No Evidence for Different Dose Responses of Commercial Methionine Sources in Broilers

Abstract

Recently, Vazquez-Anon et al. (2006) published a paper in Poultry Science suggesting evidence for DL-2-hydroxy4-(methylthio) butanoic acid (DL-HMTBA) and DL-Met (DL-M) having different dose responses in growing broilers. The authors concluded that the 2 Met sources “have a different dose-response form, DL-HMTBA could outperform DL-M at commercial levels, and DL-M could outperform DL-HMTBA at deficient levels.” Additionally, it is suggested that relevant comparisons between the 2 Met sources should be conducted “at levels of commercial use.” These conclusions and suggestions are incorrect and misleading. When comparing essential nutrients or different sources of the same nutrient (L-Met in the present case), one should always evaluate the complete dose-response curve, starting with a deficient basal diet, followed by several incremental levels of the nutrients to be tested. Depending on the data structure of the respective dose-response trial, slope-ratio (linear response) or nonlinear models should be applied. These models can be used for estimation of comparative bioefficacy or bioavailability of any nutrients, such as P (Potter et al., 1995) or Fe (Boling et al., 1998). A simple example based on broiler weight gain data by Lemme et al. (2002) shall be used to demonstrate the statements above vs. the incorrect methodology and conclusions of Vazquez-Anon et al. (2006). Two Met sources, DL-M and diluted DL-M (65%, a product representing an internal standard with a known “relative efficacy” of 65%), were added at incremental levels to a Met-deficient commercial-type diet (Table 1). Using the proper exponential regression method to analyze the data sets, diluted DL-M (65%) had a calculated relative efficacy of 60% relative to DL-M (Lemme et al., 2002), confirming the validity of the used statistical method. Using a completely different approach to evaluate the data, the relative response rate of the basal diet was 64% of the maximum response (see Table 1). The relative response rates for the 0.06, 0.12, 0.18, and 0.24% supplemented DL-M levels were 84, 94, 99, and 100%, whereas the relative response rates for diluted DLM (65%) were 76, 86, 91, and 98%. Hypothetically, several different “effectiveness figures” were generated in this ex-