Effects of Early Enteral to Parenteral Protein Ratios on Brain Volume and Somatic Growth in Very Low Birth Weight Infants

Abstract

Objective

To evaluate whether a higher proportion of enteral vs. parenteral protein (E:P ratio) in the first 28 days after birth is associated with increased brain volume and somatic growth in very low birth weight (VLBW; birth weight <1500g) infants.

Study design

This was a retrospective analysis of a sub-cohort of VLBW infants (N=256, gestational age mean 28.07 [SD 2.17] weeks, birth weight 1038.80 [SD 262.95] grams) from the Cincinnati Infant Neurodevelopment Early Prediction Study (CINEPS), a regional prospective study of infants born at ≤32 weeks’ gestation. Brain MRI was obtained at term-equivalent age. Macronutrient intake and growth metrics for the first 28 days were collected retrospectively. The primary outcome was total brain tissue volume. The relationships between E:P ratio, total and regional brain tissue volumes, and somatic growth were analyzed by multivariable linear regression models; composite variables were used to adjust for potential confounders including pregnancy risk factors and initial severity of illness.

Results

Higher E:P ratio was associated with increased total brain tissue volume but was not associated with change in head circumference z-score. In secondary analyses, higher E:P ratio was associated with increased weight velocity. There were no significant associations between E:P ratio and change in weight or length z-scores or regional brain volumes.

Conclusions

Higher E:P ratio in the first 28 days was positively associated with total brain volume and weight gain. Promoting the provision of enteral over parenteral protein may improve brain and somatic growth in VLBW infants.

Preterm infants are vulnerable to neurodevelopmental impairment (NDI) as extensive brain growth and development takes place in the third trimester.(1) Reducing NDI is an important challenge facing modern neonatologists as more preterm infants survive but are at risk for cognitive, motor, and sensory impairment.(2–4) Since poor postnatal growth is associated with worse neurologic and cognitive outcomes, optimizing early nutrition has emerged as a potentially modifiable factor in efforts to minimize NDI.(5–9) Specifically, increased early nutrient intake is associated with higher intelligence quotient in childhood,(10, 11) improved scores on developmental assessments, (12, 13) and improved MRI markers of brain development outcome such as white matter maturation,(14) larger total brain and regional tissue volumes,(15–17) and reduced brain injury.(18)

Early protein intake is particularly important for neurodevelopment because adequate protein is required to act as a substrate for neuronal growth and differentiation.(19) Animal models have demonstrated that protein-malnourished animals have reduced brain size, synapse and vesicle number, and dendritic complexity.(20–22) However, multiple studies aimed at maximizing early protein intake in premature infants have shown inconsistent effects on growth parameters and neurodevelopmental outcome.(23–30)

It may be that parenteral protein has fewer beneficial effects on neurodevelopmental outcome than enteral protein in very low birth weight (VLBW; birth weight <1500) preterm infants. This study aimed to investigate the relationship between a novel metric of the relative proportion of enteral vs. parenteral protein ratio (E:P ratio) in the first 28 days after birth with somatic growth metrics and brain tissue volumes on brain MRI at term-equivalent age (TEA). We hypothesized that higher E:P ratio would be associated with improved total brain volume and somatic growth.

METHODS

Study Design and Participants

This is a retrospective study of a sub-cohort of 269 VLBW (birth weight <1500 grams) premature infants enrolled in the Cincinnati Infant Neurodevelopment Early Prediction Study (CINEPS).(31) CINEPS is a multisite, prospective regional cohort of infants born preterm (≤32 weeks gestational age) at one of five neonatal intensive care units (NICUs) in the greater Cincinnati area from September 2016 to November 2019. Infants were included in this sub-cohort if they had a birth weight <1500 grams. Infants were excluded from CINEPS if they had known congenital anomalies affecting the neurologic systems, cyanotic heart disease, if they were on mechanical ventilation with greater than 50% oxygen at 45 weeks postmenstrual age (due to inability to tolerate transport to the MRI scanner), and if they had congenital anomalies of the gastrointestinal system affecting the ability to feed. The Cincinnati Children’s Hospital Medical Center (CCHMC) Institutional Review Board (IRB) approved this study, and the review boards of the regional hospitals approved the study based on established collaboration agreements. Parents provided written informed consent for inclusion in the CINEPS cohort. The retrospective collection of nutrition and growth data for this sub-cohort study was conducted under a waiver of consent.

Data Collection

Clinical data relating to maternal characteristics, pregnancy and birth history, neonatal characteristics, and NICU course were prospectively collected, as previously described.(31)

Total energy and individual macronutrient intake for the first 28 days after birth were collected retrospectively from the medical record and separated into enteral and parenteral components. In our regional network, enteral feedings are advanced using the same standardized protocol for VLBW infants at all sites.(32) This protocol initiates enteral feeds of maternal or donor breast milk within 48 hours of birth at 15 mL/kg/day for three days and subsequently advances by 20 mL/kg/day to a goal of 160 mL/kg/day, holding for a day at 75 mL/kg/day to fortify to 24 kcal/ounce. Infants typically achieve full enteral feeding volume by 10 to 14 days. There are variations between centers regarding products used for fortification and addition of liquid protein fortifier to donor breast milk. Parenteral protein is initiated on the day of birth; however, centers differ on starting dose and advancement. Actual enteral intake was collected from medical record flowsheets and macronutrient intake was calculated using published nutritional contents from the product manufacturers and estimated nutritional contents of maternal and donor human milk.(33–35) Parenteral nutrient intake was obtained from daily pharmacy orders and actual intake volume from intake/output flowsheets. E:P ratio was calculated as the ratio of cumulative enteral protein to cumulative parenteral protein in the first 28 days. Total and enteral protein intake (g/kg/day) were calculated as the average of daily intake for the first 28 days after birth.

Growth metrics including weight, length, and head circumference (HC) at birth and at 28 days of age were obtained from the medical record. Z-scores were calculated based on Fenton growth curves.(36) Weight gain velocity was calculated in g/kg/day using the exponential method.(37) Change in weight, length, and HC z-scores were calculated as the difference between z-scores at birth and 28 days. A negative z-score change indicates a decline and a positive z-score change indicates an increase in anthropometric status.(33)

MRI of the brain was obtained without sedation at term-equivalent age (between 39 and 44 weeks PMA) with a 3T Philips Ingenia scanner and 32-channel head coil at Cincinnati Children’s Hospital as previously described. (31) MRI variables were as follows: Axial T2-weighted image variables: echo time (TE) 166 milliseconds, repetition time (TR) 18567 milliseconds, flip angle (FA) 90°, and voxel dimensions 1.0×1.0×1.0 mm; MPRAGE T1-weighted images (Three-dimensional fast field echo): TR/TE/inversion time = 8.5/3.4/1610 milliseconds, FA 13°, in-plane resolution = 1 × 1 × 1 mm; susceptibility-weighted imaging: TE 5.4 milliseconds, TE 55 milliseconds, FA 13°, and voxel dimensions 1.0 × 1.0 × 1.5 mm. Total brain tissue and regional volumes were obtained by processing infant’s T2-weighted scans using the automated Developing Human Connectome Project (dHCP) pipeline, as previously described.(38, 39) All processed images were visually inspected for segmentation accuracy.

Composite variables were used to control for several potential confounders. The Clinical Risk Index for Babies (CRIB-II) score, a validated tool to predict initial risk of mortality in low birth weight babies that includes biological sex, gestational age, birth weight, initial base excess/deficit, and temperature at admission, was used to control for severity of illness.(40) High risk socioeconomic status (SES) was a composite of family structure, caregiver education and occupation, household income, primary language, and maternal age. Abnormal early head ultrasound included grade 3–4 intraventricular hemorrhage, cystic and non-cystic periventricular leukomalacia, cerebellar hemorrhage or atrophy, and white matter echolucencies.

Statistical Analysis

The primary outcome was total brain tissue volume on MRI at TEA. A priori selected regional brain tissue volumes including cerebellum, brain stem, cortical and deep gray matter, white matter, hippocampus, and amygdala were analyzed as secondary outcomes. Change in weight, length, and HC z-scores and weight gain velocity were analyzed as secondary outcomes. The primary exposure was E:P ratio. Total protein and enteral protein intake were secondary exposures. Multivariable linear regression models controlling for known or potential confounders were used to examine these associations. The models used to examine total and regional brain tissue volumes were controlled for CRIB-II, hypertensive disorders of pregnancy (HDP), smoking in pregnancy, histologic chorioamnionitis, antenatal corticosteroids, postmenstrual age at MRI, SES, abnormal early head ultrasound, birth hospital, and total protein. Models examining growth metrics were adjusted for CRIB-II, HDP, smoking in pregnancy, birth hospital, and total protein intake. These confounders were selected a priori based on literature review and biological plausibility of factors that could influence brain development or somatic growth, respectively.(18, 41–44) SPSS Statistics (v. 28.0.1.1, IBM) was used for descriptive statistics and SAS version 9.4, SAS Institute Inc. Cary, NC, USA was used for all other analyses. All regression model results are presented as β-estimates and 95% confidence intervals (CI) and two-sided tests with P value < 0.05 considered as statistically significant.

RESULTS

Of the 269 VLBW infants in the CINEPS cohort, 256 (95%) had complete nutrition data available for analysis. Relevant maternal, perinatal, and NICU course characteristics, collected prospectively, are described in Table I. Mean (SD) gestational age was 28.07 (2.17) weeks, mean (SD) birth weight was 1038.80 (262.95) grams, and 55% of the VLBW cohort was biologically male. Univariate linear regressions for the association between E:P ratio and baseline clinical characteristics are shown in Table II.

Table 1.

Baseline Characteristics of Mothers and their VLBW Infants

Clinical Characteristics (N=256)	Mean (SD) or N (%)
Infant Characteristics
Gestational age (weeks)	28.07 (2.17)
Male sex	139 (55%)
Singleton	174 (68%)
Birth weight (g)	1038.80 (262.95)
Birth weight z-score*	−0.13 (0.90)
Birth length z-score*	−0.13 (1.10)
Birth head circumference z-score*	−0.20 (1.23)
Small for gestational age	23 (9%)
CRIB-II score**	7.91 (3.25)
Abnormal early head ultrasound	28 (11%)
Surgical necrotizing enterocolitis or spontaneous intestinal perforation	8 (3%)
Late onset sepsis	35 (14%)
Severe bronchopulmonary dysplasia	62 (24%)
Postnatal steroids for bronchopulmonary dysplasia	39 (15%)
Prenatal/Maternal Characteristics
Antenatal steroid exposure (at least one dose)	240 (94%)
Maternal high risk socioeconomic status	47 (18%)
Maternal smoking	38 (15%)
Maternal chorioamnionitis	85 (33%)
Maternal hypertensive disorders of pregnancy	113 (44%)

^*Growth z-scores based on Fenton growth chart

^**Clinical Risk Index for Babies

Table 2.

Univariate Linear Regression for the Association Between E:P Ratio and Baseline Clinical Characteristics

Variables	β-estimate (95% CI)	Lower	Upper	p-value
Gestational age (weeks)	0.281	0.214	0.347	<0.0001
Male sex	−0.423	−0.746	−0.099	0.011
Birth weight (grams)	0.002	0.001	0.003	<0.0001
CRIB-II score*	−0.199	−0.024	−0.155	<0.0001
Birth hospital	0.056	−0.071	0.183	NS
Abnormal early head ultrasound	−1.095	−1.596	−0.595	<0.0001
Surgical necrotizing enterocolitis or spontaneous intestinal perforation	−2.730	−3.598	−1.862	<0.0001
Late onset sepsis	−1.392	−1.831	−0.952	<0.0001
Severe bronchopulmonary dysplasia	−1.018	−1.375	−0.660	<0.0001
Maternal high risk socioeconomic status	−0.117	−0.542	0.308	NS
Maternal smoking	0.090	−0.366	0.545	NS
Maternal chorioamnionitis	0.283	−0.060	0.627	NS
Maternal hypertensive disorders of pregnancy	−0.314	−0.638	0.011	NS

^*Clinical Risk Index for Babies

Nutritional Intake and Growth Metrics in First 28 Days After Birth

Nutrition data and growth in the first 28 days after birth are summarized in Table III. The mean (SD) E:P ratio was 2.7 (1.3). The distribution of E:P ratios was not normally distributed, as shown in Figure 1, with range 0.001 to 5.92 and median (interquartile range) 2.99 (1.96–3.73). Mean (SD) average daily total protein intake was 4.0 (0.4) g/kg/day. Mean (SD) average enteral daily protein intake was 2.6 (0.9) g/kg/day. The mean (SD) number of TPN days was 14.1 (5.2), and 195 (76%) of babies received less than 2 weeks of TPN. Nearly all (91%) infants received all human milk (maternal or donor) for enteral feeding in the first 28 days. Of the infants with any exposure to formula, the majority (66%) received formula only at day 26 or greater. At term equivalent (120 days after birth or at discharge) mean (SD) weight was 3018.72 (750.30) grams, length was 46.31 (3.57) centimeters, and head circumference was 33.20 (2.30) centimeters. Mean (SD) z-scores at term equivalent were −0.60 (1.03) for weight, −1.44 (1.23) for length, and −0.66 (1.19) for head circumference.

Table 3.

Nutritional Intake and Growth in First 28 Days after Birth

Nutritional Intake and Growth	Mean (SD) or N (%)
E:P ratio	2.72 (1.31)
Total protein intake (g/kg/day)	4.01 (0.42)
Enteral protein intake (g/kg/day)	2.60 (0.88)
Total energy intake (kcal/kg/day)	106.38 (11.95)
Enteral energy intake (kcal/kg/day)	76.61 (23.35)
Total parenteral nutrition days	14.09 (5.22)
Total parenteral nutrition ≤2 weeks	195 (76%)
Change in weight z-score*	−0.53 (0.53)
Change in length z-score*	−0.84 (0.66)
Change in head circumference z-score*	−0.87 (0.90)
Weight gain velocity (g/kg/day)	14.08 (3.93)

^*Growth z-scores based on Fenton growth chart

Figure 1.

Distribution of E:P ratios.

Protein Intake, Brain MRI Tissue Volumes, and Growth Metrics

As shown in Table IV, higher E:P ratio was positively associated with the primary outcome measure, total brain tissue volume, on multivariable regression analysis (β-estimate: 5.61 mm³ (95% CI: 0.13, 11.09)). E:P ratio was not significantly associated with regional brain tissue volumes (results not presented). In secondary analyses, enteral, 7.58 (1.30, 13.86), but not total (enteral + parenteral) protein intake, 10.73 (−2.20, 23.66) was significantly associated with total brain tissue volume (Tables V and VI).

Table 4.

Multivariable Linear Regression for the Association Between E:P Ratio and Total Brain Tissue Volume (mm³⁾

Variables	β-estimate (95% CI)	Lower	Upper	p-value
E:P Ratio	5.61	0.13	11.09	0.05
CRIB-II score*	−3.86	−5.77	−1.94	0.0001
Postmenstrual age at MRI	9.91	6.29	13.54	<0.0001
High risk socioeconomic status	−16.09	−28.49	−3.70	0.01
Smoking in pregnancy	−5.73	−19.68	8.23	0.42
Antenatal steroids	2.67	−9.58	14.91	0.67
Chorioamnionitis	0.55	−10.91	12.02	0.92
Abnormal early head ultrasound	0.79	−17.28	18.86	0.93
Hypertensive disorders of pregnancy	−6.26	−16.77	4.25	0.24
Birth hospital 1	5.57	−6.73	17.87	0.37
Birth hospital 2	−4.43	−26.31	17.45	0.69
Birth hospital 3	17.89	−9.47	45.25	0.20
Birth hospital 4	3.72	−13.06	20.50	0.66
Total protein (g/kg/day)	0.99	−15.24	17.21	0.91

^*Clinical Risk Index for Babies (CRIB-II) score, a validated tool to predict initial risk of mortality in low birth weight babies that includes biological sex, gestational age, birth weight, initial base excess/deficit, and temperature at admission

Table 5.

Multivariable Linear Regression Models for the Associations Between Total Protein Intake (g/kg/day) and Primary and Secondary Outcomes of Interest

Outcome Measure	β-estimate (95% CI)	p-value
Primary Outcome
Total brain tissue volume (mm3)	10.73 (−2.20, 23.66)	0.10
Secondary Outcomes
Change in weight z-score*	0.01 (−0.16, 0.17)	0.91
Change in length z-score*	0.25 (0.04, 0.47)	0.02
Change in head circumference z-score*	−0.30 (−0.59, −0.01)	0.04
Weight gain velocity (g/kg/day)	0.38 (−0.81, 1.58)	0.53

^*Growth z-scores based on Fenton growth chart

All models were adjusted for Clinical Risk Index in Babies score, hypertensive disorders of pregnancy, smoking in pregnancy, birth hospital, and total protein intake.

Total brain tissue volume was additionally adjusted for postmenstrual age at MRI, high risk socioeconomic status, chorioamnionitis, and abnormal early head ultrasound.

Table 6.

Multivariable Regression Analyses for the Association Between Enteral Protein Intake (g/kg/day) and Primary and Secondary Outcomes of Interest

Outcome Measure	β-estimate (95% CI)	p-value
Primary Outcome
Total brain tissue volume (mm3)	7.58 (1.30, 13.86)	0.02
Secondary Outcomes
Change in weight z-score*	−0.02 (−0.10, 0.06)	0.67
Change in length z-score*	0.13 (0.03, 0.24)	0.01
Change in head circumference z-score*	−0.10 (−0.25, 0.04)	0.16
Weight gain velocity (g/kg/day)	0.08 (−0.50, 0.66)	0.80

^*Growth z-scores based on Fenton growth chart

All models were adjusted for Clinical Risk Index in Babies score, hypertensive disorders of pregnancy, smoking in pregnancy, birth hospital, and total protein intake.

Total brain tissue volume was additionally adjusted for postmenstrual age at MRI, high risk socioeconomic status, chorioamnionitis, and abnormal early head ultrasound.

In secondary analyses, higher E:P ratio was significantly associated with increased weight gain velocity on multivariable regression analysis, 0.74 (0.24, 1.24). E:P ratio was not significantly associated with change in weight, length, or HC z-score (see Table VII). In analysis of secondary exposure variables, higher total protein intake was significantly associated with a negative change in HC z-score, −0.30 (−0.59, −0.01) and positive change in length z-score, 0.25 (0.04, 0.47) (Table V). In addition, higher enteral protein intake was significantly associated with positive change in length z-score, 0.13 (0.03, 0.24) (Table VI).

Table 7.

Multivariable Regression Analyses for the Association Between E:P Ratio and Secondary Outcome Measures for Somatic Growth

Outcome Measures	β-estimate (95% CI)	p-value
Change in weight z-score*	0.06 (−0.01, 0.13)	0.09
Change in length z-score*	0.05 (−0.04, 0.142)	0.27
Change in head circumference z-score*	0.06 (−0.06, 0.19)	0.31
Weight gain velocity (g/kg/day)	0.74 (0.24, 1.24)	<0.01

^*Growth z-scores based on Fenton growth chart

All models were adjusted for Clinical Risk Index in Babies score, hypertensive disorders of pregnancy, smoking in pregnancy, birth hospital, and total protein intake.

DISCUSSION

This study investigates the nutritional E:P ratio, which is a measure of the relative proportion of enteral vs. parenteral protein intake. It is logical to investigate enteral and parenteral protein as a proportion rather than as separate variables as done in previous studies, given that all VLBW preterm infants will receive some amount of nutrition via each of these routes during their NICU course. In this sub-cohort of VLBW preterm infants, higher E:P ratio in the first 28 days after birth was associated with larger total brain tissue volume on brain MRI at term-equivalent age. Higher E:P ratio was also associated with improved weight gain velocity.

Optimizing early nutrition is an important strategy to minimize NDI in VLBW preterm infants, with early protein intake noted as particularly important for brain growth and development in prior animal and human studies. (5–9, 19–22) However, studies aimed at maximizing early protein intake in premature infants have shown inconsistent effects on growth parameters and neurodevelopmental outcome.(23–30) Previous studies primarily focused on total protein intake or exclusively enteral or parenteral protein, and were not designed to investigate the effect of the relative amount of enteral vs. parenteral protein, despite the fact that infants in the NICU receive protein by both routes in varying quantities.

It may be that the benefits of early protein intake vary by route, with parenteral intake less beneficial than enteral. In some studies, enteral intake was positively associated with brain growth, white matter maturation, and developmental outcome, but parenteral intake was not.(16, 29, 45) Longer duration of parenteral nutrition has also been associated with smaller total brain volume, cerebellar volume, and cortical gray matter volume on brain MRI at term-equivalent age.(46, 47)

Researchers have hypothesized that the potential differential effects of enteral vs. parenteral protein are confounded by the fact that preterm infants who have greater enteral intake are typically less sick overall. However, prior studies have not controlled for a validated marker of severity of clinical illness.(16, 45) This hypothesis is also not supported by an animal model investigating the effects of enteral vs. parenteral protein on growth and brain development. In a study of otherwise healthy preterm pigs given ten days of exclusively parenteral or exclusively enteral protein, the pigs given exclusively parenteral nutrition had significantly smaller brain volumes, reduced motor activity, and underdeveloped myelination as measured by diffusion tensor imaging (DTI) despite comparable gains in body weight in both groups.(48) This animal model suggests that intravenous amino acids may be suboptimal substrates for brain growth compared with enteral protein, even in the absence of confounding severe illness.

Previous clinical studies of the relationship between protein intake and brain growth and development have showed mixed results, however the majority of these studies focused on total protein intake or early parenteral protein.(23–30) Prolonged TPN has been associated with smaller brain tissue volumes on MRI, suggesting that maximizing early parenteral protein intake may not be the optimal method to provide substrate for brain growth in preterm infants.(46, 47) All VLBW and very preterm infants will receive some parenteral nutrition, but increasing the proportion of early protein that is provided enterally instead of parenterally may be associated with better brain growth. In this study, higher E:P ratio was associated with larger total brain tissue volume on MRI at TEA. In addition, enteral but not total average daily protein intake in the first 28 days after birth was associated with larger total brain tissue volume on MRI at TEA. Taken together, these results indicate that early enteral protein may be more beneficial for total brain growth than parenteral protein intake.

Early protein intake by both enteral(12) and parenteral(27) routes has also been associated with improved in-hospital growth. In our sub-cohort, mean protein intake (4.01g/kg/day, SD 0.42) was in line with current nutrition recommendations from the American Academy of Pediatrics and the European Society of Paediatric Gastroenterology, Hepatology and Nutrition Committee on Nutrition.(49, 50) This is unlike some studies relating early protein, growth, and brain development in which intake was less than recommended by current guidelines, limiting the generalizability of results of prior studies.(13, 15, 27, 47) In our study, higher E:P ratio in the first 28 days after birth was associated with higher weight gain velocity. Higher total and enteral average daily protein intake were also associated with positive change in length z-score on secondary analyses, meaning an increase in length z-score from birth to 28 days. This finding is in keeping with previous studies that have shown protein to be particularly important for linear growth.(51, 52) E:P ratio was not significantly associated with change in HC z-score, and higher average daily total protein intake was associated with a negative change in HC z-score. Accurate HC measurements can be used as a proxy for brain size on MRI,(53) and higher E:P ratio was associated with larger total brain tissue volume in this cohort as discussed previously. Therefore, this unexpected result for association between HC and E:P ratio and total protein is possibly due to variability in measurement technique that limited HC accuracy in this sub-cohort, as measurements were collected retrospectively from chart review of clinical data and were not obtained using standardized approaches for research.

Our results are strengthened by the large cohort, which allowed us to control for multiple confounding variables in our multivariable linear regression model. As enteral intake may be limited by a patient’s clinical status, our study is also strengthened by controlling for a validated measure of severity of illness, CRIB-II score, in an effort to account for critical illness that may lead to differences in enteral vs. parenteral intake. This method controls for severity of illness without controlling for morbidities that are themselves post-exposure outcomes, such as diagnoses of necrotizing enterocolitis (NEC) or bronchopulmonary dysplasia (BPD).

This study has several key limitations. This is an observational study, therefore residual confounding, mostly from unknown confounders, is always possible. There are multiple factors that affect clinical illness in the NICU and a clinician’s assessment of safety to feed, and the CRIB-II score may not have adequately captured a patient’s true severity of illness. However, the rate of NEC was very low in our cohort, so the number of patients who were unable to feed for extended periods of time was likely minimal. This is further supported by 76% of the cohort having 2 weeks or less of TPN, indicating ability to complete the standard feeding protocol used in the centers in this study.

The study is also limited by the retrospective collection of growth and nutrition data from the medical record, as it is possible that some data were charted incompletely or inaccurately. Nutritional intake calculations were based on assumptions about nutrient contents of maternal and donor milk based on established guidelines, as breastmilk analysis data were not available. It is possible that the true nutrient composition of human milk given to patients in the cohort was higher or lower than assumed, or that there were benefits of human milk in addition to their specific nutrient components. The vast majority (91%) of infants in this study received human milk for the full 28-day study time period, and the majority (66%) of those who received formula did so only for the last 2–3 days of the time period of nutrition data collection in this study, so any additional benefits of human milk were likely similar across the cohort. We plan future studies to investigate the associations between growth, brain development, and protein in maternal vs. donor human milk specifically to address this limitation.

In conclusion, our results suggest that providing nutrition enterally in the first 28 days after preterm birth may be important for both brain and somatic growth. Future analyses will also investigate if higher E:P ratio and enteral protein are associated with advanced MRI brain connectivity measures and neurodevelopmental outcome at 2 years corrected age. If these results hold in future analyses, they will highlight the importance of providing nutrition via enteral routes and facilitating earlier transition off parenteral nutrition. This may be an important strategy for neonatologists aiming to reduce NDI in VLBW preterm infants.

ABBREVIATIONS:

CINEPS

Cincinnati Infant Neurodevelopment Early Prediction Study

CRIB-II

Clinical Risk Index for Babies

E:P ratio

Enteral vs. parenteral protein ratio

HC

Head circumference

HDP

Hypertensive disorders of pregnancy

IRB

Institutional review board

MRI

Magnetic resonance imaging

NDI

Neurodevelopmental impairment

NEC

Necrotizing enterocolitis

NICU

Neonatal intensive care unit

SD

Standard deviation

SES

Socioeconomic status

SIP

Spontaneous intestinal perforation

TEA

Term equivalent age

TPN

Total parenteral nutrition

VLBW

Very low birth weight