Skip to main content

Scientific Opinion on the public health risks related to mechanically separated meat (MSM) derived from poultry and swine

EFSA Journal logo
Wiley Online Library

Meta data

Abstract

The purpose of this assessment was to identify public health risks linked to mechanically separated meat (MSM) types from pork and poultry and compare them with fresh meat, minced meat and meat preparations (non-MSM); and to select, rank and suggest objective measurement methods and values for parameters to distinguish MSM types. Microbial hazards in MSM are expected to be similar to those in non-MSM, although the risk of microbial growth increases with the degree of muscle fibre degradation, thus with the separation pressure. For the distinction between the different types of MSM and non-MSM chemical, histological, molecular, textural and rheological parameters were considered as potential indicators. The analysis of available published data suggested that calcium and, if confirmed cholesterol content, was the only appropriate chemical parameters which could be used to distinguish MSM from non-MSM products. On the basis of published data, a model was developed to derive probabilities for a product to be classified as MSM based on the calcium content. Calcium content of 100 mg/100 g, as specified in the Reg. (EC) No. 2074/2005, corresponds to probability of 93.6% for a product to be classified as MSM, according to the model developed. Calcium content alone does not allow differentiation between low pressure MSM and other meat products, and other validated tests would be necessary. Histological parameters considered include microscopic detection of different tissues and their changes. The latter is a promising method for distinction of MSM types, but further validation is needed. In order to improve methods for MSM identification, specifically designed studies for the collection of data obtained by standardised methods on indicators such as calcium and cholesterol should be undertaken, while studies based on combinations of different parameters could also be useful.