| Summary: | There are wide variations in the published estimates of essential amino acid (EAA) requirements. Variationsare thought to originate from different choices in mode of expression, response variable, and mathematicalmodel. Here we conduct a meta-analysis of the growth-based, dose-response trials of 10 EAA in 22 teleostspecies: 249 studies were reviewed. Published data were entered in a spreadsheet and re-calculated across in astandard and systematic manner to allow comparisons. The considered unit of requirement were percentage ofdry diet, g of EAA per MJ of digestible energy (DE), and g of ingested EAA per kg of metabolic body weight(MBW) per day. Response variables included growth in g per kg MBW per day and thermal-unit growthcoefficient (TGC). Four mathematical models were also compared: broken-lime model (BLM), quadratic model(QM), broken-quadratic model (BQM), and saturation kinetic model (SKM). Results first indicate importantdifferences in study quality, as 54% of the reviewed papers were excluded from the meta-analysis, often timesbecause of poor growth or missing information. Additionally, the final dataset was greatly fragmented: 31% ofthe studied concerned rainbow trout, and lysine was the focus of 29% of all studies, leaving some species andEAA poorly covered. Comparisons of the requirement estimates show important variations between studies,even within species. With such variability there was no difference in requirement estimates calculated withdifferent response variables. Similarly, this variability was not different between the three modes of expression,nor was it between mathematical models. However, there were significant effects of experimental design on thequality of fit of the models. Specifically, experiments that failed to produce a clear, plateauing dose-responsecurve had greatly increased probability to yielding absurd results (e.g. negative requirement). Finally, thepresent study emphasizes the critical need of a global, standard and systematic system to report and captureresults from nutrition trials.
|
|---|
