Submitted to: Food Control
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: December 26, 2003
Publication Date: January 29, 2004
Citation: Whitaker, T.B. 2004. Standardization of mycotoxin sampling procedures: an urgent necessity. Food Control 14:233-237. Interpretive Summary: Mycotoxins are both toxic and cancer-causing compounds produced by molds that grow on some agricultural commodities. The U.S. Food and Drug Administration along with regulatory agencies from about 90 other countries have set maximum mycotoxin limits in foods and have defined sampling plans to detect contaminated foods destined for consumer use. However, mycotoxin limits and sampling plans differ widely from country to country, which makes international trade between exporters and importers difficult. The Food and Agriculture Organization and World Health Organization (FAO/WHO) have requested that United Nation member countries, working through UN committees, agree on a uniform mycotoxin limit and sampling plan to improve international trade and improve consumer safety. Because of the errors associated with each step of a mycotoxin sampling plan, the true mycotoxin concentration in a food shipment cannot be determined with 100 percent certainty and consequently some food shipments are misclassified as acceptable or unacceptable. Research by the U.S. Department of Agriculture has measured the magnitude of the errors associated with mycotoxin sampling plans and demonstrated how to best reduce these errors for available resources. The FAO/WHO have used these scientific results to establish a uniform maximum limit for aflatoxin and design a uniform aflatoxin sampling plan for raw shelled peanuts traded in the export market.
Technical Abstract: A mycotoxin-sampling plan is defined by the mycotoxin test procedure (sample size, sample preparation method, and analytical method) and the accept/reject limit. Because of the variability associated with each step of the mycotoxin test procedure, the true mycotoxin concentration of a bulk lot cannot be determined with 100% certainty. As a result, the sampling program will misclassify some lots. Some good lots will be rejected by the sampling plan (seller's risk or false positives) and some bad lots will be accepted by the sampling plan (buyer's risk or false negatives). The magnitude of these risks is directly related to the magnitude of the variability associated with the mycotoxin test procedure. It is difficult for an exporter to have an effective control program when regulatory limits and sample designs differ greatly among trading countries. In order to facilitate trade and provide protection for the consumer, it would be desirable for all trading countries to have the same mycotoxin limit and sample plan. While standardization of sampling plans among trading nations is important, any standardized sampling plan must be designed to minimize both the seller's and buyer's risks to the lowest possible levels that resources will allow. Reducing the variability of the mycotoxin test procedure will reduce both the buyer's and seller's risks. It is important to understand the sources of error in the mycotoxin test procedure so the errors can be effectively reduced. The sampling step usually is the largest source of error due to the extreme mycotoxin distribution among kernels in the lot. As an example, sampling (5 kg), sample preparation (USDA subsampling mill and 250-g subsample), and analysis (TLC) accounted for 83, 9, and 8% of the total aflatoxin testing error, respectively, when testing raw shelled peanuts for aflatoxin. Examples are given to show how increasing sample size reduces sampling error; increasing the fineness of grind and using larger subsamples reduces sample preparation error, and increasing the number of aliquots analyzed and using improved technology (HPLC versus TLC) decreases analytical error. International organizations such as FAO/WHO have used scientific techniques to evaluate and design aflatoxin sampling plans for raw shelled peanuts traded in the export market.