Reading the milk after processing

The molecular evidence on processed donor human milk is the focus here: not a search for one overall “quality score,” but an examination of which components change, which remain relatively stable and what those findings can legitimately mean. Multi-omics studies measure several layers—such as proteins, carbohydrates, lipids and metabolites—because donor milk is a mixture of functions, not one substance that rises or falls as a whole. The practical judgment is to match each result to its level of evidence: a concentration supports a compositional claim, an activity assay supports a narrower functional claim, and neither automatically predicts an infant outcome.

This distinction matters after Holder pasteurization because the same treatment can leave bulk nutrients largely recognizable while changing selected immune proteins, enzymes or molecular profiles. It also matters when an alternative method appears to preserve a component more effectively. “Preserved” must always mean preserved compared with a stated comparator, under stated conditions, for a particular component or activity.

One treatment, several molecular stories

The most consistent processing differences in the Frontiers review appear in the proteome—the collection of proteins in the milk. Holder pasteurization is associated with reductions in immunoglobulin A, lactoferrin and lysozyme in the cited comparisons. These molecules are not interchangeable. IgA helps protect mucosal surfaces, lactoferrin has antimicrobial and immunomodulatory roles, and lysozyme contributes to antimicrobial activity. A reduction in one does not tell us that all immune functions have been lost, but it does show why total protein is an incomplete measure of biological integrity.

The comparison with other methods is relative rather than absolute. HTST generally preserved more immunoglobulin and lysozyme than longer thermal treatments in the reviewed studies. HPP and freeze-drying also showed better retention of selected bioactive proteins in the small number of comparisons available. Those findings give a plausible reason to investigate alternatives, but they do not establish that every protein responds in the same way or that the method preserving one protein will preserve all other functions.

Bile salt–stimulated lipase, or BSSL, makes the difference between presence and activity clearer. BSSL supports fat digestion in newborns whose pancreatic systems are immature. The review reports large or near-total losses of BSSL in some Holder comparisons, with more activity retained after HTST, HPP or freeze-drying. One cited comparison found that Holder treatment destroyed 66% of BSSL while HTST destroyed 39%. Those figures describe particular study conditions and an activity measure; they are not a universal percentage for every banked batch. More importantly, a lower enzyme activity did not necessarily produce a detectable difference in triglyceride hydrolysis during one in-vitro digestion comparison, suggesting that other mechanisms may compensate.

The glycome—the collection of complex carbohydrates in milk—tells a different story. Human milk oligosaccharides, or HMOs, are often relatively resistant to processing. The review reports that total HMO concentration and overall composition were not significantly affected by Holder pasteurization or freeze-drying, and that HPP preserved HMO concentration and profile in the cited study.