Department of Anthropology Associate Professor Dr. Masako Fujita and co-author Eleanor Brindle recently published in the Cambridge University Press’ Experimental Results. The article, titled, “Comparing the creamatocrit of human milk before and after long-term freezing” evaluates the effect of long-term cryogenic storage on the creamatocrit, a technique for estimating the milk lipid content. This study found that the creamatocrit of human milk undergone 10 years of storage at ultra-low temperatures can provide values in high correlation with creamatocrit values obtained prior to storage. However, the results suggest a systematic bias that may vary with the amount of lipids that were in the milk in the first place. This bias may result in a subtle but systematic underestimation in the lower creamatocrit range and overestimation in the higher creamatocrit range. The authors call for future research to evaluate the correctability of this bias – if correctable, milk specimens in deep freezers of universities and milk banks can facilitate opportunities for research on human milk variation across time and space.
Read the full article at: https://doi.org/10.1017/exp.2022.9
Abstract:
Objectives: The creamatocrit is a simple technique for estimating the lipid content of milk, widely adopted for clinical and research purposes. We evaluated the effect of long-term cryogenic storage on the creamatocrit for human milk.
Methods: Frozen and thawed milk specimens (n = 18) were subjected to the creamatocrit technique. The specimens were reanalyzed after long-term cryogenic storage (10 years at <70°C). The correlation between pre- and post-storage values was tested, and their differences were analyzed using the Bland–Altman plot.
Results: The pre- and post-storage values were highly correlated (r = 0.960, p < .0001). The Bland–Altman plot revealed a positive association between their differences and means (Pitman’s test r = 0.743, p < .001), suggesting the presence of nonconstant bias across the creamatocrit range. Long-term storage of human milk may introduce subtle bias to the creamatocrit in replicating pre-storage values. Further research should evaluate whether this bias is statistically correctable.