A 2025 peer-reviewed study in the journal Food Control tested 120 plant-based milk samples — oat milk, soy milk, and almond milk — purchased from major UK supermarkets and detected mycotoxins in every single sample.
2025 UK Study: Mycotoxins Found in All Plant-Based Milk Samples Tested
Researchers from the University of Parma and Cranfield University purchased 120 plant-based beverages from the top five UK retailers in January and February 2024. Each product was tested for 19 different mycotoxins using ultra-high performance liquid chromatography coupled to mass spectrometry (UHPLC-MS/MS) — one of the most sensitive detection methods available for food contaminant analysis. The full results were published in Food Control in 2025.
The 120 beverages covered three categories: 57 oat-based beverages, 32 almond and nut-based beverages, and 31 soy-based beverages. Tested mycotoxins included regulated compounds — aflatoxins, ochratoxin A, fumonisins, deoxynivalenol (DON), HT-2, T-2, and zearalenone — and several “emerging” toxins including beauvericin, enniatins, and Alternaria toxins that carry no established legal limits in food.
Key Finding: 100% of Plant-Based Milk Samples Contained Mycotoxins
100% of all 120 plant-based milk samples contained at least one mycotoxin. The majority contained several mycotoxins simultaneously. The researchers described this co-occurrence of multiple mycotoxins as a consistent feature across all product types and called on food producers to treat mycotoxin management as a core component of food safety practice.
The simultaneous presence of multiple mycotoxins in a single product carries health implications beyond the sum of individual contamination levels. Laboratory research cited by the study’s authors has shown that certain mycotoxin combinations — including ochratoxin A with fumonisin B1, and Fusarium toxins with Alternaria toxins — can produce additive or synergistic toxic effects. Current food safety regulations set limits for individual mycotoxins and do not account for these combined exposures.
What Are Mycotoxins? Why They Matter in Plant-Based Milk
Mycotoxins are chemically stable toxic compounds produced by molds — primarily from the Aspergillus, Penicillium, Fusarium, and Alternaria genera — during crop growth, harvest, and storage. They are not destroyed by standard food processing methods including cooking, roasting, or pasteurization.
The International Agency for Research on Cancer (IARC) classifies aflatoxin B1 as a Group 1 carcinogen — the highest risk category, shared with tobacco smoke. Ochratoxin A and fumonisins are classified as Group 2A probable human carcinogens. Emerging toxins such as beauvericin and the enniatins, which appeared in nearly every plant-based milk sample in this study, have no established regulatory limits — meaning there are currently no legal thresholds against which to measure their presence in food.
Mycotoxins in Oat Milk: The Most Contaminated Plant-Based Milk
Oat milk showed the broadest mycotoxin contamination profile of the three plant-based milk categories tested. Oats are well documented in agricultural science as particularly susceptible to Fusarium mold infection, and this study’s findings reflect that directly. Three unregulated emerging toxins — beauvericin, enniatin B1, and enniatin A1 — were detected in 100% of oat milk samples. HT-2 toxin, a type A trichothecene linked to immune system effects, appeared in 94.7% of oat milks, with a maximum level of 5.90 µg/L.
Notably, aflatoxin B1 levels in some oat milk samples approached the maximum regulatory limits set for raw, unprocessed cereal grains — despite oat content in the finished beverages ranging from just 1.3% to 13%. The study’s authors indicate this points to low-quality source oat ingredients in some products.
Mycotoxin Contamination in Oat Milk — 57 Samples (UK, 2024)
| Mycotoxin | % Positive | Average Level | Maximum Level |
|---|---|---|---|
| Beauvericin (BEA) — emerging, no legal limit | 100% | 0.55 µg/L | 1.68 µg/L |
| Enniatin B1 (ENNB1) — emerging, no legal limit | 100% | 0.52 µg/L | 1.77 µg/L |
| Enniatin A1 (ENNA1) — emerging, no legal limit | 100% | 0.33 µg/L | 0.68 µg/L |
| Enniatin B (ENNB) — emerging, no legal limit | 98.2% | 1.50 µg/L | 6.17 µg/L |
| Enniatin A (ENNA) — emerging, no legal limit | 98.2% | 0.26 µg/L | 0.30 µg/L |
| HT-2 Toxin — type A trichothecene | 94.7% | 1.18 µg/L | 5.90 µg/L |
| Alternariol monomethyl ether (AME) — Alternaria toxin | 73.7% | 0.06 µg/L | 0.12 µg/L |
| Tentoxin (TEN) — Alternaria toxin | 68.4% | 0.27 µg/L | 0.38 µg/L |
| T-2 Toxin — type A trichothecene | 59.6% | 0.66 µg/L | 3.14 µg/L |
| Ochratoxin A (OTA) — IARC Group 2A probable carcinogen | 52.6% | 0.20 µg/L | 0.35 µg/L |
| Aflatoxin B2 (AFB2) — IARC Group 1 carcinogen family | 49.1% | 0.26 µg/L | 0.36 µg/L |
| Zearalenone (ZEN) — estrogenic mycotoxin | 42.1% | 0.24 µg/L | 0.38 µg/L |
| Aflatoxin B1 (AFB1) — IARC Group 1 carcinogen | 33.3% | 0.25 µg/L | 0.57 µg/L |
| Deoxynivalenol (DON) — type B trichothecene | 33.3% | 6.99 µg/L | 19.71 µg/L |
| Fumonisin B1 (FB1) — IARC Group 2A probable carcinogen | 33.3% | 0.39 µg/L | 0.52 µg/L |
| Fumonisin B2 (FB2) — IARC Group 2A probable carcinogen | 24.6% | 0.36 µg/L | 0.38 µg/L |
Mycotoxins in Soy Milk: Ochratoxin A in 9 Out of 10 Samples
Soy milk was distinguished by an exceptionally high prevalence of ochratoxin A: the nephrotoxic compound — classified by IARC as a probable human carcinogen known to accumulate in tissue with repeated exposure — was detected in 90.3% of soy milk samples. Beauvericin, the unregulated emerging Fusarium toxin, was present in 100% of samples. The enniatins were not detected in soy milk, making this one of the few categories where that group of toxins was absent.
Mycotoxin Contamination in Soy Milk — 31 Samples (UK, 2024)
| Mycotoxin | % Positive | Average Level | Maximum Level |
|---|---|---|---|
| Beauvericin (BEA) — emerging, no legal limit | 100% | 0.11 µg/L | 0.31 µg/L |
| Ochratoxin A (OTA) — IARC Group 2A probable carcinogen | 90.3% | 0.18 µg/L | 0.70 µg/L |
| Fumonisin B1 (FB1) — IARC Group 2A probable carcinogen | 67.7% | 0.30 µg/L | 0.37 µg/L |
| Aflatoxin B2 (AFB2) — IARC Group 1 carcinogen family | 61.3% | 0.11 µg/L | 0.14 µg/L |
| Tentoxin (TEN) — Alternaria toxin | 61.3% | 0.07 µg/L | 0.15 µg/L |
| Fumonisin B2 (FB2) — IARC Group 2A probable carcinogen | 54.8% | 0.27 µg/L | 0.30 µg/L |
| Alternariol (AOH) — Alternaria toxin | 54.8% | 0.09 µg/L | 0.10 µg/L |
| T-2 Toxin — type A trichothecene | 48.4% | 0.10 µg/L | 0.21 µg/L |
| Aflatoxin B1 (AFB1) — IARC Group 1 carcinogen | 35.5% | 0.18 µg/L | 0.18 µg/L |
| Zearalenone (ZEN) — estrogenic mycotoxin | 22.5% | 0.34 µg/L | 0.87 µg/L |
| HT-2 Toxin | 9.7% | 0.19 µg/L | 0.36 µg/L |
| Enniatins (ENNB, ENNB1, ENNA, ENNA1) | Not detected | — | — |
Mycotoxins in Almond Milk: Lowest Contamination, but Not Toxin-Free
Almond and nut-based beverages had the least extensive mycotoxin contamination profile among the three categories. Aflatoxins B2, G1, and G2 were not detected. However, beauvericin appeared in 100% of almond milk samples, enniatin B was found in 96.9%, and fumonisins were present in over half of all samples.
Mycotoxin Contamination in Almond/Nut Milk — 32 Samples (UK, 2024)
| Mycotoxin | % Positive | Average Level | Maximum Level |
|---|---|---|---|
| Beauvericin (BEA) — emerging, no legal limit | 100% | 0.25 µg/L | 2.15 µg/L |
| Enniatin B (ENNB) — emerging, no legal limit | 96.9% | 0.06 µg/L | 0.67 µg/L |
| Enniatin A (ENNA) — emerging, no legal limit | 93.7% | 0.04 µg/L | 0.52 µg/L |
| Alternariol (AOH) — Alternaria toxin | 75.0% | 0.08 µg/L | 0.56 µg/L |
| Fumonisin B2 (FB2) — IARC Group 2A probable carcinogen | 62.5% | 0.25 µg/L | 0.28 µg/L |
| Fumonisin B1 (FB1) — IARC Group 2A probable carcinogen | 59.4% | 0.37 µg/L | 0.49 µg/L |
| HT-2 Toxin — type A trichothecene | 43.7% | 0.88 µg/L | 2.16 µg/L |
| T-2 Toxin — type A trichothecene | 25.0% | 0.21 µg/L | 0.59 µg/L |
| Deoxynivalenol (DON) | 21.9% | 0.43 µg/L | 0.78 µg/L |
| Ochratoxin A (OTA) — IARC Group 2A probable carcinogen | 9.4% | 0.30 µg/L | 0.42 µg/L |
| Aflatoxin B1 (AFB1) — IARC Group 1 carcinogen | 9.4% | 0.18 µg/L | 0.18 µg/L |
| Aflatoxins B2, G1, G2 | Not detected | — | — |
Why Pasteurization Does Not Remove Mycotoxins from Plant-Based Milk
A critical fact about mycotoxin contamination in plant-based beverages is that the compounds persist through every stage of manufacturing. The study’s authors confirm that standard food processing methods — including cooking, roasting, baking, and pasteurization — have little to no effect on mycotoxin levels in cereals and pulses. This is consistent with the broader food safety literature on mycotoxin stability.
Plant-based beverage production involves soaking the raw ingredient in water before grinding and filtering. The authors identify this water-soaking step as a mechanism for transferring mycotoxins from the solid plant material directly into the liquid phase of the finished product. Once present, the compounds are not eliminated by the heat treatment used in pasteurization.
The study further notes that despite oat content in finished oat milks ranging from just 1.3% to 13%, AFB1 levels in some samples still approached the regulatory ceiling for raw, unprocessed grain — pointing to poor ingredient quality at the source.
The Regulatory Gap: No Legal Mycotoxin Limits for Plant-Based Milks
At the time of the study’s 2025 publication, no specific maximum limits for mycotoxins in plant-based beverages as finished food products existed in either the EU or the UK. The study’s authors state that plant-based beverages are not included in finished-product food regulations. EU and UK regulations set mycotoxin limits for raw cereal grains and certain other commodities, but those thresholds do not apply to plant-based beverages on supermarket shelves.
The emerging toxins — beauvericin and the enniatins — which appeared in virtually every sample across all three plant-based milk categories have no legal maximum levels in any food category anywhere. The authors identified expanded monitoring and formal risk assessment as urgent priorities, noting that current dietary exposure estimates are likely incomplete.
Dairy Milk vs. Plant-Based Milk: The Mycotoxin Facts
The mycotoxin profile of conventional dairy milk and raw milk differs from plant-based beverages in several fundamental, documented ways.
One Primary Compound vs. Up to 15 Simultaneously
The only mycotoxin of established primary concern in dairy milk is aflatoxin M1 (AFM1) — a metabolite produced when a cow’s body processes aflatoxin B1 from contaminated feed. This is a well-characterised, single-compound pathway studied extensively for decades. The multi-toxin contamination documented in plant-based milks — involving up to 15 or more different compounds simultaneously — has no equivalent in the dairy supply chain.
AFM1 in Dairy Milk: One of the Strictest Food Safety Limits in the World
Aflatoxin M1 in dairy milk is governed by one of the most stringent regulatory thresholds applied to any food product. The EU and UK maximum permitted level for AFM1 in raw milk is 0.05 µg/kg — fifty times stricter than the limit for AFB1 in certain raw cereal categories (2 µg/kg). This limit is enforced through routine, mandatory testing at the farm, bulk collection, and processing stages.
Bovine Metabolism as a Biological Buffer
When a dairy cow ingests aflatoxin B1, its metabolic processes convert only approximately 1–3% of the ingested amount into AFM1, which can appear in milk. The remainder is metabolized and excreted by other pathways. This biological transformation — combined with dilution across total milk output — structurally limits the amount of mycotoxin that can reach the consumer, even under imperfect feed conditions. No equivalent biological buffer exists between a contaminated oat or soy crop and the finished plant-based milk product.
Mandatory Multi-Point Supply Chain Monitoring
Dairy supply chains in the EU and UK operate under mandatory mycotoxin monitoring at multiple points: feed quality control at the farm, bulk milk testing at collection, and finished product testing at the processor. The contamination pathway is tracked end to end. Plant-based beverage production has no equivalent mandatory monitoring regime — a gap the 2025 study directly identifies.
Unregulated Emerging Toxins: Not a Dairy Concern
Beauvericin — detected in 100% of oat milk, 100% of soy milk, and 100% of almond milk samples in this study — is not a documented mycotoxin concern in dairy milk. Enniatins, found in up to 100% of oat milk samples, are similarly not a known pathway in dairy. These compounds originate from Fusarium molds that infect grain and nut crops. The bovine digestive and metabolic system does not pass these compounds into milk in any meaningful documented quantity.
Dairy Milk vs. Plant-Based Milk — Mycotoxin Profile Comparison
| Factor | Plant-Based Milks (2025 UK Study) | Dairy Milk / Raw Milk |
|---|---|---|
| Mycotoxins detected | Up to 15 compounds simultaneously; co-occurrence in all 120 samples | One primary compound of concern (AFM1); single well-characterised pathway |
| Sample contamination rate | 100% of all samples contained at least one mycotoxin | AFM1 monitored continuously; exceedances tracked and rare |
| Regulatory limits | No finished-product regulations; several prevalent toxins have no legal limits in any food category | AFM1 limited to 0.05 µg/kg — among the strictest food safety thresholds globally |
| Biological buffer | None — mycotoxins transfer directly from contaminated raw crop to finished liquid | Bovine metabolism converts only ~1–3% of ingested AFB1 to AFM1; remainder excreted |
| Effect of pasteurization | Little to no effect on mycotoxin levels | AFM1 partially reduced by heat treatment; levels monitored pre- and post-processing |
| Unregulated emerging toxins | Beauvericin in 100% of all samples; enniatins in up to 100% of oat milks — no legal limits exist | Not a documented concern or known contamination pathway in dairy milk |
| Supply chain monitoring | No mandatory finished-product monitoring programme exists | Mandatory monitoring at farm, collection, and processing stages |
| Long-term exposure data | Limited, especially for multi-toxin combinations and unregulated emerging compounds | Decades of epidemiological and toxicological data on AFM1 exposure from dairy consumption |
What the Study Did Not Test: Pea Milk and Bioengineered Dairy Alternatives
The 2025 study was limited to oat, soy, and almond/nut-based beverages — three of the most established categories on the UK market at the time of sampling. Two growing categories were not included: pea protein milk and bioengineered dairy alternatives.
Pea Milk: A Gap Worth Noting
Pea milk — made from yellow split pea protein — has grown considerably in market presence since 2024 and was not represented in the study’s sample set. This absence is significant because the study’s authors specifically flag legumes, including peas, as an under-researched category for mycotoxin contamination. Peas, lentils, chickpeas, and other pulses are listed in the paper as crops that can support fungal growth both in the field and during long-term storage, and the authors identify the lack of monitoring data for legume-derived products as an urgent gap in food safety science. Based on the study’s findings from other legume-based products, pea milk would be a logical candidate for future mycotoxin surveillance.
Bioengineered Dairy Alternatives: A Different Question Entirely
Precision fermentation milks — products that use genetically engineered microorganisms to produce whey or casein proteins without cows, such as those made by companies like Remilk or The Every Company — were not tested and fall outside the scope of this study entirely. Their mycotoxin risk profile is a structurally different question: rather than depending on crop contamination from field molds, it depends on the fermentation substrate, the microbial strains used, and the downstream processing of the fermentation output. No equivalent peer-reviewed mycotoxin surveillance data exists for these products at present. As precision fermentation dairy alternatives expand toward mainstream retail, they represent another category where independent food safety monitoring has not yet kept pace with market growth.
Summary
- All 120 plant-based milk samples contained at least one mycotoxin — a 100% contamination rate
- Most samples contained multiple mycotoxins simultaneously, with potential for synergistic effects not addressed by current regulations
- Beauvericin — an unregulated emerging toxin — was present in 100% of oat milk, 100% of soy milk, and 100% of almond milk samples
- Oat milk had the broadest contamination profile, with 16 of 19 tested mycotoxins detected across samples
- Soy milk had ochratoxin A — an IARC Group 2A probable carcinogen — in 90.3% of samples
- No specific regulatory limits for mycotoxins in plant-based beverages as finished products currently exist in the EU or UK
- Pasteurization and standard food processing do not eliminate mycotoxins from plant-based milks
- The study’s authors called for urgent expanded monitoring and risk assessment, noting current exposure estimates are likely incomplete
Dairy milk is governed by one of the most tightly controlled mycotoxin regulatory regimes in the global food supply — a strict legal limit of 0.05 µg/kg for its single primary compound of concern, enforced through mandatory multi-point supply chain monitoring, and supported by the natural biological buffering of the bovine metabolic system.
Frequently Asked Questions
Do plant-based milks contain mycotoxins?
According to a 2025 peer-reviewed study published in Food Control, all 120 plant-based milk samples tested — including oat milk, soy milk, and almond milk — contained at least one mycotoxin. Most samples contained multiple mycotoxins simultaneously.
Which plant-based milk has the most mycotoxins?
The 2025 UK study found that oat milk had the broadest mycotoxin contamination profile. Sixteen of the 19 tested compounds were detected across oat milk samples. Three unregulated emerging toxins — beauvericin, enniatin B1, and enniatin A1 — were each found in 100% of oat milk samples tested.
Is soy milk safe from mycotoxin contamination?
In the 2025 Food Control study, 100% of soy milk samples contained beauvericin and 90.3% contained ochratoxin A, a compound the IARC classifies as a probable human carcinogen (Group 2A). Enniatins were not detected in soy milk, but soy milk was not free of mycotoxin contamination overall.
How does dairy milk compare to oat milk for mycotoxins?
Dairy milk has one primary mycotoxin of concern — aflatoxin M1 (AFM1) — subject to an EU/UK legal limit of 0.05 µg/kg, one of the strictest food safety thresholds in the world. The 2025 study found up to 15 different mycotoxins simultaneously in plant-based milks, including several unregulated compounds with no legal limits in any food category. No equivalent multi-toxin contamination profile is documented in dairy milk.
Does pasteurization remove mycotoxins from plant-based milk?
No. The study’s authors confirm that standard food processing methods including pasteurization have little to no effect on mycotoxin levels. Mycotoxins are chemically stable compounds that are not destroyed by heat. The water-soaking step used in plant-based milk production is identified as a likely mechanism by which mycotoxins transfer from the raw ingredient into the finished liquid product.
Are there legal limits for mycotoxins in oat milk, soy milk, or almond milk?
No. As of the study’s 2025 publication, no specific regulatory maximum limits for mycotoxins in plant-based beverages exist as finished food products in the EU or UK. Several of the most prevalent toxins found — beauvericin and the enniatins — have no legal limits in any food category.
Source: Torrijos, R., Mihalache, O.A., Patriarca, A., Medina, A., & Dall’Asta, C. (2025). “Mycotoxin contamination in plant-based beverages and meat alternatives: A survey of the UK market.” Food Control. doi: 10.1016/j.foodcont.2025.111910.
