Letter to the Editor
Comments on ``Effects ofDL-methionine hydroxyanalogue (MHA) orDL-methionine
(DL-Met) on N-retention in broiler chickens and pigs [A. RoÈmer, H. Abel, Anim. Feed
Sci. Technol. 81 (1999) 193±203]''
The above-mentioned paper has recently been published in your journal. The major conclusion stated was that ``Both supplements showed equal effects on N-balances of chickens and pigs''. The authors also stated that ``. . .there was equal utilization of MHA
andDL-Met in. . .pigs'' and that ``... results indicate similar utilization of both sources of
methionine in growing chickens.''
Evaluation of the relative bioef®cacy for different sources of the same nutrient is essential for nutritionists and animal producers in order to use the sources correctly in everyday feed formulation. Especially amino acid sources are of great economical importance. Bioef®cacy of one nutrient source relative to another can only be determined based on a well replicated dose±response study carried out in a sensitive range of the investigated nutrient. Regression analysis is the proper tool to determine relative effectiveness of different sources of the same nutrient. Either the slope-ratio or a multi-expontial regression model is generally accepted as the appropriate method to derive relative ®gures for biological effectiveness.
The present study suffers from major weaknesses with regard to both the study design and the statistical analysis. In this letter we address these items brie¯y.
1. Study design and dose±response
The data presented in Fig. 1 do not show a clear dose±response relationship between ``methionine intake'' and ``N-retention'' for both broilers and pigs. Without a dose-dependent gradual response it is dif®cult to make a statement about differences between the bioef®cacy of nutrient sources. Especially in pigs, all of the supplemented groups show a very similar N-retention. Additionally, there is also a considerable variation within each treatment group. The authors evaluate the N-retention data with a quadratic regression model. However, it remains entirely unclear, how the two nutrient sources would be compared with such a model. The conclusion ``The regression analysis (Fig. 1(a)) showed no signi®cant difference. . .'' therefore, appears dif®cult to justify.
Moreover, at a previous conference in GoÈttingen/Germany the same broiler data were presented as N-retention as % of intake, which is a much more sensitive criterion for effectiveness than N-retention, g/d. There, the authors concluded that a lower bioef®cacy of MHA-FA relative to DL-methionine cannot be excluded. In particular, the ®rst three
data points are crucial in the present paper (Fig. 1, chickens): as given in Table 6, N
Animal Feed Science and Technology 88 (2000) 129±131
intake of treatments Met and MHA 1 was even lower than that of the basal treatment (2.49 and 2.51 g N/animal/day versus 2.97 g N/animal/day, respectively). It appears that the authors interpreted identical data in a completely different way, which is scienti®cally dif®cult to accept. It also raises doubt whether the quality of the data allows for interpretation at all. In a re-evaluation of the same data through exponential regression
Fig. 1. N-balance trials on the supplemental effect of MHA andDL-methionine in growing broilers. Biological
effectiveness of different methionine sources in broiler chickens.
analysis, our estimate of relative bioef®cacy of MHA-FA was 52% with an extremely large con®dence interval of 3±102% (see attached ®gure).
2. N-excretions in pigs
Similar comments apply to Fig. 2 (Effects of increasing methionine supplementation on urinary N-excretions in pigs). There is also no clear dose±response relationship, all of the six supplemented treatments have very similar urinary N-excretions. The lack of a sensitive dose±response function does not allow any conclusions with regard to bioef®cacy of essential nutrients.
3. Other comments
The methionine content of the basal diet should not be given as ``DL-Met'', Table 3.
Methionine hydroxyanalogue is not an amino acid, therefore, the term ``methionine equivalents'' should be used consistently, when referring to intake of different methionine sources (Figs. 1 and 2).
The paper, ``Walz and Pallauf, 1996'' in the discussion section was quoted incorrectly. In this trial, a supplemental ratio of 75:100 ofDL-methionine to MHA-FA (equimolar)
was tested in pigs. This ratio resulted in no differences in performance, although conducted in a suboptimal range. This experiment in fact suggests that the MHA-FA is of less biological effectiveness thanDL-methionine!
In summary, we believe that the data as such is not sensitive due to lack of a dose± dependent response and high within-treatment variation. On top of this, there is no adequate statistics to estimate relative biological effectiveness of the two nutrient sources to be compared. The present work therefore, is not suitable to conclude about the relative value of methionine sources.
We are looking forward to your response with best regards.
Received 6 December 1999; accepted 3 August 2000 M. Pack*, D. HoÈhler Feed Additive Division, Degussa-HuÈls AG, Postfach 1345 D-63403 Hanau (Wolfgang), Germany
*
Tel.:49-6181-59-2785; fax: 49-6181-59-2192.