Chemical Contaminants in Plant Based Protein Survey

Food Standards Scotland (FSS) provided the detailed list of suggested products. Sample purchase was carried out by staff of FSS. Samples were purchased in Scotland from supermarkets and on-line retailers. Samples were collected, packaged appropriately to maintain chilled temperature during transport and sent by courier to the laboratory. The sampling plan requested 25 samples, but an additional sample of vegan sausages was purchased which meant a total of twenty-six samples were included in the survey. 

Samples were purchased in November 2024. A small number of samples (5 out 26) were purchased from on-line retailers, the remainder were purchased from supermarkets. For most products three retail packs from one batch of each product were purchased, although in some cases either depending on availability or pack size, 2 packs or 4 packs were purchased, resulting in sample sizes from 500 g to 1000 g. A full detailed list of the samples collected has been provided separately, and a summarised list is given in Annex B (Table B1). 

The second phase of sampling was carried out in August 2025. The samples were selected and purchased by FSS from stores in Scotland and on-line retailers. As for Phase 1 replicates of each sample were purchased to obtain a target composite sample size of 500 g to 1000 g. Samples were packaged with ice packs to maintain chilled temperature during transport and sent by courier to the laboratory.

Twenty-six samples were purchased. A full detailed list of the samples collected has been provided separately, and a summarised list is given in Annex B (Table B2).

The samples comprised the following categories – protein bars (4 samples), ready-to-eat protein shakes (3), protein powders (9) and meat alternatives (10). The products were made from soy, pea and wheat protein, or combinations of these. One sample was a mixture of rice and pea protein and one was a mixture of whey and soy protein. 

This was a limited survey that only included small numbers of each type of product in each sampling phase and as such was not fully representative of the market, therefore it was not appropriate to name brands.

For both sample collection periods, samples were immediately logged into the laboratory information management system (LIMS) on receipt and given a unique identifying number. Storage conditions on the package were adhered to. Any samples with short expiry dates or sold as frozen products were stored in the freezer until preparation and analysed according to standard practice and in-house quality procedures to preserve sample integrity and prevent sample spoilage. These samples tended to have high moisture content so freezing also prevented any potential changes to mycotoxin contaminant levels that could have occurred if the samples became spoiled or mouldy.

Multiple packs with the same batch code / use by date or other identifier were purchased to make up each individual sample. In most cases, depending on pack size, a minimum of two units were purchased for each sample.

For all samples, all sample packs received were combined, and homogenised using laboratory equipment to produce a smooth, homogenous sample. 

For meat replacements, protein bars and other solid samples a laboratory blender (GM Mill, Retsch) was used. Liquids were combined in a large container and mixed using a magnetic stirrer for approximately 30 minutes. After homogenisation the samples were split into at least 3 aliquots and were stored in the freezer after homogenisation. 

All mixing equipment was thoroughly cleaned between each sample to prevent cross contamination. 

For protein powders, individual packs for each sample were combined and shaken and mixed manually before dividing into aliquots for analysis. All samples were stored in the freezer after opening. 

Samples were divided into product categories with specific analytical requests for each product. These had been specified by FSS, the number of each type of sample and the analyses required are set out in Table 4. The analytes requested included Type A trichothecene mycotoxins (T-2 toxin and HT-2 toxin), Type B trichothecene (DON), as well as other Fusarium mycotoxins (ZON, fumonisins), and other mycotoxins of significant health concern that are regulated in other food products (aflatoxins and OTA). Samples were analysed for the tropane alkaloids atropine and scopolamine, and samples that contained wheat protein were also analysed for ergot alkaloids. 

Table 4. Phase 1 product categories and requested analyses

Samples were chosen based on availability of products in each category of commodities where data gaps exist for testing for these contaminants. Samples were analysed for mycotoxins using a method based on multi-mycotoxin IAC clean-up followed by LC-MS/MS analysis giving full coverage of all mycotoxins requested by FSS as well as allowing determination of sterigmatocystin. 

The analytes requested for phase 2 and the methods used to carry out those analyses are listed in Table 5.

Table 5. List of methods and analytes for Phase 2 of the study

* α-Zearalanol and β-Zearalanol* were requested but are not included in the method FSG 818