Genetic Switch in Breast Milk: Indonesian Study Redefines Secretor Status Cut-Off

Human milk is far more than a simple source of nutrition. Suspended within it are complex sugars called human milk oligosaccharides, or HMOs, which babies cannot digest themselves but which act as fuel for beneficial gut bacteria and as decoys that block disease-causing microbes from attaching to the infant intestinal lining. A new study from Indonesia, published in Pediatric Research, has now mapped how a mother’s genes shape this sugar profile, and in doing so has challenged one of the field’s most widely used laboratory thresholds. The work, led by Klara Yuliarti of Universitas Indonesia and colleagues in Jakarta and Groningen, combined detailed sugar chemistry with full gene sequencing to reveal that the standard cut-off for classifying mothers as so-called secretors may be badly miscalibrated for populations outside Europe.

The biological logic behind the study rests on two genes, FUT2 and FUT3, which encode fucosyltransferase enzymes responsible for adding fucose sugars to the oligosaccharide backbone. Women who carry at least one functional copy of FUT2, known as secretors, produce abundant amounts of alpha-1,2-fucosylated HMOs, most notably 2′-fucosyllactose, or 2’FL, which is typically the most abundant oligosaccharide in human milk. Women who carry two non-functional copies of FUT2, known as non-secretors, produce milk that is nearly devoid of these fucosylated structures and instead contains higher relative levels of other HMO types. This genetic difference creates one of the most striking examples of inter-individual variation in human milk composition, and it matters clinically because the health benefits attributed to specific HMOs, from protection against diarrhea to shaping the infant microbiome, differ depending on which structures are present.

To understand how this plays out in an Indonesian population, the researchers recruited 120 healthy mothers aged at least 18 years who were exclusively breastfeeding, and collected breast milk samples when their infants were between 2 and 5 weeks old. The team analyzed the oligosaccharide profile using Ultra High-Performance Liquid Chromatography with Fluorescence Detection, a sensitive analytical technique capable of resolving and quantifying individual HMO structures. In parallel, they sequenced the entire coding region of the FUT2 gene from each mother to identify all genetic variants present, rather than relying on targeted testing of only the most common mutations known from European populations.

The results revealed that 65.8 percent of the Indonesian mothers were secretors, a proportion notably lower than the roughly 70 to 80 percent typically reported in Caucasian populations. This difference is not merely a curiosity of population genetics. It has direct implications for how studies of infant health outcomes should be interpreted in different parts of the world, and it suggests that the contribution of less common HMOs, which are proportionally more important in non-secretor milk, may deserve greater attention in populations where non-secretor status is more frequent.

Among the genetic findings was a novel FUT2 variant, designated c.851C>G, which had not previously been reported in the literature. The identification of new variants in underrepresented populations underscores how much genetic diversity remains to be catalogued, and it reinforces the argument that diagnostic tools and reference ranges developed in one population cannot simply be transplanted to another without validation.

Perhaps the most consequential finding of the study concerns the threshold used to classify secretor status from milk measurements alone. In many laboratories and clinical studies around the world, a 2’FL concentration of 25 to 50 milligrams per liter has been used as the boundary between secretor and non-secretor classification. The Indonesian team applied receiver operating characteristic, or ROC, analysis to their combined dataset of FUT2 genotypes and measured 2’FL concentrations, a statistical approach that identifies the cut-off value maximizing both sensitivity and specificity. The result was striking: the optimal threshold came out at 425.9 milligrams per liter, roughly an order of magnitude higher than the conventional range, with 98.7 percent sensitivity and 100 percent specificity for correctly classifying secretor genotype.

The discrepancy between the old and new cut-offs is not simply a matter of units or measurement technique. It reflects a genuine biological reality that 2’FL concentrations in the milk of true secretors are typically hundreds of milligrams per liter, while non-secretors produce almost none. A threshold set at 25 to 50 milligrams per liter would classify many non-secretors with trace amounts of 2’FL as secretors, blurring the very distinction the test is meant to capture. The authors argue that the widely used lower cut-off needs to be re-evaluated, ideally in light of clinical health impact data and genotype findings, rather than being perpetuated through citation alone.

The statistical association between FUT2 genotype and secretor phenotype in the study was highly significant, with a p-value below 0.001, confirming that the genetic variants identified through sequencing reliably predict the oligosaccharide profile measured in milk. This tight genotype-phenotype correlation provides a strong foundation for using genetic testing as a reference standard against which biochemical cut-offs can be calibrated, and it demonstrates that the relationship holds across diverse populations even as the absolute frequencies of secretor and non-secretor genotypes vary.

The broader significance of this work extends beyond Indonesia. Human milk oligosaccharide research has intensified in recent years as these sugars have been added to infant formula and investigated as therapeutic agents, yet the reference values and classification schemes used in that research have been built largely on data from European and North American cohorts. Studies from Vietnam, China, Malaysia, and now Indonesia have consistently shown that HMO profiles differ geographically, driven in part by differing frequencies of FUT2 and FUT3 polymorphisms. The Indonesian findings add to this growing body of evidence that a one-size-fits-all approach to HMO classification is inadequate, and that population-specific validation is essential for both research and any future clinical applications.

The authors themselves note that the proposed cut-off of approximately 425 milligrams per liter needs further evaluation in longitudinal studies that capture the dynamic changes in milk composition across the full lactation period, since oligosaccharide concentrations shift over time and the present cross-sectional design captured only a narrow window at 2 to 5 weeks postpartum. They also emphasize that the health consequences of secretor status for infants, including the role of less common HMOs that dominate in non-secretor milk, remain incompletely understood and warrant deeper investigation. As the science of human milk moves from descriptive chemistry toward precision nutrition, studies like this one make clear that the genetic diversity of mothers worldwide must be built into the foundations of that field from the start.

Subject of Research: FUT2 genotype-phenotype correlation and human milk oligosaccharide secretor status in Indonesian mothers

Article Title: Profile of human milk oligosaccharides and maternal FUT2 polymorphism: genotype-phenotype correlation and proposed cut-off for secretor status

Article References: Yuliarti, K., van Leeuwen, S. S., Mansyur, M., Timan, I. S., Ariani, Y., Sidiartha, I. G. L., Simanjuntak, E., & Sjarif, D. R. (2026). Profile of human milk oligosaccharides and maternal FUT2 polymorphism: genotype-phenotype correlation and proposed cut-off for secretor status. Pediatric Research. https://doi.org/10.1038/s41390-026-05579-z

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05579-z

Keywords: human milk oligosaccharides, FUT2, secretor status, 2'-fucosyllactose, breast milk, genotype-phenotype correlation, ROC analysis, Indonesia, pediatric research, fucosyltransferase, infant nutrition, genetic variant