The molecular story begins with structure. Human hemoglobin is a tetrameric protein that makes up roughly 96 percent of the dry weight of a red blood cell, and it comes in developmentally staged varieties. Fetal hemoglobin, or HbF, is built from two alpha-globin and two gamma-globin chains, giving it the designation alpha-2-gamma-2. Adult hemoglobin, HbA, swaps the gamma chains for beta chains, producing alpha-2-beta-2. The difference sounds trivial, a matter of two protein subunits, but it fundamentally alters the molecule’s biophysical behavior. HbF binds oxygen more avidly than HbA, a property that allows fetal red blood cells to snatch oxygen from maternal blood across the placenta and deliver it to developing tissues at precisely the right level, neither starving the fetus nor flooding it with excess oxygen.

What makes this system even more striking is that the switch from fetal to adult hemoglobin production runs on an internal developmental clock. It proceeds on schedule whether the fetus remains in the womb or arrives months early into a neonatal intensive care unit. At term birth, HbF typically constitutes 70 to 90 percent of total hemoglobin, with most of the remainder being HbA. In the weeks and months after birth, that balance steadily shifts as the bone marrow ramps up adult hemoglobin synthesis. But for infants born extremely preterm, this transition unfolds in the alien environment of the NICU, punctuated by blood draws, illness, and often repeated transfusions of adult donor blood, and until now clinicians have had no validated reference intervals to tell them what a normal HbF percentage should look like at any given gestational age.

That gap matters because a growing body of research has linked fetal hemoglobin levels to the outcomes that define preterm survival. Observational studies have associated a low fraction of HbF with bronchopulmonary dysplasia, the chronic lung disease of prematurity, and with retinopathy of prematurity, the disordered eye vessel growth that can threaten the vision of very preterm infants. A 2026 analysis further reported that the fetal hemoglobin fraction correlates with the risk of complications of prematurity more broadly. The logic is biologically plausible: fetal red cells carry oxygen with different kinetics and may behave differently in the microvasculature, so the composition of a preterm infant’s blood could influence how oxygen-sensitive organs develop. Yet without reference intervals, an individual laboratory value could not be confidently classified as normal or abnormal.

The Utah-led team attacked the problem with a statistical strategy that has quietly been transforming laboratory medicine. Rather than recruiting a pristine cohort of healthy volunteers, which is nearly impossible in neonatology, the researchers mined deidentified clinical records and applied the refineR algorithm, a computational method designed to extract reference intervals from real-world laboratory data. The algorithm works by identifying the apparently healthy portion of a mixed distribution, mathematically separating the central body of routine results from the pathological tail contributed by sick patients. The same team had previously used refineR to establish improved reference intervals for circulating nucleated red blood cell counts in neonates, and the approach sidesteps the selection biases that plague conventional reference interval studies.

To build the fetal hemoglobin intervals, the investigators focused on non-transfused infants spanning 23 to 40 weeks of gestation, deliberately excluding babies whose red cell populations had been altered by donor blood. This exclusion is critical, because a transfusion of adult blood instantly dilutes a preterm infant’s HbF fraction and would contaminate any attempt to define the natural developmental curve. The retrospective, deidentified records review was approved by the Intermountain Health Institutional Review Board and the University of Utah Institutional Review Board with a waiver of informed consent, and the work was conducted in compliance with federal human subjects regulations including HIPAA. Collaboration with neonatal intensive care medical directors at McKay-Dee Hospital and Utah Valley Hospital enabled collection of HbF values across the participating sites.

The resulting reference intervals, presented as curves across the gestational age spectrum from 23 to 40 weeks, give clinicians for the first time a population-based expectation for what percentage of hemoglobin should be fetal at each stage of preterm development. The clinical implications ripple outward in several directions. A HbF percentage falling below the expected interval might flag an infant at elevated risk for the oxygen-dependent complications of prematurity, prompting intensified surveillance of the lungs or retina. Conversely, the intervals provide an essential denominator for the growing interest in fetal red cell-based therapies, including a multicenter Italian randomized trial now testing cord red blood cell transfusions for severe retinopathy in preterm neonates. Interpreting any such intervention requires knowing where the natural baseline lies.

The study also highlights how much diagnostic power is locked inside routine laboratory measurements that hospitals already collect. Modern blood gas analyzers and automated hematology instruments can quantify hemoglobin derivatives and fractions at the push of a button, generating streams of data that historically were used only for individual patient decisions. Computational tools like refineR convert that accumulated clinical exhaust into population-level knowledge, a pattern now being repeated across neonatal hematology. For a field in which enrolling thousands of healthy newborns into reference studies is ethically and practically untenable, data-driven interval estimation may be the only realistic path forward, and the fetal hemoglobin curves are among its most consequential products to date.

There remain important caveats and next steps. Reference intervals describe populations, not prognoses, and the authors are careful to frame their curves as a foundation rather than a finished diagnostic test. Whether a low HbF fraction is a cause of bronchopulmonary dysplasia or retinopathy of prematurity, a marker of shared underlying vulnerability, or something in between, will require prospective studies that track infants against these new intervals over time. The investigators note that inquiries regarding data access can be addressed to the corresponding author, and the statistical code is available upon written request, inviting replication and extension by other centers. Still, the conceptual shift is hard to overstate: a molecule long celebrated for its role in the placental oxygen handshake now has a quantitative yardstick spanning the entire preterm period.

For the roughly one in ten infants born preterm worldwide, the stakes of oxygen management could hardly be higher, and fetal hemoglobin sits at the center of that physiology. By defining what normal looks like from 23 to 40 weeks of gestation, the Utah team has converted a century-old biological insight into a usable clinical benchmark. The next chapter, testing whether HbF percentage can stratify risk, guide transfusion strategy, or measure the benefit of fetal-cell therapies, now has the reference frame it needs. In the quiet arithmetic of reference intervals, neonatal medicine may have gained one of its most valuable new vital signs.

Subject of Research: Reference intervals for fetal hemoglobin percentage in preterm and term neonates from 23 to 40 weeks of gestation

Article Title: Establishing reference intervals for fetal hemoglobin percentage from 23 to 40 weeks of gestation using refineR

Article References: Zeiler, B. B., Bahr, T. M., Doyle, K., Maxfield, R. E., Stone, E. F., Addams, J. L., Ohls, R. K., & Christensen, R. D. (2026). Establishing reference intervals for fetal hemoglobin percentage from 23 to 40 weeks of gestation using refineR. Journal of Perinatology. https://doi.org/10.1038/s41372-026-02918-0

Image Credits: AI Generated

DOI: 10.1038/s41372-026-02918-0

Keywords: fetal hemoglobin, HbF, preterm infants, reference intervals, refineR algorithm, neonatology, neonatal intensive care, bronchopulmonary dysplasia, retinopathy of prematurity, oxygen transport, hemoglobin switching, transfusion medicine