Can Swallowing a Hemp Cosmetic Produce a Reactive Urine THC Screen When Skin Application Does Not?

The forensic distinction: reactive screen versus confirmed THC exposure

A urine immunoassay is designed to classify a specimen relative to a screening signal threshold. A reactive result is not the same analytical proposition as identification and quantitation of a specified cannabinoid metabolite by a selective confirmatory method. In the Jain study, the urine strips were calibrated against carboxy-THC (also written COOH-THC or THC-COOH) at a nominal 50 ng/mL cutoff. The investigators reported the number of reactive strips, but did not retain those urine specimens for an independent chemical determination of THC-COOH. That is the central boundary on interpretation.

The study also performed a different, more selective analysis on a different matrix: venous blood was processed into plasma and tested by validated liquid chromatography–tandem mass spectrometry (LC–MS/MS). Investigators could quantify cannabidiol (CBD) and its metabolite 7-hydroxy-CBD after oral administration. The delta-9-THC, delta-8-THC, 11-hydroxy-THC, THC-COOH and cannabinol signals in plasma were reported as undetectable or below the assay’s lower limit of quantification. That statement does not establish that urine contained no THC metabolite: plasma and urine are different specimens, with different test targets, concentrations and detection windows. Nor does it validate the reactive urine immunoassays.

A forensic reader must keep four separate facts apart: the composition of the administered formulation; directly assayed plasma cannabinoids; a presumptive urinary screening signal; and a person’s actual function or behavior. None can be silently substituted for another.

What did the researchers actually give the participants?

Twenty healthy adults aged 18–50 enrolled in an open-label, randomized-sequence, two-route crossover study. Eighteen completed both periods; two did not return for the second period. Among completers, 13 were women and five were men, with a reported mean age of 29.8 years. Participants received the same natural hemp extract-based cosmetic (NHEC) formulation on separate study visits, with a 15-day washout between routes.

The experimental dose used 200 microliters of CBD-rich hemp blend containing approximately 198 mg cannabidiol, mixed with 0.8 mL medium-chain triglyceride oil. At one visit, the 1-mL formulation was swallowed; at the other, the formulation was applied to the forearm. Participants fasted overnight, received a meal one hour after administration, and ate lunch four hours afterward. These are controlled research procedures, not demonstrations of everyday accidental exposure. In particular, the product was a cosmetic formulation and not established by this small trial as safe for routine oral use.

The product’s reported THC content was below the study’s product-testing quantification limit. “Below quantification” does not mean a molecular proof of zero THC, does not establish the composition of unrelated retail hemp cosmetics, and does not identify the substance responsible for a later urinary strip reaction. The study was exploratory rather than a large pre-specified diagnostic accuracy trial.

Why a crossover design matters

The same participants could contribute observations after each route, reducing some between-person variation. Yet a crossover does not itself identify the chemistry behind a screening reaction. Investigators also had to consider sequence, carryover and unmeasured pre-existing exposure. The stated 15-day washout and nonquantifiable plasma predose measurements limit some carryover concerns, but they do not substitute for a measured predose urine THC-COOH concentration, which was not obtained in this study.

What was directly measured, and in which biological specimen?

Venous blood was collected through a catheter at baseline, 15 and 30 minutes, and 1, 2, 3, 4, 5, 6, 8 and 24 hours. EDTA blood tubes were processed, and plasma was separated and stored at −80°C pending mass-spectrometric analysis. The measured drug concentrations reported in the pharmacokinetic figures are therefore plasma values, not venous whole-blood THC concentrations and not breath alcohol concentration (BrAC) or blood alcohol concentration (BAC).

The LC–MS/MS method targeted CBD, 7-hydroxy-CBD (7-OH-CBD), delta-9-THC, delta-8-THC, 11-hydroxy-THC, THC-COOH and cannabinol. The reported lower limit of quantification for CBD and 7-OH-CBD was 0.5 ng/mL. The THC-COOH calibration range started at 2.5 ng/mL; this is an assay characteristic of the plasma method, not a urine confirmatory threshold. A result below an assay’s lower quantifiable concentration must not be reported as proven chemical absence.

The urine testing was less specific: at the 24-hour visit, each participant’s urine was examined with a commercially available Easy@Home single-panel THC immunoassay strip, using a nominal COOH-THC screening cutoff of 50 ng/mL. The article states that the urine samples were discarded after testing. Accordingly, no retained same-sample urine confirmatory LC–MS/MS or GC–MS results can be compared to the reactive strips.

How different were CBD exposure levels after swallowing versus skin application?

The difference in measured systemic exposure was large under the tested conditions. For CBD, the reported geometric-mean area under the plasma concentration–time curve from 0 to 24 hours (AUC_0–24) was 281 h·ng/mL after oral administration, compared with 19 h·ng/mL after topical application. These are summaries of exposure over time—not concentrations present at any single moment, and not numbers that can be entered as a THC or BAC result.

The oral CBD geometric-mean peak plasma concentration (C_max) was 52.13 ng/mL, with a very wide reported 95% geometric confidence interval of 3.31–820.30 ng/mL. The oral median time to peak was three hours (reported range two to eight hours). The investigators did not compute a reliable topical CBD C_max, because quantifiable topical values were too sparse. That missing estimate should not be replaced with zero, a guessed number, or an unjustified percentage of the oral peak.

Oral 7-OH-CBD was also quantifiable: its reported geometric-mean C_max was 22.06 ng/mL, with a 95% geometric confidence interval of 2.17–224.27 ng/mL, and its AUC_0–24 was 166.77 h·ng/mL. Topical 7-OH-CBD values were not sufficiently detected or quantifiable to calculate corresponding parameters. Thus exposure after application to intact forearm skin was limited for this specific vehicle, but that does not establish that every transdermal formulation behaves similarly.

Were the published bioavailability figures direct measurements?

No. The paper presents estimated absolute CBD bioavailability of 10.18% by mouth and 0.69% via skin, calculated using previously published intravenous pharmacokinetic information. This study did not give participants an intravenous CBD reference dose for that comparison. Those percentages are model-derived cross-study estimates, not independently measured intravenous control arms within this crossover. The directly observed plasma AUC values and the modeled absolute bioavailability estimates must be reported as different kinds of evidence.

Why is 7 of 18 more precise than saying the cosmetic “caused a positive drug test”?

The study’s Table 3 reports one reactive oral-period urine result at the first visit and six at the third visit, totaling seven of the 18 completers. It reported zero reactive results after skin application. These are meaningful observed contrasts. But the evidentiary claim “caused a confirmed positive drug test” adds two unproved propositions: that the urine genuinely contained the confirmatory target at a particular concentration, and that administration caused its appearance rather than another exposure or analytical mechanism.

Neither proposition was resolved. The article says urine was discarded after the strip examination, so no analytically specific confirmatory result exists for those seven samples. The methods describe baseline blood measurements but do not establish baseline urinary THC-COOH testing before each intervention. Participants were asked to abstain from hemp products and marijuana, yet instructions alone are not an assay for prior or intervening exposure.

The commercial strip’s stated 50 ng/mL threshold is a calibrator-based screening threshold. Reading a reactive strip is not equivalent to measuring a person’s urine THC-COOH concentration at exactly 50 ng/mL, much less at a value above that in a legal evidentiary sense. It is also not equivalent to measured plasma THC, which remained below the reported quantitation capabilities.

Does a nonquantifiable plasma THC result contradict a reactive urine strip?

It identifies an unresolved difference between methods and matrices, not necessarily a contradiction. A selective plasma assay and a competitive urine immunoassay answer different questions. The compounds measured, practical limits, expected timing and sample handling differ. Without a confirmatory test of the same urine specimens, one cannot determine whether the strips correctly detected a THC metabolite, reacted to some other contributor, or reflected a testing anomaly. The study provides an important question for further research rather than a definitive biochemical explanation.

Did the investigators test whether cannabidiol itself fooled the urine strip?

Yes, on a restricted laboratory scale. To explore cross-reactivity, the researchers fortified three separate blank urine lots with selected concentrations of CBD or CBD metabolites, THC, and combinations. Under the selected conditions, CBD, 7-hydroxy-CBD and carboxy-CBD alone did not produce reactive THC strip results. That finding does not support a simple statement that these pure compounds alone made the seven participant strips positive.

There was also a revealing interference finding: when a low THC concentration was challenged in the presence of a high CBD concentration, the investigators observed nonreactive strips. That experiment suggests a possible false-negative interference under those artificial laboratory conditions—the opposite direction from the participant’s unexplained reactive strips. Selected spiked specimens, however, cannot establish the sensitivity, specificity or interference behavior of every urine matrix, product metabolite mixture or testing platform.

The original authors discussed several potential explanations, including limited product assay sensitivity and chemical transformations. None was experimentally established as the cause of the seven reactive urine specimens. In particular, this study did not demonstrate human gastric conversion of CBD to THC: no before-and-after THC formation was directly identified by an appropriately specific urinary confirmatory assay. A scientifically defensible report retains that uncertainty.

What would a confirmatory forensic investigation require?

A rigorous assessment of a disputed result would start with the actual test report and the test manufacturer’s details. The reviewer should identify the sampling date, collection conditions, specimen type, screening platform, nominal cutoff, quality controls and any laboratory-specific acceptance criteria. Crucially, determine whether a second selective method—typically mass spectrometry appropriately validated for urine—was performed on the same specimen. If confirmation was performed, the actual identified analyte, method, reporting threshold and concentration matter more than generalizations about a consumer product.

Additional scientific questions may include whether an authentic product sample from the relevant lot was analyzed by an appropriate method with a known detection/quantification limit; whether the person’s prior exposures and collection timeline are documented; whether specimen integrity was preserved; and whether pre-dose or repeat results exist. None of these may be assumed from this study alone. A result from one hemp formulation administered in a research clinic does not predict the impurity profile of a different batch of cream, oil, capsule or tincture.

For an explanation of the methodological difference between screening and confirmation, see confirmatory drug testing. The distinction among plasma, serum, whole blood and urine is addressed further in biological matrices in toxicology. Collection and preservation issues are discussed in specimen collection for toxicology. These contextual resources do not replace the original study or the actual case laboratory records.

Does the study support a personal impairment conclusion?

No. No driving task, psychomotor examination, accident reconstruction, eyewitness test, sobriety assessment or measured real-world impairment endpoint appears in the experiment. Plasma CBD exposure is not a numerical substitute for measured THC or demonstrated effects on behavior. The oral cosmetic period’s strip reactivity does not establish a level of intoxication, time of cannabis use, the credibility of a witness, or the cause of a collision.

Even a properly confirmed urinary THC-COOH result would be evidence about an analyte in a urine specimen, not by itself a direct measurement of a person’s contemporaneous central nervous system performance. Questions about causal exposure, the identity and dose of ingested material, and an individual’s functional state require additional evidence matched to the question. The present work offers no validated patient-level decision threshold for such conclusions.

Can the topical findings be generalized to all hemp creams or cannabinoid patches?

No. The topical intervention placed one particular medium-chain-triglyceride-based cosmetic formulation onto forearm skin. It was not designed as a specialized systemic transdermal drug-delivery patch. Different formulations may contain penetration enhancers, varied cannabinoids and concentrations; skin location, dose, duration, damage, dressing and application frequency can alter exposure. The small study does not establish a universal “topical CBD never enters blood” principle.

The topical plasma CBD values were frequently below quantification, but some were measurable and the reported AUC_0–24 estimate was not zero. At the same time, no urine strips were reactive after this particular single topical exposure among 18 completers. Those findings describe this experimental formulation and protocol, not every imaginable skin product or repeated high-dose exposure. The absence of a reactive strip is also not synonymous with a quantified demonstration that a metabolite was absent from urine.

What are the study’s main limitations for forensic use?

Sample size and purpose: only 18 participants completed both interventions, and the experiment was explicitly exploratory. It was not powered or designed to calculate forensic false-positive rates for the public or for different brands of urine testing.

Open-label routes: investigators and participants could distinguish swallowed oil from skin application. Although the sequence was randomized and visits separated, no masking removed all procedural differences.

Missing urine confirmation: the principal endpoint of forensic interest—the seven reactive strips—was not accompanied by urine mass-spectrometric confirmation, and specimens were discarded. The study could not identify the actual urine compound triggering those reactions.

Missing baseline urine target measurements: without measured predose urinary THC-COOH, the study cannot conclusively assign a reactive result to newly formed analyte from the product. Predose plasma measurements do not close that gap.

Product and matrix constraints: tested product composition was reported relative to assay limits. Plasma LC–MS/MS results cannot establish urine chemistry; modeled bioavailability drew on external intravenous literature; the topical data were too sparse for several pharmacokinetic parameters.

Behavioral scope: participants were not studied as drivers, employees undergoing forensic confirmation or impaired people. Any extrapolation from a consumer urinary strip to individual clinical or legal function would require independent evidence.

There were also two reported adverse-event observations, including a vasovagal episode associated with intravenous blood sampling and elevated liver enzymes associated with the product in one participant. With the available design, these observations neither prove that oral ingestion of cosmetics is safe nor quantify a general adverse-event risk.

Seven frequently asked questions about hemp cosmetics and THC urine screens

1. Did approximately 39% of volunteers have a confirmed THC-positive urine test?

No. Seven of 18 completing participants had reactive presumptive immunoassay strips after oral administration. The investigators did not confirm the analyte or quantify THC-COOH in those urine specimens.

2. Were any urine strips reactive after using the formulation on the forearm?

No. The study reported zero of 18 reactive strips following that specific single topical application. It cannot rule out different findings with other products, doses, schedules, damaged skin or analytical platforms.

3. Was THC found in the participants’ blood?

The separately tested plasma THC-related analytes were undetectable or below the method’s lower quantification thresholds. That is not the same as demonstrating absolute absence in venous whole blood or identifying what caused a urinary screen reaction.

4. Did swallowing CBD prove that the body converted CBD into THC?

No. The study did not directly establish that conversion pathway in people. The urinary screening observations are unexplained without specific confirmatory tests and adequate baseline controls.

5. Does a 50 ng/mL urine strip mean the participant’s THC-COOH concentration was exactly 50 ng/mL?

No. It is a nominal screen calibrator cutoff used to distinguish strip reactions, not a precise mass-spectrometric measurement of the participant’s urine concentration.

6. Can the reported oral CBD plasma peak be used as a THC or impairment threshold?

No. The 52.13 ng/mL geometric-mean oral peak described CBD in plasma. It was not a THC concentration, whole-blood measurement, BAC, driving score or validated cutoff for impairment.

7. What is the strongest lesson for a disputed employment or forensic screen?

Confirm the specimen and method before interpreting the result. A presumptive urine THC reaction after hemp product exposure may warrant a specific, independently validated confirmatory analysis, product composition review and chronology assessment. The research does not permit a blanket inference either that all such screens are false or that they establish THC use and impairment.

Bottom line for scientific and litigation analysis

Jain and colleagues observed substantially greater systemic CBD exposure after swallowing the tested hemp cosmetic than after applying it to the forearm. They also observed seven reactive 24-hour urine THC strips after the oral route and none after the topical route, while plasma THC-related analytes were below the reported quantification limits. The unresolved analytical issue is precisely what a selective same-urine confirmation could have answered—but was not performed. Accordingly, the study supports careful distinction between administration route, measured plasma CBD, unconfirmed urine immunoassay response, and actual impairment, not a universal rule about cosmetics and drug tests.

Original primary study: Jain M, Hudson R, Tarrell A, Green DJ, Clifford JJ, Watt K, Mihalopoulos N, Rower JE, Yellepeddi V, Enioutina EY. Results of an Exploratory Crossover Pharmacokinetic Study Evaluating a Natural Hemp Extract-Based Cosmetic Product: Comparison of Topical and Oral Routes of Administration. Pharmaceuticals. 2026;19(2):231. doi:10.3390/ph19020231.