Quality Assessment of Raw Human Milk Under Prolonged Household Freezing: Titratable Acidity, Energy Content, and Microbiological Safety

This dataset was generated from a prospective observational cohort study designed to evaluate the quality of raw human milk (RHM) stored under household freezing conditions for up to 150 days. The study was conducted in July 2025 at the Nutrition and Dietetics Laboratory of the Faculdade Israelita de Ciências da Saúde Albert Einstein (FICSAE), São Paulo, Brazil. The research was motivated by the discrepancy between Brazilian regulations, which recommend a maximum freezing period of 15 days for expressed human milk, and international guidelines that allow storage for three to twelve months for healthy full-term infants. The dataset was created to provide empirical evidence on the behavior of selected milk quality indicators during prolonged domestic freezing. Milk samples were obtained from 28 lactating adults and analyzed at seven storage time points: 1, 15, 30, 60, 90, 120, and 150 days after collection. Three quality indicators were assessed at each time point: titratable acidity (°D), measured by the Dornic titration method; energy content (kcal/L), estimated by the crematocrit technique; and total coliform occurrence, determined by the presumptive phase of the Brilliant Green Bile Lactose (BGBL) broth method. The dataset includes raw analytical measurements, derived variables, participant-level summaries, sociodemographic and lactation-related characteristics, and a complete data dictionary. Data are available in both wide and long formats to facilitate descriptive, longitudinal, and inferential analyses. Statistical output files generated in Jamovi are also provided to support reproducibility. The data show longitudinal variation in titratable acidity during storage, while energy content remained measurable throughout the 150-day period. Total coliform occurrence was infrequent and recorded as a binary outcome (Absent/Present). Summary statistics for all three indicators are included in the dataset documentation. Missing observations (21.9%) reflect variability in expressed milk volume among participants and are documented at the individual level. This dataset can be used to investigate changes in human milk quality during domestic frozen storage, validate statistical models of acidity and energy trajectories, perform secondary analyses involving donor characteristics and milk quality parameters, and support evidence-based discussions regarding human milk storage recommendations. To the authors’ knowledge, it represents the first Brazilian longitudinal dataset evaluating raw human milk quality at seven storage time points extending to 150 days under household freezing conditions.