We often answer questions from health care providers about our detailed and evidence-based safety protocols, including donor screening modeled after blood banking and milk storage, and testing and pasteurization techniques compliant with FDA food preparation regulations and the Human Milk Banking Association of North America (HMBANA).
Recently, we have received a number of queries about milk analysis. Milk analysis is a relatively new addition to the tools available to neonatologists and NICUs. As a new technology, it takes time to conduct research and to determine best practices.
The Mothers’ Milk Bank Northeast (MMBNE) Medical and Research Advisory Boards have considered this issue carefully. The state of the current technology does not support a decision to analyze our milk for nutritional content. We and other colleagues in the medical, research, and milk banking community are concerned about the lack of evidence on the accuracy and clinical utility of the analyzers now available and in use.
The primary source of carbohydrates in both cow milk and human milk is lactose. However, human milk also contains large amounts of other complex carbohydrates, such as oligosaccharides and glycoproteins. The current milk analyzers were designed to analyze cow’s milk, whose simpler composition and lower variation allows the measured total carbohydrates to be the surrogate measure for lactose, with the assumption that the non-lactose carbohydrate is both constant and small in amount. However, in human milk, many carbohydrates are in the form of glycoproteins and oligosaccharides, whose composition varies widely among mothers. When the analyzers are modified for measuring human milk, the assumptions used in the new algorithm may not account for the wide individual variation in human milk.
Neonatal Intensive Care Units (NICUs) want analysis of milk to account for individual variation when fortifying donor human milk, but the analyzers do not measure that individual variation well. To address this problem, HMBANA guidelines require milk banks to pool milk from multiple donors to achieve a good mix – an average – of components of mothers’ milk. This reduces the variation among batches of pasteurized donor milk. This is designed to make each batch of donor milk as close to the literature values for mature milk (milk produced from one month parturition onward) as is feasible. MMBNE’s policy is that it is better to provide representative (averaged) milk than to provide potentially inaccurate and unreliable nutritional information.
To date, the FDA has not approved any milk analyzer for use on human milk to inform clinical decision-making. In discussion with FDA and researchers in the area of human milk, it is clear that reliable and accurate analysis of human milk is very difficult to achieve, but research is underway. Currently, several research studies are evaluating alternative milk analyzers for use on human milk for FDA approval. When these research efforts are complete and the FDA makes a ruling on one or more analyzers for use on human milk, MMBNE will revisit this issue and, per HMBANA Guidelines requirements, we will inform all of the hospitals using MMBNE milk which analyzer we are using and the protocols for measuring and calibrating our milk analysis.
Until there is an approved standardized method for the analysis of individual human milk samples, using the literature values for mature human milk as the basis for supplementation to achieve the needs of an individual neonate seems to be the best and most conservative practice.
Robert Insoft, MD
Medical Director, MMBNE
Chair, MMBNE Medical Advisory Board
Chief Medical Officer, Women and Infants Hospital, Providence, RI
David S. Newburg, PhD
Chair, MMBNE Research Advisory Board
Chair, HMBANA Microbiology Committee
Naomi Bar-Yam, PhD
Executive Director, MMBNE
On behalf of the Mothers’ Milk Bank Northeast (MMBNE) Medical and Research Advisory Boards