Are infant formulas biomimetic of human milk digestive behavior? A dynamic in vitro study
Résumé
Human milk is the ideal food for infant nutrition and optimal growth. When breastfeeding is not available, infant formulas are often administrated. The chemical composition of infant formulas has been optimized but human milk structure and bioactive compounds contents are not mimicked1. These initial differences may modulate the kinetics of digestion and of disintegration of the emulsion in the digestive tract of the newborns. To check this hypothesis, the digestive behaviors of a commercial infant formula (IF) and raw human milk (HM) were compared using a dynamic in vitro digestion system (DIDGI®)2. Supported by an exhaustive literature review3, the dynamic digester was set to mimic as closely as possible the digestive conditions of term newborns. Pooled term HM (n=5 donors)4 and a liquid commercially available IF were digested in triplicate. Samples were
collected from the gastric and intestinal compartments throughout digestion. Initial emulsions and digesta were characterized. Lipolysis and proteolysis kinetics were monitored by SDSPage, thin-layer and gas chromatography methods. Structural changes were evaluated by confocal microscopy and laser light scattering.
The initial differences in structure, composition and pre-hydrolysis state of the two emulsions strongly impacted digestion kinetics and deconstruction. Lipolysis levels were higher in HM than in IF before digestion and during gastric phase. This order was reversed at the beginning of the intestinal phase. Gastric proteolysis of caseins was accelerated in IF compared to HM. As observed by confocal microscopy and in coherence with previous results on model infant emulsions5, the differences in initial structure of the emulsions were maintained during the gastric phase of digestion: presence of native human milk fat globule in HM (5.1 ± 0.2 μm) and submicronic lipid droplets in IF (0.2 ± 0.0 μm). HM presented a structural destabilization from 60 min of gastric digestion, with a bimodal distribution of the particles size (mode 1 = 76.5 μm, mode 2 = 6.5 μm). IF gastric digestion was marked by average mode diameter changing drastically from 90 min (16.6 ± 2.2 μm). A progressive disruption of these aggregates was observed in the intestinal phase for both HM and IF. Despite in constant evolution, infant formulas are not yet biomimetic of human milk digestive behavior. These differences may have important physiological implications and underline the need for developing soft processes with minimal impact on milk native emulsion structure in the near future.