Protein efficiency ratio: AACC/ASTM collaborative study.

Eight laboratories (7 of the laboratories conducted animal experiments) participated in a collaborative study to standardize some of the methodology associated with animal bioassays for determining protein efficiency ratios and to suggest improvements which would reduce the variation among laboratories. One-, 2-, 3-, and 4-week protein efficiency ratios (PER) with 0-, 2-, or 4-day adaptation periods were obtained from each laboratory, respectively, for 6 protein sources: casein, lean beef, lactalbumin, textured vegetable protein, peanut flour, and wheat flour. Analyses were computed for PER and adjusted PER (APER). From the analysis of variance for PER and APER, significant (P less than 0.05) effects were observed due to laboratories, adaptation length, protein sources, and/or interactions among these variables. In general, APER values show much less variation among laboratories than PER values. The reproducibility and repeatability variances were significantly (P less than 0.05) greater for an assay length of 2 weeks than they were for 3- or 4-week assays. Two protein sources, casein and textured vegetable protein, were fed at both high (10%) and low (6%) levels of protein. Analysis of variance of PER values shows a significant (P less than 0.05) laboratory by protein level by assay length interaction.