Artificial intelligence news.
Straight from the original announcements.

Human milk composition reveals mammary gland function and links to infant growth

October 8, 2026, 20:09Source: PR Newswire
Human milk composition reveals mammary gland function and links to infant growth
Image: PR Newswire

Researchers at NYU Grossman School of Medicine have identified shared molecular patterns in human milk that connect maternal nutrition, mammary gland biology and infant growth. The study, published online October 8 in the journal Science, points to seven candidate biomarkers.

Key points
  • NYU Grossman researchers analyzed 1,543 human milk samples from cohorts in Canada, Pakistan and Burkina Faso.
  • Seven candidate biomarkers link maternal nutrition, mammary gland function and infant growth: C2, C3, C4, C5, lysine, 6′ sialyllactose and selenium.
  • C5 was the most prominent signal, and the team proposes low vitamin B5 may constrain CoA-dependent metabolism.
  • A companion Science paper found supplementation during lactation raised several B vitamins in milk while many components stayed stable.
  • The study was published October 8 in Science as part of the IMiC Consortium.

What the study examined

The team analyzed thousands of molecular measurements from 1,543 human milk samples collected in three cohorts in Canada, Pakistan and Burkina Faso. The samples covered metabolites, proteins, micronutrients such as vitamins, macronutrients such as carbohydrates, and human milk oligosaccharides. The researchers combined these data with infant growth data and computational approaches including machine learning.

According to the university, the goal was to ask not only what is present in milk, but also what its composition may reveal about the function of the gland that produces it. The study looked at three connected aspects of that function: milk synthesis and secretion, epithelial barrier and tissue remodeling, and immune and repair activity.

In two randomized nutritional trials, maternal supplementation promoted recovery or stability among infants at higher risk for poor growth. Multi-omic analyses also revealed a shared pattern of milk composition associated with both maternal supplementation and infant growth.

Seven candidate biomarkers

The analysis identified seven candidate biomarkers linked to maternal nutritional status, infant growth and biological programs related to mammary gland function. According to the researchers, these markers captured aspects of mammary biology that extend beyond established measures of blood-milk barrier permeability.

  • Four are short-chain acylcarnitines (C2, C3, C4 and C5), molecules that help cells process nutrients for energy.
  • Lysine, an amino acid.
  • 6′ sialyllactose, a human milk oligosaccharide.
  • Selenium, a mineral.

Short-chain acylcarnitines showed the strongest links across mammary functional domains, and C5 was a particularly prominent signal. Higher C5 levels were associated with reduced milk synthesis signatures and with patterns related to tissue remodeling and immune activity.

A possible role for vitamin B5

Because C5 stood out, the team looked at nutritional factors that might explain variation in acylcarnitine levels. Acylcarnitine metabolism depends on coenzyme A (CoA), which is synthesized from pantothenic acid, or vitamin B5. Drawing on associations across cohorts and established metabolic biology, the researchers propose that low vitamin B5 availability may constrain CoA-dependent metabolism and contribute to altered acylcarnitine profiles, including accumulation of C5.

First author Dr. April Jauhal said one plausible route linking maternal nutrition to mammary metabolism in the data is "via the pantothenic-CoA-C5 metabolic pathway," and that the group is investigating its functional role in mammary biology and lactation.

A companion study and next steps

The work is one of two complementary papers in Science from the International Milk Composition (IMiC) Consortium, conducted with Meghan Azad of the University of Manitoba and funded by the Gates Foundation. IMiC brought together more than 1,000 mother-infant pairs and profiled more than 25,000 molecular features in human milk. The companion study asked which milk components can be changed through maternal nutrition, and found that supplementation during lactation increased several B vitamins in milk while many other components remained comparatively stable.

Senior author Liat Shenhav, PhD, said the studies show that human milk is "neither fixed nor simply a reflection of what a mother eats." She said understanding which components respond to nutrition is essential for designing interventions that support maternal and infant health.

The Shenhav Lab is now building a computational and experimental framework to test nutritional bottlenecks that may limit mammary metabolism and milk production, including the vitamin B5, coenzyme A and C5 pathway. It is also studying colostrum and transitional milk to see how mammary functional states emerge between pregnancy and mature lactation.

Why it matters

Human milk contains thousands of nutrients, metabolites, proteins and immune factors, but research has often examined them one at a time. Because milk can be collected repeatedly, the authors say it offers a non-invasive way to follow mammary gland biology over time. The findings are described as candidate markers and a proposed mechanism, and the lab says further work is under way to test the vitamin B5 pathway.

Related news

AI tool PREVIA predicts PTSD risk from emergency room narratives
ResearchNovember 9, 2026, 11:07

AI tool PREVIA predicts PTSD risk from emergency room narratives

An AI tool developed by NYU Langone researchers predicted which emergency room patients would later develop post-traumatic stress disorder by analyzing their…

Inductive Wins OpenADMET CYP Challenge for Fourth Straight Time
ResearchNovember 6, 2026, 17:00

Inductive Wins OpenADMET CYP Challenge for Fourth Straight Time

Inductive placed first in the OpenADMET CYP Direct Inhibition Blind Challenge, finishing ahead of more than 270 participants from biopharma, academia and AI…

MIT’s CheMeleon predicts molecular properties from limited data
ResearchNovember 6, 2026, 20:00

MIT’s CheMeleon predicts molecular properties from limited data

MIT researchers have developed CheMeleon, a machine-learning model designed to predict molecular properties when only limited data are available. The open-source system…