Enteral nutrition with unfortified human milk during the first 2 postnatal weeks often leads to cumulative protein and energy deficits among preterm infants. Fortified human milk administered soon after birth could increase fat-free mass (FFM) and improve growth in these infants.
This was a masked, randomized trial. Starting on feeding day 2, extremely preterm infants 28 weeks or younger fed maternal or donor milk were randomized to receive either a diet fortified with a human-based product (intervention group) or a standard, unfortified diet (control group). This practice continued until the feeding day when a standard bovine-based fortifier was ordered. Caregivers were masked. The primary outcome was FFM-for-age z score at 36 weeks of postmenstrual age (PMA).
A total of 150 infants were randomized between 2020 and 2022. The mean birth weight was 795±250 g, and the median gestational age was 26 weeks. Eleven infants died during the observation period. The primary outcome was assessed in 105 infants (70%). FFM-for-age z scores did not differ between groups. Length gain velocities from birth to 36 weeks PMA were higher in the intervention group. Declines in head circumference-for-age z score from birth to 36 weeks’ PMA were less pronounced in the intervention group.
In infants born extremely preterm, human milk diets fortified soon after birth do not increase FFM accretion at 36 weeks’ PMA, but they may increase length gain velocity and reduce declines in head circumference-for-age z scores from birth to 36 weeks’ PMA.
Partial enteral nutrition with unfortified human milk during the first 2 postnatal weeks can be insufficient to prevent cumulative nutritional deficits in infants born extremely preterm (28 weeks of gestation or less).
This trial indicates that early human milk fortification increases length gain velocity and reduces declines in head circumference z scores from birth to term equivalent age among infants born extremely preterm, but it appears not to increase fat-free mass accretion.
The first 2 postnatal weeks delineate a period of a great opportunity to prevent energy and protein deficits in critically ill infants born extremely preterm at 28 weeks of gestation or less.1,2 During this critical period for growth and development, provision of protein-enriched human milk diets could prevent nutritional deficits and mitigate the effect of acute critical illness on the risk of adverse growth outcomes in these infants.3
A sufficient intake of enteral protein during the first 2 postnatal weeks could increase protein synthesis, prevent excessive weight loss, and promote growth.2,4 Because fat-free mass (FFM) accounts for muscle, bone, brain, and other tissues built on a protein matrix, a protein-enriched human milk diet initiated soon after birth could also favor FFM accretion.5 In recent years, several studies that measured body composition with noninvasive methods such as air-displacement plethysmography (ADP)6,7 have shown that most extremely preterm infants reach term-equivalent age with decreased FFM. Observational studies have also reported that FFM gains may lower the risk of long-term complications such as neurodevelopmental impairment,8 obesity, and chronic diseases in these infants,9–12 but previous randomized trials have not reported the effects of early human milk fortification on FFM accretion.13,14
It is necessary to determine the effectiveness of human milk fortification during the phase of acute critical illness among extremely preterm infants. To assess the effect of improved human milk diets on the growth of preterm infants, we conducted this double-masked, randomized controlled trial to test the hypothesis that human milk diets fortified with a human milk–based product soon after birth increase FFM accretion in infants born extremely preterm.
Patients and Methods
The Increased Milk Protein to Accrue Critical Tissue (IMPACT) trial was a parallel group, masked, randomized controlled trial with a 1:1 allocation ratio. Extremely preterm infants with a gestational age of 28 weeks or less admitted to the neonatal unit at the University of Alabama at Birmingham Hospital were eligible. Infants with major congenital anomalies and infants with a terminal illness in whom decisions to withhold or limit life support have been made were excluded. The trial was registered in ClinicalTrials.gov as NCT04325308 on March 27, 2020. The University of Alabama at Birmingham institutional review board approved this trial in July 2020 (IRB-300005089).
We screened all extremely preterm infants admitted to our neonatal unit to determine trial eligibility. To enable treatment allocation before or on feeding day 2, we obtained written parental informed consent soon after birth, preferentially within the first 48 hours after birth. We assigned infants to 1 of the study groups using computer-generated random-block sequences and numbered, opaque, sealed envelopes. We randomized twins individually. Nutrition room staff not involved in patient care opened the sealed envelopes in sequential order after receiving notification of informed consent, allocated the study intervention, and dispensed feeding syringes with labels that did not show the treatment allocation. Thus, we masked the intervention to clinicians, parents, and outcome evaluators.
Extremely preterm infants fed human milk were randomized to receive either a fortified human milk diet (intervention group) or a usual, unfortified human milk diet (control group) within the first 96 hours after birth. Infants in the intervention group received unfortified maternal or donor milk on feeding day 1 (within the first 96 hours after birth). On feeding day 2, a human milk–based fortifier that increases the energy and protein content of human milk was added (Prolact, Prolacta Bioscience, Inc, City of Industry, CA). Adding this product to human milk increased the caloric density from 20 to 24 kcal/oz and the protein content by approximately 1.2 g/dL. This practice continued until the day when a standard bovine-based human milk fortifier was ordered (Enfamil Liquid Human Milk Fortifier High Protein, Mead Johnson Nutrition, Evansville, IN). Infants in the control group received maternal or donor milk from feeding day 1 until the feeding day on which the standard bovine-based human milk fortifier was ordered.
This trial compared the 2 human milk diets during the first 2 postnatal weeks under ordinary clinical circumstances without trying to control other interventions strictly. Therefore, all other intensive care and nutrition aspects were provided at clinicians’ discretion. The feeding protocol in our neonatal unit recommends early administration of oropharyngeal colostrum; administration of enteral feeds as intermittent bolus gavage every 3 hours; initiation of enteral feeds with either maternal or donor milk via orogastric tube within the first 96 hours after birth with 20 to 25 mL/kg/d; progression of enteral feeds with daily increments of 20 to 25 mL/kg/d on feeding day 2 until full enteral nutrition is established (>120 mL/kg/d); and addition of bovine-based fortifiers at approximately postnatal day 14 after full enteral nutrition has been established. In our unit, if maternal milk supply is insufficient, infants receive donor milk as an alternative until 32 or 33 weeks of postmenstrual age (PMA). Subsequently, they receive formula if their mothers can no longer supply their milk.
The primary efficacy outcome was FFM-for-age z score at 36 weeks’ PMA or hospital discharge (whichever occurred first) using ADP (PeaPod, Cosmed USA, Concord, CA). Body composition measurements were performed at 36 weeks’ PMA or hospital discharge if an infant no longer required significant respiratory support with mechanical ventilation, continuous positive airway pressure, or high-flow nasal cannula. Body composition measurements were converted into z score values using updated, sex-specific reference curves of body composition in preterm infants.15 Anthropometric measurements from birth to 36 weeks’ PMA were converted into z score values using the Fenton growth curves. Secondary efficacy outcomes included significant weight loss during the first 14 days after birth (decline in weight-for-age z score from birth to 14 days >0.8),16 weight gain velocity in grams per kilogram per day from birth to 36 weeks’ PMA calculated using the exponential method, postnatal growth failure (weight <10th centile at 36 weeks’ PMA), moderate to severe malnutrition (decline in weight-for-age z score from birth to 36 weeks’ PMA >1.2),17 FFM in kilograms and percentage at 36 weeks’ PMA, fat mass in kilograms and percentage at 36 weeks’ PMA, and anthropometric measurements (weight, head circumference, and length) at 36 weeks’ PMA. Length at 36 weeks’ PMA was measured with length boards. Head circumference was measured with a flexible tape measure.
The primary safety outcomes included spontaneous intestinal perforation (SIP), necrotizing enterocolitis (NEC) stage 2 or 3, and death. A data safety and monitoring committee examined individual infant data at 25% and 50% enrollment to rule out the possibility of a temporal association between the study intervention and the primary safety outcomes.
Sample Size
We used data from a previous enteral feeding trial to calculate the sample size for this trial.5 To detect a 0.5 difference in FFM-for-age z scores between groups with SD of 1, 0.05 level of significance, and 80% power for a t test that compares means from 2 independent samples, we estimated that a sample size of 126 patients would be necessary for this superiority trial. Anticipating that approximately 20% of study participants would be lost to follow-up for assessment of the primary outcome at 36 weeks’ PMA, we added 12 patients to each group and increased the sample size to 150, 75 patients in each group (n = 150).
Statistical Analyses
This trial recorded core data on nutrition as recommended by consensus groups.18 We used mean and SD or median and interquartile ranges (IQR) to summarize continuous variables. To summarize categorical variables, we used frequencies and proportions. We compared categorical variables with either χ2 or Fisher exact tests and continuous variables with t test. We estimated mean differences, relative risks, and their respective 95% confidence intervals to report differences between groups. For analysis of the primary outcome, an unadjusted t test comparison of the mean FFM-for-age z score between control and intervention groups was performed, assuming a normal distribution of this variable.19,20 All trial outcomes were analyzed using JMP Pro 16.1 following the intention-to-treat principle.18
Results
Of 230 extremely preterm infants assessed for eligibility between August 2020 and October 2022, 150 were randomized (Fig 1). Baseline characteristics of the study participants are shown in Table 1. The mean birth weight was 795 g (SD, 250), and the median gestational age was 26 weeks (IQR, 24-27). Of 150 infants included, 31 had a gestational age of 23 weeks or less (21%). Approximately one-half of infants were of non-Hispanic Black race/ethnicity. Nearly 90% of infants were exposed to at least 1 dose of antenatal steroids. The median postnatal age at consent was 48 hours (IQR, 24-72).
Enrollment, randomization, and outcomes. NEC, necrotizing enterocolitis; PMA, postmenstrual age; RA, room air; SIP, spontaneous intestinal perforation.
Enrollment, randomization, and outcomes. NEC, necrotizing enterocolitis; PMA, postmenstrual age; RA, room air; SIP, spontaneous intestinal perforation.
Baseline Characteristics
| Characteristics | Intervention Group (n = 75) | Control Group (n = 75) |
|---|---|---|
| Birth weight in g, mean ± SD | 770 ± 254 | 820 ± 245 |
| Gestational age at delivery in wk, median (IQR) | 26 (24–27) | 26 (24–27) |
| Small for gestational age at birth, n (%) | 15 (20) | 8 (11) |
| Female sex, n (%) | 43 (57) | 29 (39) |
| Black race, n (%) | 45 (60) | 38 (51) |
| Apgar score at 5 min, median (IQR) | 8 (7–9) | 8 (7–9) |
| Exposure to antenatal steroids, n (%) | 66 (88) | 66 (88) |
| Postnatal age at initiation of enteral feeding in days, median (IQR) | 2 (2–3) | 2 (2–3) |
| Postnatal age at full enteral feeding (120 mL/kg/d) in days, median (IQR) | 8 (7–12) | 8 (7–11) |
| Postnatal age at initiation of bovine-based fortifier in days, median (IQR) | 15 (14–20) | 15 (12–18) |
| Characteristics | Intervention Group (n = 75) | Control Group (n = 75) |
|---|---|---|
| Birth weight in g, mean ± SD | 770 ± 254 | 820 ± 245 |
| Gestational age at delivery in wk, median (IQR) | 26 (24–27) | 26 (24–27) |
| Small for gestational age at birth, n (%) | 15 (20) | 8 (11) |
| Female sex, n (%) | 43 (57) | 29 (39) |
| Black race, n (%) | 45 (60) | 38 (51) |
| Apgar score at 5 min, median (IQR) | 8 (7–9) | 8 (7–9) |
| Exposure to antenatal steroids, n (%) | 66 (88) | 66 (88) |
| Postnatal age at initiation of enteral feeding in days, median (IQR) | 2 (2–3) | 2 (2–3) |
| Postnatal age at full enteral feeding (120 mL/kg/d) in days, median (IQR) | 8 (7–12) | 8 (7–11) |
| Postnatal age at initiation of bovine-based fortifier in days, median (IQR) | 15 (14–20) | 15 (12–18) |
IQR, interquartile range.
More than 80% of infants who participated in this trial achieved full enteral nutrition within the first 2 weeks after birth. Maternal milk intake did not differ between the intervention and control group at postnatal day 7 (91 ± 55 vs 96 ± 54 mL/kg/d; P = .58) or postnatal day 14 (127 ± 58 vs 135 ± 60 mL/kg/d; P = .37).
The primary outcome was measured in 105 of 150 infants randomized (70%) (Table 2). Approximately 90% of infants who did not require significant respiratory support with mechanical ventilation, continuous positive airway pressure, or high-flow nasal cannula at 36 weeks’ PMA had a body composition measurement. FFM-for-age z scores did not differ between groups. Length gain velocities from birth to 36 weeks’ PMA were higher in the intervention group (Table 3). Declines in head circumference-for-age z score from birth to 36 weeks’ PMA were less pronounced in the intervention group. Two post hoc exploratory analyses were performed to account for unexpected imbalances in baseline characteristics. One excluded small-for-gestational-age infants (Supplemental Table 4) and the other assessed the interaction between the study intervention and sex. The analysis adjusted for sex revealed that the direction and magnitude of the effect sizes reported in the primary analysis were not significantly different in female and male infants.
Body Composition Outcomes at 36 Weeks PMA
| Outcome | Intervention Group (n = 52) | Control Group (n = 53) | Mean Difference (95% CI) |
|---|---|---|---|
| FFM-for age z score, mean ± SD | −1.7 ± 1.4 | −1. 6 ± 1.5 | MD: −0.1 (−0.7 to 0.4); P = .67 |
| FFM in kg, mean ± SD | 2.13 ± 0.56 | 2.19 ± 0.55 | MD: −0.1 (−0.3 to 0.2); P = .60 |
| FFM in percentage, mean ± SD | 83 ± 4 | 82 ± 5 | MD: 0.5 (−1.2 to 2.2); P = .55 |
| FM-for-age z score, mean ± SD | 1.0 ± 1.1 | 1.3 ± 1.2 | MD: −0.2 (−0.7 to 0.2); P = .32 |
| FM in kg, mean ± SD | 0.47 ± 0.23 | 0.51 ± 0.30 | MD: 0 (−0.1 to 0.1); P = .42 |
| FM in percentage, mean ± SD | 17 ± 4 | 18 ± 5 | MD: 0.5 (−2.2 to 1.2); P = .55 |
| Outcome | Intervention Group (n = 52) | Control Group (n = 53) | Mean Difference (95% CI) |
|---|---|---|---|
| FFM-for age z score, mean ± SD | −1.7 ± 1.4 | −1. 6 ± 1.5 | MD: −0.1 (−0.7 to 0.4); P = .67 |
| FFM in kg, mean ± SD | 2.13 ± 0.56 | 2.19 ± 0.55 | MD: −0.1 (−0.3 to 0.2); P = .60 |
| FFM in percentage, mean ± SD | 83 ± 4 | 82 ± 5 | MD: 0.5 (−1.2 to 2.2); P = .55 |
| FM-for-age z score, mean ± SD | 1.0 ± 1.1 | 1.3 ± 1.2 | MD: −0.2 (−0.7 to 0.2); P = .32 |
| FM in kg, mean ± SD | 0.47 ± 0.23 | 0.51 ± 0.30 | MD: 0 (−0.1 to 0.1); P = .42 |
| FM in percentage, mean ± SD | 17 ± 4 | 18 ± 5 | MD: 0.5 (−2.2 to 1.2); P = .55 |
CI, confidence interval; FFM, fat-free mass; FM, fat mass; MD, mean difference.
Secondary Outcomes
| Outcome | Intervention Group (n = 65) | Control Group (n = 63) | RR or MD (95% CI) |
|---|---|---|---|
| Declines in weight-for-age z score from birth to postnatal day 14, mean ± SD | −0.8 ± 0.6 | −1.1 ± 0.6 | MD: 0.3 (0.1–0.5); P ≤ .01 |
| Declines in weight-for-age z score from birth to 36 wk PMA, mean ± SD | −1.0 ± 0.7 | −1.2 ± 0.8 | MD: 0.2 (–0.1 to 0.5); P = .13 |
| Declines in weight-for-age z score from birth to 36 wk PMA greater than 1.2, n (%) | 23 (35) | 30 (48) | RR: 0.74 (0.49–1.13); P = .16 |
| Postnatal growth failure (weight <10th percentile) at 36 wk PMA, n (%) | 31 (48) | 30 (48) | RR: 1.00 (0.70–1.44); P = .99 |
| Weight gain velocity in g/kg/d from birth to 36 wk PMA, mean ± SD | 14.7 ± 2.5 | 13.9 ± 2.5 | MD: 0.7 (–0.2 to 1.6); P = .11 |
| Length gain velocity in cm/wk from birth to 36 wk PMA, mean ± SD | 0.9 ± 0.2 | 0.8 ± 0.3 | MD: 0.1 (0–0.2); P = .04 |
| Declines in length-for-age z score from birth to 36 wk PMA, mean ± SD | −1.5 ± 1.0 | −1.9 ± 1.3 | MD: 0.3 (–0.1 to 0.7); P = .10 |
| Head circumference gain velocity in cm/wk from birth to 36 wk PMA, mean ± SD | 0.7 ± 0.1 | 0.7 ± 0.2 | MD: 0 (0–0.1); P = .14 |
| Declines in head circumference-for-age z score from birth to 36 wk PMA, mean ± SD | −0.9 ± 0.8 | −1.3 ± 1.1 | MD: 0.4 (0.1–0.8); P = .01 |
| NEC, n (%)a | 1 (1) | 3 (4) | RR: 0.33 (0.04–3.13); P = .31 |
| SIP, n (%)a | 5 (7) | 4 (5) | RR: 1.25 (0.35–4.47); P = .73 |
| Death, n (%)a | 6 (8) | 5 (7) | RR: 1.20 (0.38–3.76); P = .75 |
| NEC, SIP, or death, n (%)a | 9 (12) | 10 (13) | RR: 0.90 (0.39–2.09); P = .81 |
| Outcome | Intervention Group (n = 65) | Control Group (n = 63) | RR or MD (95% CI) |
|---|---|---|---|
| Declines in weight-for-age z score from birth to postnatal day 14, mean ± SD | −0.8 ± 0.6 | −1.1 ± 0.6 | MD: 0.3 (0.1–0.5); P ≤ .01 |
| Declines in weight-for-age z score from birth to 36 wk PMA, mean ± SD | −1.0 ± 0.7 | −1.2 ± 0.8 | MD: 0.2 (–0.1 to 0.5); P = .13 |
| Declines in weight-for-age z score from birth to 36 wk PMA greater than 1.2, n (%) | 23 (35) | 30 (48) | RR: 0.74 (0.49–1.13); P = .16 |
| Postnatal growth failure (weight <10th percentile) at 36 wk PMA, n (%) | 31 (48) | 30 (48) | RR: 1.00 (0.70–1.44); P = .99 |
| Weight gain velocity in g/kg/d from birth to 36 wk PMA, mean ± SD | 14.7 ± 2.5 | 13.9 ± 2.5 | MD: 0.7 (–0.2 to 1.6); P = .11 |
| Length gain velocity in cm/wk from birth to 36 wk PMA, mean ± SD | 0.9 ± 0.2 | 0.8 ± 0.3 | MD: 0.1 (0–0.2); P = .04 |
| Declines in length-for-age z score from birth to 36 wk PMA, mean ± SD | −1.5 ± 1.0 | −1.9 ± 1.3 | MD: 0.3 (–0.1 to 0.7); P = .10 |
| Head circumference gain velocity in cm/wk from birth to 36 wk PMA, mean ± SD | 0.7 ± 0.1 | 0.7 ± 0.2 | MD: 0 (0–0.1); P = .14 |
| Declines in head circumference-for-age z score from birth to 36 wk PMA, mean ± SD | −0.9 ± 0.8 | −1.3 ± 1.1 | MD: 0.4 (0.1–0.8); P = .01 |
| NEC, n (%)a | 1 (1) | 3 (4) | RR: 0.33 (0.04–3.13); P = .31 |
| SIP, n (%)a | 5 (7) | 4 (5) | RR: 1.25 (0.35–4.47); P = .73 |
| Death, n (%)a | 6 (8) | 5 (7) | RR: 1.20 (0.38–3.76); P = .75 |
| NEC, SIP, or death, n (%)a | 9 (12) | 10 (13) | RR: 0.90 (0.39–2.09); P = .81 |
CI, confidence interval; MD, mean difference; NEC, necrotizing enterocolitis; PMA, postmenstrual age; RR, relative risk; SIP, spontaneous intestinal perforation.
n = 150; 75 in the intervention group and 75 in the control group.
The risk of postnatal growth failure and the risk of moderate to severe malnutrition at 36 weeks’ PMA were not significantly lower in the intervention group. The risk of SIP, NEC, death, and the combined outcome of SIP, NEC, or death did not differ between groups (Table 3). The highest blood urea nitrogen values were observed at postnatal day 24 (95% confidence interval, 22-26 days). The highest blood urea nitrogen values during the first 50 days after birth did not differ between groups (33 vs 37 mg/dL; P = .37). None of the 4 cases of NEC occurred during the transition from a human-based to a bovine-based fortifier. Two infants unexposed to bovine-based fortifiers developed NEC before reaching full enteral feeding. The other 2 infants developed NEC after several weeks of exposure to bovine-based fortifiers.
Discussion
In this pragmatic randomized trial, we recruited a representative sample of infants born extremely preterm, assigned the study intervention randomly, masked clinicians to the study intervention, monitored anthropometric measurements at regular intervals until term-equivalent age, and compared the mean differences in FFM accretion between the 2 groups at term equivalent age. Our analysis indicated that FFM at term equivalent age did not differ between groups. We also found that early human milk fortification reduced declines in weight-for-age z scores from birth to postnatal day 14. Additionally, we established that early human milk fortification was associated with increased length gain velocities and reduced declines in head circumference-for age z scores from birth to term-equivalent age. To our knowledge, this is the largest randomized controlled trial of early human milk fortification that includes only extremely preterm infants, a high-risk population often underrepresented in enteral feeding trials.
This trial that provides the highest level of evidence for causality addresses 2 important limitations described in the 2 most recent meta-analyses of early fortification practices: the risk of bias from lack of masking and the imprecision in efficacy outcomes observed in previous trials with smaller sample sizes.13,14 Masked randomization ensured that the groups were similar in terms of known and unknown factors. Masking also avoided differential noncompliance and minimized surveillance and ascertainment biases. Our trial had a sample size with sufficient power to detect a 0.5 mean difference in z scores between groups, a growth outcome that could be considered clinically meaningful.
The main limitations of this trial were the insufficient power to assess the potential harms of early human milk fortification and the single-center design. This trial provided valuable safety and feasibility data regarding the practice of early human milk fortification with a human-based product and the subsequent transition to a bovine-based fortifier around postnatal day 14, but our sample size had insufficient power to detect significant differences in NEC between groups. Nevertheless, the number of NEC cases reported in this trial was similar to those reported in trials of early human milk fortification that included preterm infants born at older gestational ages.21,22 This finding could be attributed to the provision of exclusive human milk diets during the first 2 postnatal weeks and human milk diets up to 32 weeks’ PMA in all infants, a feeding practice that reduces the risk of NEC.23
The single-center design is a limitation because feeding practices may differ substantially across neonatal units. Based on current clinical evidence from meta-analyses of randomized trials designed to establish full enteral nutrition soon after birth,24,25 many neonatal units have already implemented recent recommendations regarding the early and rapid progression of enteral nutrition.1 In these units, full enteral nutrition is achieved within the first 2 weeks after birth in more than 80% of all extremely preterm infants.26,27 Both randomization groups in this trial experienced a rapid transition to full enteral nutrition and likely benefited from the early and rapid progression of enteral feeding volumes. This finding implies that the observed differences in weight, length, and head circumference following early human milk fortification were not related to fluid intake, but likely because of the early provision of additional enteral protein and renal solutes, such as sodium, chloride, calcium, and phosphorus, while establishing full enteral nutrition. In circumstances in which delays establishing full enteral nutrition are expected,21,27 particularly in growth-restricted infants,28 early human milk fortification could still favor FFM accretion. Maternal to donor milk feeding ratios may also differ across neonatal units. In infants predominantly fed donor milk, early human milk fortification with human-based products that increase energy and protein content could be more beneficial.29,30 Recent observational studies suggest that maternal milk produced during the first 14 days after preterm birth could meet energy and protein requirements without additional fortification in extremely preterm infants.31 Thus, future trials should carefully consider the issue of protein source and quality. Assuming that the effects of enteral protein intake on growth and neurodevelopment are comparable to those of parenteral protein intake could lead to inaccurate conclusions.32,33
One of the main strengths of this trial was the intention-to-treat analysis performed. This analytic approach preserves the randomization effect and increases generalizability of our findings. The other strength was using FFM as a primary outcome measure, which is rarely reported in enteral feeding trials. Measuring this outcome at term-equivalent age could help identify nutritional practices that prevent excessive weight gain from fat mass gains, especially with the increasing availability of reference values for FFM accretion at different gestational ages.34,35 However, it is important to acknowledge that ADP-measured FFM can only be assessed in infants who do not require significant respiratory support. Therefore, information on FFM accretion in the most immature and sickest infants, who are most likely to develop nutritional deficits, is often missing. This limitation could potentially explain the poor correlation between growth and body composition outcomes reported in observational studies that included preterm infants.36,37
This trial is the first to report significant length and head circumference gains resulting from early human milk fortification. Our finding of reduced weight z score declines during the first 2 weeks after birth is consistent with previous trials of early human milk fortification that have shown the same direction of effect in preterm infants born at older gestational ages.21,22 It also provides evidence that including the most immature infants in nutritional studies that assess growth outcomes requires careful consideration. Although including these infants in trials may help identify strategies to prevent significant weight loss during the phase of acute critical illness, these trials may not always detect significant benefits on weight or FFM gains at 36 weeks’ PMA or discharge. In this trial, more than 20% of infants included were born at 23 weeks of gestation or less. Despite not identifying a significant reduction in the risk of moderate to severe malnutrition following early human milk fortification, we reported length and head circumference gains per week that are comparable to those reported in observational studies that demonstrated successful prevention of postnatal growth failure after the gradual introduction of multiple innovative nutritional practices in a neonatal unit.38,39
Gains in length and head circumference are critical indicators of overall growth. Length is a measure of overall body size. A larger head circumference can indicate that the brain is growing, which is crucial for cognitive and motor development. Future studies should analyze the long-term effects of early human milk fortification on growth outcomes. Our adjusted analysis revealed that early human milk fortification had similar short-term effects on FFM, weight, length, and head circumference gains in both female and male infants, but observational studies have reported slightly slower head growth40 and faster weight gain after term equivalent age in extremely preterm female infants.41 Studying the long-term effects of early human milk fortification on neurodevelopmental outcomes is also highly recommended. If the positive effects on length and head circumference observed in this trial translate into potential benefits for neurodevelopment at 2 years of age, early human milk fortification could still be justified, even if this nutritional practice does not have an impact on FFM at 36 weeks PMA.
Conclusions
This trial reveals that human milk diets fortified soon after birth in infants born extremely preterm do not increase FFM accretion at term-equivalent age. Early provision of fortified human milk within the first 96 hours after birth may increase length gain velocity and reduce declines in head circumference-for-age z scores from birth to 36 weeks’ PMA.
Dr Salas conceptualized and designed the study, obtained funding, carried out the initial analysis, drafted the initial manuscript, and approved the final manuscript as submitted; Dr Gunawan carried out the statistical analysis, reviewed and revised manuscript, and approved the final manuscript as submitted; Ms Nguyen and Ms Reeves designed the data collection instruments, monitored patient enrollment and compliance, collected data, performed body composition assessments, reviewed and revised manuscript, and approved the♮final manuscript as submitted; Ms Finck and Ms Argent performed the randomization, monitored compliance, reviewed the manuscript, and approved the final manuscript as submitted; Dr Carlo helped design the study, critically reviewed the manuscript, and approved the final manuscript as submitted; and all authors approved the final manuscript as submitted♮and agree to be accountable for all aspects of the work.
The Increased Milk Protein to Accrue Critical Tissue (IMPACT) trial has been registered at www.clinicaltrials.gov (identifier NCT04325308). Deidentified individual participant data will be made available on publication to researchers who provide a methodologically sound proposal for use in achieving the goals of the approved proposal. Proposals should be submitted to [email protected].
COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2023-062391.
FUNDING: This trial was supported by a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), grant number K23HD102554. The funder did not participate in the work. The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no potential conflicts of interest relevant to disclose. Dr Salas patented an instrumented feeding bottle and received consulting fees for participation in Mead Johnson Nutrition advisory board meetings. Prolacta Bioscience provided the human-based fortifiers at cost and had no role in the design of the study, collection, analysis, interpretation of data, or writing of the manuscript.
Comments