Meta-analysis to predict the effects of metabolizable amino acids on dairy cattle performance

Meta-analytic methods were used to determine statistical relationships between metabolizable AA supplies and milk protein yield, milk protein percentage, and milk yield in lactating dairy cows. Sixty-three research publications (258 treatment means) were identified through a search of published literature using 3 search engines and met the criteria for inclusion in this meta-analysis. The Cornell Net Carbohydrate and Protein System (CNCPS) version 6.5 was used to determine dietary nutrient parameters including metabolizable AA. Two approaches were used to analyze the data. First, mixed models were fitted to determine whether explanatory variables predicted responses. Each mixed model contained a global intercept, a random intercept for each experiment, and data were weighted by the inverse of the SEM squared. The second analysis approach used classical effect size meta-analytical evaluation of responses to treatment weighted by the inverse of the treatment variance and with a random effect of treatment nested within experiment. Regardless of the analytical approach, CNCPS-predicted metabolizable Met (g/d) was associated with milk protein percentage and yield. Milk yield was positively associated with CNCPS-predicted metabolizable His, Leu, Trp, Thr, and nonessential AA (g/d). Milk true protein yield was also associated with CNCPS-predicted metabolizable Leu (g/d). Predicted metabolizable Lysine (g/d) did not increase responses in production outcomes. However, mean metabolizable Lys supply was less than typically recommended and the change with treatment was minimal (157 vs. 162 g; 6.36 vs. 6.38% metabolizable protein). Experiments based solely on Lys or Met interventions were excluded from the study database. It is possible that the inclusion of these experiments may have provided additional insight into the effect of these AA on responses. This meta-analysis supports other research indicating a positive effect of Met and His as co-limiting AA in dairy cows and suggests Leu, Trp, and Thr be given greater consideration in future research.