A retail prevalence study to investigate microbiological contamination levels in ready-to-eat (RTE) plant-based dairy and meat substitutes

The prevalence study investigate the microbiological quality of ready-to-eat (RTE) plant-based dairy and meat substitutes in the European Union. It provode data on the numbers of these products potentially contaminated with microbial hazards at retail level. The analytical testing corried oit at the end of shelf-life in order to determine the worst-case scenario for contamination with L. monocytogenes and other foodborne bacterial pathogens at the end of the foods shelf-life duration. The information generated contributes relevant prevalence data to the evidence base, and be useful further microbial risk assessment of these products. The main objectives/goals of the project are as follows: - to determine the microbial quality of RTE plant-based dairy and meat substitutes at the end of the foods labelled shelf-life duration - to generate relevant prevalence data taht can be used for future microbial risk assessment of this projects - to collaborate with other Member States of the EU in order to generate relevant scientific evidence, and - to raise awareness regarding the possible food safety risks for consumers of RTE plant-based dairy and meat substitutes

A total of 104 samples from plant-based substitutes for milk (n=23) and cream (n=7), cheese (n=33), meat (n=30) and fish (n=11) were included in the study. Samples were collected from supermarkets in Berlin, including discount supermarkets, regular supermarkets and stores specializing in organic food. Some products were bought from a wholesaler and from two online shops. Various plant-derived proteins or oils contributed the main ingredients. In most products, except for plant-based milk substitutes, oils, starches or other plant-based ingredients supplemented the protein sources, or blends of proteins were used. The pH value varied between the product categories. Plant-based milk substitutes had an almost neutral pH, with the 25-75 % quantile ranging from 6.6 to 7.4 (mean 7.0, median 7.0). Other food categories were slightly acidic, with plant-based meat substitutes (25-75 % quantile 4.9-4.9, mean 5.5, median 5.6) and plant-based fish substitutes (25-75 % quantile 5.0-5.8, mean 5.4, median 5.6) showing similar pH ranges. Plant-based cheese substitutes were more acidic (25-75 % quantile 4.4-5.2, mean 4.9, median 4.7). Plant-based substitutes for meat, fish and cheese had mean and median aW of 0.95 (25-75 % quantile 0.94-0.96), with some products showing higher or lower values. Plant-based milk substitutes had higher aW (25-75 % quantile 0.96-0.97, mean 0.97, median 0.96).The retail prevalence study investigating the microbiological contamination levels in ready-to-eat (RTE) plant-based dairy, meat and fish substitutes indicated no microbiological concern for samples taken in Germany with regard to Listeria monocytogenes, Salmonella spp., Shigatoxin-producing E. coli (STECshort forShiga toxin-producing Escherichia coli), Clostridium perfringens and Staphylococcus aureus. As several Salmonella outbreaks in Germany in recent years have been associated with products containing e. g. cashew or coconut, which are also used as main ingredient in plant-based RTE foods, the absence of Salmonella in this study is reassuring.None of the samples tested positive for L. monocytogenes, Salmonella spp. or C. perfringens. Coagulase-positive Staphylococci (CPS) were detected in one of 104 products (1,0 %). This positive sample was a plant-based milk substitute, and CPS levels were low (<10 cfu/mlshort formillilitre). Species identification by MALDI-TOF-MS assigned the isolate to S. Aureus, and the isolate showed no hemolytic activity on sheep blood agar.For E. coli, one plant-based milk substitute tested positive for stx2 gene by real-time PCR but could not be confirmed by culture and was therefore classified as presumptive STECshort forShiga toxin-producing Escherichia coli positive in accordance with ISO/TS 13136:2012. As the sample was labelled as ultra-high temperature treated, it was assumed that residual stx2-DNA or unspecific matrix effects accounted for the signal.General E. coli were enumerated at levels >20 cfu/g in 2 of the 104 samples, with counts of 720 and 910 cfu/g, respectively, in two fermented soft cheese alternatives with cashew as the main ingredient. Currently, there are no general microbiological limits for E. coli in RTE food in Germany; guidance values are provided by the German Society for Hygiene and Microbiology (DGHM) for specific products or categories, but no limits are defined for E. coli in soft cheese or vegan cheese substitutes.In contrast to these largely favourable findings for classical hygiene indicators and major foodborne pathogens, signals of concern emerged for other hazards. Although L. monocytogenes was not detected, the non-pathogenic species L. innocua was isolated from two plant-based cheese substitutes and one plant-based meat substitute, with one sample showing counts above 100 cfu/g (3100 cfu/g). L. innocua frequently occurs alongside L. monocytogenes in food production environments, and both species exhibit similar environmental survival strategies, so its detection can serve as an indicator of inadequate hygiene during food production and a warning that hygiene concepts in the manufacturing plants should be improved to prevent future contamination with Listeria spp., including L. monocytogenes.In addition, a high prevalence of Bacillus cereus (sensu lato) was observed, beeing detected in 37 of 104 products (36 %). Prevalence within the different food categories was 0 % for plant-based dairy milk substitutes, 14 % for plant-based dairy cream substitutes, 45 % for plant-based cheese substitutes, 50 % for plant-based meat substitutes and 55 % for plant-based fish substitutes. However, the colony counts were generally low. Only two plant-based cheese substitute samples showed B. cereus (s.l.) counts of 2100 and 3000 cfu/g, respectively. In total, 39 isolates were obtained from the 37 positive samples (two samples yielded two distinct isolates). All isolates carried the enterotoxin genes nheA/B/C, partly in combination with hblC/D/A or cytK-2. The cytK-1 gene, associated with a more cytotoxic variant of cytotoxin K, was not detected, but the sample with the highest colony count contained an isolate harbouring the ces gene cluster responsible for production of the emetic toxin cereulide. None of the isolates was able to form parasporal crystals, a characteristic attributed to the species B. thuringiensis.Taken together, these findings indicate the generally satisfactory microbiological quality of RTE plant-based dairy, meat and fish substitutes available in Germany. However, risk-based monitoring of foodborne pathogens such as Salmonella or STECshort forShiga toxin-producing Escherichia coli in these very heterogenous RTE foods appears warranted, given that raw materials have previously been associated with pathogen outbreaks, processing may introduce contamination, and the variable physicochemical properties of these products do not necessarily preclude the presence and growth of harmful bacteria.Finally, microbiological results of our study specifically highlight the need for continued vigilance and targeted hygiene improvements to mitigate risks posed by B. cereus and Listeria.