Prepared by: Dr. Nurul Izzah Khalid

Senior Lecturer

Department of Food Technology, Faculty of Food Science and Technology, UPM

Human milk is not simply a mixture of protein, fat and carbohydrates. It is a complex biological system containing nutritional, immunological and bioactive components that can be affected by how the milk is handled, processed and stored. From a food engineering perspective, the challenge is therefore not only to extend shelf life, but also to preserve as much of the original quality as possible while maintaining microbiological safety.

One technology attracting increasing attention is freeze-drying, also known as lyophilisation. In this process, the milk is first frozen, and the frozen water is subsequently removed under vacuum through sublimation, in which ice changes directly from a solid to vapour without passing through the liquid phase.

The concept sounds simple: remove the water, produce a powder, and add water again when the milk is needed. In practice, however, the process is considerably more complex.

Why freeze-dry human milk?

Donor human milk (DHM) normally requires refrigerated or frozen storage. Dependence on freezers and cold-chain transportation can create logistical challenges, particularly when milk needs to be transported over long distances or supplied in resource-limited settings. Cheema et al. (2025) highlighted these challenges as one reason freeze-drying could be useful in human milk banking.

Drying also converts the milk into a more compact form. Meng et al. (2021), in their review of the composition and physicochemical properties of human milk, noted that drying may save storage space, prolong storage time and reduce refrigeration requirements.

The more important question, however, is: what happens to the milk after the water is removed?

Are the nutrients still preserved?

Current evidence is encouraging, but the findings need to be interpreted carefully.

Cheema et al. (2025) studied donor human milk that had already undergone routine milk-bank processing, including pasteurisation and microbiological screening, before freeze-drying. In that study, the concentrations of fat, protein, lactose, solid-non-fat and salts were not significantly altered by freeze-drying. These macronutrients also remained stable during storage for up to 12 months, including under ambient-temperature conditions.

Protein, lipid and metabolite fingerprints also remained broadly consistent across the storage conditions evaluated.

However, this does not mean that every component of human milk remains completely unchanged.

Dávila-Caraballo et al. (2024) compared milk preserved at −20°C, −80°C and by freeze-drying. Total protein did not differ significantly among the three preservation methods. In the same study, in vitro protein digestibility was 77.91% for freeze-dried milk, compared with 77.55% for milk stored at −80°C and 75.55% for milk stored at −20°C under the experimental conditions used.

For the immunological components examined, several cytokines and secretory immunoglobulin A (sIgA) also showed no significant differences among the three preservation methods.

The carbohydrate findings were more complex. Several human milk oligosaccharides (HMOs) did not differ significantly, but lactose, total HMOs and some fucosylated HMOs differed among the preservation methods.

Therefore, a more accurate conclusion is:

Freeze-drying can preserve many important components of human milk well, but its effects are not necessarily identical for every constituent.

The most important point: freeze-drying is not sterilisation

This is perhaps the most important issue from a food safety perspective.

A dry powder may appear stable, but dry does not mean sterile.

Blackshaw et al. (2021) tested donor human milk that had been deliberately inoculated with bacteria. Freeze-drying alone resulted in only about a 0.9-log reduction in Staphylococcus aureus and a 0.5-log reduction in Salmonella typhimurium.

These reductions show that freeze-drying may affect bacterial survival, but they are not sufficient for the process to be regarded as pasteurisation or sterilisation.

Jarzynka et al. (2021) reached a similarly important conclusion. Their study showed that freeze-drying alone did not ensure microbiological purity. In their inoculation experiments, Staphylococcus aureus, Listeria monocytogenes and Cronobacter sakazakii remained detectable after lyophilisation and storage.

This leads to a fundamental food engineering principle:

Freeze-drying is primarily a drying and preservation technology; it should not automatically be considered a process that makes contaminated milk safe.

Why did one study report microbiological stability for 12 months?

At first glance, this may appear inconsistent with the microbiological studies above.

Cheema et al. (2025) reported no microbial growth before or after freeze-drying and observed microbiological stability during storage.

However, one experimental detail is crucial: the donor milk had already been pasteurised and microbiologically screened before freeze-drying.

The findings are therefore not contradictory.

Cheema et al. demonstrated that freeze-drying can help preserve milk that has already undergone an appropriate safety treatment, whereas Blackshaw et al. and Jarzynka et al. demonstrated that freeze-drying should not be expected to eliminate microorganisms already present in contaminated milk.

This distinction is critical when translating laboratory technology into real-world processing.

Safety may require more than one processing step

The selected studies also illustrate the use of combined processing approaches.

Jarzynka et al. (2021), for example, applied high-pressure processing (HPP) at 450 MPa, 21°C for 15 minutes, followed by lyophilisation. For the bacterial strains evaluated, samples treated with HPP and HPP plus lyophilisation showed no pathogen growth during storage for up to six months.

Blackshaw et al. (2021) evaluated another approach: freeze-drying followed by gamma irradiation. In their experimental model, 2 kGy gamma irradiation after freeze-drying produced bacterial reductions similar to Holder pasteurisation for the tested organisms, while at 5 kGy and above, no viable colonies of S. aureus or S. typhimurium were detected in the inoculated samples.

These findings are experimental and should not be interpreted as processing instructions for home use. They do, however, illustrate how food engineers approach safety: not by relying on one machine, but by designing an entire processing system.

Freeze-drying is not a single standardised recipe

The six selected papers also show that freeze-drying conditions can vary considerably.

Blackshaw et al., for example, froze their samples at −80°C before freeze-drying, with the overall procedure lasting approximately 48 hours.

In the study by Cheema et al., the milk was cooled to approximately −40°C before the primary drying stage was carried out at −45°C, and the samples were removed from the freeze-dryer after 48 hours.

This is another important point from a food engineering perspective: the performance of freeze-drying cannot be judged simply by the name of the technology. Freezing conditions, pressure, drying time, the condition of the starting material and post-drying handling all form part of the process design.

Packaging after drying is also part of the process

Cheema et al. provide a useful example.

After freeze-drying, their samples were placed in Mylar bags together with a 5 g silica desiccant packet and then heat-sealed before storage at ambient temperature, approximately 4°C or approximately −20°C.

Therefore, when that study reports compositional stability for up to 12 months at ambient temperature, the finding must be understood in the context of the complete experimental system, including the condition of the donor milk before drying and the packaging used afterward.

The figure of “12 months at room temperature” should not therefore be generalised to every freeze-dried human milk product produced under different conditions.

Reconstitution also needs to be precise

Another engineering issue is reconstitution.

In the Cheema et al. study, the amount of water removed during freeze-drying was determined by weighing. During reconstitution, the same amount of sterile water that had been lost was added back at approximately 40°C, followed by gentle mixing.

Blackshaw et al. similarly used a specified powder-to-water ratio and sterile water warmed to 40°C for reconstitution.

From a food engineering perspective, this is essentially a mass-balance problem. If the amount of water added back is inaccurate, the resulting nutrient concentration may no longer match that of the original milk.

Has freeze-dried human milk already been used clinically?

There are promising developments, but the clinical evidence remains preliminary.

Rochow et al. (2025) evaluated a fortifier produced from high-temperature short-time pasteurised and freeze-dried donor human milk. The study involved 32 preterm infants in the intervention group and 32 in a comparison group. Infants included in the study had a gestational age of at least 30 weeks.

More than 3,100 feedings were evaluated. Feeding tolerance was good in the study population, and no significant differences were observed in several growth outcomes between the groups. However, the authors emphasised that the study was observational, involved a relatively small sample and used retrospective controls. Larger studies are still required.

Importantly, the product investigated was a human milk fortifier, not ordinary freeze-dried breast milk intended for routine home use. The researchers also noted that the protein-to-energy ratio of the freeze-dried donor-milk fortifier may be inadequate for very-low-birth-weight infants.

The findings therefore should not be interpreted as proof that all freeze-dried human milk products are appropriate for all infants.

So, is freeze-dried human milk a technology for the future?

Based on these six selected studies, the technology is highly promising.

Freeze-drying has shown the ability to preserve many nutritional and bioactive components, convert milk into a powdered form and, under carefully controlled processing and packaging conditions, maintain compositional stability for up to 12 months at ambient temperature.

At the same time, the evidence makes one point very clear: freeze-drying should not be viewed simply as “put the milk into a freeze-dryer and it becomes shelf-stable”.

From a food engineering perspective, the more appropriate questions are:

What was the condition of the milk before drying? Has its microbiological safety already been established? What processing parameters were used? How was the powder packaged and stored? And how will it be accurately reconstituted before use?

Taken together, the six studies show that the success of freeze-dried human milk depends on the entire processing chain, not on freeze-drying alone.

Freeze-drying may create new possibilities for donor human milk management, human milk banking and the development of human milk-derived fortifiers. However, based on the current evidence from these studies, it is more accurate to describe the technology as a promising and evolving preservation approach, rather than an automatic replacement for established human milk safety and storage procedures.

References

Blackshaw, K., Wu, J., Valtchev, P., Lau, E., Banati, R. B., Dehghani, F., & Schindeler, A. (2021). The effects of thermal pasteurisation, freeze-drying, and gamma-irradiation on the antibacterial properties of donor human milk. Foods, 10(9), 2077. 

Cheema, S. K., Grimwade-Mann, M., Weaver, G., Collins, B., Shenker, N., & Cameron, S. (2025). Freeze-drying donor human milk allows compositional stability for 12 months at ambient temperatures. Journal of Food Composition and Analysis, 137, 106936. 

Dávila-Caraballo, G. J., Serrato-Márquez, E., Grimaldo-Rivas, M. D., Chuck-Hernández, C., Vega-Cantú, Y. I., Ortega-Alonzo, S. E., Coronado-Cerda, E., & Urrutia-Baca, V. H. (2024). Evaluation of freezing, ultra-cold freezing, and freeze-drying on the main components of human breast milk. Journal of Food Composition and Analysis, 136, 106712. 

Jarzynka, S., Strom, K., Barbarska, O., Pawlikowska, E., Minkiewicz-Zochniak, A., Rosiak, E., Oledzka, G., & Wesolowska, A. (2021). Combination of high-pressure processing and freeze-drying as the most effective techniques in maintaining biological values and microbiological safety of donor milk. International Journal of Environmental Research and Public Health, 18(4), 2147. 

Meng, F., Uniacke-Lowe, T., Ryan, A. C., & Kelly, A. L. (2021). The composition and physico-chemical properties of human milk: A review. Trends in Food Science & Technology, 112, 608–621. 

Rochow, N., Weiss, G. A., Knab, K., Prothmann, I., Schäfer, S., Zimmermann, J. L., Meis, A., Lohmüller-Weiß, S., Simon, K., Schäfer, S., Welsch, J., & Fusch, C. (2025). Freeze-dried donor milk for fortification of mother’s own milk in preterm infants: A preliminary observational study. Nutrients, 17(19), 3057.

Date of Input: 06/09/2026 | Updated: 06/09/2026 | nurulizzah