Written by: Jyoti Madhusoodanan, Ph.D. | Issue # 131 | 2026
- The nutritional value of human milk is largely studied in the context of mother-infant pairs, but donor milk may differ in its composition.
- In 600 samples from milk banks in five countries, macronutrients such as protein and fat were similar.
- A mother’s lactation stage had a strong impact on nutritional composition.
- Micronutrients such as vitamin B12 and trace minerals were much more variable than protein and fats.
Donated breastmilk supports the lives of some of the most fragile humans: low birth weight and pre-term infants. Donated breastmilk is especially important to protect babies from necrotizing enterocolitis when a mother’s own milk may be unavailable or insufficient.
Donated human milk must adhere to certain safety standards, such as pasteurization, to ensure it will not transmit infections. But beyond these safety standards, little is known about its nutritional value. Studies of the nutrients in breastmilk have largely relied on samples gathered from mothers feeding their own babies. But a mother’s milk changes over time and is shaped by her diet, the baby’s age, infections in mother or infant, and many other factors.
Donated milk is produced by women who typically produce enough extra milk to be able to feed multiple other babies. “We can’t assume that this milk looks like what’s happening within the maternal-infant dyad,” says Maryanne Perrin, a nutrition researcher at the University of North Carolina in Greensboro.
As a graduate student in 2015, Perrin studied how mother’s milk changes in the second year after a baby is born. Comparing samples from donor milk banks, she found that the energy content of the samples ranged from 13 to 36 calories in each ounce of milk. “I was like, holy cow, this is quite a bit of variation,” she said, “particularly when you’re thinking about feeding a very nutritionally vulnerable preterm infant.”
In two new studies, Perrin and her colleagues examined the nutritional composition of donated breastmilk in milk banks across the world, including samples from low-, middle- and high-income countries. They also examined the effects of pooling samples – a common practice to reduce variability in donor milk.
For the first study, the researchers collected 600 samples from 8 milk banks in Poland, Kenya, Chile, Vietnam, and the United States [1]. Each sample represented milk collected over at least three days from a single approved donor and then pooled together. This is typically how milk banks collect and process samples. The researchers collected both samples that had been pasteurized and cooled and raw samples prior to pasteurization to test whether the process affected the nutrients in milk.
Across different countries, levels of macronutrients, such as protein and fat, and micronutrients, including vitamins and trace minerals, were similar across milk samples. Pasteurization did not significantly alter the nutritional composition of donated milk, although it led to a significant decrease in bioactive components such as lactoferrin.
Only one milk bank based in Kenya was significantly different in almost every way. When the researchers examined their collection practices, they found that most milk banks began collecting between 2-3 weeks after mothers gave birth. But the Kenyan milk bank was located within the hospital and collected milk about one and a half weeks after mothers had given birth. The data suggest that the milk bank’s practices – whether they recruit donors at early stages of lactation or mature stages can “have a dramatic impact on the nutritional signature” of milk, Perrin said. “That was a big takeaway.”
The team also found that a mother’s age, birth term, lactation stage and the length of time a sample had been stored were strongly correlated with differences in milk composition. But these factors only explained about 38 percent of variation. Other differences could not be attributed to clear causes. “It really confirmed that there’s a lot of unknown variability in what is driving the composition of milk from donors,” Perrin said.
In their next study, the researchers created a model to calculate how milk pooling practices might affect variability between donor samples [2]. They used data only from U.S. donors to eliminate any differences that might exist due to environmental factors, geography or different dietary trends. They tested different random pooling scenarios, ranging from 3 donors up to 9 donors. They found that macronutrients were much less variable than micronutrients – while lactose, for example, remained at consistent levels whether 3 donors were pooled or 9, vitamin and mineral levels showed 2-3-fold level differences until higher numbers of donors were pooled together. Statistically, this is unsurprising, Perrin said. “What’s new is really showing that the patterns look quite different in the micronutrients compared to the macronutrients.”
The data could help clinicians make better informed choices about what nutrients a pre-term infant may be receiving. “We want them to have a strong understanding of how much variation they should expect in milk based on how many donors are in the pool,” Perrin said.
The team is working to understand the reasons underlying the nutritional variation they observed. Their research is a step toward optimizing donor milk to improve infants’ growth and health. “There are standards for almost every other food you would be feeding this population, but there are really no standards other than the safety standards around pasteurization,” Perrin said. “This is the kind of data we need to try and address that.”
References
- Perrin MT, Mansen K, Israel-Ballard K, Bode L, Hampel D, Shahab-Ferdows S, Allen LH, Maggio FC, Njuguna EM, Tran HT, Wesolowska A. A cross-sectional study of macronutrients, vitamins, minerals, and bioactive components in human milk of approved milk bank donors from low-, middle-, and high-income countries. The American Journal of Clinical Nutrition. 2025 Sep 18.
- Smith RM, Richter S, Iwayemi EF, Mansen K, Israel‐Ballard K, Hampel D, Shahab‐Ferdows S, Allen LH, Bode L, Perrin MT. Evaluating the Effect of Random Multi‐Donor Pooling on the Nutritional Variability in Donor Human Milk Using Computer Modeling. Maternal & Child Nutrition. 2026 Mar;22(1):e70158.