic plasticity, long-term potentiation, and spatial learning [34,35].Brain-derived neurotrophic factor (BDNF) has many potential roles in development and memory [36].Galectins 1 and 3 are involved in proliferation of neural stem cells and are expressed in response to ischemic and traumatic brain injury [37,38].
The insulin-like growth factor (IGF) axis has a crucial role in fetal growth and alterations in the expression of different aspects of this axis may be involved in the pathophysiology of FGR [44].Two studies investigated aspects of the IGF axis and reported alterations in IGFbinding protein (IGFBP)-1 and IGF-1 levels [27,43].IGF-1 was found to only be decreased on P1 in severely growth restricted compared to AGA neonates [43].However, IGFBP-1 was increased on P1 for both severely growth restricted and growth limited neonates and P14, 21, and 28 for severely growth restricted neonates compared to AGA neonates [27,43].There were some conflicting findings for P7 IGFBP-1 levels with one study only reporting increased levels in growth limited neonates [27] while another only reported increases in severely growth restricted neonates [43].
Leptin levels were measured in one study with an increase in maternal serum at delivery and decreased levels in umbilical vein blood at delivery reported [42].Other biomarkers examined included matrix metalloproteinase (MMP)-1, MMP-9, thyroid stimulating hormone, and erythropoietin (EPO).Although no significant changes were observed for MMP-1 on P1, 7, or 14, lower levels of MMP-9 were observed in growth limited neonates on P1 and 7 and increased levels were observed in severely growth restricted neonates on P14 [27].Growth limited neonates were also observed to have increased levels of MMP-9 at later timepoints on P21 and 28 [28].Thyroid stimulating hormone was found to be increased in severely growth restricted neonates on P21 and 28, although no differences were identified for EPO levels at these timepoints [28].

Summary
Only BDNF, IGFBP-1, and MMP-9 levels have been investigated in more than one study [26-28, 41, 43].Therefore, findings discussed here require further exploration for validation.Results for BDNF are conflicting [26,41].Although the findings for IGFBP-1 and later time points for MMP-9 are relatively consistent [27,28,43], there is little overlap in the timepoints assessed and the study populations are all from the Extremely Low Gestational Age Newborns (ELGAN) study.

Structural/Support Proteins
Structural support proteins of the central nervous system (CNS) S100B and neuron-specific enolase (NSE) have been considered potential biomarkers of CNS damage.S100B modulates cell proliferation, apoptosis, and differentiation of a number of cell types in the CNS [45][46][47].Expressed in the later stages of neural differentiation, NSE is a highly specific marker for neurons and peripheral neuroendocrine cells [48].Amniotic fluid samples from preterm infants with abnormal neurosonograms had higher NSE concentrations [49].Increased P1 serum levels of S100B have been reported in term newborn infants with hypoxicischemic encephalopathy [50].
Seven studies have examined S100B as a biomarker of FGR and SGA with conflicting findings.Four studies examining S100B levels in maternal blood samples found no differences; however, one study observed significant increases (see Table 3; [29,[51][52][53][54]).Three studies found no differences in S100B levels in cord blood samples [29,51,52]; however, two studies observed increases in S100B levels in cord blood in small for date (SFD) neonates (using the definition from Campbell and Thoms [55] and postnatal confirmation of a birth weight <10th centile) with abnormal fetal Doppler findings [56] and SGA neonates [57].Velipaşaoğlu et al. [54] observed increased S100B levels in umbilical arterial blood at delivery for FGR compared to AGA neonates but no significant alterations in umbilical venous blood.One study examined S100B levels in neonatal blood on P1 and 4 but observed no differences [26] while another study did observe increases in S100B levels in neonatal blood of FGR compared to AGA neonates at birth [53].
Four studies have investigated NSE as a biomarker for FGR and SGA with inconsistent results.While Mazarico et al. [52] observed significantly higher levels of NSE in fetal umbilical arterial and maternal blood at birth in FGR, other studies have not observed this alteration [54,58].Tayman et al. [59] observed increased NSE levels in peripheral venous blood 12 h after birth for SGA compared to AGA neonates; however, levels were decreased in asymmetric SGA compared to symmetric SGA neonates.
Tau has an integral role in axonal maintenance [60] and has also been a protein of interest as a biomarker of FGR.Increases in Tau levels were observed in FGR neonates in cord blood at birth [26].However, no differences were observed in neonatal serum at P1 to 5 in FGR neonates compared to non-FGR [26].
Glial fibrillary astrocytic protein (GFAP), as a biomarker for CNS injury [61], has also been considered a potential biomarker for FGR.However, no significant differences in GFAP levels were identified between FGR and non-FGR neonates at birth or P1 to 5 [26].One study also examined levels of alpha-foetoprotein but observed no significant alterations [54].

Summary
These findings are insufficient to confirm or refute the hypothesis that the pathophysiology of FGR results in alterations in the levels of structural/support proteins of the CNS in blood.However, some studies have investigated the potential of these biomarkers to detect brain injury in FGR/SGA and indicate that structural proteins of the CNS may be promising candidates as prognostic indicators of brain outcomes [26,62].Yue et al. [26] found that GFAP levels on P3 and 4 were increased in FGR neonates with brain injury (defined as neonates who had either periventricular white matter injury, grades 3 or 4 intraventricular hemorrhage, seizures, or death) compared to FGR neonates without brain injury.In another study, Mazarico et al. [62] observed statistically significant inverse relationships for all infants between both the concentrations of S100B and NSE with the results of the Bayley-III cognitive assessment at 2 years of age.Although S100B shows promise as a biomarker for brain injury, it has a relatively short half-life which might impact its clinical utility as a biomarker [63].Yet these studies indicate that structural/support proteins may be effective and clinically relevant blood biomarkers for detecting risk of brain injury in FGR and SGA.

Vasculature Markers
The neurovascular unit has an integral role in maintaining the BBB to ensure healthy brain development [17].Many cell types form the neurovascular unit, including vascular endothelial cells, glial cells, and neurons.Large animal models of FGR report neurovascular unit alterations with reduction in number of endothelial cells, reduced cerebral vascularity, and altered glial cell interaction with cerebral blood vessels [18-20, 22, 23].This breakdown of the neurovascular unit and effective BBB may result in systemic immune cells negatively impacting the brain [20].
One such potential marker of alterations to the vasculature system is vascular endothelial growth factor (VEGF).VEGF is a potent angiogenic factor and stimulates the formation of blood vessels [64,65].Decreased immunoreactivity of VEGF is reported in the FGR lamb brain with reduction in blood vessel density in brain injury regions [18].Therefore, varied levels of VEGF in the blood may indicate neurovascular disruption.Four studies have examined VEGF as a potential biomarker for FGR with reported decreased levels at varying time points for each study (see Table 4; [26][27][28]66]).One study observed decreases in VEGF levels in FGR neonates from birth to P5 [26] with these findings confirmed in other studies on P1 [27,66] and P7 [66].However, no differences were observed at later timepoints: P14, 21, and 28 [28,66].
Other vasculature markers may also indicate disruption to the BBB with a number of studies considering such markers as potential markers of the FGR phenotypical white matter and neuronal disruption [27,29,51,58,66].Specifically, increased levels of Troponin T have been reported in FGR umbilical cord blood at birth [58].Ang-1 levels were increased on P1 and 21 in FGR neonates but not on P7, 14, and 28 while Ang-2 levels were not altered at any of these timepoints [66].Plasminogen activator inhibitor-1, ischemia-modified albumin, phosphatidylinositol-glycan biosynthesis class F, and myeloperoxidase have also all been examined with no alterations observed at any of the timepoints considered in the respective studies [27-29, 51, 66].

Summary
Although some vasculature biomarkers show promise, there are many inconsistencies in the findings [27,66].The most consistent results across studies were decreased levels of VEGF [26][27][28]66].There is a pathophysiologic rationale to continue VEGF investigation in this population given multiple studies show abnormal VEGF levels in dysfunctional placental tissue [67].However, decreased levels of VEGF were evident mostly at early time points, with some studies not detecting any early changes past day of birth [27].Ultimately, the results indicate that vasculature related blood biomarkers such as VEGF and its receptors could be associated with indication of brain injury in FGR.Furthering our understanding of the potential mechanisms involved in causing these alterations in expression and their relationship with the pathophysiological neuronal and white matter injury of FGR would be an important area of future research.

Limitations
Few of the studies examined the same potential biomarkers and when they did, often results were conflicting.This may be due to the heterogeneity within and between cohorts -FGR v SGA, early-onset v late-onset FGR, GA, timing of sample collection, and cohort size.
Although there is consensus that FGR and SGA are distinct and different diagnoses, the studies considered here do not have consistent definitions of FGR.There have been attempts to standardize the definition of FGR [6].Unfortunately, none of the articles included here conformed to this definition although some used aspects of it [29, 41, 42, 52-54, 56, 58].Although only two studies report their cohorts as SGA or SFD [56,57], many studies have described the cohort as FGR but use a definition of birth weight <10th percentile for GA (online supplementary Table 1; for all online suppl.material, see https://doi.org/10.1159/000530492), which without further in utero measurements of growth, aligns more with a definition of SGA [26,39,40,51].Furthermore, FGR can be early-or lateonset which has different associated risks of adverse neurodevelopmental outcomes [68].Early-onset or symmetrical FGR, occurring prior to 32 weeks gestation, characteristically results in global growth restriction throughout the pregnancy and represents 20-30% of FGR fetuses [69].In contrast, late-onset or asymmetrical FGR occurs in 70-80% of FGR fetuses and results in fetal circulatory redistribution with preferential blood flow to the heart and brain compared with peripheral organs [69].This results in a decreased liver-to-brain ratio [7].Late-onset FGR is more likely to be associated with uteroplacental insufficiency from maternal factors such as preeclampsia or pregnancy induced hypertension.Although initially considered to be brain sparing, recent studies have shown worse neurodevelopmental outcomes in asymmetric FGR neonates compared with symmetric FGR cohorts [70][71][72][73].
GA is a factor which may impact the findings.Differences in GA were observed between studies and groups within studies (online suppl.Table 1) [26,29,41,52,54,57,58].For example, Yue et al. [26] investigated a number of biomarkers of interest; however, the median GA for FGR and non-FGR neonates was significantly different (29.9 weeks and 27.3 weeks, respectively).Furthermore, preterm FGR findings may be different from term FGR findings.This could be due to the differences in maturation and health status of the infant.Furthermore, comparing both preterm and term FGR introduces pathophysiological confounding factors due to the broad range and often undiagnosed reasons for prematurity.In addition, FGR pregnancies that make it to term are more likely to be associated with uteroplacental insufficiency than those born prior to 32 weeks.Among many other concerns, the inflammatory state associated with preterm birth [74] could particularly alter the circulating levels of inflammatory biomarkers.Therefore, GA-matching is crucial for appropriately identifying blood biomarkers for FGR.Further largescale studies accounting for GA and with only term cohorts would be beneficial.
Aside from GA, there are many other confounding variables that must be considered when comparing NG with FGR infants.Although most studies provided description for group differences for GA, maternal age, birth weight, mode of delivery and gender, very few studies provided details on confounding neonatal variables such as sepsis, mechanical ventilation, neonatal admission, or severity of illness.This information is important to understand the etiology and impact of FGR for each study and therefore interpret the findings.
The variation in and lack of details for sample types from these studies is another crucial limitation.Although some studies described how protein levels in samples were analyzed, not all studies provided details of this process.Further, sample type varied between studies and some studies included few details regarding sample type.For example, not all studies investigated umbilical cord blood, and if included, it was often not specified whether venous or arterial blood was sampled [26, 29, 39-41, 51, 52].Consistency in sample types and details on sampling methods are crucial for determining the best approach for future studies.
A further limitation is the relatively small cohort size in some studies (online suppl.Table 1) [29,39,41,42,51,53,56,58]. Without adequate power in such a heterogeneous population, it is difficult to draw definitive conclusions from reported results.Large cohort multi-institutional studies examining multiple biomarkers would be useful in the search for effective biomarkers to detect brain injury in FGR/SGA neonates.Furthermore, future research into the underlying pathophysiology of FGR through examining placental factors is also warranted.
A further limitation is the paucity of long-term neurodevelopmental follow-up in these studies.Mazarico et al. [62] reported a significant inverse relationship for all infants between both the concentrations of s100B and NSE with the results of the Bayley-III cognitive assessment at 2 years of age.However, no other studies examined in this review reported neurodevelopmental follow-ups.Examining associations of blood biomarkers with long-term neurodevelopmental follow-up would provide strong evidence for biomarkers of choice as predictors of adverse neurological outcomes in these infants.

Conclusions
Blood biomarkers are becoming an important clinical diagnostic tool for many conditions.This diagnostic tool would be particularly beneficial for earlier diagnosis of brain injury in FGR and SGA neonates.Current screening for FGR largely relies on ultrasounds during pregnancy and can be particularly effective at detecting early-onset FGR/SGA.However, especially in cases of late-onset FGR, many remain undiagnosed until close to birth.This diagnosis lacks further indication of risk of brain injury for the neonate.Despite the clinical importance of earlier detection of brain injury in FGR, consistent and reliable biomarkers are lacking.Even though FGR complications are partly driven by placental factors where the fetus is vulnerable to the chronic condition, this study is not looking at what causes it but how to detect brain injury.However, future research into placental factors associated with FGR brain injury would be beneficial.
With large variations in GA within and between studies, inconsistent definitions of FGR/SGA, small cohorts, as well as variable health status, it is difficult to draw conclusions on potential blood biomarkers for FGR/SGA reported in the current literature.The search for potential biomarkers of FGR and brain injury should continue.This combination would not only assist in consistent and early diagnosis of FGR but may also allow individualized detection of brain injury and open opportunities for targeted therapeutic neuroprotective trials as well as earlier access to neurodevelopmental support for the infant.Despite evidence indicating that FGR/SGA infants have long-term outcomes comparable to preterm infants [75] most FGR/SGA neonates (especially later gestation) are not routinely followed for neurodevelopmental outcomes (i.e., high-risk infant follow-up clinic).Blood biomarkers may help identify which of these babies need this kind of close follow-up and which may not.Although the blood biomarkers discussed here have been considered due to their potential association with brain injury, this is insufficient evidence of a correlation with risk of adverse neurodevelopmental outcomes in FGR/SGA infants.Future studies should assess neurodevelopmental outcomes to determine correlations with blood biomarkers of interest in FGR and SGA neonates.

Table 1 .
Summary of inflammatory biomarkers investigated in FGR/SGA neonates

Table 2 .
Summary of growth and developmental biomarkers investigated in FGR/SGA neonates

Table 3 .
Summary of structural and support proteins as biomarkers investigated in FGR/SGA neonates date; AGA, appropriate for gestational age; arrows indicate significant increases or decreases in FGR samples; NSE, neuron-specific enolase; GFAP, glial fibrillary astrocytic protein; AFP, alpha-foetoprotein.