High-throughput <i>in vitro</i> prediction of available energy in feed ingredients for pigs using a novel computer-controlled digestion system

This study aimed to establish predictive equations for digestible energy (DE) and metabolizable energy (ME) of feed ingredients using a novel computer-controlled simulated digestion system (CCSDS) and validate the additivity of predicted energy values in complete diets for growing pigs. An in vivo experiment was conducted to determine the DE and ME of 30 experimental diets. A total of 60 barrows (initial BW of 37.3 ± 4.7 kg) were used, divided into two replicate blocks of 30 pigs each. Within each block, a 30 × 3 Youden square design was implemented across three experimental periods. Two pigs were allocated per diet during each period, resulting in six replicates per diet. Experimental diets included 20 feed ingredients, and 10 validation diets were formulated with the above feed ingredients to assess energy additivity and accuracy. The in vitro digestible energy (IVDE) was determined using CCSDS with five replicates for each diet. Strong correlations were observed between IVDE and in vivo DE (DE = 1.001 × IVDE + 180, R2 = 0.85, RSD = 310 kcal/kg of DM, P < 0.01) and ME (ME = 1.015 × IVDE-29, R2 = 0.89, RSD = 254 kcal/kg of DM, P < 0.01), with predicted values closely aligning with determined values across all feed ingredients. The mean IVDE: DE and IVDE: ME were approximately 0.95 and 0.99, respectively, highlighting the high predictive accuracy of CCSDS. Validation diets demonstrated consistent energy additivity. Single sample t-tests revealed no difference was observed between predicted and determined DE values in eight out of ten diets, and between predicted and determined ME values in nine out of ten diets. In conclusion, these findings underscore the utility of CCSDS as a cost-effective and reliable alternative to in vivo methods, offering significant potential for precise energy assessment in swine feed formulation.