Vegan vs Meat-Rich Diet: Mold Toxins in Urine After 8 Weeks

On September 18, 2026, researchers from the University of Freiburg published a randomized controlled trial in npj Science of Food evaluating mycotoxin exposure across vegan and meat-rich diets. While creatinine-adjusted urinary biomarkers showed higher levels of deoxynivalenol, ochratoxin A, and zearalenone in the vegan group, unadjusted crude urine measurements revealed no significant differences. The authors stated that direct health outcomes were not measured and wrote that the results are not meant to discourage plant-forward diets.

Key takeaways:

• An 8-week German RCT found higher creatinine-adjusted urinary levels of deoxynivalenol (DON), ochratoxin A (OTA), and zearalenone (ZEN) in vegan participants compared to those on a meat-rich diet.
• T-2 toxin was significantly higher in the meat-rich group.
• Crude, non-creatinine-adjusted concentrations for DON, OTA, and ZEN showed no statistically significant differences between the two diets, suggesting creatinine adjustments may distort biomarker calculations in people eating no meat.
• The trial did not assess clinical health outcomes or safety thresholds, and the researchers continue to encourage plant-forward dietary patterns.

What happened

On September 18, 2026, researchers at the Centre for Complementary Medicine, Department of Internal Medicine II, Medical Center - University of Freiburg in Germany published findings from a dietary intervention in npj Science of Food. The open-access paper, accessible via DOI: 10.1038/s41538-026-01152-4 and PMC13594211, analyzed secondary urine samples from a randomized controlled trial registered under DRKS00031541. The underlying trial was registered on March 27, 2023, and ethics approval for mycotoxin analysis was granted under an amendment in December 2025.

The study enrolled 63 healthy adult omnivores with a median age of 23, comprising 37 men and 26 women. Following a one-week run-in phase on a mixed diet adhering to German Nutrition Society advice, participants were randomized to an isocaloric vegan diet (n = 32) or a meat-rich diet (n = 31) for eight weeks. The meat-rich group consumed at least 150 grams of meat daily (fish was not counted), while the vegan group consumed no animal foods and was encouraged to choose whole foods. Meals were not provided directly to participants.

Researchers collected morning spot urine at baseline and at week 8 to measure six fungal metabolites: aflatoxin, deoxynivalenol (DON), fumonisin, ochratoxin A (OTA), T-2 toxin, and zearalenone (ZEN). Aflatoxin and fumonisin were essentially below detection limits. Dietary intake was monitored using weighed food diaries.

What changed

At the end of the eight-week intervention, creatinine-adjusted urinary medians showed distinct divergences between the two cohorts. Deoxynivalenol measured 7.15 ug/g (IQR 8.10) in the vegan group compared to 2.85 ug/g (IQR 6.04) in the meat-rich group (p = 0.004). Ochratoxin A reached 0.45 ug/g (IQR 0.37) in the vegan group versus 0.20 ug/g (IQR 0.28) in the meat-rich group (p = 0.003). Zearalenone was also higher in the vegan cohort at 1.34 ug/g (IQR 1.52) versus 0.49 ug/g (IQR 1.01) for meat eaters (p = 0.001).

Conversely, creatinine-adjusted T-2 toxin was higher in the meat-rich group at 5.15 ug/g (IQR 5.81) compared to 3.44 ug/g (IQR 3.67) in the vegan group (p = 0.047). In mixed-model repeated measures, OTA showed a predicted increase in the vegan cohort from baseline of +0.24 ug/g (95% CI 0.13 to 0.35, p < 0.001) with no significant change in the meat group. For DON, predicted week-8 values were 5.99 ug/g (95% CI 3.85 to 8.13) in meat eaters versus 14.70 ug/g (95% CI 9.62 to 19.79) in vegans (p = 0.002). The mixed model for T-2 toxin was not statistically significant overall (p = 0.105).

A crucial technical nuance emerged from unadjusted measurements. When evaluated as crude, non-creatinine-adjusted concentrations, urinary levels at week 8 did not differ significantly between groups for DON (p = 0.271), OTA (p = 0.649), or ZEN (p = 0.138). Crude T-2 toxin remained significantly higher in the meat-rich cohort (p = 0.005). The study authors noted that creatinine adjustment can be distorted in vegans, who consume no creatine from meat or fish, and that neither 24-hour urine nor specific-gravity adjustment was available.

Who is affected

This trial examined healthy young adults in Germany who shifted to a plant-based diet over a relatively short, eight-week duration. Because the cohort was young, healthy, and omnivorous prior to the trial, the findings cannot be generalized automatically to lifelong vegans, children, older demographics, or people consuming different regional food supplies outside Germany.

The findings add to a broader set of studies on how diet shapes exposure to contaminants that are not added on purpose.

Other recent coverage looks at unintended residues and contaminants in everyday diets, such as the FDA's latest pesticide residue report and a study linking PFAS in children's blood to diet.

What it means for shoppers

Shoppers do not need to discard plant-forward habits based on these biomarker findings. The study authors explicitly stated that commercial reference-range upper limits from testing laboratories do not reflect established clinical safety thresholds, nor did the trial measure any direct health outcomes or clinical endpoints.

Furthermore, the paper did not compare individual urinary amounts against tolerable daily intake limits established by international authorities. The researchers noted that food-group associations were challenging to interpret. While moderate positive correlations appeared between vegetable intake and creatinine-adjusted DON, OTA, and ZEN, whole grains correlated with ZEN, and plant yoghurt correlated with most mycotoxins, no multiple-testing corrections were applied. Plant-based milk intake did not differ between the groups at week 8.

Consumers should also recognize that mold toxins are naturally occurring substances rather than intentional additives. Regulatory guidance from the European Food Safety Authority defines mycotoxins as toxic compounds naturally produced by fungi, commonly found in cereals, dried fruits, nuts, and spices. In parallel, the FDA actively monitors toxins like aflatoxins, DON, fumonisins, patulin, ochratoxin A, T-2/HT-2, and zearalenone, maintaining advisory levels such as those for DON in finished wheat products.

The background

Mycotoxins represent a long-standing agricultural and food-safety challenge. According to the EFSA, mycotoxins may cause effects ranging from gastrointestinal and kidney disorders to immune deficiency and cancer.

Prior international biomonitoring studies assessing vegans versus omnivores across the United Kingdom, Germany, France, Norway, and Iceland have yielded inconsistent outcomes. For instance, an earlier Icelandic biomonitoring study identified comparable DON concentrations between dietary groups and reported a ten-fold lower level of OTA in vegans. The Freiburg trial tested the question in a randomized setting, but its findings reflect one short dietary intervention rather than a settled consensus.

Diet-pattern studies have raised similar questions elsewhere, such as Ultra-Processed Food, Crohn's Risk and Low DHA: New Study. Understanding food safety requires differentiating between direct toxicological harm, chemical additives, and naturally occurring biological byproducts.

What to watch next

The authors list modest sample size, spot urine only, short-term vegans and no multiple-testing correction among the study's limitations.

Researchers also noted that spices were poorly recorded in participant diaries, food groupings were broad, diaries did not cover all days immediately preceding urine collection, and specific foods or brand-name packaged items were never chemically tested for mycotoxin presence. Moreover, the trial could not explain why T-2 toxin levels were higher among participants assigned to the meat-rich regimen.

The authors of the Freiburg trial explicitly concluded: "The herein presented results are not meant to discourage individuals to adopt such a diet. To the contrary, we encourage plant-forward diets." What evidence cannot establish at this time is whether the observed biomarker shifts carry any long-term clinical significance for healthy individuals.