Does Oral CBD Turn Into THC in the Stomach? What a Three-Minipig Study Actually Measured

Short answer: In a small, original animal experiment, oral administration of synthetic cannabidiol (CBD) produced measurable CBD in the plasma and gastrointestinal contents of three male minipigs, but neither delta-9-tetrahydrocannabinol (THC) nor 11-hydroxy-THC (11-OH-THC) was quantifiable in the specimens tested. The findings argue against substantial detectable CBD-to-THC conversion under that particular dosing and sampling protocol. They do not establish that conversion is chemically impossible, that every CBD product is free of THC, or that an individual’s positive urine or blood result can be explained—or excluded—by this experiment alone.

The original investigation was Wray, Stott, Jones and Wright, Cannabidiol Does Not Convert to Δ9-Tetrahydrocannabinol in an In Vivo Animal Model, published in Cannabis and Cannabinoid Research in 2017 (DOI: 10.1089/can.2017.0032). Its value to forensic toxicology lies in the difference between a reaction demonstrated in simulated laboratory fluid and analytes actually identified in separately collected biological specimens after administration to a living organism.

Why does the CBD-to-THC conversion question arise?

CBD and THC are distinct cannabinoids. Under some acidic laboratory conditions, CBD can undergo chemical cyclization to products that include THC. That reaction is a legitimate chemical observation, but a vessel containing simulated gastric fluid is not interchangeable with a functioning gastrointestinal tract. Time, formulation, fluid movement, food, absorption and metabolism can alter both whether a chemical species forms and whether sufficient material survives to reach a measurable biological compartment.

This distinction matters when a laboratory finding is used to support a legal inference. A person may report using only a product labeled CBD, while subsequent testing detects a THC-related analyte. Plausible explanations must be investigated against the product composition, exact analyte, actual biological matrix, collection chronology and analytical method. A laboratory demonstration of acid-catalyzed conversion is not evidence that the reported individual’s stomach created the detected THC. Conversely, a negative small-animal experiment cannot prove that conversion never occurs under any conceivable conditions.

What was the original experiment—and what was not tested?

The research used three individual male Göttingen minipigs, not human volunteers. The animals were acclimatized before dosing. Investigators administered an oral sesame-oil formulation containing synthetic CBD at 100 mg/mL. Synthetic material was chosen specifically to reduce the complication of small quantities of THC that might accompany some plant-derived CBD preparations.

Each scheduled gavage delivered approximately 15 mg CBD per kilogram of body weight, equivalent to about 0.15 mL of the formulation per kilogram. The animals were dosed twice daily for four days, then once in the morning of Day 5: nine scheduled administrations per animal. That protocol describes administered mass, not a measured circulating cannabinoid concentration. A nominal 15 mg/kg exposure should not be re-described as 15 mg/kg of absorbed CBD or as any equivalent THC blood concentration.

No placebo group, THC-treated positive-control group or human comparison group was described as part of this experiment. The principal question was analytical: after repeated oral CBD dosing, would CBD, THC or 11-OH-THC be found in the selected plasma and gastrointestinal specimens?

How were the samples collected and analyzed?

On Days 1 and 5, jugular venous blood was sampled immediately before the morning dose and at 1, 2, 4 and 6 hours afterward. Each collected blood specimen was centrifuged to separate plasma. The reported cannabinoid concentrations therefore refer to plasma, not whole blood. Five scheduled time points on two days across three animals correspond to 30 planned animal-time blood collections, but there are still only three independent animals.

Six hours after the final morning dose on Day 5, the animals were euthanized and the contents of the stomach and small intestine were collected separately. Those contents are gastrointestinal material, not peripheral blood, urine, breath, tissue drug distribution or serum. The authors used liquid chromatography–tandem mass spectrometry (LC–MS/MS) to assay CBD, THC and 11-OH-THC in the two different specimen types.

The published lower limits of quantification (LLOQs) were analyte- and matrix-specific. A result below the LLOQ means it was not quantifiable under the stated method conditions. It is more precise than saying that a molecule was absent from the entire animal.

What cannabinoids were actually detected?

CBD was quantifiable in plasma after administration, and CBD was present in both sampled gastrointestinal compartments. In contrast, parent delta-9-THC and 11-OH-THC were reported below the stated quantification limits in the sampled plasma and gastrointestinal contents. The paper does not report measurable conversion products at those endpoints. That is an analytically bounded negative result, rather than a statistical test demonstrating an exactly zero chemical conversion rate.

Importantly, 11-OH-THC is a metabolite that can arise following THC exposure; it is not the same measurement as parent THC. The study’s inclusion of both analytes strengthened the search for evidence of circulating THC-related material within its testing window. Yet the investigators did not directly assay THC-COOH, its glucuronide or delta-8-THC. The absence of quantifiable THC and 11-OH-THC cannot be extended to untested compounds.

Evidence table 1: actual sample types and analytical limits

Numbers from different rows cannot be exchanged as though they came from one chemical matrix. For additional context on the distinction between plasma and other specimens, see the published guide to biological matrices in forensic toxicology.

How much CBD reached plasma after the first and ninth doses?

The experiment did not simply fail to expose the animals to CBD. The study observed postdose plasma CBD and found substantial CBD concentrations in the gastrointestinal contents at the final sampling time. The first-day and fifth-day pharmacokinetic summaries, however, have different effective group sizes, an important feature easily lost in a quick abstract summary.

One animal partially regurgitated its Day 1 dose, and its unusually low first-day CBD values were excluded from the published Day 1 means. Thus the reported first-day mean peak of 328 ng/mL and 0–6 hour area under the curve (AUC) of 921 h·ng/mL are based on two animals, while corresponding Day 5 summaries use all three animals. The experiment cannot support a rigorous claim that an independent three-animal mean changed by a specified percentage between the first and last day without acknowledging this mismatch.

Evidence table 2: plasma CBD pharmacokinetics, with correct denominators

The reported C_max is a maximum among sampled plasma observations, while AUC summarizes measured exposure over an interval. Neither is a direct observation of intestinal CBD-to-THC conversion. The study’s limited pharmacokinetic coverage should not be used to construct an individualized last-dose or last-use estimate. See the published pharmacokinetics interpretation guide for why models and time windows require caution.

Why does the partially regurgitated first dose matter?

The first animal’s Day 1 plasma CBD concentrations were far lower than those of the other two animals. The study recorded partial regurgitation and excluded that animal from first-day summary calculations. Its individual values remained available in the published table, so the exclusion can be evaluated rather than overlooked.

A methodological consequence is that the Day 1 table does not describe three equally exposed animals. A comparison of the printed first-day and fifth-day group means should therefore be framed as a descriptive observation under different inclusion rules, not as a tested statement that CBD accumulates or fails to accumulate. Day 5 predose CBD also remained detectable, which is compatible with prior administration but does not establish a particular individual’s disposition parameters.

The dose-retention issue does not erase the main negative THC observation, which was reported for plasma samples from the study animals. It does limit claims about how precisely the intended oral CBD dose was delivered and how tightly the measured CBD values should cluster.

Why did the researchers collect stomach and intestinal contents?

If acid conversion of CBD into THC occurred extensively in the gastrointestinal environment during this regimen, gastrointestinal contents were relevant places to look for a chemical signal. The investigators did not rely solely on circulating plasma: they sampled material from both the stomach and small intestine six hours after the last dose.

CBD was present in those compartments at concentrations much higher numerically than the mean plasma CBD values, but those materials are different matrices. The mean gastric CBD concentration was 84,500 ng/mL; the small-intestinal mean was 43,900 ng/mL. These were concentrations in contents, not drug masses recovered from entire organs and not direct measurements of fractional CBD absorption. The authors reported THC and 11-OH-THC below their 1 ng/mL GI LLOQs.

The snapshot is informative but finite. It does not repeatedly sample stomach contents immediately after every dose, document all transient intermediates, or estimate mass balance throughout the intestinal tract. A hypothesized conversion signal occurring transiently or below the quantitative method limits is different from a quantified THC finding.

Does a negative result prove that chemical conversion is impossible?

No. An analytical negative must be stated at the resolution of the experiment. The conclusion supported here is that quantifiable parent THC and 11-OH-THC were not observed at the tested plasma and gastrointestinal endpoints after the specified oral synthetic-CBD regimen in three male minipigs.

That is more informative than a theoretical assertion based on acid chemistry alone, but it is narrower than a universal prohibition. The study used one CBD formulation, one species and sex, one repeated-dose schedule, a limited number of sampling times and specifically selected analytes. The source does not provide a human interventional conversion rate, a large-animal frequency estimate with narrow confidence bounds, or validated results for every possible acidic product, solvent and biological compartment.

Nor does it establish a statistically nonsignificant between-group conversion difference: there was no independent randomized comparator group or conversion-rate hypothesis test. The null findings are measurements below analytical reporting limits. This distinction matters when experts translate an instrument’s negative finding into language suitable for testimony.

What is different about in-vitro simulated gastric fluid?

In-vitro chemistry can isolate specific variables such as acidity, reaction duration and solvent composition. It is useful for showing whether a transformation is chemically feasible under those conditions. An intact animal introduces moving fluids, changing concentrations, complex formulations, absorption through the intestinal wall and metabolism. A compound may disappear from one compartment without ever yielding a measurable amount of a particular downstream analyte in another.

The Wray experiment therefore tests a bridge between these two settings rather than rejecting laboratory chemistry. The authors chose minipigs because aspects of their gastrointestinal function resemble human physiology; however, similarity is not species identity. The paper expressly calls for additional human studies and other measured metabolites to test generality. A defensible forensic conclusion should preserve that asymmetry: conversion in principle is not the same as demonstrated production of blood THC after real-world CBD ingestion.

Could a THC-positive result after a CBD-labeled product have other explanations?

Yes, and this animal study does not rank those possibilities for an actual case. A CBD-labeled commercial preparation may differ from the synthetic CBD used by Wray and colleagues. A product’s label is not a comprehensive certificate of its true composition, batch-to-batch reproducibility or exposure history. If a case turns on product contents, the exact lot, a representative retained sample and an appropriately validated cannabinoid assay become crucial.

The laboratory also needs to identify which cannabinoid or metabolite was reported. Parent THC in collected whole blood, THC-COOH in urine, THC in separated plasma and immunoassay screening signals are different findings. A presumptive screening signal may require specific confirmatory chemical identification. The absence of parent THC and 11-OH-THC above quantification limits in minipig plasma does not tell us whether a particular patient’s urine immunoassay recognized THC-COOH, another cannabinoid, or an interfering compound.

Nor should a product’s reported CBD administration automatically be reclassified as a controlled THC administration. To support a conversion hypothesis in an individual case, an analyst would need a scientific basis linking formulation, known exposures, serial measurements, metabolite profiles and verified specimen conditions. For the general distinction between a preliminary screen and a specifically identified analyte, consult confirmatory drug-testing principles.

Why can plasma, whole blood and urine not be substituted?

Wray et al. isolated plasma from venous blood after centrifugation. A report that says “plasma THC below 0.5 ng/mL” is not a legal whole-blood THC determination at 0.5 ng/mL. Matrix distributions, collection conditions and reporting units have to be addressed before comparing findings across methods, especially when applying statutory or policy thresholds based on whole blood.

The gastrointestinal CBD values are yet another compartment. The paper contains no paired quantitative urine THC-COOH results, no human oral-fluid THC measurements and no exhaled-breath result. A THC-COOH-positive urine sample in another study may reflect a metabolite and detection window different from the sampled parent plasma THC here. It cannot be treated as the same biomarker simply because both findings are called “cannabinoid testing.”

Similarly, 11-OH-THC is pharmacologically and analytically distinct from THC-COOH. Testing the former but not the latter prevents extrapolation to the latter’s detection frequency, threshold or persistence. If a forensic case requires last-use timing, the actual specimen collection time and complete assay panel must be obtained rather than filled in from a different animal experiment.

What can a forensic expert say about individual impairment or causation?

Only a limited inference follows. The experiment was not a driving, psychomotor, cognition, clinical-intoxication or collision-reconstruction study. Three minipigs receiving synthetic CBD were tested for selected cannabinoid analytes in plasma and GI contents. There were no measured human actions, standardized impairment tests, eyewitness accounts, real-road events or accident outcomes.

The reported absence of quantifiable THC is relevant when someone proposes a universal mechanism in which oral CBD necessarily becomes circulating THC. It does not establish the behavior or mental state of a specific person. A positive result from someone claiming CBD use may have several competing source explanations; a negative finding in this minipig model does not choose between them. Neither does a measured CBD plasma concentration, by itself, provide an individualized impairment threshold.

The appropriate legal-scientific use is to identify the mechanism tested, its analytic coverage, the size and biological context of the experiment, and the assumptions that would be needed to extrapolate to a different person or product. Those assumptions should be articulated, not smuggled into an apparently categorical yes-or-no answer.

What additional evidence would make the CBD-conversion question more decisive?

A stronger investigation would directly characterize the administered product for THC and other cannabinoids; measure verified dosing and exposure; sample clinically relevant time points in appropriate human studies; quantify parent CBD, delta-9-THC, 11-OH-THC, THC-COOH and potentially delta-8-THC with separately validated methods; and distinguish unchanged CBD from true new products and their metabolites. Controls for carryover, sample handling, background intake, and assay specificity would matter as much as a nominal instrument threshold.

Prospective measurements could then ask how often products lacking analytically detectable THC generate detectable THC-related species after oral use, and at what levels in specified biological compartments. Replication across formulations and practical dosing conditions would be necessary before extrapolating to a particular commercial product or individual. Wray et al. contribute a specific negative animal observation to this larger question; their study does not complete that validation sequence.

Frequently asked questions about CBD-to-THC conversion

1. Did the minipigs receive THC or only CBD?

The scheduled administered formulation contained synthetic CBD at 100 mg/mL, delivered by gavage at 15 mg/kg per dose. The investigators selected synthetic CBD to avoid a plant-extract THC contamination problem. The experiment did not deliberately dose THC as a positive-control arm.

2. Was THC completely absent from every animal?

The study reported THC below 0.5 ng/mL in sampled plasma and below 1 ng/mL in sampled stomach and small-intestinal contents. Those are analytical limits for specific samples. They do not prove that literally no THC molecule occurred anywhere in an animal, or that all possible THC-related compounds were measured.

3. What does the negative 11-hydroxy-THC finding add?

11-OH-THC is a metabolite associated with THC exposure. The investigators measured it to look for a signal of THC that might already have undergone metabolism. It was below the applicable quantification limits. THC-COOH, however, was not assayed in this experiment, so a claim about negative THC-COOH would exceed the record.

4. How many independent experimental subjects were there?

Three male minipigs. Multiple blood collections from the same animals do not enlarge that biological sample size. For printed Day 1 plasma CBD means, only two were included because one animal partially regurgitated its dose; Day 5 summaries included all three.

5. Can this study disprove a urine THC-COOH result after CBD use?

No. The work did not measure human urine THC-COOH, test the relevant urine immunoassay, inspect the individual’s CBD product or reproduce that person’s collection chronology. Separate, appropriately interpreted laboratory evidence is required to assess the urine finding and its possible sources.

6. Does CBD in stomach contents mean CBD entered whole blood at that concentration?

No. Stomach and intestinal contents are distinct from circulating blood, and the investigators reported separately measured plasma CBD values. The gastrointestinal numbers cannot be substituted for plasma or whole-blood concentrations or used to calculate an absorbed dose without additional validated information.

7. Is the study enough to establish a human impairment or driving rule?

No. It provides no human driving performance, impairment classification, last-use clock or collision evidence. Its legitimate contribution is a narrowly specified negative chemical finding after repeated oral CBD administration to three minipigs.

Bottom line for scientific and courtroom interpretation

In this original animal study, repeated oral synthetic CBD produced quantifiable CBD in plasma and GI contents while parent THC and 11-OH-THC remained below the applicable analytical quantification limits in those particular specimens. The study therefore weighs against a claim that the tested oral CBD regimen necessarily generated detectable THC in this intact animal system. The result is not a general declaration about every cannabinoid product, every unmeasured metabolite or every human exposure. Its most defensible forensic application is to separate chemical possibility, measured specimen evidence and person-specific legal conclusions.

Primary research: Wray L, Stott C, Jones N, Wright S. Cannabidiol Does Not Convert to Δ9-Tetrahydrocannabinol in an In Vivo Animal Model. Cannabis and Cannabinoid Research 2017;2(1):282–287. https://doi.org/10.1089/can.2017.0032.