Before processing: the milk is already changing
Donor human milk banking begins with a fact that the word “donor” can hide: the milk entering the bank is not a standardized substance. Its composition reflects when it was produced, the pregnancy and lactation context in which it was produced, and the choices made while collecting and combining it. Understanding that starting variation is essential for interpreting any later comparison of pasteurization methods, because a difference observed after treatment may have begun before the milk reached the treatment equipment.
The focus here is the state of donor milk before processing and the relationships that shape it. The practical judgment is straightforward: do not attribute every difference between batches to heat, and do not define a high-quality bank only by the consistency of the final bottle. A bank has to preserve the donor’s own infant’s access to milk, support safe and sustainable donation, and decide when pooling improves a product and when it erases information or expands the cost of failure.
A biological starting material, not a fixed recipe
Human milk changes over time. WHO’s evidence synthesis reports that protein, vitamins A and E, and some minerals and trace elements generally decrease across lactation, while lactose and some B vitamins increase. The Frontiers review likewise identifies lactation stage, gestational age and maternal diet as sources of variation in donor human milk. These are not minor differences in a manufactured ingredient. They are part of the biology of milk, which responds to the needs of a growing infant and to the circumstances of the donor.
That does not mean that milk from one stage is simply superior to milk from another. It means that the word “milk” covers multiple starting states. Milk collected early after birth may have a different balance of protein and micronutrients from mature milk. Milk from a donor who delivered preterm may differ from milk associated with a term birth. A donor’s diet and other characteristics can contribute further variation. The relevance of each difference depends on the recipient, the amount consumed, the rest of the feeding plan and the strength of the evidence connecting the difference to an outcome.
The direction of a change is not the same as its clinical importance. A component may be more concentrated at one stage of lactation without proving that selecting that stage improves the health of every recipient. Conversely, a component that changes little on average may still vary substantially among donors. This is why a bank or a study needs to distinguish a population pattern from a rule for individual donations. WHO notes that early-lactation milk can contain more protein, zinc, copper, selenium and vitamins A, B12 and E, but the standards do not turn that observation into a requirement that only early-lactation donors be accepted.
This distinction matters when researchers compare processing methods. Suppose one batch treated by Holder pasteurization contains more early-lactation milk and another batch treated by a different method contains more mature milk. If the two batches differ in protein, immunoglobulins or other components, the treatment may not be the only explanation. The same problem occurs when one study compares pooled milk and another analyzes samples from individual donors. A result can be real within a study while still being difficult to generalize beyond its particular donor mix.
The comparison can also be distorted by what the study chooses to measure. A broad nutritional average may hide a change in a smaller bioactive fraction; a difference between banks may reflect their donor populations rather than their treatment equipment. The Frontiers review’s discussion of donor and lactation variables is therefore more than background description. It explains why processing studies need clear comparators and why a reported post-treatment difference should be read as the combined result of starting material and processing conditions unless the design separates them.
The sources support a cautious interpretation of composition. WHO notes that milk banks collecting mainly early milk can differ from banks collecting predominantly mature milk in nutrients and bioactive compounds such as immunoglobulin A (IgA) and lactoferrin. But WHO also reports no evidence that limiting donors to an early postpartum group improves recipient outcomes. The biological difference is supported more clearly than a universal clinical preference for one source.
Pooling is one response to this variability. Combining milk from several donors can reduce variation in protein and fat and may reduce variation in IgA activity. A pooled product can be easier to describe and administer because its composition is less dependent on one donor’s particular sample. For a milk bank serving vulnerable infants, that consistency has practical value: the product is less likely to swing sharply from one bottle to the next simply because a different contribution was used.
But consistency has a price. Pooling averages differences that might matter biologically, making it harder to know which donor or lactation stage contributed to a particular characteristic. It can also increase the size of the batch exposed to a single contamination event. If one contribution causes a pooled batch to fail screening, more milk may be discarded than if that contribution had remained separate. Combining donors can therefore reduce variation in what reaches the infant while increasing the scale of a possible loss and the complexity of tracing its origin.
The choice is not between a perfectly individualized product and a perfectly uniform one. It is between different ways of managing variation. Keeping contributions separate preserves more information but may produce greater bottle-to-bottle differences. Pooling can make the administered product more consistent but can hide meaningful differences and create larger batch-level consequences when something goes wrong. The appropriate balance depends on the bank’s safety controls, demand, infrastructure and cultural context, not on a slogan about naturalness or standardization.
The lifecycle in the diagram has two important features. First, the milk does not become a bank product merely because someone has expressed it. Eligibility, support and storage determine whether a contribution can move forward. Second, there are two legitimate endpoints: release to a recipient or discard after a control detects an unacceptable problem. A bank that reports only the amount distributed, without considering the milk lost and the reasons for loss, misses part of the system’s quality story.
The donor relationship is part of the product
A donor is not a replaceable source of raw material. WHO recommends that milk banks recruit only mothers who can express more milk than their infant needs or who have stored surplus milk. The bank should protect continued breastfeeding, support donor health, prevent excessive donation and hyperlactation, and train donors in safe expression and hygiene. These requirements recognize that the donor’s infant has a prior claim on the mother’s milk and that the physical work of expression can affect the donor’s health and ability to continue breastfeeding.
“Surplus” is therefore a relationship, not just a volume measurement. A donation that looks large enough in isolation may still be inappropriate if producing it displaces feeding for the donor’s infant or creates pressure to maintain an unhealthy level of expression. Conversely, a smaller contribution may be valuable if support and safe handling allow it to be collected without disrupting breastfeeding. The bank’s responsibility is to create conditions in which donation remains compatible with the donor–infant dyad, rather than treating higher collection totals as an automatic sign of success.
The evidence behind some of these recommendations is limited. WHO reports no direct systematic-review evidence that education in expression prevents harm to donors, their infants or the donated product. It also reports no clear evidence that a particular expression method causes worse outcomes. The Standards Development Group nevertheless judged support important because lack of guidance could contribute to poor breastfeeding outcomes, oversupply, mastitis or breast trauma. This is a case where a responsible standard is not identical to a claim that a large controlled trial has already proved every benefit.
The same evidence-calibrated approach applies to the method of expression. Hand expression, manual pumps and electric pumps can all be acceptable. Some reviewed evidence suggested that expression method or equipment can affect protein or fat content, but it was not strong enough for WHO to change general donation instructions. Another review found no difference in contamination according to the expression method in the limited available evidence. The sensible conclusion is not that technique never matters, nor that one device guarantees superior milk. It is that banks should teach safe, acceptable practices while avoiding unsupported rules that create barriers for donors.
Recruitment and compensation also shape the material that enters the system. WHO recommends inclusive recruitment across lactation stages and does not offer financial incentives for milk donation. Actual costs—such as transport, equipment or expression supplies—may be reimbursed or supported in kind. The distinction is intended to preserve financial neutrality. It reduces the burden of donation without making volume the primary reason to continue.
The concern about incentives is technical as well as ethical. WHO’s narrative evidence review describes a possibility that donors motivated mainly by financial gain may prioritize quantity over quality, conceal a health or storage problem, or adulterate milk to increase volume. It also notes the risk that a rapid increase in donations could overwhelm screening, testing and pasteurization capacity. These are documented risks considered in the standards, not evidence that every compensated donation is contaminated. The broader point is that a supply intervention can change the behavior of the people and institutions responsible for quality.
Practical support can improve access in a different way. Providing containers, pumps or transport assistance can make donation possible for people who would otherwise face excessive cost or logistical difficulty. It may also reduce waste by helping donors follow safer storage and expression practices. Inclusive recruitment expands the potential supply, while support helps that supply remain sustainable. But neither works if the bank lacks enough staff, cold storage, laboratory capacity and processing time to absorb the additional volume.
Capacity behaves like a limit on the whole pathway. More donations are useful only when the bank can screen them, keep them cold, process them, test them and trace them without lowering standards. If collection grows faster than those controls, the result can be longer delays, more failed batches and more discarded milk rather than more safe milk for recipients. A decision that appears to improve access at the collection stage can therefore reduce access at the point of delivery. This is why the WHO standards weigh resource implications and supply effects alongside donor ethics and product safety.
This is why donor protection and access cannot be separated from product quality. Excluding too many donors can restrict supply and may communicate that milk from people who do not qualify is unsafe for their own infants, even when the exclusion is only a bank-eligibility decision. Poorly handled rejection can discourage breastfeeding. On the other hand, accepting more milk without maintaining eligibility checks and processing controls can increase contamination, discard and risk to recipients. The bank has to widen access without allowing the safety boundary to dissolve.
Pooling makes these pressures visible. A bank may pool milk to create a more consistent nutritional product, but the decision also determines how much milk is placed at risk together. Track-and-trace systems can reduce the consequences of that choice by recording the movement and processing of milk from donor to recipient and back again. They do not prevent every failure, but they give the bank a way to investigate an adverse event and identify the affected product rather than treating the entire supply as unknowable.
By the time donor milk reaches a treatment step, it has already passed through a series of biological and institutional transitions. Lactation stage and donor characteristics have shaped its composition. Recruitment and support have influenced whether donation is safe and sustainable. Pooling may have reduced variation while increasing batch exposure. Storage and handling may have preserved or damaged the starting material. Processing will add another transition, but it will not be acting on a blank slate. That is the reason a meaningful comparison must follow the whole bank pathway—and why the next safety question is how its controls work together as the milk moves toward a vulnerable recipient.