Human milk vs. Infant formula digestive fate: In vitro dynamic digestion and in vivo mini-piglet models lead to similar conclusions

Infant formula (IF), the only nutritionally adequate substitute for human milk (HM), still needs to be improved to be more biomimetic with HM, including in terms of digestive fate. The latter can be explored using different digestion models. The present study aimed to compare IF and HM digestion using in vivo (mini-piglet) and in vitro (dynamic system, DIDGI®) models. Fresh mature HM was collected and compared with a standard bovine IF. In vivo, 18 Yucatan mini-piglets (24-day-old) received HM or IF and were euthanized 30 min after the last meal. The entire digestive content was collected from the stomach to the colon. In vitro, the same meals were fed to an in vitro dynamic digestion model simulating the term infant at four weeks of age. Digesta were sampled regularly in the gastric and intestinal compartments. Structure (confocal microscopy and laser light scattering) and proteolysis (SDS-PAGE for residual intact proteins, OPA for hydrolysis degree, LC-MS/MS for peptides) were investigated along digestion. The digesta microstructure differed between HM and IF in a similar way between in vitro and in vivo digestion. In vitro gastric proteolysis of caseins and α-lactalbumin was significantly slower for HM than for IF, such as for the early intestinal proteolysis degree. In vitro bioaccessibility of free AAs explained only 30 % of the true ileal digestibility of AAs. Peptide mapping of caseins differed between HM and IF along their digestion. The relative peptide mapping data over six proteins from HM and IF were highly correlated between in vitro and in vivo digestion, particularly at 80 and 120 min of in vitro gastric digestion vs. in vivo stomach data and at 20 and 40 min of in vitro intestinal digestion vs. in vivo proximal jejunum data (r = 0.7-0.9, p < 0.0001, n = 1604). 40 to 50 % of the bioactive peptides identified in vivo were also found in vitro, with a good correlation of their abundances (r = 0.5, p < 0.0001, n = 61). Overall, in vitro and in vivo digestion were in good agreement, both indicating a different digestive fate for HM and IF.