Dynamic Remodeling of the Human Milk Serum Proteome Across Lactation: A Paired Two-Stage DIA Proteomic Study in Term...

Human milk composition changes across lactation, but paired within-subject proteomic analyses comparing longitudinal trajectories in term and preterm milk remain limited. We aimed to characterize stage-associated proteomic changes within each cohort and determine whether longitudinal remodeling is shared or divergent between term and preterm lactation.

In this single-center prospective study conducted at the Neonatal Intensive Care Unit, Jagiellonian University Medical College, Kraków, Poland (October 2020-November 2021), 40 lactating mothers (20 preterm, <32 weeks' gestation, mean age 29.4 ± 6.1 years; 20 term, 37-42 weeks, mean age 30.2 ± 5.5 years) provided paired milk samples at ≤10 days postpartum and week 5. Milk serum proteomes were analyzed by quantitative data-independent acquisition mass spectrometry; differential abundance was assessed using two-sample t-tests with Storey false discovery rate correction (q < 0.05) and fold-change >1.5, followed by ClueGO pathway enrichment.

Stage-associated differential abundance was identified for 108 proteins in term milk (58 increased, 50 decreased) and 103 in preterm milk (64 increased, 39 decreased). Of these, 87 were shared between cohorts (80.6% of term, 84.5% of preterm set) with concordant directionality. Shared upregulated pathways included oxidative stress response and glycolysis (e.g., PRDX5, fold change 2.49, q = 0.044); shared downregulated pathways related to mucosal immunity (e.g., tenascin, fold change 9.61-11.41, q < 0.0001). Cohort-specific pathway signals were limited relative to shared remodeling.

The human milk serum proteome undergoes substantial longitudinal remodeling in both term and preterm lactation, with most changes following a common temporal pattern; prematurity-related differences appear selective rather than global. These findings support lactation stage as a key determinant of milk proteomic composition and underscore the value of longitudinal, stage-aware study designs, although formal time-by-group interaction testing was not performed.