Intercomparison Study

The aims of the intercomparison study were to optimize the methods in a wide range of laboratories, and assess the laboratories, before carrying out the certification analysis. A meeting of all the participants was held to agree the methods that should be investigated, to identify sources of error for each method and to provide suggestions for their avoidance and control. The participants in the study agreed that five methods should be used for the assessment. These were the Uppsala method (9), the AOAC Prosky method 985.29 (2), both versions of the En- | glyst method (gas chromatographic and colorimetric) (8) and a recent version of j the AOAC method 991.43 by Lee, with MES-TRIS buffer (10). -a

The participants agreed that some changes or recommendations should be incorporated into the protocol of each method to optimise their performance.

These were a combination of the participants' experience, some recent improve- |

ments to methods and some results from previous studies.

The changes in the Englyst procedure were all subsequently published by Englyst, Quigley and Hudson (8). They were presented to the meeting in advance of publication. The recommendations for the AOAC procedures were mainly to highlight points in the existing method that needed standardizing to avoid sources of variability. The only change to the Uppsala method was the inclusion of a

New Food Reference Materials

factor to compensate for the difference in degradation during the hydrolysis procedure between free galacturonic acid in the standard solution and polymeric galacturonic acid in the sample. The use of this factor resulted in better agreement with other similar methods. It is now included in the published method (9).

The intercomparison study was designed to eliminate as many potential variables as possible, and some materials were supplied by the coordinator. It had originally been intended to supply the Uppsala participants with the same enzymes as the AOAC participants. However, it was found that the amyloglucosi-dase issued to the AOAC participants contained some unwanted hemicellulase activity. This caused losses in the Uppsala procedure, which has longer incubation stages. The Uppsala participants were therefore supplied from a purer source. This illustrates the importance of checking the activity of the enzymes in the method to be used.

Preliminary Intercomparison Analyses

Before starting the main intercomparison analysis, the participants were asked to analyze a mixture of four polysaccharide materials arabinogalactan (larch gum), fibrous cellulose, sugar free citrus pectin, and galactomannan (alcohol washed guar gum), and to obtain results within the known value ±2 SD (based on the performance in the intercomparison study). Participants who were using gas chro-matography were asked, in addition, to carry out a similar exercise on a mixture of sugars. These two preliminary tests proved to be very useful, as they highlighted problems in some laboratories with the filtration step of the AOAC procedures, miscalculations, misunderstandings of the role of blanks, problems with response factors in GC, a problem with the makeup of a sugar standard and of the unknown sugar mixture, and uncovered minor errors in the protocol.

Because of this exercise, many laboratories subsequently produced very good data for the main intercomparison study by all the methods.

Table 2 shows the very low coefficients of variation between laboratories achieved with all the methods, and the good agreement between the mean values for dietary fiber. This agreement using the polysaccharide mix contrasts with the large difference in dietary fiber values found between the Englyst method and AOAC methods for many food materials and underlines its applicability for initial quality control purposes.

Main Intercomparison Analyses

The choice of suitable test materials for the study was limited. The five new reference materials could not be used because the participants must eventually carry out the analysis for certification without knowing the expected results. Alternate materials had to be found. Four dried reference materials that had already been tested for homogeneity and stability were kindly supplied by the Food Re-


Table 2 The Dietary Fiber Levels and Coefficients of Variation Between Laboratories Found in the Preliminary Intercomparison Test with a Mixture of Polysaccharides

Method # of Labs g/100 g db (between labs)

Englyst GC 7 92.5 3.7

Englyst Color 9 91.0 4.0

Uppsala 5 91.2 3.3

search Institute at Norwich. They were white bread (RM 416), wholemeal bread (RM 417), a mixture of wholemeal and white bread (RM 418) and cornflakes (RM 419). These materials were analyzed by all five methods (Table 3).

The dietary fiber levels given by methods which do not exclude residual starch and lignin (Uppsala and AOAC) were higher than the values by the Englyst methods that only measure non-starch polysaccharides.

The standard deviation between laboratories was low for all methods, except the Uppsala method. This is an improvement with respect to the Englyst methods, because the uncertainty of past results was too high to permit certification. This fulfilled a major purpose of the intercomparison study which was to check that the methods, when applied by the participating laboratories, were sufficiently repeatable and reproducible to be used for certification.

The results by the AOAC 985.29 Prosky method and the AOAC 991.43

Table 3 Dietary Fiber Values of Four Cereal Based Reference Materials Analyzed by Five Methods During the Intercomparison Study

Reference material

AOAC 985.29 Englyst GC Uppsala AOAC 991.43 Englyst Color

# of labs White bread Wholemeal


8.0 ± 0.6 6.3 ± 0.5 8.2 ± 0.6 8.1 ± 0.5 6.8 ± 0.4

Dietary fiber values, g/100 g dry basis Mean ±1 SD


New Food Reference Materials

Lee method were compared statistically. The mean results for each reference material by the two methods were not significantly different (p > 0.05 t-test).

The participants met to discuss the results of the intercomparison study and to agree upon the criteria which would be applied to the results of each laboratory to decide which laboratories should take part in the final certification study. They also agreed upon limits that would be applied to the results of the certification study to decide whether the data was sufficiently repeatable and reproducible to be submitted for certification. They agreed upon the details to be incorporated in the instructions for the certification study to give the maximum chance of success. The project was funded on the basis of using three methods for certification and the participants agreed on the AOAC 985.29 Prosky method, the Englyst GC method and the Uppsala method. However, several laboratories had obtained very good results with the AOAC 991.43 method of Lee with MES-TRIS buffer and with the Englyst colorimetric method and they volunteered to undertake the certification analyzes without payment.

Lose 10 Pounds Naturally

Lose 10 Pounds Naturally

Studies show obesity may soon overtake tobacco as the leading cause of death in America. Are you ready to drop those extra pounds you've been carrying around? Awesome. Let's start off with a couple positive don't. You don't need to jump on a diet craze and you don't need to start exercising for hours each day.

Get My Free Ebook

Post a comment