Can we mimic in vivo digestion of the preterm newborn by an in vitro dynamic model?
Résumé
Understanding human milk digestion is a key step for developing biomimetic infant formulae. Ethical and technical reasons limit the possibility of in vivo trials. Relevant in vitro models are thus necessary. Our aim was to compare a dynamic in vitro digestion system developed by the French National Institute for Agricultural Research1 (DIDGI ®) against in vivo data collected in preterm newborns.
Supported by an exhaustive literature review2, the dynamic digester parameters were set in terms of types and amounts of enzymes, secretions, pH decrease and emptying rate to mimic as closely as possible the gastric digestive conditions of preterm newborns (1.9 kg). Raw or pasteurized (62.5°C, 30 min) preterm human milks (n=5 donors) were digested in vitro in triplicate. An in vivo study was conducted on hospitalized preterm newborns at Rennes Hospital (NCT02112331; n=12; mean ± SD: 28 ± 17 days of life, 1.5 ± 0.2 kg of bodyweight). In vitro and in vivo gastric digesta were sampled regularly (30 or 35 min, 60 min, 90 min). Structural changes were evaluated by confocal microscopy and laser light scattering. Lipolysis and proteolysis were determined by SDS-Page, thin-layer and gas chromatography methods. Gastric volume, pH and emptying rates were monitored.
During in vitro and in vivo digestions, gastric digestive kinetics were in overall similar and were not affected by pasteurization (p> 0.05), although this impacted the structural evolution of digesta. At 90 min of digestion, there was a similar degree of lipolysis (p>0.05) in vitro (11.3 ± 3.8%) and in vivo (11.5 ± 5.9%) and a similar proportion of residual proteins for β-casein and serum albumin (25.8 and 34%, in average). The latter was somewhat lower in vitro for lactoferrin (-11%) and for serum albumin (-20%). Curves of pH decrease and emptying rates could be improved to adjust closer in vivo gastric conditions.
Overall, our model appears as a relevant tool to study digestive kinetics of human milk and could also be applied for infant formula digestion. This will be useful to the scientist community and food manufacturers who focus on neonatal gastric digestion and infant formula optimization.