Casein infusion rate influences feed intake differently depending on metabolizable protein balance in dairy cows: A multilevel meta-analysis.

The effects of casein infusion have been investigated extensively in ruminant species. Its effect on responses in dry matter intake (DMI) has been reviewed and indicated no significant effect. The literature reviewed in the current meta-analysis is more extensive and limited to dairy cows fed ad libitum. A total of 51 studies were included in the meta-analysis and data were fitted to a multilevel model adjusting for the correlated nature of some studies. The effect size was the mean difference calculated by subtracting the means for the control from the casein-infused group. Overall, casein infusion [average of 333 g of dry matter (DM)/d; range: 91 to 1,092 g of DM/d] tended to increase responses in DMI by 0.18 kg/d (n=48 studies; 3 outliers). However, an interaction was observed between the casein infusion rate (IR) and the initial metabolizable protein (MP) balance [i.e., supply minus requirements (NRC, 2001)]. When control cows were in negative MP balance (n=27 studies), responses in DMI averaged 0.28 kg/d at mean MP balance (-264 g/d) and casein IR (336 g/d), and a 100g/d increment in the casein IR from its mean increased further responses by 0.14 kg/d (MP balance being constant), compared with cows not infused with casein. In contrast, when control cows were in positive MP balance (n=22 studies; 2 outliers), responses in DMI averaged -0.20 kg/d at mean casein IR (339 g/d), and a 100g/d increment in the casein IR from its mean further decreased responses by 0.33 kg/d, compared with cows not infused with casein. Responses in milk true protein yield at mean casein IR were greater (109 vs. 65 g/d) for cows in negative vs. positive MP balance, respectively, and the influence of the casein IR on responses was significant only for cows in negative MP balance. A 100g/d increment in the casein IR from its mean increased further responses in milk true protein yield by 25 g/d, compared with cows not infused with casein. Responses in blood urea concentration increased in casein studies (+0.59 mM) and the influence of the casein IR was greatest for cows in positive MP balance (0.26 vs. 0.11 mM per 100g/d increment). Responses in DMI were also correlated negatively with responses in blood urea concentration only for cows in positive MP balance. Together, these results suggest an association between satiety and deamination and oxidation of AA supplied in excess of requirements for cows in positive MP balance. Therefore, casein stimulated appetite in cows fed MP-deficient diets possibly via the supply of orexigenic AA or through a pull effect in response to an increased metabolic demand. Conversely, casein induced satiety in cows fed diets supplying MP in excess of requirements. Not precluding other factors involved in satiety (e.g., insulin, gut peptides), casein could have increased the supply of AA (e.g., Ser, Thr, Tyr), which might depress appetite at the brain level or increase the deamination and the oxidation of AA in oversupply in agreement with the hepatic oxidation theory.