Quantifying the effects of dietary fibres on protein digestibility in pigs - A review

Summary

Dietary fibres may negatively affect apparent ileal digestibility (AID) of crude protein (CP) in different ways. Insoluble dietary fibres (IDF) typically limit AID of CP through physical encapsulation of proteins by the cell wall matrix and increasing digesta bulk, whereas soluble dietary fibres (SDF) may alter digesta physicochemical characteristics, such as viscosity, to various extents particularly when originating from cereal ingredients, consequently reducing protein hydrolysis and amino acid absorption. Both IDF and SDF may increase endogenous losses, but the underlying mechanisms can differ. Hence, we hypothesised that the extent of reduction in protein digestibility will differ for IDF vs. SDF, and cereal vs. non-cereal fibre sources. An analysis was performed to quantify the relation between total dietary fibres (TDF), SDF, or IDF and AID of CP. Data were collected from peer-reviewed publications (n = 58) and feed ingredient tables (Centraal Veevoeder Bureau, CVB, n = 40; National Research Council, NRC, n = 18; China Nutrient Requirements of Swine, MAFIC, n = 33), and analysed by simple ordinary least squares regression. Across databases, AID of CP decreased by 0.6 to 0.7%-units per g/100 g increase of TDF for cereal fibre sources (P < 0.01) and by 0.5 to 0.9%-units for non-cereal fibre sources (P < 0.01), except for NRC (P = 0.220). This was mainly explained by SDF for both cereal (−1.4 to −4.2%-units per g/100 g increase, P < 0.01) and non-cereal fibre sources (−0.6 to −3.3%-units, P < 0.01), except in the peer-reviewed publications database (P = 0.482). A greater reduction in AID of Thr (0.6 to 0.9%-units per g/100 g increase in TDF, P < 0.01) vs. Met (0.3 to 0.7%-units, P < 0.01), except for NRC (P = 0.62), illustrates the role of endogenous losses with incremental TDF inclusion. In conclusion, AID of CP decreased by 0.5 to 0.9%-units per g/100 g increase in TDF. In cereal ingredients this was more pronounced for SDF, as expected based on the type of fibres present. Finally, we propose regression coefficients obtained from the four databases that can be used to correct AID of CP for both ingredients and complete diets with varying contents of dietary fibres of different composition for use in practical feed formulation.