4. Executive Summary
Food Standards Scotland (FSS) commissioned Fera Science Ltd. to carry out a survey of chemical contaminants in plant-based protein products.The aim was to provide evidence on levels of mycotoxins, plant alkaloids, acrylamide and erucic acid that can be found in plant-based protein products made from wheat, soy, pea protein, or a mixture of these proteins.
Study Design
The samples analysed were products containing plant-based protein. This included protein powders, drinks and bars, tofu, and vegetarian burgers, mince and similar products. Samples were made from soy, pea or wheat protein or combinations of these (the most commonly consumed categories of plant protein products). Products were purchased online or from retail stores in Scotland, with 26 products purchased in Phase 1 in 2024 and a further 26 purchased in Phase 2 in 2025. Phase 1 samples concentrated on products with minimal ingredients and Phase 2 included more complex samples.
Phase 1 of the study tested for the Fusarium mycotoxins (deoxynivalenol, zearalenone, T-2 and HT-2 toxin and fumonisins B1, B2 and B3), aflatoxins (B1, B2, G1 and G2) and ochratoxin A and the plant toxins, tropane alkaloids (atropine and scopolamine). While the mycotoxins tested are regulated in other foods, there are no specific Maximum Levels (MLs, legally enforceable limits) established in The Contaminants in Food (Scotland) Regulations 2013 and Commission Regulation (EC) No 1881/2006 (1, 2) for plant-based protein products. However, MLs for other foods (e.g. cereals, groundnuts) can be used for comparison. Ergot alkaloids, which are not regulated within GB but are within the EU and Northern Ireland, were also analysed in a subset of samples that contained wheat. The samples were also analysed for acrylamide.
Phase 2 of the project added further analytes, namely the mycotoxins, citrinin, enniatins, beauvericin, fusaric acid, Alternaria toxins and the plant toxins pyrrolizidine alkaloids. Products that contained rapeseed oil were also analysed for erucic acid.
Mycotoxins
For the samples tested for Phase 1, at least one mycotoxin residue above the Limit of Quantification (LOQ) was detected in 14 of the 26 samples (54%). One sample of soy protein powder contained 6 different mycotoxins. Ochratoxin A (OTA) was the most frequently detected mycotoxin, it was present in each of the protein types, comprising nine samples in total (35%) at concentrations ranging from 0.11 to 9.33 µg/kg. The highest level (9.33 µg/kg) was found in a pea protein powder, the next highest concentration (4.15 µg/kg) was found in soy protein powder. Aflatoxins were detected in three samples. One soy protein powder contained 2.24 µg/kg AFB1, as well as AFB2, AFG1 and AFG2 resulting in a total aflatoxin level of 4.94 µg/kg. This sample also contained OTA (4.15 µg/kg) and zearalenone (ZON) (1.13 µg/kg). Four of 26 (15%) samples contained DON at levels from 2.85 to 7.35 µg/kg. Three samples (12%) contained ZON at 1.1 to 1.75 µg/kg, one sample contained fumonisins FB1, FB2 and FB3 and one sample contained sterigmatocystin (STG).
The 26 samples for Phase 2 were also analysed using this method. The results were broadly similar to Phase 1, although the concentrations measured were lower. OTA was the most commonly detected mycotoxin, 7 out of 26 (27%) samples contained OTA at levels ranging from 0.13 to 0.55 µg/kg. Five samples contained DON at levels from 2.32 to 14.55 µg/kg, and five samples (19%) contained ZON at 1.28 to 4.85 µg/kg. Three samples each contained AFB1 (0.15 – 0.23 µg/kg) and STG (0.22 – 0.54 µg/kg). One sample contained HT-2 toxin at 4.16 µg/kg. Fumonisins and T-2 toxin were not detected in any of the samples.
The OTA results for the pea protein and soy protein powders in Phase 1 are higher than MLs in other foods such as cereals (3 and 5 µg/kg),and wheat gluten (8 µg/kg), and similar to the MLs for dried vine fruit (10 ug/kg) and coffee (5 and 10 ug/kg). The soy protein powder sample also contained AFB1 and total aflatoxins just above the MLs (2 and 4 µg/kg respectively) for nuts and cereals.
In Phase 1 five samples that contained wheat protein were analysed for ergot alkaloids. Two samples were found to contain ergot alkaloids, eight ergot alkaloids were detected in one sample giving a sum of ergot alkaloids of 5.71 µg/kg. The other sample contained 9 compounds with a sum of 6.9 µg/kg.
All 26 Phase 2 samples were analysed for ergot alkaloids, three samples (12%) contained the sum of ergot alkaloids at levels of 2.81, 3.41 and 11.01 µg/kg. These three products contained wheat (flour and or gluten) as ingredients. There are no MLs for ergot alkaloids in GB. There are MLs in force in the EU (4), and these apply for cereals and cereal products. Plant-based protein products are not mentioned specifically. The lowest ML is 20 µg/kg for processed cereal based food for infants and young children.
Plant alkaloids (tropane and pyrrolizidine)
All Phase 1 samples were analysed for the tropane alkaloids atropine and scopolamine. Four samples contained tropane alkaloids above the individual LOQs of 0.1 µg/kg. Two samples contained only atropine (at 0.19 and 0.2 µg/kg) while two contained both atropine and scopolamine. All four samples were soy based, the highest concentration of tropane alkaloids (1.71 µg/kg) was found in a tofu (soy) sample.
All Phase 2 samples were analysed for atropine and scopolamine, 3 out of 26 samples contained atropine at levels of 0.52 to 1.15 µg/kg. One sample contained scopolamine at 0.38 µg/kg.
There are no MLs for tropane alkaloids in plant-based protein products, the only MLs in force in GB are for baby food and processed cereal-based food for infants and young children, where MLs are set at 1.0 µg/kg each for atropine and scopolamine (2). MLs for the sum of atropine and scopolamine (tropane alkaloids) are set in the EU for cereal products at 5 to 15 µg/kg, and 25 and 50 µg/kg for dry herbal infusions (3). All of the samples in this survey contained tropane alkaloid levels well below the MLs for cereals (4).
Phase 2 samples were also analysed for pyrrolizidine alkaloids (PAs), none were detected in any of the 26 samples.
Other Mycotoxins
Two additional LC-MS/MS methods (MM2 and multi-mycotoxin) were used to analyse an extended range of Fusarium mycotoxins for the Phase 2 samples. Four out of 26 samples (15 %) contained ZON, three just below the LOQ (two at 2.3 µg/kg and one at 1.7 µg/kg), the other contained 4.6 µg/kg. These results were very similar to the ZON results found for the same samples using the 11+ IAC method. DON was found in one sample at 12.6 µg/kg, compared to 11.22 µg/kg found by the 11+ IAC method.
Beauvericin was found in 8 samples (31 %) at levels of 0.8 – 1.8 µg/kg. Enniatin A was found in 6 out of 26 samples (0.7 – 1.2 µg/kg) and enniatin A1 was found in 7 samples (0.6 -1.4 µg/kg). Enniatin B was found in 14 out of 26 samples at levels from 0.6 – 11.8 µg/kg and enniatin B1 was found in 9 out 26 samples at levels from 0.7 – 29.1 µg/kg. There are no MLs for enniatins and beauvericin. Sixteen samples contained levels of fusaric acid from 4.1 – 118.4 µg/kg. The highest level quantified was 118.4 µg/kg, but there was also a sample that flagged as >1000 µg/kg. This is indicative as the initial extract result was outside the calibration range and the recovery was low which increases the uncertainty about the result. There are no MLs for fusaric acid. All Phase 2 samples were also analysed for citrinin, no residues above the LOQ of 2.5 µg/kg were detected.
Alternaria toxins
The Phase 2 samples were analysed for Alternaria toxins. Tenuazonic acid was most frequently detected and was found in 14 out of 26 (54 %) of samples at levels from 3.8 – 62.7 µg/kg. Alternariol was the next most frequently found, five samples contained levels from 1.1 – 13.5 µg/kg, four samples contained alternariol monomethylether at levels from 2.1 – 6.1 µg/kg. Altenuene and tentoxin were not detected in any samples. There are no regulations in force for Alternaria toxins, so there are no MLs to compare these results to. However, these findings are similar to results reported in other published studies (6-10).
Acrylamide
Phase 1 samples were analysed for acrylamide, 2 of out 26 samples contained acrylamide above the LOQ of 30 µg/kg. One sample, a plant based burger, contained 36.8 µg/kg and the other product, a vegetarian sausage, made from soy and wheat, contained 41.7 µg/kg. There are no MLs for acrylamide but there are benchmark levels (BMLs) in place in GB. BMLs are guidance values, not legal limits, that are used as reference points for monitoring and risk management. The BMLs for acrylamide range from 40 to 4000 µg/kg for baby food and chicory based coffee substitutes respectively (5).
Erucic acid
A subset of 10 Phase 2 samples were analysed for erucic acid. The selected samples contained rapeseed as an ingredient. The levels of erucic acid measured ranged from 0 to 6.0 g/kg erucic acid in the oil / fat fraction of the sample. These were all below the ML of 20 g/kg erucic acid for oils used as an ingredient in food or for direct consumption.
Summary
All of the samples analysed were compliant with GB MLs where they exist. Two samples in Phase 1 contained aflatoxins and OTA at concentrations at or above MLs for foods that might be considered comparable such as wheat gluten, nuts and cereals.