In vitro starch digestion kinetics of diets varying in resistant starch and arabinoxylan compared with in vivo portal appearance of glucose in pigs

- The in vitro method ranked the three diets in a similar relative manner as in vivo.
- The rate of glucose release was much faster in vitro than in vivo.
- The in vitro assay was able to mimic the relative absorption in vivo when the starch structure was the limiting factor.
- The in vitro assay was less able to account for the influence of either the soluble AX or the effect of the protein structure.

The current study was undertaken to investigate the relationship between the in vitro starch digestion kinetics and the in vivo portal glucose appearance in pigs used as models for humans. In vivo data were obtained from a previous study where the portal glucose appearance was obtained from six catheterised pigs equipped with permanent catheters in an artery and the portal vein and with a flow probe attached to the portal vein for monitoring the blood flow rate. Three experimental diets were studied - a low dietary fibre (DF), western-style diet (WSD) and two high-DF diets containing resistant starch (RSD) or arabinoxylan (AXD). A modified Englyst-assay involving gradual glucose hydrolysis over a time frame of 6 h was used in vitro. The in vitro starch digestion kinetics was modelled using a mechanistic growth model (R2 > 0.995), whereas the in vivo data were better described by a sigmoid Gompertz model (R2 > 0.997). The estimated plateau values were higher in vitro than in vivo but the diets were similarly ranked; ~ 95% for AXD and WSD and 81.8% for RSD in vitro and ~ 86% and 76.6% for the same diets in vivo. The rate of glucose release in vitro was much faster than the portal glucose appearance in vivo (0.0347-0.0705 versus 0.0136-0.0197% starch/min) with the starch in RSD being the most slowly degradable. This difference was most likely an effect of gastric retention. In conclusion, the in vitro method ranked the three diets in a similar relative manner as in vivo but the rate of glucose release was much faster in vitro than in vivo. It was only when the starch structure set the limit for the starch hydrolysis, however, that similar relative results were obtained.