Skip to main content

Individually Targeted Human Milk Fortification: A Randomized Clinical Trial

In brief

Individual milk analysis shows no growth or brain benefit for very preterm infants

In a double-blind trial of 130 infants born 24-30 weeks, point-of-care milk analysis guided extra protein or fat but did not improve weight, length, head-circumference z scores, fat-free mass, or near-term brain MRI compared with standard fortification. Both groups met nutrient targets, suggesting routine targeted fortification is unnecessary.

Journal
JAMA pediatrics (Q1)
Published
14 September 2026
Study design
Randomized controlled trial
Evidence level
Level 1, High (CEBM 1b)
Authors
Mandy Brown Belfort, Banu Ahtam, Katherine A Bell, Paige Berger, Sara Cherkerzian, Deirdre Ellard, et al.
PMID
42734908
DOI
10.1001/jamapediatrics.2026.4220

Why clinicians should know about it

  • Picked for Neonatology (top studies of the week, 20 September 2026): Double‑blind RCT of individualized human milk fortification
  • Picked for Pediatric Surgery (top studies of the week, 20 September 2026): Ranked by evidence level and journal quartile

Abstract

IMPORTANCE: Human milk is recommended for hospitalized very preterm infants, but the optimal macronutrient fortification strategy is unknown. OBJECTIVE: To compare individually targeted fortification using point-of-care human milk analysis with standard-of-care fortification. DESIGN, SETTING, AND PARTICIPANTS: This double-blind randomized clinical trial enrolled participants from February 2020 to March 2025 at a single academic level III neonatal intensive care unit (NICU), with outcomes assessed through NICU discharge. Very preterm infants born at 24 to 30 completed weeks' gestation were eligible for inclusion. Data were analyzed from June 2025 to March 2026. INTERVENTION: All participants received standard of care comprising maternal and/or pasteurized donor human milk (termed base milk) fortified with commercial multicomponent human milk fortifier and protein modular, with additional protein and/or fat provided for clinical growth faltering. The intervention group received protein and/or fat beyond standard of care when indicated by milk analysis with midinfrared spectroscopy (Miris Human Milk Analyzer [HMA] version 08-2-02-107 [Miris AB]), targeting minimum true protein of 1 g/dL and energy of 67 kcal/dL in base milk. MAIN OUTCOMES AND MEASURES: Primary outcomes were z scores of body weight and length at study diet end; fat-free mass measured with air displacement plethysmography; and quantitative metrics from brain magnetic resonance imaging (MRI) near term-equivalent age. RESULTS: Of 251 eligible infants, 206 were approached and 130 consented, were randomized, and received the study diet (65 each in the control and intervention groups); 113 infants (87%) completed the study diet though 36 weeks' postmenstrual age. Mean (SD) gestational age was 28.5 (1.9) weeks, and mean (SD) birth weight was 1165 (298) g. Median (IQR) true protein content of base milk (control: 1.24 [1.13-1.34] g/dL vs intervention: 1.18 [1.12-1.32] g/dL) and total protein intake from base milk and all fortifiers (control: 4.58 [4.44-4.74] g/kg/d vs intervention: 4.47 [4.31-4.60] g/kg/d) were unexpectedly higher in the control group; fat and energy content and intakes were similar between groups. Nutrient intakes all met or exceeded recommended levels in both groups. In intention-to-treat analyses, there was no evidence that z scores of weight, length, and head circumference, body composition, or brain MRI indices differed between groups. Results were similar in analyses limited to infants who completed the study diet. CONCLUSIONS AND RELEVANCE: Findings in this randomized clinical trial do not support routine use of point-of-care human milk analysis to inform individually targeted human milk fortification in the NICU. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT03977259.

Abstract as published, via PubMed.

View on PubMedFull text at the publisherOpen in the app

For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.